A node importance evaluation method integrating static and dynamic indicators with management decision requirements Authors: Zhengfeng Ma, Haoyun Yang Published: 2026 ABSTRACT Focusing on the needs of road network node importance evaluation and the limitations of existing methods, a novel method to comprehensively evaluate road network node importance considering the influence of operating state changes is proposed. First, an evaluation index system combining static and dynamic indexes is constructed. Therefore, an impact evaluation model of the topological potential is proposed to reflect the operating state of the road network. Second, the integrated weighting method based on multi-objective programming is proposed to determine the final weight of the evaluation index and the sorting result of node importance; further, the particle swarm optimization algorithm is employed to solve this problem. Third, a node grading algorithm based on various sorting results is proposed. Finally, an experimental scheme is designed which verifies the rationality of our proposed methods, algorithms, and theories based on a case study in Xining City. A change in the evaluation index influences the importance of nodes and their sorting, and the dynamic index of the change in road network operation state should be considered. The change in the index weight significantly influences the evaluation of node importance and its sorting and the weight of the index should be reasonably determined; important nodes are more reliable when graded according to the internal relationship and rules of various sorting results. Our proposed method can determine the index weight, node importance sorting, and evaluate the importance of nodes in road networks. Introduction As an important part of the urban road network, intersections (hereinafter referred to as nodes) play an important role in the operation of urban road traffic. In the actual traffic scene, once the traffic congestion has been triggered by accidents, formulating targeted traffic control measures for important nodes to effectively prevent or significantly alleviate the traffic congestion is a solution to ensure the normal operation of urban traffic. The accurate and reasonable determination of important nodes of transportation systems by the traffic administration department is critical. Therefore, the use of a reasonable method to evaluate the importance of road network nodes has increasingly attracted the attention of traffic researchers and administrators. In recent years, studies on the evaluation of the importance of road network nodes mainly focus on three categories: the simple method of a single index model, the comprehensive evaluation method based on multiple indexes, and the attack model method based on a complex network. The simple method based on a single index model mainly reflects the importance of nodes by constructing representative evaluation indexes. The main results are as follows: In 2015, the inverse of generalized travel time was presented as a utility function to quantify the influence of external disturbances on the transportation system and used to determine key transportation facilities in1; Zeng et al.2 proposed a novel node importance evaluation method based on attraction between nodes. According to their algorithm, the importance of a node may be measured by the product of the node trust value and the sum of the node’s attractiveness to other nodes in the network; Liu et al.3 studied the influence of the number of lanes on the importance of the node and proposed an improved spatial weighting model. In addition, by considering the correlation coefficient of traffic flow between intersections, a method of measuring the importance of urban road network nodes with H index was proposed; Xu et al.4 proposed a method to sort the importance of nodes in the urban road network through data mining of trajectories of numerous urban taxis; Lu et al.5 started from the node, combined with the topological structure around the node, and identified the node importance and its propagation ability based on the node clustering H-index mixture centrality. Kunt et al.6 adopted a node shrinking method based on network cohesion, which comprehensively considered the degree and location of nodes to determine their importance. Deng et al.7 proposed the concepts of generalized edge betweenness and generalized edge clustering coefficients based on node betweenness and node coefficient, and then proposed an evaluation method on node and edge-based node importance. Zhao et al. 8 regarded financial institutions as a node in the financial system network and measured the importance of financial institutions in the financial network by calculating the adjacency and information entropy of each node. The method is relatively simple and is easy to operate; however, it has few index factors and a certain one-sidedness. To overcome the shortcomings of the first method, some scholars introduced the method of comprehensive evaluation to study the importance of nodes. In 2011, Wang et al.9 selected the node degree, node betweenness, and peak hour traffic flow as the indexes to evaluate the importance of nodes in a traffic network. Through these indexes, information such as network topology, shortest path, and traffic flow status may be integrated to grade the importance of nodes using C cluster analysis. Meanwhile, Song et al.10 incorporated the urban flow intensity into the node importance evaluation index system according to the characteristics of urban agglomerations and adopted factor analysis as an objective method to avoid the randomness of subjective values in the calculation of node importance. Here, the importance level of each node is distinguished owing to the adoption of the K-means clustering. Consequently, Cao et al.11 used the analytic hierarchy process (AHP) to determine the weight of each index, in which the node traffic flow was defined as the sum of the traffic flow on the roads connecting the nodes by selecting the degree and betweenness of nodes in the topology network, urban road level, and node traffic flow as the evaluation indexes of node importance. Lu et al.12 improved the traditional neighborhood structure hole index from the perspective of topology and network characteristics between nodes and their neighbors combined with the improved traditional TOPSIS method using relative entropy. The key of this method is to calculate the objective weight of the evaluation index using the Gini and Kendall coefficients. Finally, the improved index may be applied to implement the importance evaluation of network nodes and calculate the node importance. Fan et al.13 analyzed the nodes in the power communication network from the physical topology, traffic distribution, and service importance distribution. Further, a key node identification algorithm was constructed by adopting a multilayer network to identify important nodes. Despite the differences in the vulnerability indexes of different layers of the communication network, the method which comprehensively included multiple attribute indexes, that influences the evaluation of node importance, such as node degree, node betweenness, and traffic flow, may overcome the one-sidedness of the first method and has certain advantages. Unfortunately, neither of them considered the influence of node change on the cascading failures on the traffic network. Consequently, some scholars have introduced the concept of complex network theory into various networks and determined the importance of nodes based on the impact of node changes on the network with cascading failures. In 2013, Wang et al.14 established and presented the node importance evaluation method of a transportation network considering cascading failure of the transportation network based on node deletion. Candelieri et al.15 used network connectivity and efficiency loss to determine important nodes and used multi-graph to model public transportation networks and determined the attack strategy based on node degree and betweenness. Tang et al.16 proposed a weighted K-order spreading number algorithm based on network topology to evaluate the importance of nodes. Particularly, the nodes were set as infection sources, the total number of infected nodes was defined as the K-order spreading number after the spreading time K, and the important nodes were determined by the change in the network efficiency caused by attacking the node (removing all connected edges of the node). Wang et al.17 used the average weighted efficiency of the network after cascading failures to measure the consequences of cascading failures (efficiency model) to determine key nodes in wireless sensor networks. Yu et al.18 proposed a method to evaluate the importance of nodes in the case of cascading failures, which used the degree of a node and its neighbor nodes to define the node load and provided the transmission method of the node load in the case of node failure. In this algorithm, the change in the total number of edges before and after the network was used to measure the importance of the node. In addition, the duality theory was introduced to construct a road network topology map with road segments as nodes19. The underlying principle of this method is that traffic congestion has been regarded as a node attack mode and the precision of the state vulnerability assessment is improved by considering the impact of congestion. Meanwhile, Cao et al.20 determined the attack mode according to the node degree and betweenness. Accordingly, the road network efficiency model was established to determine the network change before and after node removal and the vulnerability and node importance of the bus network were studied. References16–18 mainly discussed local impact evaluation models, which reflect the impact of node changes on the network through self-defined evaluation indexes using the network topology as input. The other part14,15,19,20 belongs to the global impact evaluation model, which mainly reflects the impact of node changes on the network through changes in network efficiency. The common denominator of the two methods is based on the network topology and can reflect the impact of node changes on the network. However, they did not consider other node attributes. Therefore, the determination of a reasonable algorithm to evaluate the importance of the nodes by combining the comprehensive evaluation method is a key problem in the attack modes environment. Therefore, Du et al.21 proposed an improved topological potential model to calculate the importance of subway network nodes. A network adjacency matrix was constructed in the method based on the passenger flow data, and the in-degree, out-degree, closeness, maximum eigenvalue, and betweenness centralities of the nodes were calculated. The result of the comprehensive weighting of the five indexes was used as a fixed attribute of the topological potential model, and the topological potential value and importance sorting of the nodes were calculated. Finally, the proposed model was verified using the node removal method and the change in network efficiency. In summary, scholars in various fields have performed extensive studies on the evaluation of node importance and have achieved a series of research results. However, the evaluation of node importance of urban road networks has its particularity and none of the existing methods can meet the actual needs. The main reasons are: first, it takes a process for nodes to change from “unimpeded” to “congested” and the congestion levels also differ. Existing methods do not include the impact of non-physical damage on nodes when studying the impact of node changes on the road network. It is inappropriate to use the method of direct removal. Second, the traffic congestion of the node influences the entire road network and the local impact evaluation model is apparently unreasonable. Regarding the global impact evaluation model, based on studies, with an increase in the number of road network nodes, the network efficiency model cannot properly reflect the impact of node changes on the road network. Third, owing to the surrounding environment of the node (such as restricted areas) and traffic control measures (such as traffic restrictions), the location in the road network and topology are similar and the actual traffic operation of nodes with the same impact on the road network also differs. Therefore, other indices should be comprehensively considered to evaluate the importance of nodes. Although the comprehensive evaluation and attack models have been combined by scholars and used to evaluate the node importance of the subway network, in-depth studies on the impact of index selection and index weight on the evaluation of node importance are still lacking. Contributions of this study To meet the needs of evaluating the importance of nodes in an urban road network and simultaneously overcome the problems existing in the current methods, a novel method is proposed to comprehensively evaluate the importance of nodes in a road network considering the influence of changes in the operating state. The main contributions of this study are as follows:This method addresses a long-standing problem in the field of node importance evaluation, specifically how to quantitatively model changes in node operational status and integrate them with the ranking and weighting processes, which has not been systematically addressed in previous research.The influence of the change in the operating state of nodes on the road network is used as an index to evaluate the importance of nodes and a static and dynamic evaluation index system is established which considers the inherent topological properties of the road network and also considers traffic attributes. Further, the influence of node changes on the road network is described in detail.A road network operation state evaluation model based on the topological potential is proposed, which overcomes the failure of the traditional efficiency evaluation model as the number of nodes increases.An integrated weighting method based on multi-objective programming is proposed, which overcomes the problem that the existing combined weighting methods lead to different evaluation results of node importance owing to different principles.Considers the internal relationship between different sorting results; a node grading method based on different sorting results is proposed to increase the reliability of important nodes. This method addresses a long-standing problem in the field of node importance evaluation, specifically how to quantitatively model changes in node operational status and integrate them with the ranking and weighting processes, which has not been systematically addressed in previous research. This method addresses a long-standing problem in the field of node importance evaluation, specifically how to quantitatively model changes in node operational status and integrate them with the ranking and weighting processes, which has not been systematically addressed in previous research. The influence of the change in the operating state of nodes on the road network is used as an index to evaluate the importance of nodes and a static and dynamic evaluation index system is established which considers the inherent topological properties of the road network and also considers traffic attributes. Further, the influence of node changes on the road network is described in detail. The influence of the change in the operating state of nodes on the road network is used as an index to evaluate the importance of nodes and a static and dynamic evaluation index system is established which considers the inherent topological properties of the road network and also considers traffic attributes. Further, the influence of node changes on the road network is described in detail. A road network operation state evaluation model based on the topological potential is proposed, which overcomes the failure of the traditional efficiency evaluation model as the number of nodes increases. A road network operation state evaluation model based on the topological potential is proposed, which overcomes the failure of the traditional efficiency evaluation model as the number of nodes increases. An integrated weighting method based on multi-objective programming is proposed, which overcomes the problem that the existing combined weighting methods lead to different evaluation results of node importance owing to different principles. An integrated weighting method based on multi-objective programming is proposed, which overcomes the problem that the existing combined weighting methods lead to different evaluation results of node importance owing to different principles. Considers the internal relationship between different sorting results; a node grading method based on different sorting results is proposed to increase the reliability of important nodes. Considers the internal relationship between different sorting results; a node grading method based on different sorting results is proposed to increase the reliability of important nodes. Organization of this paper To describe the novel method of the comprehensive evaluation of the importance of nodes in a road network, the remainder of the paper is divided as follows. In Section II, the concept, model, data acquisition method, and the relationship between the importance of nodes and the related static indexes of road network node importance evaluation are introduced. Second, the concept of dynamic indexes and index data acquisition method based on the topological potential model are explained. Next, for the follow-up study of the impact of the evaluation index on the node importance evaluation, a set of evaluation index alternatives is provided. Finally, the preprocessing method of the evaluation index data is introduced. In Section III, based on the subjective and objective weighting method and the combined weighting method, an integrated weighting method based on the multi-objective programming model is proposed and its solution method is provided. In Section IV, the node sorting and grading methods are introduced; a node grading algorithm based on various sorting results is proposed in particular. In Section V, based on the case of the intersection in the core area of the urban road network in Xining City, Qinghai Province, China, an experimental scheme is designed to verify the rationality of the methods, algorithms, and theories proposed in this study. Section VI provides the conclusions of this study and future research perspectives. Evaluation indexes and data preprocessing Evaluation indexes Undoubtedly, the importance of road network nodes is not only related to static indexes such as node degree, clustering coefficient, and betweenness but also has a greater relationship with the impact of changes in the node traffic operation state on the road network. To scientifically evaluate the importance of road network nodes and to include static indexes such as node degree, clustering coefficient, betweenness and average traffic volume, and average saturation of nodes during peak periods, the dynamic index (process evaluation index) of the impact of the change in the traffic operation state of a node on the road network is also considered. Next, the concept of each index and the evaluation models are explained in detail. Static evaluation index To better distinguish static and dynamic indexes, the indexes whose evaluation data acquisition are independent of the change value in road network operation state are defined as static evaluation indexes, such as node degree, clustering coefficient, betweenness and average traffic volume, and average saturation during peak hours. (1) Node Degree. The node degree of a road network refers to the number of edges connected to a node, which is composed of out-degree and in-degree. The relevant definitions are as follows:1\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$K_{i}^{out} = \sum\limits_{j = 1}^{n} {e_{ij} }$$\end{document}2\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$K_{i}^{in} = \sum\limits_{j = 1}^{n} {e_{ij} }$$\end{document}3\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{array}{c}{K}_{i}=ceil\left[\frac{{K}_{i}^{out}+{K}_{i}^{in}}{2}\right]\end{array}$$\end{document}where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$e_{ij}$$\end{document} is the adjacency matrix element, 1 means node i is connected to node j, and 0 means the node is not connected; \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\sum\limits_{j = 1}^{n} {}$$\end{document} is the summation symbol (similarly hereinafter), where j = 1, 2, … , n(n is the total number of network nodes); ceil means rounding up to a whole number. The greater the node degree of the road network, the more the road segments connected to it, and the greater the impact of the node on the road network. Therefore, for road network traffic operation, the greater the node degree, the more important the intersection. Practically, the out-degree of a node is the sum of the elements in the adjacency matrix with the value of 1 in row i and the in-degree is the sum of the elements in the adjacency matrix with the value of 1 in column j. (2) Clustering Coefficient. The clustering coefficient of nodes in a road network refers to the ratio of the number of edges that exist among node neighbors to the number of possible edges. The specific definition is as follows:4\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{array}{c}{C}_{i}=\frac{2E\left(i\right)}{{K}_{i}\left({K}_{i}-1\right)}\end{array}$$\end{document}where E(i) refers to the number of edges that actually exist between the neighbor nodes of node i and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$K_{i}$$\end{document} is the degree of node i . The larger the node clustering coefficient, the smaller the degree of dependence of adjacent nodes on it, and the smaller the impact on the road network. Therefore, the smaller the node clustering coefficient, the greater the importance of the node. (3) Betweenness. The betweenness of a road network node is defined as the ratio of the number of shortest paths passing through the node to the total number of shortest paths in the network. The specific definition is as follows:5\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{array}{c}{B}_{i}=\sum_{i,j\in V\left(j\ne k\right)} \frac{{n}_{j\dot{k}}\left(i\right)}{{n}_{j\dot{k}}}\end{array}$$\end{document}where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$n_{jk} (i)$$\end{document} is the number of shortest paths passing through node i and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$n_{jk}$$\end{document} is the total number of shortest paths in the network. The betweenness reflects the role and influence of the node in the entire network. The larger the value, the greater the influence of the node on the entire road network, and the more important it is. Practically, the adjacency matrix can be used as the input and the Floyd algorithm in python can be used to solve the problem; consequently, the betweenness value of the node can be further obtained by statistics. (4) Average Traffic Volume during Peak Hours. The traffic volume of the urban road network is characterized by a typical bimodal curve. Owing to factors such as commuting and school time, travel is too concentrated during the morning and evening rush hours, and the traffic demand exceeds the supply, resulting in a traffic jam. Conversely, the traffic volume of the urban road network during non-peak hours is relatively small, and traffic congestion does not occur. Therefore, to improve the efficiency of traffic operation and reduce traffic congestion, studies should be focused on peak hours. Therefore, the average traffic volume during peak hours is selected as an evaluation index of node importance, which is defined as follows:6\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$V_{i} = \frac{1}{{N_{t} }}\sum\limits_{t} {\sum\limits_{l} {V_{ilt} } }$$\end{document}where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$V_{ilt}$$\end{document} is the expanded hourly traffic volume of node i from entry l(l = 1,2, ⋯ ,L) in the statistical period \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$t(t = 1,2, \cdots ,N_{t} )$$\end{document},L is the total number of onwards of node i , and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$N_{t}$$\end{document} is the total number of peak statistical periods (usually 5 min). Compared with other intersections, if an intersection with a large average traffic volume during peak hours fails or develops traffic congestion, it will have a greater impact on the entire road network. Therefore, the intersection with higher traffic volume is more important. (5) Average Saturation during Peak Periods. Although the average traffic volume during peak hours can reflect the role of nodes in the traffic operation of the road network, it cannot directly reflect the congestion level of nodes. Therefore, the average saturation index of the peak period of the node is introduced. Saturation, also known as the intersection traffic load factor, represents the ratio of the actual traffic volume to the traffic capacity at the intersection. The average saturation index during peak periods can be defined as follows:7\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\alpha_{i} = \frac{{V_{i} }}{{CAP_{i} }} = \frac{{\frac{1}{{N_{t} }}\sum\limits_{t} {\sum\limits_{l} {V_{ilt} } } }}{{\sum\limits_{l} {\sum\limits_{d} {S_{ld} \frac{{g_{ld} }}{{C_{t} }}} } }}$$\end{document}where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$CAP_{i}$$\end{document} represents the traffic capacity of intersection i ; \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$S_{ld}$$\end{document} represents the saturated flow rate of lane d of entry l of the intersection; \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$g_{ld}$$\end{document} represents the effective green light time of lane d of entry j ; \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$C_{t}$$\end{document} represents the intersection cycle time. The average saturation attribute during peak periods reflects the operation of the intersection and effectively distinguishes the congestion of the intersection. The larger the saturation value of the intersection, the more congestion, and the greater the impact on the surrounding road network. Therefore, for traffic administration, the larger the saturation value, the more important the intersection. Because the average saturation during peak hours is strongly correlated with the average traffic volume during peak hours, the two attributes are optimized in this study. Dynamic evaluation index Contrary to the static evaluation index, the dynamic evaluation index mainly reflects the influence of the change of the node operation state on the road network operation state. The traditional efficiency evaluation model reflecting the impact of node changes on the network is as follows:8\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{array}{c}E=\frac{1}{n\left(n-1\right)}\sum_{i\ne j} \frac{1}{{d}_{ij}}\end{array}$$\end{document} In the equation, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\mathrm{d}}_{\mathrm{i}\mathrm{j}}$$\end{document} is the shortest path length from node i to node j in the network. As observed, with an increase in the number of network nodes, the influence of node changes on the network cannot be properly distinguished with the efficiency evaluation model. To solve this problem, an impact evaluation model based on topological potential is proposed. Here, the operation state of the node is determined by the saturation reference value, the operation state of the road network is reflected by the topology potential of the road network, and the influence of the change in the operation state of the node on the operation state of the road network is determined using the change value of the operation state of the road network caused by the node switching between different saturation reference values. The above three aspects will be explained in detail. a) Saturation Reference Value of Node Operating State Determining the saturation reference value of the operation state of nodes in a road network is the basis for the study on the impact of node operation state changes on the road network. However, most of the current methods are based on norms and experience and are not sufficiently scientific backed. Owing to the differences in urban spatial layout, road structure, traffic source distribution, and residents’ travel habits, the changing trends of traffic operation state between different cities also differ. Therefore, when studying the impact of changes in node operating status on the road network, the changing trend of the average saturation of the nodes in the urban road network under study should be determined first. Because the urban residents’ travel time is mainly around the commute time (on and off school time), a method to determine the saturation reference value based on the regularity of residents’ travel time is proposed in this study. The specific steps are as follows: Step 1: the peak period (7:00–9:00 and 17:30–19:00) is selected and the value of the saturation level of the node at a certain time interval (5 min) is investigated and determined; Step 2: taking each time interval during peak hours as the horizontal axis, and taking the average saturation of each node in each time interval as the vertical axis, the time sequence diagram of the average saturation change of road network nodes is drawn. Step 3: The unimpeded and slightly congested saturation reference values of \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\alpha}_{i0}$$\end{document} and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\alpha}_{i1}$$\end{document}, respectively, of the node are determined according to the principle of "change in growth rate." Step 4: Further, the mean value of the saturation of road network nodes at each time interval is taken as the saturation reference value \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\alpha_{i2}$$\end{document} of congestion. Step 5: Finally, the saturation reference value \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\alpha_{i3}$$\end{document} where the mean value of the average saturation value is greater than the assumed saturation reference value is severely congested. To study the impact of changes in the node operating state on the traffic operation of the road network, after the saturation reference value of the road network node operating state is determined, the representation and evaluation method of the road network operating state should be clarified. b) Evaluation of Road Network Operation State The concept of particle topological potential in the microscopic world is introduced to better represent the operation state of the road network, treating the nodes of the road network as microscopic particles, and using the sum of the topological potential values of all nodes to evaluate the operating state of the road network. (1) Definition of Topological Potential. Definition 1: Given a network G = (V, E), where V = (v1, v2,…, vn), is a non-empty finite set of nodes, E ∈ V × V is a set of node pairs or edges, and ‖E‖ = m. Taking the network as a physical system containing numerous nodes, each node interacts with each other, has its own action field around it, and is not only influenced by itself but also the joint action of other nodes. A field is thus determined on the entire network topology, called the topology potential field22. From the definition of the potential function of the data field, the topological potential of any node vi ∈ V can be expressed as:9\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{array}{c}\varphi \left({v}_{i}\right)=\sum_{j=1}^{n} \left({m}_{j}\times {e}^{-(\frac{{d}_{j\to i}}{\sigma }{)}^{2}}\right)\end{array}$$\end{document}where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$d_{j \to i}$$\end{document} represents the shortest distance from node j to node i , σ represents the influence factor, meaning the influence range of each node, and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$m_{j}$$\end{document} represents the quality of node j , which can be used to show the inherent attribute of node j . (2) Definition of Potential Entropy. The accuracy of the topological potential is influenced by the value of the influence factor, and the influence factor can be optimized according to the potential entropy of the nodes in the network. The potential entropy is defined below. Definition 2: Given a network G = (V, E) and its topological potential field corresponding to a certain value , where V = (v1, v2,…, vn) is a non-empty finite set of nodes, E ∈ V × V is the set of node pairs or edges, let the topological potential of v1, v2,…, vn be \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\varphi (v_{1} )$$\end{document} , \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\varphi (v_{2} )$$\end{document} ,… \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\varphi (v_{n} )$$\end{document}, the potential entropy of the corresponding topological potential field is defined as:10\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{array}{c}H=-\sum_{i=1}^{n} \frac{\varphi \left({\nu}_{i}\right)}{Z}ln\left(\frac{\varphi \left({\nu}_{i}\right)}{Z}\right)={\sum}_{i=1}^{n}ln(\frac{Z}{\varphi \left({v}_{i}\right)}{)}^{\frac{\varphi \left({v}_{i}\right)}{Z}}=\sum_{i=1}^{n} \mathit{ln}{\left(\frac{\sum_{i=1}^{n} \sum_{j=1}^{n} \left({m}_{j}\times {e}^{-(\frac{{d}_{j\to i}}{\sigma }{)}^{2}}\right)}{\sum_{j=1}^{n} \left({m}_{j}\times {e}^{-(\frac{{d}_{j\to i}}{\sigma }{)}^{2}}\right)}\right)}^{\frac{\sum_{j=1}^{n} \left({m}_{j}\times {e}^{-(\frac{{d}_{j\to i}}{\sigma }{)}^{2}}\right)}{\sum_{i=1}^{n} \left({m}_{j}\times {e}^{-(\frac{{d}_{j\to i}}{\sigma }{)}^{2}}\right)}}\end{array}$$\end{document}where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$Z = \sum\limits_{i = 1}^{n} {} \varphi (v_{i} )$$\end{document} is a standardized factor. from the above equation: (1) If \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\sigma \to 0;\varphi \left(j\to i\right)\to 0;\varphi \left(i\right)={m}_{i}=M$$\end{document} , the potential entropy approaches \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$H_{\max } { = }\ln N$$\end{document}; (2) If \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\sigma \to \infty ,\varphi \left(j\to i\right)\to {m}_{j},\varphi \left(i\right)=NM$$\end{document} , after normalization by Z , the potential entropy again approaches \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$H_{\max } { = }\ln N$$\end{document}. (3) With the monotonous increase in the influence factor, the change in the potential entropy reaches a maximum value at both ends, and there will be a minimum value in the middle (representing a concave function characteristic); at this time, the potential distribution of the nodes is the most uneven and the uncertainty is the smallest. As shown in the Figure 1, it is the relationship diagram between the potential entropy and the influence factor for a network containing 5 nodes.Fig. 1Diagram of the Relationship between Potential Entropy and the Influence Factor. Diagram of the Relationship between Potential Entropy and the Influence Factor. (3) Determination of Topological Potential and Potential Entropy. Considering the characteristic that the potential entropy H exhibits a concave curve and takes the value \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$H_{\max } { = }\ln N$$\end{document} when \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\sigma \to 0$$\end{document} and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\sigma \to \infty$$\end{document} , the Triple Iteration Method is specifically selected, to determine the optimal influence factor value. The initial interval lower bound is set as a = \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\upsigma}_{1}$$\end{document}=0. Furthermore, a reasonable step size \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\vartriangle {\upsigma }$$\end{document} is taken (e.g.,\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${ }\vartriangle {\upsigma }$$\end{document} = 5), and we set \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\upsigma}_{2}$$\end{document} = \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\vartriangle {\upsigma }$$\end{document} and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\upsigma}_{3}$$\end{document} = \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\upsigma}_{2}$$\end{document} + \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\vartriangle {\upsigma }$$\end{document}. If H(σ_2) > H(σ_3) , then we set \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\upsigma}_{2}$$\end{document} = \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\upsigma}_{3}$$\end{document} and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\upsigma}_{3}$$\end{document} = \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\upsigma}_{2}$$\end{document} + \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\vartriangle {\upsigma }$$\end{document}, until the condition H(\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\upsigma}_{2}$$\end{document}) < H(\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\upsigma}_{3}$$\end{document}) is met. Then we take the upper limit of the triple iteration interval as b = \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\upsigma}_{3}$$\end{document}=\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\upsigma}_{3}^{*}$$\end{document}. Once the lower and upper bounds of the initial interval are determined, the optimal influence factor value \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\upsigma }^{*}$$\end{document} and the topological potential H can be further calculated according to the following Triple Iteration Algorithm flowchart (Fig. 2).Fig. 2Flowchart of the Triple Iteration Method for Calculating the Influence Factor. Flowchart of the Triple Iteration Method for Calculating the Influence Factor. (4) Evaluation Model of Road Network Operation State. The traffic congestion of a road network node is successively spread to its adjacent nodes. This has an impact on the traffic operation of the entire road network. That is, the change in the congestion situation of a local intersection not only changes its topological potential but also influences the topological potential of other intersections in the road network. To scientifically evaluate the impact of changes in the operating state of road network nodes on the road network, according to the definition of the network topology potential field, an evaluation model for the overall operating level of road network nodes is proposed, as shown in (18).11\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{array}{c}\varphi =\sum_{i=1}^{n} \varphi \left({\nu}_{i}\right)=\sum_{i=1}^{n} \sum_{j=1}^{n} \left[{m}_{j}\times {e}^{-(\frac{{d}_{j\to i}}{\sigma }{)}^{2}}\right]\end{array}$$\end{document} After the issues on the grading of node congestion level and evaluation of road network operation state are solved, the impact of node operation state changes on the road network operation state can be studied. c) Evaluation of the influence of node operation state change on road network operation state In part a) of the Dynamic Evaluation Index, urban traffic operational status is categorized into four states: unimpeded, slightly congested, congested, and severely congested. Considering that the transition of traffic status is a progressive process, theoretically, there should be six scenarios of status changes (unimpeded to slightly congested, slightly congested to congested, congested to severely congested, severely congested to congested, congested to slightly congested, and slightly congested to unimpeded). In practical traffic management, these six scenarios should be investigated separately to identify the critical nodes for each situation in a targeted manner. Due to manuscript space limitations, and to demonstrate the impact of bidirectional status changes (e.g., unimpeded to slightly congested and slightly congested to unimpeded) as well as different scenarios of same-direction changes (e.g., unimpeded to slightly congested and congested to severely congested) on the identification of critical nodes, three specific transitions (unimpeded to slightly congested, slightly congested to unimpeded, and congested to severely congested) were selected for illustration. (1) The impact \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\Delta \varphi_{{_{1i} }}$$\end{document} on the road network when the operating state of the node changes from unimpeded to slightly congested. This process is mainly used to study the influence on the topological potential of the entire road network after the operation state of a single node is changed from unimpeded to slightly congested. The importance of each node can be sorted by comparing the influence of each node. The more important the node, the greater the impact on the road network after congestion; thus, the congestion of the node should be totally avoided. The major study methods are as follows: d) The 0–1 adjacency matrix \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$A = \left( {a_{ij} } \right)_{n \times n}$$\end{document} of the road network intersection is input and the optimal influence factor σ of the topological potential model is calculated through the ternary iterative algorithm. e) The intersection standard saturation matrix \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${B}_{1}={\left({b}_{ij}\right)}_{n\times n}$$\end{document} is input, when \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$a_{ij}$$\end{document} is a non-zero element, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$b_{ij}$$\end{document} assumes the initial value \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\alpha_{i0}$$\end{document} of the unimpeded saturation of the node, and when the off-diagonal element \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$a_{ij}$$\end{document} is 0 element, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$b_{ij}$$\end{document} assumes ∞. The shortest path weight matrix \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$D_{1} = \left( {d_{ij} } \right)_{n \times n}$$\end{document} is calculated using the Floyd algorithm and the topological potential φ of the benchmark road network is calculated using the topological potential model. f) The saturation of intersection i is changed from the unimpeded saturation reference value \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\alpha_{i0}$$\end{document} to the slight congestion saturation reference value \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\alpha_{i1}$$\end{document} , and the element in the standard saturation matrix that involves intersection and whose value is the initial value of the unimpeded saturation to ( \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\alpha_{i0}$$\end{document} + \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\alpha_{i1}$$\end{document} )/2 is changed, the changed saturation matrix is denoted as \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$B_{2}$$\end{document} . g) Using the optimal influence factor σ and saturation matrix \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$B_{2}$$\end{document} , the topology potential model is used to further calculate the topological potential \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\varphi{\prime}$$\end{document} of the road network after the operation state of node i is changed from unimpeded to slightly congested. h) Because \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\Delta \varphi_{{_{1i} }} = \left| {\varphi{\prime} - \varphi } \right|$$\end{document}, the topological potential change of the road network is further calculated after the operation state of node is changed from unimpeded to slightly congested. i) Repeat ③–⑥, until all the topological potential changes of the road network after the operation state of all nodes is changed from unimpeded to slightly congested are obtained. (2) The impact \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\Delta \varphi_{{_{2i} }}$$\end{document} on the road network when the node operation state changes from slightly congested to unimpeded. This process is mainly used to study the influence on the topological potential of the entire road network after the operation state of a single intersection is changed from slightly congested to unimpeded. By comparing the impact of each intersection, its importance can be sorted. The more important the intersection, the greater the impact on the road network after it is restored to be unimpeded; thus, the state of congestion should be swiftly restored. The study method is similar to that when the node operation state is changed from unimpeded to slightly congested. The difference is that when the element \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$a_{ij}$$\end{document} of the 0–1 adjacency matrix is a non-zero element, the standard saturation matrix element \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$b_{ij}$$\end{document} assumes the initial value \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\alpha_{i1}$$\end{document} of the slightly congested saturation, and the state change process involves changing the saturation of intersection i from \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\alpha_{i1}$$\end{document} to \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\alpha_{i0}$$\end{document}. (3) The impact \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\Delta \varphi_{{_{3i} }}$$\end{document} on the road network when the node operating state changes from congestion to severe congestion. This process is mainly used to study the impact on the topological potential of the entire road network after the operation state of a single intersection changes from congestion to severe congestion. By comparing the impact of each intersection, its importance can be sorted. This result is significant in that the more important the intersection, the greater the impact on the road network after its operation status is changed to severe congestion; thus, severe congestion should be totally avoided. This study method is similar to that when the node operation state is changed from unimpeded to slightly congest. The difference is that when the element \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$a_{ij}$$\end{document} of the 0–1 adjacency matrix is a non-zero element, the element \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$b_{ij}$$\end{document} of the standard saturation matrix assumes the initial value \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\alpha_{i2}$$\end{document} of the congestion saturation, the state change process involves changing the saturation of intersection i from the initial value \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\alpha_{i2}$$\end{document} of congestion saturation to the initial value \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\alpha_{i3}$$\end{document} of severe congestion saturation. Evaluation index scheme set According to the actual needs, five evaluation index combination schemes are proposed, as shown in Table 1. The main reasons for proposing five scenarios (schemes) in this paper include:① Schemes 1 and 2 are designed to compare the different effects of saturation degree and traffic volume on node importance evaluation, primarily to better illustrate that selecting saturation degree is superior to selecting traffic volume if the focus is on changes in node operational status, and to serve as the basis for subsequent evaluation indicator schemes;② Schemes 3, 4, and 5 are further proposed based on Scheme 2 mainly to illustrate the impact of the aforementioned three scenarios of different node state changes on the results of node importance evaluation, thereby demonstrating that different groups of important nodes can be selected for different traffic management objectives.Table 1Evaluation index combination schemes.Index Scheme NoEvaluation Index SchemeIndex Scheme NoEvaluation Index Scheme1\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$K_{i}$$\end{document} + \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$C_{i}$$\end{document} + \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$B_{i}$$\end{document} + \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$V_{i}$$\end{document}4\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$K_{i}$$\end{document} + \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$C_{i}$$\end{document} + \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$B_{i}$$\end{document} + \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\alpha_{i}$$\end{document} + \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\Delta \varphi_{2i}$$\end{document}2\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$K_{i}$$\end{document} + \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$C_{i}$$\end{document} + \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$B_{i}$$\end{document} + \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\alpha_{i}$$\end{document}5\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$K_{i}$$\end{document} + \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$C_{i}$$\end{document} + \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$B_{i}$$\end{document} + \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\alpha_{i}$$\end{document} + \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\Delta \varphi_{3i}$$\end{document}3\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$K_{i}$$\end{document} + \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$C_{i}$$\end{document} + \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$B_{i}$$\end{document} + \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\alpha_{i}$$\end{document} + \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\Delta \varphi_{1i}$$\end{document} Table 1Evaluation index combination schemes.Index Scheme NoEvaluation Index SchemeIndex Scheme NoEvaluation Index Scheme1\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$K_{i}$$\end{document} + \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$C_{i}$$\end{document} + \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$B_{i}$$\end{document} + \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$V_{i}$$\end{document}4\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$K_{i}$$\end{document} + \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$C_{i}$$\end{document} + \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$B_{i}$$\end{document} + \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\alpha_{i}$$\end{document} + \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\Delta \varphi_{2i}$$\end{document}2\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$K_{i}$$\end{document} + \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$C_{i}$$\end{document} + \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$B_{i}$$\end{document} + \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\alpha_{i}$$\end{document}5\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$K_{i}$$\end{document} + \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$C_{i}$$\end{document} + \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$B_{i}$$\end{document} + \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\alpha_{i}$$\end{document} + \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\Delta \varphi_{3i}$$\end{document}3\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$K_{i}$$\end{document} + \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$C_{i}$$\end{document} + \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$B_{i}$$\end{document} + \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\alpha_{i}$$\end{document} + \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\Delta \varphi_{1i}$$\end{document} Evaluation index combination schemes. Evaluation index data preprocessing In actual engineering, data needs to be preprocessed owing to the differences in type and size, to facilitate quantitative calculation. Consequently, the data needs to be consistent and normalized. For the convenience of explanation, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$x_{ij0}$$\end{document} is used to represent the original evaluation data of the indexes (node degree \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$K_{i}$$\end{document}, clustering coefficient \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$C_{i}$$\end{document}, betweenness \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$B_{i}$$\end{document}, traffic volume \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$V_{i}$$\end{document}, saturation \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\alpha_{i}$$\end{document}, impact of changes in the operating state of the intersection on the road network (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\Delta \varphi_{1i}$$\end{document},\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\Delta \varphi_{2i}$$\end{document},\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\Delta \varphi_{3i}$$\end{document})), and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$x_{ij}{\prime}$$\end{document} is used to represent pre-processed index data (node degree \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$K_{i}{\prime}$$\end{document},, clustering coefficient \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$C_{i}{\prime}$$\end{document}, betweenness \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$B_{i}{\prime}$$\end{document}, traffic volume \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$V_{i}{\prime}$$\end{document}, saturation \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\alpha_{i}{\prime}$$\end{document}, and the impact of changes in the operating state of the intersection on the road network (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\Delta \varphi^{\prime}_{1i}$$\end{document},\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\Delta \varphi^{\prime}_{2i}$$\end{document},\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\Delta \varphi^{\prime}_{3i}$$\end{document})). Consistent processing Smaller indexes (such as clustering coefficients) need to be uniformly transformed into larger indexes, as follows:12\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{array}{c}{x}_{ij}=max\left({x}_{ij0}\right)-{x}_{ij0}\end{array}$$\end{document}\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$x_{ij0}$$\end{document} is the original value of the evaluation object i of the smaller index j and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\max (x_{ij0} )$$\end{document} represents the maximum value of all the evaluation objects of the smaller index j, it can also assume the theoretical maximum value. Normalization Based on the consistent processing of the index values, to avoid the influence of non-uniform index dimensions, the index values should be further normalized. The specific methods are as follows:13\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{array}{c}{x}_{ij}^{{^{\prime}}^{ }}=\frac{{x}_{ij}-\mathrm{min}\left({x}_{ij}\right)}{\mathrm{max}\left({x}_{ij}\right)-\mathrm{min}\left({x}_{ij}\right)}\end{array}$$\end{document}where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\max (x_{ij} )$$\end{document} represents the maximum value of all the evaluation objects of the index j after the normalized processing and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\min (x_{ij} )$$\end{document} represents the minimum value of all the evaluation objects of the index j after the normalized processing. Evaluation index weight The impact of evaluation indexes on the evaluation of node importance has its own emphasis. Therefore, if the importance of nodes is to be reasonably evaluated, the calibration of each index weight factor is particularly critical. Combination weights have been favored in recent years owing to their combination of the advantages of subjective and objective weights. However, owing to different principles, in the face of the same evaluation issue, the evaluation results obtained by different combined weighting methods also differ, which leads to difficulties in decision-making. To solve this problem, an integrated weighting method based on a multi-objective programming model is proposed. First, the G1 (order relationship method) and the CRITIC methods are selected to determine the subjective and objective weights of the evaluation indexes, respectively. Second, the product method and the maximum deviation sum of squares method are selected to integrate the subjective and objective weights, respectively, and then the combined weights of the evaluation indexes are determined. Finally, based on the combined weights and the multi-objective programming model, the final weights of the evaluation indexes are obtained. For the convenience of explanation, the index weights obtained using the G1, the CRITIC, the product, the maximum deviation sum of squares methods, and the integrated weighting method based on the multi-objective programming model are used as \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$w^{11}_{j}$$\end{document}, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$w^{12}_{j}$$\end{document}, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$w^{21}_{j}$$\end{document}, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$w^{22}_{j}$$\end{document}, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$w^{*}_{j}$$\end{document} (j = 1,2, ⋯ n,n is the total number of evaluation indexes), respectively. Next, the aforementioned weighting methods will be elaborated. Introduction to relevant subjective and objective weighting methods G1 method (order relationship method) For the AHP method, the consistency of the judgment matrix should be checked when determining the index weight and the process is more complicated. Therefore, the G1 method (order relationship method) is used23 to determine the subjective weight of the index. The main steps of the method are as follows: Assuming an evaluation index \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$x_{{j_{0} }} ,j_{0} = 1,2, \cdots ,n$$\end{document} , the results are as follows in descending order of importance:14\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{array}{c}{x}_{1}^{*}>{x}_{2}^{*}>L>{x}_{n}^{*}\end{array}$$\end{document} For the G1 method, the weight \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$w_{n}^{11}$$\end{document} of the index corresponding to \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$x_{n}^{*}$$\end{document} is first calculated as follows:15\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${w}_{n}^{11}=(1+\sum_{j=2}^{n} \prod_{i=j}^{n} {r}_{j}{)}^{-1}$$\end{document}where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$r_{j}$$\end{document} represents the weight ratio of adjacent elements calculated in descending order of importance, as shown in (16).16\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{array}{c}{r}_{j}=\frac{{w}_{j-1}^{11}}{{w}_{j}^{11}},j=n,n-1,n-2,L,2\end{array}$$\end{document}where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$w_{j}^{11}$$\end{document} represents the subjective weight of index j and the value of \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$r_{j}$$\end{document} can be referred to in Table 2. When n is larger, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$r_{n} = 1$$\end{document} can be taken.Table 2\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$r_{k}$$\end{document} Assignment reference.\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$r_{k}$$\end{document} valueValue determination principle1.0Index \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$x_{k - 1}^{*}$$\end{document} has the same importance as index \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$x^{*}_{k}$$\end{document}1.2Index \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$x_{k - 1}^{*}$$\end{document} is slightly more important than index \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$x^{*}_{k}$$\end{document}1.4Index \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$x_{k - 1}^{*}$$\end{document} is significantly more important than index \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$x^{*}_{k}$$\end{document}1.6Index \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$x_{k - 1}^{*}$$\end{document} is strongly more important than index \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$x^{*}_{k}$$\end{document}1.8Index \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$x_{k - 1}^{*}$$\end{document} is extremely more important than index \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$x^{*}_{k}$$\end{document} Table 2\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$r_{k}$$\end{document} Assignment reference.\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$r_{k}$$\end{document} valueValue determination principle1.0Index \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$x_{k - 1}^{*}$$\end{document} has the same importance as index \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$x^{*}_{k}$$\end{document}1.2Index \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$x_{k - 1}^{*}$$\end{document} is slightly more important than index \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$x^{*}_{k}$$\end{document}1.4Index \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$x_{k - 1}^{*}$$\end{document} is significantly more important than index \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$x^{*}_{k}$$\end{document}1.6Index \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$x_{k - 1}^{*}$$\end{document} is strongly more important than index \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$x^{*}_{k}$$\end{document}1.8Index \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$x_{k - 1}^{*}$$\end{document} is extremely more important than index \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$x^{*}_{k}$$\end{document} \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$r_{k}$$\end{document} Assignment reference. CRITIC method The CRITIC method24,25 is considered a calculation method that can reflect the objective weight of indexes and it mainly calculates the weight by evaluating the contrast strength and conflict between indexes. The main principles of this method are as follows: m schemes and n indexes exist, the corresponding standardized evaluation matrix is \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$Y = (y_{ij} )_{m \times n}$$\end{document} , and the objective weight of the evaluation indexes can be calculated as follows:17\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{array}{c}{R}_{j}=\sum_{i=1,i\ne j}^{n} \left(1-{r}_{ij}\right)\end{array}$$\end{document}18\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{array}{c}{C}_{j}={\sigma}_{j}{R}_{j}\end{array}$$\end{document}19\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{array}{c}{w}_{j}^{12}=\frac{{C}_{j}}{\sum_{j=1}^{n} {C}_{j}}\end{array}$$\end{document}where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$r_{ij}$$\end{document} represents the correlation coefficient between indexes i and j ; \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$R_{j}$$\end{document} represents the conflict of index j ; \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$C_{j}$$\end{document} represents the amount of information contained in index j ; \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\sigma_{j}$$\end{document} represents the standard deviation of the values of each scheme of index j ; \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$w_{{_{j} }}^{12}$$\end{document} represents the objective weight of index j using the CRITIC method. The subjective weight of the index reflects the will of the decision-maker; however, it is significantly influenced by the knowledge level of experts, whereas the objective weight is fully based on the evaluation data and calculated through a more rigorous mathematical formula. Both have their own advantages. To effectively integrate the advantages of subjective and objective weights, according to different principles, two combined weighting methods including the product and the maximum deviation sum of squares methods are introduced and are described in detail below. Introduction to relevant combination weighting methods Product method The product method26 is based on the evidence theory integrating an algorithm to determine the index weight. The specific method is as follows:20\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{array}{c}{w}_{j}^{21}=\frac{\prod_{i=1}^{2} {w}_{j}^{1i}}{\sum_{j=1}^{n} \prod_{i=1}^{2} {w}_{j}^{1i}}\end{array}$$\end{document}where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$w_{{_{j} }}^{1i}$$\end{document} represents the weight value of the index j in the i weighting scheme. j) Maximum deviation sum of squares. Compared to the product method, the determination of the weight of the index using the maximum deviation sum of squares method is more complicated. The main process is as follows: First, the relationship between the combined weight and the subjective and objective weights is established as follows:21\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{array}{c}\begin{array}{c}{w}_{j}^{22}={\theta}_{1}{w}_{j}^{11}+{\theta}_{2}{w}_{j}^{12}\\ s.t.\left\{\begin{array}{cc}{\theta}_{k}\ge 0,k=\mathrm{1,2}& \\ \sum_{k=1}^{2} {\theta}_{k}^{2}=1& \end{array}\right.\end{array}\end{array}$$\end{document} Second, to obtain \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$w^{22}$$\end{document} in Eq. (22), the maximum deviation sum of squares of all evaluation objects can be assumed as the goal:22\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$maxJ({w}^{22})=\sum_{i=1}^{m} \sum_{{i}_{1}=1}^{m} [\sum_{j=1}^{n} \left({b}_{ij}-{b}_{{i}_{1}j}\right)\left({\theta}_{1}{w}_{j}^{11}+{\theta}_{2}{w}_{j}^{12}\right){]}^{2}$$\end{document}where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$w_{j}^{22}$$\end{document} is the weight value of index j determined using the maximum deviation sum of squares method, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$b_{ij}$$\end{document} and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$b_{{i_{1} j}}$$\end{document} represent the standardized evaluation data of evaluation objects i and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$i_{1}$$\end{document} in index j , respectively, n represents the total number of evaluation indexes, and m represents the total evaluation objects. To determine the optimal solution of the objective function of (23), the following matrix can be constructed based on the standardized evaluation data \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$b_{ij}$$\end{document} :23\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{array}{c}B=\left[\begin{array}{ccc}\sum_{i=1}^{m} \sum_{{i}_{1}=1}^{m} \left|{b}_{{i}_{1}}-{b}_{{i}_{1}1}\right|\left|{b}_{{i}_{1}}-{b}_{{i}_{1}1}\right|& \mathrm{\dots}& \sum_{i=1}^{m} \sum_{{i}_{1}=1}^{m} \left|{b}_{{i}_{1}}-{b}_{{i}_{1}1}\right|\left|{b}_{{i}_{m}}-{b}_{{i}_{1}n}\right|\\ \mathrm{\dots}& \mathrm{\dots}& \mathrm{\dots}\\ \sum_{i=1}^{m} \sum_{{i}_{1}=1}^{m} \left|{b}_{in}-{b}_{{i}_{1}n}\right|\left|{b}_{{i}_{1}}-{b}_{{i}_{1}1}\right|& \mathrm{\dots}& \sum_{i=1}^{m} \sum_{{i}_{1}=1}^{m} \left|{b}_{in}-{b}_{{i}_{1}n}\right|\left|{b}_{in}-{b}_{{i}_{1}n}\right|\end{array}\right]\end{array}$$\end{document} The matrix \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$B{\prime}$$\end{document} is built again:24\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{array}{c}{B}^{{^{\prime}}^{ }}=WB{W}^{T}\end{array}$$\end{document}where W is obtained as follows:25\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{array}{c}W=\left(\begin{array}{c}{w}^{11}\\ \\ {w}^{12}\end{array}\right)\end{array}$$\end{document} Further, the maximum eigenvalue \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\lambda_{\max }$$\end{document} of matrix \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$B{\prime}$$\end{document} and its corresponding unitized eigenvector \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\theta^{*}$$\end{document} can be obtained and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\theta^{*}$$\end{document} is the optimal solution of the objective function of (23). Finally, by normalizing \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\theta^{*}$$\end{document} and substituting it into (21), \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$w_{j}^{22}$$\end{document} can be solved. Integrated weighting method based on a multi-objective programming model Although the combined weighting method better extracted the advantages of the subjective and objective weighting methods, owing to different principles, this method leads to inconsistent evaluation and sorting results; thus, decision-making is difficult. To solve this problem, an integrated weighting method based on a multi-objective programming model is proposed. The index weights obtained using the combined weighting method are further integrated through the method, which aims to maximize the sum of the correlation coefficients of the sorting results, to ensure the superiority of the final weights. The main contents and steps of the integrated weighting method based on the multi-objective programming model are as follows: If two combined weight vectors \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$w_{j}^{21} ,w_{j}^{22}$$\end{document} are known and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$w_{j}^{*}$$\end{document} is the weight to be calculated, they are expressed in vector form as follows:26\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{array}{c}W=\left[\begin{array}{ccc}{w}_{j}^{21}& {w}_{j}^{22}& {w}_{j}^{*}\end{array}\right]\end{array}$$\end{document} A linear correlation exists between \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$w_{j}^{*}$$\end{document} and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$w_{j}^{21} ,w_{j}^{22}$$\end{document} in (27), which is described as follows:27\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{array}{c}{w}_{j}^{*}=a{w}_{j}^{21}+b{w}_{j}^{22}\end{array}$$\end{document}where a and b are coefficients and a + b = 1. If the \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$w_{j}$$\end{document} and the preprocessed data \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$x_{ij}{\prime}$$\end{document} of the evaluation index are known, the importance of node i can be calculated as follows:28\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{array}{c}{Z}_{i}^{k}=\sum_{j=1}^{n} {w}_{j}^{2k}{x}_{ij},k=\mathrm{1,2},3,{w}_{j}^{23}={w}_{j}^{*}\end{array}$$\end{document}where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$Z_{i}^{k}$$\end{document} represents the importance of node i and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$w_{j}^{2k}$$\end{document} represents the 2 k-th weight scheme value. The sorting result can be obtained by sorting the importance \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$Z_{i}^{k}$$\end{document} of the nodes in descending order, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$p_{{w_{j}^{2k} }}$$\end{document} is used to represent the node sorting result obtained by the 2 k-th weighting scheme. To realize the effective integration of the known combination weight vectors and ensure that the obtained new weight \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$w_{j}^{*}$$\end{document} is better than the known combination weight \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$w_{j}^{21} ,w_{j}^{22}$$\end{document}, the correlation coefficient of the evaluation sorting results is introduced and the specific definition is as follows:29\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{array}{c}{r}_{{k}_{1}{k}_{2}}=\frac{\mathrm{c}\mathrm{o}\mathrm{v}\left({p}_{{w}_{j}^{2{k}_{1}}},{p}_{{w}_{j}^{2{k}_{2}}}\right)}{\sqrt{D\left({p}_{{w}_{j}^{2{k}_{1}}}\right)}\sqrt{D\left({p}_{{w}_{j}^{2{k}_{2}}}\right)}},{k}_{1}=\mathrm{1,2},3;{k}_{2}=\mathrm{1,2},3\end{array}$$\end{document}where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$r_{{k_{1} k_{2} }}$$\end{document} represents the correlation coefficient of the evaluation sorting results calculated from the combined weight \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${w}_{j}^{2{k}_{1}}$$\end{document} and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${w}_{j}^{2{k}_{2}}$$\end{document}. Further, the objective function used in building the multi-objective programming model is as follows:30\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{array}{*{20}c} {minf_{1} = \mathop \sum \limits_{k = 1}^{2} \left\| {w^{*} - w^{2k} } \right\|_{2} } \\ \end{array}$$\end{document}31\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$max{f}_{2}=\left|\sum_{i=1}^{3} {r}_{i3}-max\left\{\sum_{i=1}^{3} {r}_{i1},\sum_{i=1}^{3} {r}_{i2}\right\}\right|$$\end{document}32\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{array}{c}min{f}_{3}=\left|(\sum_{i=1}^{3} {r}_{i3}>max\left\{\sum_{i=1}^{3} {r}_{i1},\sum_{i=1}^{3} {r}_{i2}\right\})-1\right|\end{array}$$\end{document}e model comprises three main objective functions: the first objective is to effectively integrate the information of the known combination weights, which is mainly guaranteed by the minimum distance between the weight vector and the known combination weight vectors \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$w_{j}^{*}$$\end{document} . The second objective is to ensure the superiority of the weight vector \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$w_{j}^{21} ,w_{j}^{22}$$\end{document} to be obtained by evaluating the sorting result with the largest correlation coefficient; the third objective is also to ensure the superiority of the weight vector \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$w_{j}^{*}$$\end{document} to be sought. It differs from the second objective in that this objective is reflected by the difference in the sum of the correlation coefficients of the sorting results. The constraints of the model include (30)-(32), and the purpose is to obtain the final index weight \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$w_{j}^{*}$$\end{document} by solving a and b in (30). Because the above model is a multi-objective nonlinear programming model, the following formula can be used to convert it into a single-objective model and the solution is obtained using an intelligent optimization algorithm.33\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{array}{c}minf={f}_{1}+\frac{{f}_{3}}{{f}_{2}}\end{array}$$\end{document} Solve with integrated weighting method based on multi-objective programming model Fundamental principles of the particle swarm optimization (PSO) algorithm The solution to the multi-objective nonlinear programming problem is a difficult problem, and PSO has been used in solving it. PSO is characterized by fewer parameters, fast convergence speed, high precision, and easy implementation. Standard PSO27,28 is introduced to solve the multi-objective programming model of the integrated weighting method proposed in this study. PSO is used to solve optimization problems in two stages: initializing the population (several initial solutions to the problem) and population optimization iteration. The basic principle is described below. Suppose in a search space of D dimension ( D is the number of unknown variables included in the problem to be solved; we selected two), a population \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$X = (X_{1} ,X_{2} , \cdots ,X_{n} )$$\end{document} composed of n particles, of which the i-th particle is represented as a D dimensional vector \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${X}_{i}=({x}_{i1},{x}_{i2},\mathrm{L},{x}_{iD}{)}^{T}$$\end{document} , indicating the position of the i-th particle in the D-dimensional search space, and also a potential solution to the problem. The fitness value corresponding to each particle position \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$X_{i}$$\end{document} can be calculated according to the objective function. The velocity of i-th particle is \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$V_{i} = (V_{i1} ,V_{i2} , \cdots ,V_{iD} )^{T}$$\end{document}, its individual extreme value is \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$P_{i} = (P_{i1} ,P_{i2} , \cdots ,P_{iD} )^{T}$$\end{document}, and the global extremum of the population is \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$P_{g} = (P_{g1} ,P_{g2} , \cdots ,P_{gD} )^{T}$$\end{document}. In each iteration process, the particle updates its speed and position through the individual and the global extrema. The updated equation is as follows:34\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{array}{c}{V}_{id}^{k+1}=\chi \left[{V}_{id}^{k}+{c}_{1}{r}_{1}\left({P}_{id}^{k}-{X}_{id}^{k}\right)+{c}_{2}{r}_{2}\left({P}_{gd}^{k}-{X}_{id}^{k}\right)\right]\end{array}$$\end{document}35\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{array}{c}{X}_{id}^{k+1}={X}_{id}^{k}+{V}_{id}^{k+1}\end{array}$$\end{document}36\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{array}{c}\chi =\frac{2}{\left|2-\varphi -\sqrt{{\varphi }^{2}-4\varphi }\right|},\varphi ={c}_{1}+{c}_{2}\end{array}$$\end{document}where d = 1,2, ⋯ ,D, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$i{ = }1,2, \cdots ,n$$\end{document} , k , k is the current iteration number, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$k{ = }1,2, \cdots ,N$$\end{document}, where N is the maximum iteration number of the algorithm; c1 and c2 are non-negative acceleration factors; r1 and r2 are random numbers distributed between [0,1]. χ is a constraint factor, which ensures the global convergence of the algorithm. To prevent the blind search of particles, their positions and velocities within a certain interval should be limited, Considering that the values solved in this paper are weights, it is constrained that X ∈ [0,1], V ∈ [ - 0.1,0.1]. Influence and determination of parameters in the PSO algorithm When using the Particle Swarm Optimization (PSO) algorithm to solve the integrated weighting method based on a multi-objective programming model, the determination of relevant parameters is particularly important. The population size (Ns) determines the algorithm’s global search capability and convergence speed; a larger population enhances global search, but increases the computational cost of each iteration, whereas a smaller population may converge faster but is prone to local optima. The acceleration factors c1 and c2 control the individual (cognitive) and social components, respectively; an excessively high c1 may lead to over-dispersion, and an excessively high c2 may cause premature convergence. The maximum number of iterations (N) directly impacts computation time; a value that is too small may result in a non-optimal outcome, while a value that is too large increases unnecessary computational burden. To determine reasonable parameter values, one can initially set the range for each parameter based on relevant literature, and then analyze and determine the values experimentally using a combinatorial comparison approach, with the convergence speed and final value of the fitness function as evaluation metrics. Importance sorting and grading of road network node Importance sorting of road network node Undoubtedly, the integrated weighting method based on the multi-objective programming model has included the superiority of sorting results when determining the index weights; thus, the weight \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$w_{j}^{*}$$\end{document} determined using the model is used to calculate the importance value \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$Z_{i}^{*}$$\end{document} of the node, as shown in (28), \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$Z_{i}^{*}$$\end{document} is arranged in in descending order, and the sorting result \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$p_{{w_{j}^{*} }}$$\end{document} of the nodes is finally determined. Corresponding to the evaluation index scheme set in Table 3, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$p_{{_{{w_{j}^{*} }} }}^{1}$$\end{document}, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$p_{{_{{w_{j}^{*} }} }}^{2}$$\end{document}, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$p_{{_{{w_{j}^{*} }} }}^{3}$$\end{document}, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$p_{{_{{w_{j}^{*} }} }}^{4}$$\end{document} and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$p_{{_{{w_{j}^{*} }} }}^{5}$$\end{document} are used to represent the node importance sorting results corresponding to each index system.Table 3the evaluation index scheme set. Table 3the evaluation index scheme set. the evaluation index scheme set. Node grading based on different sorting results Node sorting results \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$p_{{_{{w_{j}^{*} }} }}^{3}$$\end{document}, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$p_{{_{{w_{j}^{*} }} }}^{4}$$\end{document} and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$p_{{_{{w_{j}^{*} }} }}^{5}$$\end{document} differ owing to different operation state change patterns. In node grading, to comprehensively consider the sorting results, a node grading algorithm is proposed based on various sorting results. The algorithm maintains the original sorting results while ensuring the same set of nodes in each sorting result. The main process of the algorithm is described as follows: Case study Case study and experimental scheme Case introduction To verify the rationality of the methods, algorithms, and theories proposed in this study, a specific case is introduced for illustration. The case study considers the road network nodes in the core area of the urban road of Xining City, Qinghai Province, China (Fig. 3) as the object. To facilitate the description of the problem, the topology diagram of the road network in the study area is drawn according to the actual connection of the road network nodes and the research needs, as shown in Fig. 4.Fig. 3Schematic of the core area of urban road in Xining.Fig. 4Topological map of a road intersection in an urban core area. Schematic of the core area of urban road in Xining. Topological map of a road intersection in an urban core area. Xining City, Qinghai Province is a typical “Valley Cross” city. The traffic transformation of the main traffic flow (east–west, south-north) in the city needs to pass through the roads in the core area shown in Fig. 3, which increases the traffic burden in this area. To ensure unimpeded traffic in this area, the importance of nodes should be evaluated to formulate a targeted traffic control plan based on the evaluation results. The data on node traffic volume and saturation used in this case study originates from a large-scale traffic survey organized by Qinghai Nationalities University from 7:30 to 9:00 on November 26, 2018. During the specific survey, the data were recorded every 5 min as a time series, with 18 time series. The node saturation sequence diagram is shown in Fig. 5.Fig. 5Grading map of road network node operation state (Xining City, Qinghai Province). Grading map of road network node operation state (Xining City, Qinghai Province). Experimental scheme The evaluation of node importance and the sorting and grading of nodes are influenced by the determination of the reference value of node operating state saturation, selection of the dynamic index evaluation model, preprocessing of the evaluation index data, determination of the evaluation index weight, and selection of the evaluation index (Fig. 6). To verify the rationality of the algorithms and theoretical expressions proposed in this study, an experimental scheme is designed according to the logical relationship between the contents of each part, as shown in Fig. 7. Next section elaborates the contents of each part.Fig. 6Sequence diagram of intersection saturation in morning rush hour.Fig. 7Experimental scheme. Sequence diagram of intersection saturation in morning rush hour. Experimental scheme. Saturation reference value of the operating state of a node According to the node saturation value shown in Fig. 6, the changed trend of node saturation in the peak period and the saturation reference value corresponding to each operating state can be further obtained, as shown in Fig. 6. As shown in Fig. 5, the average saturation value of road network nodes during the period of 7:30 to 7:35 is 0.64, whereas it changed abruptly to 0.73 from 7:35 to 7:40, and the rate of change is the largest in Fig. 5. The saturation reference values corresponding to each operating state of the node are: unimpeded (= 0.64), basic congestion (= 0.73), congestion (= 0.81), and severe congestion (= 0.85). Evidently, the actual saturation changed trend of the nodes in the studied city is included in the determination method of the saturation reference value based on the residents’ travel time rule, particularly when the residents’ travel time regularity is reflected when determining the unimpeded and slight congestion saturation reference values; that is, the closer to commute time (on and off school time), the more concentrated the travel, and there will be a sudden change in saturation. Dynamic index evaluation model optimization To illustrate the superiority of the road network impact evaluation model based on topological potential, the change in the road network operating state when the operating state of the selected node changes from unimpeded (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\alpha_{i0}$$\end{document} = 0.64) to slightly congested (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\alpha_{i1}$$\end{document} = 0.73) is compared. Table 4 shows the impact values of changes in node operating state on the road network calculated using the traditional efficiency evaluation model and the evaluation model based on topological potential (the value of the optimal influence factor σ of the topological potential model calculated using the ternary iterative algorithm is 3.2594). As shown in Table 4,\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\Delta \varphi_{1ia}$$\end{document} is the change value of the road network operation state obtained using the traditional efficiency evaluation model and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\Delta \varphi_{1ib}$$\end{document} is the change value of the road network operation state obtained using the evaluation model based on the topological potential.Table 4Influence of node operation status change on road network.ID\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\Delta \varphi_{1ia}$$\end{document}\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\Delta \varphi_{1ib}$$\end{document}ID\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\Delta \varphi_{1ia}$$\end{document}\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\Delta \varphi_{1ib}$$\end{document}ID\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\Delta \varphi_{1ia}$$\end{document}\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\Delta \varphi_{1ib}$$\end{document}10.0000.697210.0011.465410.0011.50720.0001.391220.0015.103420.0012.09530.0012.879230.0015.463430.0012.21240.0013.417240.0014.135440.0011.76150.0000.958250.0014.246450.0014.65760.0001.016260.0013.218460.0014.56970.0000.956270.0012.252470.0011.72480.0001.316280.0015.213480.0011.47590.0011.519290.0011.694490.0011.402100.0012.161300.0012.179500.0013.158110.0013.023310.0014.136510.0000.972120.0001.081320.0001.250520.0011.619130.0011.659330.0016.822530.0000.844140.0011.788340.0011.439540.0011.444150.0012.900350.0013.483550.0011.629160.0014.061360.0012.250560.0001.024170.0001.254370.0014.727570.0001.367180.0012.532380.0014.459580.0000.698190.0011.986390.0015.307590.0012.361200.0012.045400.0013.707600.0014.692 Table 4Influence of node operation status change on road network.ID\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\Delta \varphi_{1ia}$$\end{document}\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\Delta \varphi_{1ib}$$\end{document}ID\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\Delta \varphi_{1ia}$$\end{document}\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\Delta \varphi_{1ib}$$\end{document}ID\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\Delta \varphi_{1ia}$$\end{document}\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\Delta \varphi_{1ib}$$\end{document}10.0000.697210.0011.465410.0011.50720.0001.391220.0015.103420.0012.09530.0012.879230.0015.463430.0012.21240.0013.417240.0014.135440.0011.76150.0000.958250.0014.246450.0014.65760.0001.016260.0013.218460.0014.56970.0000.956270.0012.252470.0011.72480.0001.316280.0015.213480.0011.47590.0011.519290.0011.694490.0011.402100.0012.161300.0012.179500.0013.158110.0013.023310.0014.136510.0000.972120.0001.081320.0001.250520.0011.619130.0011.659330.0016.822530.0000.844140.0011.788340.0011.439540.0011.444150.0012.900350.0013.483550.0011.629160.0014.061360.0012.250560.0001.024170.0001.254370.0014.727570.0001.367180.0012.532380.0014.459580.0000.698190.0011.986390.0015.307590.0012.361200.0012.045400.0013.707600.0014.692 Influence of node operation status change on road network. To compare and illustrate the difference between the two evaluation models, the concept of sample heterogeneities is proposed, which is defined as follows:37\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{array}{c}dif{f}_{i}=\frac{{N}_{i1}}{{N}_{i}}\end{array}$$\end{document}where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$diff_{i}$$\end{document} represents the heterogeneities of the sample data i, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$N_{i}$$\end{document} represents the number of samples of sample data i, and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$N_{i1}$$\end{document} represents the number of different sample data in sample data i. According to Table 4, the impact of node operating state changes on the road network obtained using the traditional efficiency evaluation model is low in distinguishability of the dataset, only 3.3%, whereas the dataset of the impact of the changes in the node operating state on the road network obtained using the evaluation model based on topological potential is almost 100% distinguishable. The latter can better evaluate the impact of changes in the operating status of each node on the road network, and their differences, and is more conducive to sorting the nodes. Therefore, the influence evaluation model based on topological potential is adopted in the study to calculate the data of the dynamic evaluation index. Evaluation Metrics preprocessing data Based on the evaluation index consistency and normalization processing methods, the original index data such as node degree \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$K_{i}$$\end{document}, clustering coefficient \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$C_{i}$$\end{document}, betweenness \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$B_{i}$$\end{document}, traffic volume \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$V_{i}$$\end{document}, saturation \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\alpha_{i}$$\end{document}, and the impact of the change in intersection operating state on the road network (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\Delta \varphi_{1i}$$\end{document},\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\Delta \varphi_{2i}$$\end{document},\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\Delta \varphi_{3i}$$\end{document}) used for evaluation are preprocessed. The results are shown in Table 5.Table 5Evaluation index data preprocessing results.ID\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$K_{i}{\prime}$$\end{document}\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$C_{i}{\prime}$$\end{document}\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$B_{i}{\prime}$$\end{document}\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$V_{i}^{\prime }$$\end{document}\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\alpha_{i}^{\prime }$$\end{document}\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\Delta \varphi_{1i}^{\prime }$$\end{document}\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\Delta \varphi_{2i}^{\prime }$$\end{document}\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\Delta \varphi_{3i}^{\prime }$$\end{document}10.4001.0000.2270.6470.8350.1020.1250.10520.6001.0000.3840.4350.8470.2040.2600.20930.6001.0000.4730.4860.8660.4220.3300.42040.4001.0000.5710.3650.8970.5010.3800.45050.6001.0000.2520.7310.9310.1400.1980.14960.8001.0000.2980.4200.9040.1490.3280.17170.6001.0000.2980.3810.8560.1400.3010.16480.6001.0000.3410.2520.8400.1930.3370.20590.8001.0000.4120.5240.8330.2230.5640.261100.8001.0000.5760.5310.8070.3170.5930.366110.8000.8330.5430.4670.8950.4430.5430.491120.6000.6670.2560.3110.9160.1580.2440.186130.6001.0000.3900.4550.9200.2430.4100.249140.8001.0000.3930.4340.8610.2620.6620.306150.8001.0000.5200.3630.8550.4250.7120.481160.8001.0000.5600.3720.9160.5950.6990.656170.6001.0000.3700.4920.9420.1840.4360.219180.6001.0000.4270.3810.7580.3710.4140.367190.8001.0000.3800.3720.9150.2910.5850.344200.6001.0000.5550.2290.7650.3000.5840.317210.6001.0000.3920.2790.8240.2150.6230.256220.8001.0001.0000.3850.8830.7480.9410.789230.8001.0000.7070.4550.8110.8010.8720.890240.8001.0000.6120.3710.8720.6060.7330.676250.6001.0000.8130.3200.7350.6220.6510.573260.6001.0000.5731.0000.8470.4720.4800.449270.8001.0000.4410.4390.8550.3300.6260.379280.8000.8330.9940.4770.9480.7641.0000.793290.6000.6670.2630.3570.8570.2480.4710.291300.6000.6670.3980.3710.8530.3190.3890.372310.6001.0000.5060.6320.8810.6060.4170.565320.6001.0000.1660.3920.8810.1830.4300.217330.8001.0000.8920.5990.9381.0000.9271.000340.6001.0000.5290.4420.8150.2110.6520.262350.8001.0000.6550.3760.8630.5110.6740.598360.6001.0000.4010.4040.9700.3300.6180.389370.6001.0000.7390.4490.9260.6930.6540.632380.6001.0000.7150.4870.8390.6540.5360.630390.8001.0000.7850.3860.8850.7780.9650.781401.0000.8000.5030.4150.8920.5430.7960.604410.6000.6670.4070.1590.7800.2210.4560.263420.6001.0000.5180.8910.8850.3070.4170.311430.8001.0000.4720.4920.8200.3240.6670.346440.8000.8330.6710.5780.8770.2580.6150.303451.0000.8000.8090.8031.0000.6830.9010.737460.6001.0000.7080.4120.8350.6700.6090.621470.6001.0000.4190.7820.8790.2530.3300.257480.8001.0000.3090.3830.9200.2160.5010.238490.8001.0000.4700.5450.7940.2060.5370.251500.8001.0000.4220.4650.9070.4630.6230.483510.4001.0000.1530.6670.8720.1420.3120.158520.6001.0000.2940.6340.8970.2370.2000.245530.6001.0000.1580.3310.9250.1240.2600.141540.8001.0000.3270.4900.8890.2120.3400.249550.6001.0000.2740.2590.8630.2390.3380.243560.4001.0000.2130.6550.8790.1500.0990.162570.6001.0000.2390.6270.8300.2000.1550.217580.4001.0000.1530.6450.8230.1020.1300.108590.4001.0000.4060.9510.8860.3460.3410.318600.8000.8330.6450.4060.8920.6880.5130.683 Table 5Evaluation index data preprocessing results.ID\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$K_{i}{\prime}$$\end{document}\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$C_{i}{\prime}$$\end{document}\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$B_{i}{\prime}$$\end{document}\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$V_{i}^{\prime }$$\end{document}\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\alpha_{i}^{\prime }$$\end{document}\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\Delta \varphi_{1i}^{\prime }$$\end{document}\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\Delta \varphi_{2i}^{\prime }$$\end{document}\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\Delta \varphi_{3i}^{\prime }$$\end{document}10.4001.0000.2270.6470.8350.1020.1250.10520.6001.0000.3840.4350.8470.2040.2600.20930.6001.0000.4730.4860.8660.4220.3300.42040.4001.0000.5710.3650.8970.5010.3800.45050.6001.0000.2520.7310.9310.1400.1980.14960.8001.0000.2980.4200.9040.1490.3280.17170.6001.0000.2980.3810.8560.1400.3010.16480.6001.0000.3410.2520.8400.1930.3370.20590.8001.0000.4120.5240.8330.2230.5640.261100.8001.0000.5760.5310.8070.3170.5930.366110.8000.8330.5430.4670.8950.4430.5430.491120.6000.6670.2560.3110.9160.1580.2440.186130.6001.0000.3900.4550.9200.2430.4100.249140.8001.0000.3930.4340.8610.2620.6620.306150.8001.0000.5200.3630.8550.4250.7120.481160.8001.0000.5600.3720.9160.5950.6990.656170.6001.0000.3700.4920.9420.1840.4360.219180.6001.0000.4270.3810.7580.3710.4140.367190.8001.0000.3800.3720.9150.2910.5850.344200.6001.0000.5550.2290.7650.3000.5840.317210.6001.0000.3920.2790.8240.2150.6230.256220.8001.0001.0000.3850.8830.7480.9410.789230.8001.0000.7070.4550.8110.8010.8720.890240.8001.0000.6120.3710.8720.6060.7330.676250.6001.0000.8130.3200.7350.6220.6510.573260.6001.0000.5731.0000.8470.4720.4800.449270.8001.0000.4410.4390.8550.3300.6260.379280.8000.8330.9940.4770.9480.7641.0000.793290.6000.6670.2630.3570.8570.2480.4710.291300.6000.6670.3980.3710.8530.3190.3890.372310.6001.0000.5060.6320.8810.6060.4170.565320.6001.0000.1660.3920.8810.1830.4300.217330.8001.0000.8920.5990.9381.0000.9271.000340.6001.0000.5290.4420.8150.2110.6520.262350.8001.0000.6550.3760.8630.5110.6740.598360.6001.0000.4010.4040.9700.3300.6180.389370.6001.0000.7390.4490.9260.6930.6540.632380.6001.0000.7150.4870.8390.6540.5360.630390.8001.0000.7850.3860.8850.7780.9650.781401.0000.8000.5030.4150.8920.5430.7960.604410.6000.6670.4070.1590.7800.2210.4560.263420.6001.0000.5180.8910.8850.3070.4170.311430.8001.0000.4720.4920.8200.3240.6670.346440.8000.8330.6710.5780.8770.2580.6150.303451.0000.8000.8090.8031.0000.6830.9010.737460.6001.0000.7080.4120.8350.6700.6090.621470.6001.0000.4190.7820.8790.2530.3300.257480.8001.0000.3090.3830.9200.2160.5010.238490.8001.0000.4700.5450.7940.2060.5370.251500.8001.0000.4220.4650.9070.4630.6230.483510.4001.0000.1530.6670.8720.1420.3120.158520.6001.0000.2940.6340.8970.2370.2000.245530.6001.0000.1580.3310.9250.1240.2600.141540.8001.0000.3270.4900.8890.2120.3400.249550.6001.0000.2740.2590.8630.2390.3380.243560.4001.0000.2130.6550.8790.1500.0990.162570.6001.0000.2390.6270.8300.2000.1550.217580.4001.0000.1530.6450.8230.1020.1300.108590.4001.0000.4060.9510.8860.3460.3410.318600.8000.8330.6450.4060.8920.6880.5130.683 Evaluation index data preprocessing results. The influence of index weight on node importance Index scheme set 1 (node degree \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$K_{i}$$\end{document} + clustering coefficient \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$C_{i}$$\end{document} + betweenness \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$B_{i}$$\end{document} + traffic volume \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$V_{i}$$\end{document}) is selected to study the influence of index weight on node importance. The main process is as follows. Index weight calculation Determining subjective weight using the G1 method. \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$x_{i1} ,x_{i2} ,x_{i3} ,x_{i4}$$\end{document} represent four indexes: node degree \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$K_{i}$$\end{document}, clustering coefficient \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$C_{i}$$\end{document}, betweenness \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$B_{i}$$\end{document}, and traffic volume \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$V_{i}$$\end{document}. After experts consultation, the sorting results and ratios of each index determined according to G1 (order relationship analysis method) are as follows:38\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{array}{c}{x}_{i4}>{x}_{i3}>{x}_{i2}>{x}_{i1}\end{array}$$\end{document}39\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{array}{c}{r}_{2}=\frac{{x}_{i4}}{{x}_{i3}}=1.2;{r}_{3}=\frac{{x}_{i3}}{{x}_{i2}}=1.2;{r}_{4}=\frac{{x}_{i2}}{{x}_{i1}}=1.2\end{array}$$\end{document} First, the weight value of \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$x_{i1}$$\end{document} is 0.3956 and the weights of \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$x_{i2} ,x_{i3} ,x_{i4}$$\end{document} are obtained as follows:40\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{array}{c}{w}^{11}=\left[\begin{array}{cccc}0.1863& 0.2235& 0.2683& 0.3219\end{array}\right]\end{array}$$\end{document} Determining the objective weight using the CRITIC method. The objective weights determined using the CRITIC method are as follows:41\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{array}{c}{w}^{12}=\left[\begin{array}{cccc}0.2441& 0.3111& 0.2250& 0.2198\end{array}\right]\end{array}$$\end{document} Determining combined weights using the product method. The subjective and objective weights are combined with the product method to obtain:42\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{array}{c}{w}^{21}=\left[\begin{array}{cccc}0.1848& 0.2825& 0.2453& 0.2874\end{array}\right]\end{array}$$\end{document} Combined weghts are determined using the maximum deviation sum of squares method. The coefficients of the subjective and objective weights determined using the maximum deviation sum of squares method are 0.4907 and 0.5093, respectively, and the combined weights of the subjective and objective weights can be further weighted and averaged:43\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{array}{c}{w}^{22}=\left[\begin{array}{cccc}0.2156& 0.2678& 0.2464& 0.2702\end{array}\right]\end{array}$$\end{document} Index weights are determined using the integrated weighting method based on the multi-objective programming model. Taking w21 and w22 as the input of the multi-objective-based integrated weighting method, and using MATLAB, PSO is used to solve this problem , The ranges of the relevant parameters are as follows: Values for the maximum number of iterations:\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$N = \left[ {\begin{array}{*{20}c} {100} & {150} & {200} & {250} & {300} \\ \end{array} } \right]$$\end{document} Values for the population size:\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$N_{s} = \left[ {\begin{array}{*{20}c} {30} & {40} & {50} \\ \end{array} } \right]$$\end{document} Values for the acceleration coefficients:\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$c_{1} = c_{2} = \left[ {\begin{array}{*{20}c} {1.0} & {1.2} & {1.4} & {1.6} & {1.8} & {2.0} \\ \end{array} } \right]$$\end{document} Based on the parameter values listed above, 540(5 × 3 × 6 × 6) sets of parameter combination experiments were designed using the MATLAB software platform. The fastest convergence speed and the minimum value of the fitness function were used as evaluation metrics to ultimately determine the final values for each parameter. The convergence curve of the fitness function corresponding to N = 300, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$N_{s} = 50$$\end{document},\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$c_{1} = c_{2} = 1.4$$\end{document} is shown in the Fig. 8.Fig. 8Fitness function pattern. Fitness function pattern. The results obtained by solving the aforementioned PSO model the weighting coefficients of w21 and w22 are obtained as 0.5686 and 0.4314, respectively. After further weighting calculation, the final index weight \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$w^{*}$$\end{document} is obtained as:44\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{array}{c}{w}^{*}=\left[\begin{array}{cccc}0.1981& 0.2762& 0.2457& 0.2800\end{array}\right]\end{array}$$\end{document} Node importance calculation and sorting First, the importance value \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$Z_{i}^{*}$$\end{document} of nodes is calculated according to the index weight and preprocessing data and then sorted in descending order. The results are shown in Table 6.Table 6Sorting results.IDW21SortingW22SortingW*SortingIDW21SortingW22SortingW*SortingIDW21SortingW22SortingW*Sorting133333321465050413621822626262225274642601836322222223501525434036604454545242725274420111154242422515461545112020623392326141414464447392339275454544717181010102854454856215694735492944544497171049494730341919505156211159283531334345121515112354728325233525885813382438331965253858814284359346526543232321537384335574848555555551643924364857575653535317243737371717175730303018311693813401358292929199593139213259121212201631164018604060414141 Table 6Sorting results.IDW21SortingW22SortingW*SortingIDW21SortingW22SortingW*SortingIDW21SortingW22SortingW*Sorting133333321465050413621822626262225274642601836322222223501525434036604454545242725274420111154242422515461545112020623392326141414464447392339275454544717181010102854454856215694735492944544497171049494730341919505156211159283531334345121515112354728325233525885813382438331965253858814284359346526543232321537384335574848555555551643924364857575653535317243737371717175730303018311693813401358292929199593139213259121212201631164018604060414141 Sorting results. According to the three sorting results in Table 6, the correlation coefficient bubble chart is further drawn, as shown in Fig. 9.Fig. 9Correlation coefficient diagram of sorting results. Correlation coefficient diagram of sorting results. As shown in Fig. 11, the sum of the correlation coefficients between the node sorting results calculated according to weight \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$w^{*}$$\end{document} and the node sorting results obtained by other combined weights is the largest, reaching 2.04, whereas the sum of the correlation coefficients between the node sorting results calculated according to weights \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$w^{21} ,w^{22}$$\end{document} and the node sorting results obtained by other combined weights are 1.79 and 1.55, respectively. Thus, the node sorting result calculated according to the weight \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$w^{*}$$\end{document} is more reliable. Therefore, the integrated weighting method based on the multi-objective programming model is adopted to determine the weight value of each evaluation index and further used to calculate the importance sorting result of the nodes. The impact of evaluation index on the importance of nodes Next, the impact of different evaluation index sets on the evaluation and sorting of node importance is studied by comprehensively using the saturation reference value of the node operating state, the road network impact evaluation model based on topological potential, the integrated weighting method based on the multi-objective programming model, and the evaluation index preprocessing data. Optimization of traffic volume and saturation Index scheme set 2 (node degree \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$K_{i}$$\end{document} + clustering coefficient \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$C_{i}$$\end{document} + betweenness \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$B_{i}$$\end{document} + saturation \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\alpha}_{i}$$\end{document} ) is selected. The index weight \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$w^{*}$$\end{document} is obtained using the integrated weighting method based on the multi-objective programming model. Its relationship with the combined weights (w21 and w22) is shown in Fig. 10, and the correlation coefficients between the sorting results corresponding to each weight are shown in 1222. As shown in Fig. 11, the node sorting result corresponding to \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$w^{*}$$\end{document} presents advantages; thus, the sorting result is taken as the final node sorting result.Fig. 10Weight comparison (index scheme 2).Fig. 11Correlation coefficient diagram of sorting results (index scheme 2). Weight comparison (index scheme 2). Correlation coefficient diagram of sorting results (index scheme 2). Further, by comparing the node sorting results corresponding to index schemes 1 and 2, Fig. 12 is obtained.Fig. 12Sorting results of node importance corresponding to two evaluation index schemes. Sorting results of node importance corresponding to two evaluation index schemes. As shown in Fig. 12, after traffic volume in index scheme 1 is replaced with saturation for index scheme 2, the sorting results of the nodes also changed. The sorting results of the following nodes have increased significantly: 4(46/30), 6 (34/17), 13(38/24), 15(25/15), 16(20/7), 17(37/16), 19(30/12), 36(42/11), 37(17/5), 50(21/10), 53(56/39), and 55(55/45). The nodes with obvious drops in sorting results are 9(18/29), 10(8/25), 18(41/52), 25(23/38), 26(2/28), 42(5/22), 43(15/27), 47(10/33), 49(9/34), 57(36/51), and 59(14/44). The changes in the sorting results are explained using nodes 6 and 10 as case studies. With respect to the position of the nodes in the road network, the distance between nodes 6 and 10 is relatively close and the number of neighbor nodes is the same; thus, the values of the node degree, clustering coefficient, and betweenness are close. Considering traffic volume, the traffic volume of node 10 (5088 pcu/h) is larger than that of node 6 (4020 pcu/h); thus, when index scheme 1 is adopted, the sorting of node 10 is higher (8th), whereas the sorting of node 6 is significantly behind (34th). However, after considering the traffic capacity of the nodes, the saturation value of the two changed. The saturation value of node 6 is 0.84, whereas the saturation value of node 10 is only 0.75. Therefore, the sorting of node 10 dropped from the original 8th to 25th, whereas the sorting of node 6 rose from 34 to 17th. Undoubtedly, the change in the evaluation index directly influences the importance sorting of nodes and sometimes the change is more significant. Therefore, the rational selection of evaluation indexes is of significance. Because the nodes with large traffic volumes are not necessarily more congested, and according to the actual demand of evaluation of the traffic operation state of road networks, index scheme 2 is considered better than index scheme 1. Therefore, in the follow-up case study and analysis, other evaluation indexes are added based on index scheme 2. The influence of dynamic indexes on the evaluation of node importance (4) Index scheme set 3 (node degree \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$K_{i}$$\end{document} + clustering coefficient \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$C_{i}$$\end{document} + betweenness \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$B_{i}$$\end{document} + saturation \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\alpha}_{i}$$\end{document}) is selected.By comparing the node ranking results corresponding to index schemes 2 and 3, Fig. 13 is obtained.Fig. 13Sorting results of node importance corresponding to two evaluation index schemes. Sorting results of node importance corresponding to two evaluation index schemes. As shown in Fig. 13, after the dynamic index is added for index scheme 3 based on index scheme 2, the sorting results of the nodes also changed. The nodes whose sorting results have increased significantly are 3(32/24), 4(30//19), 11(43/34), 25(38/21), 26(28/20), 31(26/14), 38(18/11), 40(41/26), 46(21/ 12), and 60 (36/17). The nodes with significant drops in sorting results are 5 (31/38), 6 (17/33), 13 (24/31), 17 (16/32), 48 (14/23), and 54 (19/30). The changes in the sorting results are explained using nodes 3 and 13 as examples. Regarding the node positions, nodes 3 and 13 are both located at the edge of the road network in the core area (as shown in Fig. 4), node 3 is located on the east side, whereas node 13 is located on the south edge and enjoy a certain similarity with node 3. As regards the node evaluation data, as shown in Table 7, the preprocessing results (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$K_{i}{\prime}$$\end{document},\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$C_{{_{i} }}{\prime}$$\end{document}) of the node degree and clustering coefficient of the two nodes are similar and the difference between the preprocessing results (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$B_{{_{i} }}{\prime}$$\end{document},\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\alpha_{{_{i} }}{\prime}$$\end{document}) of the betweenness and saturation is close. However, based on the weight, in index scheme 2, the sorting of node 13 is 24th, and the sorting of node 3 is 32nd. Based on the dynamic index , the sorting of node 3 reversed, rising from 32nd to 24th, and the sorting of node 13 dropped from 24th to 31st. Combine with Fig. 4, the sorting results change because when the node operation state is changed from unimpeded to slightly congested, node 3 has a more significant influence on the surrounding nodes (the mutual travel distance between surrounding nodes is significantly greater than that of node 13), this in turn makes the overall topological potential of the road network drop more sharply; thus, the betweenness plays an indirect role.Table 7Comparison of related data between nodes 3 and 13.ID\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$K_{i}{\prime}$$\end{document}\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$C_{i}{\prime}$$\end{document}\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$B_{i}{\prime}$$\end{document}\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\alpha_{i}^{\prime }$$\end{document}\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\Delta \varphi_{1i}^{\prime }$$\end{document}Sorting of index scheme 2Sorting of index scheme 330.6001.0000.4730.8660.4223224130.6001.0000.3900.9200.2432431 Table 7Comparison of related data between nodes 3 and 13.ID\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$K_{i}{\prime}$$\end{document}\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$C_{i}{\prime}$$\end{document}\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$B_{i}{\prime}$$\end{document}\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\alpha_{i}^{\prime }$$\end{document}\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\Delta \varphi_{1i}^{\prime }$$\end{document}Sorting of index scheme 2Sorting of index scheme 330.6001.0000.4730.8660.4223224130.6001.0000.3900.9200.2432431 Comparison of related data between nodes 3 and 13. Because the dynamic index influences the node importance and its sorting result, it should be included in the evaluation of the node importance of the road network. (5) The influence of different dynamic indexes on the evaluation of node importance. Similarly, index scheme sets 4 and 5 are selected. As shown in Fig. 14, the corresponding node importance sorting results differ for different index schemes. The difference between the node sorting results of index schemes 4, 3, and 5, is relatively large, whereas the difference between the node sorting results of index schemes 3 and 5 is relatively small.Fig. 14Comparison of sorting results. Comparison of sorting results. The sorting of node importance differs in the traffic flow gathering and dissipating stages. In addition to paying attention to the nodes in the top sorting, we should also focus on those nodes with a larger increase in sorting. Further, although the index schemes differ, the nodes at the top and bottom of the sorting have relatively minor changes and are stable and the nodes with relatively large changes in the sorting results are mainly located in the middle. Node grading The importance of classifying nodes in a road network into grades is of significance in traffic operation management and control. Reasonable node grading is conducive to the discovery of important nodes, which is the premise and basis for targeted management and control. To fully utilize the different node sorting results under the conditions of three dynamic indexes, our proposed node grading algorithm based on various sorting results is used to grade nodes in the road network in the core area of Xining City, Qinghai Province. The results are shown in Table 8.Table 8Final grading results of each index scheme.Classification levelGrading results(Index scheme 3)Grading results(Index scheme 4)Grading results(Index scheme 5)Very important(10 in total)33 , 22 , 39 , 45 , 28 , 37 , 16 , 23 , 24 , 3533 , 22 , 39 , 45 , 28 , 16 , 23 , 37 , 24 , 3533 , 22 , 39 , 45 , 28 , 23 , 16 , 37 , 24 , 35important(39 in total)38, 46, 50, 31, 15, 36, 60, 19, 4, 26, 25, 27, 48, 3, 42, 40, 10, 14, 43, 54, 13, 17, 6, 11, 9, 47, 52, 5, 59, 49, 44, 34, 55, 2, 53, 20, 21, 8, 3250, 15, 36, 19, 14, 27, 48, 43, 46, 40, 10, 38, 17, 9, 34, 42, 13, 25, 26, 44, 6, 31, 49, 54, 60, 21, 11, 4, 3, 47, 20, 5, 32, 53, 55, 59, 52, 8, 250, 38, 46, 15, 36, 31, 19, 60, 27, 40, 48, 26, 10, 4, 25, 14, 3, 42, 54, 43, 17, 13, 6, 11, 9, 47, 49, 52, 44, 5, 34, 59, 2, 53, 55, 20, 21, 8, 32Common(11 in total)18, 7, 57 , 56 , 51 , 1 , 58 , 30 , 12 , 29 , 417 , 18, 51 , 57 , 56 , 1 , 58 , 30 , 29 , 12 , 417, 18, 57 , 56 , 51 , 1 , 58 , 30 , 12 , 29 , 41 Table 8Final grading results of each index scheme.Classification levelGrading results(Index scheme 3)Grading results(Index scheme 4)Grading results(Index scheme 5)Very important(10 in total)33 , 22 , 39 , 45 , 28 , 37 , 16 , 23 , 24 , 3533 , 22 , 39 , 45 , 28 , 16 , 23 , 37 , 24 , 3533 , 22 , 39 , 45 , 28 , 23 , 16 , 37 , 24 , 35important(39 in total)38, 46, 50, 31, 15, 36, 60, 19, 4, 26, 25, 27, 48, 3, 42, 40, 10, 14, 43, 54, 13, 17, 6, 11, 9, 47, 52, 5, 59, 49, 44, 34, 55, 2, 53, 20, 21, 8, 3250, 15, 36, 19, 14, 27, 48, 43, 46, 40, 10, 38, 17, 9, 34, 42, 13, 25, 26, 44, 6, 31, 49, 54, 60, 21, 11, 4, 3, 47, 20, 5, 32, 53, 55, 59, 52, 8, 250, 38, 46, 15, 36, 31, 19, 60, 27, 40, 48, 26, 10, 4, 25, 14, 3, 42, 54, 43, 17, 13, 6, 11, 9, 47, 49, 52, 44, 5, 34, 59, 2, 53, 55, 20, 21, 8, 32Common(11 in total)18, 7, 57 , 56 , 51 , 1 , 58 , 30 , 12 , 29 , 417 , 18, 51 , 57 , 56 , 1 , 58 , 30 , 29 , 12 , 417, 18, 57 , 56 , 51 , 1 , 58 , 30 , 12 , 29 , 41 Final grading results of each index scheme. Very important (10 in total) important (39 in total) Common (11 in total) As shown in Table 8, although the sorting results differ, the importance grading of road network nodes still follows a certain rule. The difference is that the sorting of nodes at the very important and general levels is relatively stable and changes slightly, whereas the sorting of nodes at the important level changes significantly. Particularly, 10 nodes such as 16, 22, 23, 24, 28, 33, 35, 37, 39, and 45, are in a very important position in three index schemes and should be of significance. Conclusion Starting from the actual needs of evaluating the importance of nodes in a road network by considering the dynamic index of changes in the operating state of road networks caused by changes in the operating state of nodes, an evaluation index system integrating static and dynamic indicators is constructed. The selection of evaluation indexes, data acquisition methods, index weight determination, node importance sorting, node grading, and other contents have been studied in depth. The relevant methods and models improved the scientificity and rationality of the evaluation results of the importance of nodes in a road network and can be used as the basis for decision making by traffic management departments on traffic operation management. The main conclusions are as follows:This paper pioneeringly proposes a node importance calculation paradigm that fuses structure and attributes, performing node importance evaluation and grading while simultaneously maintaining network structural constraints and integrating multi-dimensional attributes. This "structure-attribute-decision" three-layer coupling mechanism is unattainable by any existing single method or simple integration.The change in the operation state of road networks caused by a change in the operation state of a node was included in this study; the evaluation index system comprising static indexes such as the road network node degree, clustering coefficient, betweenness, average traffic volume during peak hours/average saturation during peak hours, and dynamic index of the changes in the operation state the road network caused by the change in the operation state of nodes was constructed and five different evaluation index schemes were provided.To better obtain the evaluation data of dynamic indexes, first, a method to determine the saturation reference value based on residents’ travel time rule was proposed, which provided a numerical basis to reflect the operation status of nodes. Second, as regards the problems existing in the traditional efficiency evaluation model, an impact evaluation model based on topological potential was proposed and a ternary iterative algorithm was introduced to calculate the impact factor of the model. This model overcame the failure of the original model with an increase in the number of nodes. The changes in the operating state of road networks caused by the changes in the operating states of different nodes were effectively evaluated and distinguished by this model. Third, based on the two methods and models mentioned above, combined with three different node operating state changes, an algorithm to obtain dynamic indexes was proposed.Aiming at the problems existing in the current index weighting methods, an integrated weighting method based on multi-objective programming was proposed and the standard PSO was introduced to solve these problems. Three objective functions were used in the method to ensure its index weight and node sorting results were significantly better than other combined weight methods; the problem of inconsistent node sorting results caused by different principles of combined weight methods was solved and the decision making of importance sorting of nodes in a road network was effectively supported.The influence of different evaluation indexes on the evaluation of node importance was analyzed and compared. Meanwhile, to effectively integrate the node sorting results of different index schemes, a node sorting algorithm based on multiple sorting results was proposed. Assuming that each sorting result remains unchanged, the consistency of node sets within each grading was guaranteed, which is more conducive to discovering important nodes.Taking 60 road intersections in the core area of the urban road of Xining City, Qinghai Province as the object, the experimental process of the case study was designed and actual data was used for comparative analysis. Based on the results, the change in the evaluation index directly influences the importance of nodes and their sorting and the dynamic index of the change in the operation state of road networks caused by the change in the operation state of notes should be considered. The change in the index weight influenced the evaluation of node importance and its sorting and the weight of the index should be reasonably determined. Grading nodes by considering the internal relationship and laws of various sorting results increased the reliability of important nodes.The method described in this paper utilized a Lenovo laptop computer (Model: ThinkPad E490) and MATLAB R2020a to conduct relevant experiments. Taking 60 intersections in the core area of Xining’s urban roads as an example, statistics indicate that the average time for multi-core (quad-core) parallel computing is only 0.6 s. The computational complexity of the method proposed in this paper is mainly influenced by the calculation of the overall operational level of the road network nodes, with a time complexity of O(n3), where n is the total number of road network nodes. Considering that the proposed method is primarily applied to urban road networks, if n is taken in the order of 102 to 104, the computation can still be completed within a reasonable time (from minutes to several hours) on modern computing platforms, which is acceptable for planning-type and off-line analysis applications. This paper pioneeringly proposes a node importance calculation paradigm that fuses structure and attributes, performing node importance evaluation and grading while simultaneously maintaining network structural constraints and integrating multi-dimensional attributes. This "structure-attribute-decision" three-layer coupling mechanism is unattainable by any existing single method or simple integration. This paper pioneeringly proposes a node importance calculation paradigm that fuses structure and attributes, performing node importance evaluation and grading while simultaneously maintaining network structural constraints and integrating multi-dimensional attributes. This "structure-attribute-decision" three-layer coupling mechanism is unattainable by any existing single method or simple integration. The change in the operation state of road networks caused by a change in the operation state of a node was included in this study; the evaluation index system comprising static indexes such as the road network node degree, clustering coefficient, betweenness, average traffic volume during peak hours/average saturation during peak hours, and dynamic index of the changes in the operation state the road network caused by the change in the operation state of nodes was constructed and five different evaluation index schemes were provided. The change in the operation state of road networks caused by a change in the operation state of a node was included in this study; the evaluation index system comprising static indexes such as the road network node degree, clustering coefficient, betweenness, average traffic volume during peak hours/average saturation during peak hours, and dynamic index of the changes in the operation state the road network caused by the change in the operation state of nodes was constructed and five different evaluation index schemes were provided. To better obtain the evaluation data of dynamic indexes, first, a method to determine the saturation reference value based on residents’ travel time rule was proposed, which provided a numerical basis to reflect the operation status of nodes. Second, as regards the problems existing in the traditional efficiency evaluation model, an impact evaluation model based on topological potential was proposed and a ternary iterative algorithm was introduced to calculate the impact factor of the model. This model overcame the failure of the original model with an increase in the number of nodes. The changes in the operating state of road networks caused by the changes in the operating states of different nodes were effectively evaluated and distinguished by this model. Third, based on the two methods and models mentioned above, combined with three different node operating state changes, an algorithm to obtain dynamic indexes was proposed. To better obtain the evaluation data of dynamic indexes, first, a method to determine the saturation reference value based on residents’ travel time rule was proposed, which provided a numerical basis to reflect the operation status of nodes. Second, as regards the problems existing in the traditional efficiency evaluation model, an impact evaluation model based on topological potential was proposed and a ternary iterative algorithm was introduced to calculate the impact factor of the model. This model overcame the failure of the original model with an increase in the number of nodes. The changes in the operating state of road networks caused by the changes in the operating states of different nodes were effectively evaluated and distinguished by this model. Third, based on the two methods and models mentioned above, combined with three different node operating state changes, an algorithm to obtain dynamic indexes was proposed. Aiming at the problems existing in the current index weighting methods, an integrated weighting method based on multi-objective programming was proposed and the standard PSO was introduced to solve these problems. Three objective functions were used in the method to ensure its index weight and node sorting results were significantly better than other combined weight methods; the problem of inconsistent node sorting results caused by different principles of combined weight methods was solved and the decision making of importance sorting of nodes in a road network was effectively supported. Aiming at the problems existing in the current index weighting methods, an integrated weighting method based on multi-objective programming was proposed and the standard PSO was introduced to solve these problems. Three objective functions were used in the method to ensure its index weight and node sorting results were significantly better than other combined weight methods; the problem of inconsistent node sorting results caused by different principles of combined weight methods was solved and the decision making of importance sorting of nodes in a road network was effectively supported. The influence of different evaluation indexes on the evaluation of node importance was analyzed and compared. Meanwhile, to effectively integrate the node sorting results of different index schemes, a node sorting algorithm based on multiple sorting results was proposed. Assuming that each sorting result remains unchanged, the consistency of node sets within each grading was guaranteed, which is more conducive to discovering important nodes. The influence of different evaluation indexes on the evaluation of node importance was analyzed and compared. Meanwhile, to effectively integrate the node sorting results of different index schemes, a node sorting algorithm based on multiple sorting results was proposed. Assuming that each sorting result remains unchanged, the consistency of node sets within each grading was guaranteed, which is more conducive to discovering important nodes. Taking 60 road intersections in the core area of the urban road of Xining City, Qinghai Province as the object, the experimental process of the case study was designed and actual data was used for comparative analysis. Based on the results, the change in the evaluation index directly influences the importance of nodes and their sorting and the dynamic index of the change in the operation state of road networks caused by the change in the operation state of notes should be considered. The change in the index weight influenced the evaluation of node importance and its sorting and the weight of the index should be reasonably determined. Grading nodes by considering the internal relationship and laws of various sorting results increased the reliability of important nodes. Taking 60 road intersections in the core area of the urban road of Xining City, Qinghai Province as the object, the experimental process of the case study was designed and actual data was used for comparative analysis. Based on the results, the change in the evaluation index directly influences the importance of nodes and their sorting and the dynamic index of the change in the operation state of road networks caused by the change in the operation state of notes should be considered. The change in the index weight influenced the evaluation of node importance and its sorting and the weight of the index should be reasonably determined. Grading nodes by considering the internal relationship and laws of various sorting results increased the reliability of important nodes. The method described in this paper utilized a Lenovo laptop computer (Model: ThinkPad E490) and MATLAB R2020a to conduct relevant experiments. Taking 60 intersections in the core area of Xining’s urban roads as an example, statistics indicate that the average time for multi-core (quad-core) parallel computing is only 0.6 s. The computational complexity of the method proposed in this paper is mainly influenced by the calculation of the overall operational level of the road network nodes, with a time complexity of O(n3), where n is the total number of road network nodes. Considering that the proposed method is primarily applied to urban road networks, if n is taken in the order of 102 to 104, the computation can still be completed within a reasonable time (from minutes to several hours) on modern computing platforms, which is acceptable for planning-type and off-line analysis applications. The method described in this paper utilized a Lenovo laptop computer (Model: ThinkPad E490) and MATLAB R2020a to conduct relevant experiments. Taking 60 intersections in the core area of Xining’s urban roads as an example, statistics indicate that the average time for multi-core (quad-core) parallel computing is only 0.6 s. The computational complexity of the method proposed in this paper is mainly influenced by the calculation of the overall operational level of the road network nodes, with a time complexity of O(n3), where n is the total number of road network nodes. Considering that the proposed method is primarily applied to urban road networks, if n is taken in the order of 102 to 104, the computation can still be completed within a reasonable time (from minutes to several hours) on modern computing platforms, which is acceptable for planning-type and off-line analysis applications. Limitations and future work In the process of determining the impact of changes in node operating status on changes in the operating status of the road network, the node operating status change propagation model and algorithm adopted in this study remain relatively simple. In future work, they will be further optimized by incorporating node traffic management and control strategies, the comprehensive attributes of road sections adjacent to the nodes, and residents’ travel OD data passing through the nodes.