A Traffic Evacuation Model for Enhancing Resilience During Railway Disruption

Hangli Ge 1 , Xiaojie Yang 2 , Jinyu Chen 1 , Francesco Flammini 3,4 , Noboru Koshizuka 1 · 2025

Read A Traffic Evacuation Model for Enhancing Resilience During Railway Disruption by Hangli Ge 1 , Xiaojie Yang 2 , Jinyu Chen 1 , Francesco Flammini 3,4 , Noboru Koshizuka 1 (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2503-05078.txt.

Introduction

Hangli Ge and co-authors build a traffic evacuation model for railway disruption in a 2025 paper on resilience in the Greater Tokyo rail network. When stations or lines fail, crowds must move to nearby stations without swamping them.

The model joins lines across operators with local geography in one frame. An optimized plan is solved with nonlinear math on matrices, weighing station capacity and travel cost. It can simulate failure at any number of stations or lines. Trials on Tokyo, Shinjuku, and Shibuya stations kept flows within neighbor capacity and recovery time. Average travel time also improved. Transit planners get a tested tool for crisis routing in dense rail cities.

Catalog record

CREATORHangli Ge 1 , Xiaojie Yang 2 , Jinyu Chen 1 , Francesco Flammini 3,4 , Noboru Koshizuka 1
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE25 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2503-05078

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

A Traffic Evacuation Model for Enhancing Resilience During Railway Disruption Hangli Ge 1 , Xiaojie Yang 2 , Jinyu Chen 1 , Francesco Flammini 3,4 , Noboru Koshizuka 1 Abstract This paper introduces a traffic evacuation model for railway disruptions to improve resilience. The research focuses on the problem of failure of several nodes or lines on the railway network topology. We proposed a holistic approach that integrates lines of various operator companies as well as external geographical features of the railway system. The optimized evacuation model was mathematically derived based on matrix computation using nonlinear programming. The model also takes into account the capacity of the surrounding evacuation stations, as well as the travel cost. Moreover, our model can flexibly simulate disruptions at multiple stations or any number of stations and lines, enhancing its applicability. We collected the large-scale railway network of the Greater Tokyo area for experimentation and evaluation. We simulated evacuation plans for several major stations, including Tokyo, Shinjuku, and Shibuya. The results indicate that the evacuation passenger flow (EPF) and the average travel time (ATT) during emergencies were optimized, staying within both the capacity limits of the targeted neighboring stations and the disruption recovery time. I Introduction Resilience is a concept widely applied across fields such as economics, psychology, and engineering. It originally derived from the Latin word “resiliere”, was first defined in the ecological system [ 1 ] . Transportation systems, being at the forefront of various environmental and socioeconomic shocks, have made the study of resilience a topic of significant interest among transportation scholars. In particular, railway systems, as critical infrastructure, play a fundamental role in passenger or cargo transport, significantly contributing to economic growth, urban development, and environmental sustainability. However, under abnormal or emergency conditions—such as natural disasters, equipment failures, cyber attacks, operational disruptions, or accidents—trains may fail to arrive or depart on schedule or even experience service interruptions, leading to significant variations in railway operations. These issues can result in congestion, stoppages, or accidents, ultimately reducing passenger service quality and causing economic and productivity losses. Therefore, enhancing the resilience of transportation systems is essential for mitigating the impact of such disruptions [ 2 , 3 ] . Fig. 1: The railway network of Greater Tokyo area Improving the resilience of railways to recover to normal conditions more quickly is a critical problem in the management of railway systems. Among various recovery strategies, traffic analysis and evacuation simulation are particularly significant. However, these tasks are highly complex due to the lack of readily available large-scale data on railway networks and passenger flows—whether at the cross-city, nationwide, or cross-country level. If the model focuses solely on a single railway line or lines operated by the same company, the process is relatively straightforward. However, constructing a comprehensive rail network presents significant challenges. The involvement of multiple railway operators further complicates data collection, integration, computation, and the simulation of decision-making within a large-scale railway system. In this research, we propose a traffic evacuation model to enhance railway resilience during disruptions. The model addresses the failure of multiple nodes or lines within the network topology. We introduce an optimization model for traffic evacuation that considers delay times, overcrowded passenger flows, and the capacities of neighboring stations. We constructed a large-scale railway network dataset for the Greater Tokyo area (visualized in Fig. 1 ), integrating data from multiple operating companies and incorporating geographical features such as connectivity and distance, which were modeled in the evacuation cost matrix. Our traffic evacuation model employs an optimization algorithm based on matrix calculations, ensuring both intuitiveness and interpretability. The key contributions of this research can be summarized as follows: • Proposal of a fusion approach which integrates multiple geographic features to construct an evacuation cost matrix that dynamically adjusts based on predefined disruption durations. • Development of an optimization model using nonlinear programming that accounts for the capacity of surrounding stations and travel costs. Additionally, the proposed model can flexibly simulate disruptions affecting multiple stations or spanning cross multiple lines, enhancing its applicability. Built on a solid mathematical foundation, the model leverages straightforward matrix calculations and optimization solutions, making it intuitive, easy to understand, and simple to deploy. • Validation through a case study on Greater Tokyo’s large-scale railway network. Geographic features and passenger statistics were incorporated into simulations for several major stations. The optimization focused on evacuation passenger flow (EPF) and the average travel time (ATT), ensuring both remained within the capacity limits of the evacuated target stations and adhered to the designated disruption time. II Related Work II-A Resillience Assessment A resilient transport system is crucial for maintaining functionality and accessibility, especially in the face of unexpected events, such as natural disasters, accidents or infrastructure failures [ 4 ] . However, the operationalization of resilience practices in transport planning and found that resilience is not formally considered [ 5 ] . Researchers have developed various models and systems to address railway resillience issues and continue to explore new methods for improving passenger experience and reducing delay or congestion when abnormaly events occur on the railway system. However, regarding the resillience, little is known about the performance dynamics during disruptions [ 6 ] . Data-driven approach to quantify the resilience of the railway network through assessment, utilizing historical traffic data from the Netherlands to reconstruct resilience curves and analyze disruption dynamics using composite indicators and statistical tests, providing valuable insights for improving system resilience [ 6 ] . Similar research has evaluated the resilience of complementary networks in various disaster scenarios from the perspective of network structure and function [ 4 ] . These resilience-related studies are constructive, but one limitation in resilience measurement is that the features of the network have not been fully considered. II-B Practice for Improving Resillience Several practices are used to improve the resilience of railway systems. They put the focus on transport planning mainly include the identification of critical nodes and edges [ 7 ] , transport network design [ 8 ] , emergency response, and evacuation [ 9 , 5 ] . Digital twins can help optimize maintenance, reduce downtime, and improve safety by providing real-time monitoring and analysis of system performance [ 10 ] . In addition, the use of sensors and data analytics or AI (Artificial intelligence)-based predictions can help predict when maintenance is required, reducing the likelihood of equipment failure and improving overall system reliability [ 11 ] . Similarly, a study proposes a novel method to optimize the resilience of urban rail systems, focusing on the combined impact of delayed trains and overcrowded passenger flows [ 12 ] . Additionally, the use of complex network principles can help analyze the network structure and identify potential bottlenecks, enabling targeted improvements. “KUTTY” simulator, which simulates the flow of passengers quantitatively in all segments of the network [ 13 ] . The model has also been used to simulate the effects of a virtual accident at Otemachi Station, showing that the flow pattern of passengers will change significantly [ 13 ] . Traffic evacuation models also have been proposed for optimizing the evacuation process, minimize travel delays, and ensure the safety of passengers. An approach is to use a resilience-based optimization model, which

Read the complete text