A Nonlocal Aw-Rascle-Zhang system with linear pressure term

D. Amadori, F. A. Chiarello, G. Cipollone · 2025

Read A Nonlocal Aw-Rascle-Zhang system with linear pressure term by D. Amadori, F. A. Chiarello, G. Cipollone (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-2509-20973.txt.

Introduction

Traffic models that treat each driver as reacting only to the car ahead miss how drivers look further down the road. Debora Amadori, Felisia Angela Chiarello, and Gianmarco Cipollone study a nonlocal traffic model in a 2025 mathematics paper. Their system extends the Aw-Rascle-Zhang model of traffic flow.

In the nonlocal version, the pressure-like term becomes a convolution of vehicle density with a kernel, spreading interactions over a stretch of road. The structure parallels the Euler-alignment systems used for flocking. The authors build entropy solutions by a sticky particle approximation. They prove the approximations converge to weak solutions, with well-posedness and stability estimates included. The analysis gives the nonlocal traffic model a firm mathematical footing.

Catalog record

CREATORD. Amadori, F. A. Chiarello, G. Cipollone
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE72 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2509-20973

Browse the full library · About the Tower of Babel

Full text

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

A Nonlocal Aw-Rascle-Zhang system with linear pressure term D. Amadori, F. A. Chiarello, G. Cipollone Abstract . In this paper, we study a nonlocal extension of the Aw-Rascle-Zhang traffic model, where the pressure-like term is modeled as a convolution between vehicle density and a kernel function. This formulation captures nonlocal driver interactions and aligns structurally with the Euler-alignment system studied in [ 23 ] . Using a sticky particle approximation, we construct entropy solutions to the equation for the cumulative density and prove convergence of approximate solutions to weak solutions of the nonlocal system. The analysis includes well-posedness, stability estimates, and an entropic selection principle. Keywords : Nonlocal Aw-Rascle-Zhang system, sticky particle dynamics, weak solutions 1991 Mathematics Subject Classification Primary: 35D30, 35L65, 35Q92; Secondary: 76N10, 35B30. 1. Introduction Traffic flow models have been studied since many decades, following the growing scientific interest raised by the applications. Within the class of macroscopic models, the classical LWR model [ 24 , 27 ] is based on a closure relation that expresses the velocity as a function of the density and has been widely used for applications in engineering because of its simplicity. However, it shows some limitations, as for instance it is not suitable to predict some real-life complex traffic phenomena such as phantom jams. Second order models have been introduced to overcome the limitations for first order models. One of the well-known second order traffic models is the so-called Aw-Rascle-Zhang model (ARZ), introduced in [ 3 , 31 ] , that reads as (1.1) \begin{cases}\partial_{t}\rho+\partial_{x}(\rho u)=0,\\ \partial_{t}(u+p(\rho))+u\partial_{x}(u+p(\rho))=0\,.\end{cases} Here \rho(x,t)\geq 0 represents the macroscopic density of vehicles, i.e. the number of vehicles per unit length, u(x,t) the mean velocity and p(\rho) is a smooth increasing pressure-like function representing the attitude of drivers to adapt to the traffic conditions . In this paper, motivated by some recent results about nonlocal traffic models, see for example [ 11 , 16 , 10 ] , we introduce a nonlocal version of ( 1.1 ) where the pressure term is replaced by a convolution term between the density \rho and a kernel function \omega . In particular, we focus our analysis on the case p(\rho)=\rho . The system reads as (1.2) \begin{cases}\partial_{t}\rho+\partial_{x}(\rho u)=0,\\ \partial_{t}u+u\partial_{x}u=-(\partial_{t}+u\partial_{x})(\omega\ast\rho)\,.\end{cases} The second equation describes the acceleration of drivers along their motion, and implies that the quantity (1.3) \psi\mathrel{\mathop{\mathchar 58\relax}}=u+\omega\ast\rho satisfies the equation (1.4) \partial_{t}\psi+u\partial_{x}\psi=0\,. Multiplying the first equation of the system by \psi and the second one by \rho , the system can be written in conservation form as follows: (1.5) \begin{cases}\partial_{t}\rho+\partial_{x}(\rho u)=0,\\ \partial_{t}(\rho\psi)+\partial_{x}(\rho u\psi)=0.\end{cases} Throughout the paper we will assume that the initial total mass equals to 1 and then, by conservation of total mass, \int_{\mathbb{R}}\rho(x,t)\,dx=1 for all t>0 . The convolution kernel \omega in ( 1.2 ) is assumed to satisfy \textbf{(H)}\qquad\omega\mathrel{\mathop{\mathchar 58\relax}}\mathbb{R}\to[0,+\infty)\,,\quad\omega\in W^{1,1}(\mathbb{R})\cap W^{1,\infty}(\mathbb{R})\,,\qquad\int_{\mathbb{R}}\omega\,dx=1\,. Moreover, we will denote by \phi the derivative of \omega : (1.6) \phi\mathrel{\mathop{\mathchar 58\relax}}=\omega^{\prime}\in L^{1}(\mathbb{R})\cap L^{\infty}(\mathbb{R})\,. It is interesting to compare ( 1.2 ) with the Euler-alignment system, coming from the theory of collective behavior, see [ 23 ] , (1.7) \begin{cases}\partial_{t}\rho+\partial_{x}(\rho u)=0,\\ \partial_{t}(\rho u)+\partial_{x}(\rho u^{2})=\rho(\phi\ast(\rho u))-\rho u(\phi\ast\rho)\end{cases} where \rho and u represent the density and velocity, respectively, and the function \phi is called communication protocol governing the strength of the interactions between the agents. Indeed, we reformulate the second equation of ( 1.2 ) using the evaluations of the partial derivatives of the convolution products, i.e. \displaystyle\partial_{t}(\omega\ast\rho)(x,t) \displaystyle=\int_{-\infty}^{\infty}\partial_{t}\rho(y,t)\omega(x-y)\,dy=-\int_{-\infty}^{\infty}\partial_{x}(\rho u)(y,t)\omega(x-y)\,dy \displaystyle=-\int_{-\infty}^{\infty}(\rho u)(y,t)\phi(x-y)\,dy=-(\phi\ast(\rho u))(x,t), and \displaystyle u\partial_{x}(\omega\ast\rho)(x,t) \displaystyle=u(x,t)\int_{-\infty}^{\infty}\omega(x-y)\rho_{y}(y,t)\,dy \displaystyle=u(x)\omega(x-y)\rho(y,t)|_{-\infty}^{\infty}+u(x,t)\int_{-\infty}^{\infty}\phi(x-y)\rho(y,t)\,dy \displaystyle=u(\phi\ast\rho)(x,t). Now, the second equation of ( 1.2 ) reads (1.8) \partial_{t}u+u\partial_{x}u=\phi\ast(\rho u)-u(\phi\ast\rho); multiplying ( 1.8 ) by the density \rho and summing the equation of the conservation of mass, we obtain \partial_{t}(\rho u)+\partial_{x}(\rho u^{2})=\rho(\phi\ast(\rho u))-\rho u(\phi\ast\rho). Then, the nonlocal ARZ system ( 1.2 ) can be rewritten in the Euler-alignment form ( 1.7 ). System ( 1.7 ) with measure-valued density and bounded velocity has been studied in [ 23 ] , considering a communication protocol \phi\in L^{1}_{loc}(\mathbb{R}) which is non-negative and even. Appropriately chosen \phi lead to a long-time behavior referred to as flocking, see [ 28 ] . Moreover, the global regularity theory for smooth solutions has been developed in [ 8 ] via a critical threshold condition for the regular case, i.e. a communication protocol which is Lipschitz continuous and bounded. For weakly singular interaction, i.e. \phi with an integrable singularity at the origin, details about smooth solutions can be found in [ 22 ] , while for strongly singular interaction, see [ 14 , 20 , 29 ] . It is important to notice that system ( 1.7 ) is the hydrodynamic version of the Cucker-Smale system of ODE’s, see [ 12 ] . The derivation of a kinetic formulation corresponding to the Cucker-Smale system can be found in [ 18 , 7 , 26 ] . The purpose of this paper is to prove the well-posedness and stability estimates for system ( 1.5 ), and to provide an entropic selection principle. In particular, we develop a global well-posedness theory for weak solutions of ( 1.5 ) with measure-valued density and bounded velocity. Our approach is based on an adaptation of the sticky particle approximation, originally introduced by Brenier and Grenier [ 5 ] to analyze the 1D pressureless Euler equations and developed in [ 23 ] . The novelty with respect to the previous results in [ 23 ] is that we are considering an integrable and bounded kernel function \phi which is not necessarily symmetric (even) nor positive. The shape of its antiderivative \omega , see ( 1.6 ) and (H) , can be chosen closer to that one of the anisotropic kernel functions usually used in literature for nonlocal traffic flow models such as in [ 9 , 10 , 11 ] . The choice of a non-symmetric kernel function makes the analysis more involved and interesting, moreover it is coherent from the modeling point of view because drivers are supposed to adapt the velocity with respect to what happens mainly downstream in the traffic. It is worth quoting the following recent paper [ 19 ] in which the authors consider a nonlocal version of the ARZ model with nonlocality in the pressure density obtained as a convolution type term between a positive, bounded and non-increasing on [0,\infty) kernel function. In particular, [ 19 ] is focused on the well-posedness of smooth solutions and the phase-plane analysis. Other recent related papers are [ 25 ] and [ 1 ] . In the first paper the authors establish existence, uniqueness and Oleĭnik estimates, together with the singular limit, for a nonlocal version of the so-called Generalized Aw-Rascle-Zhang model (GARZ) [ 15 ] , where the drivers’ velocity depends on both the empty road velocity and the convolution of the density with an anisotropic kernel. In [ 1 ] the authors analyze the well-posedness and the singular limit for another nonlocal version of the GARZ model in which the nonlocality is present in the flux function as a convolution-type term between the velocity function and an anisotropic kernel function. The paper is organized as follows. - In Section 2 we introduce the microscopic approximation based on a nonlocal version of the sticky particle dynamics. - In Section 3 we introduce a

Read the complete text