Efficient Origin-Destination Estimation Using Microscopic Traffic Simulation with Restricted Rerouting  

作　　者：Kazuki Abe Hideki Fujii Shinobu Yoshimura 

机构地区：[1]Vector Research Institute,Inc.,Shibuya-ku,Tokyo,150-0002,Japan [2]School of Engineering,The University of Tokyo,Bunkyo-ku,Tokyo,113-8656,Japan

出　　处：《Computer Modeling in Engineering & Sciences》2023年第5期1091-1109,共19页工程与科学中的计算机建模（英文）

基　　金：supported by JSPS KAKENHI (Grant Nos.15H01785 and 19H02377).

摘　　要：Traffic simulators are utilized to solve a variety of traffic-related problems.For such simulators,origin-destination(OD)traffic volumes as mobility demands are required to input,and we need to estimate them.The authors regard an OD estimation as a bi-level programming problem,and apply a microscopic traffic simulation model to it.However,the simulation trials can be computationally expensive if full dynamic rerouting is allowed,when employing multi-agent-based models in the estimation process.This paper proposes an efficient OD estimation method using a multi-agent-based simulator with restricted dynamic rerouting to reduce the computational load.Even though,in the case of large traffic demand,the restriction on dynamic rerouting can result in heavier congestion.The authors resolve this problem by introducing constraints of the bi-level programming problem depending on link congestion.Test results show that the accuracy of the link traffic volume reproduced with the proposed method is virtually identical to that of existing methods but that the proposed method is more computationally efficient in a wide-range or high-demand context.

关 键 词：ODestimation microscopic traffic simulation dynamic rerouting bi-level programming multi-agent basedmodel 

分 类 号：TP3[自动化与计算机技术—计算机科学与技术]