基于多智能体仿真的城市轨道交通限流策略

Passenger Flow Control Strategy of Urban Rail Transit Based on Multi-Agent Simulation

摘要:

针对高峰时段城市轨道交通线路过饱和客流的运输效率问题，基于地铁线路的网络拓扑结构与客流需求矩阵，建立了以客运周转量最大为目标的城市轨道交通线网限流的多智能体仿真模型。该模型以Anylogic软件为建模基础，通过构建列车类、线路类、路网类、乘客类等4类智能体，实现了乘车限流、站台限流和闸机限流等客流控制手段，提出了针对单线多站地铁线路的全线协调的客流控制策略。规模为23个车站、122 933对OD（起讫点）的某条地铁线路的客流控制实验结果表明，限流提升了1.01%的总客运周转量。

Abstract:

Aimed at the transportation efficiency problem of oversaturated passenger flow of urban rail transit lines during peak hours， a multi-agent simulation model of passenger flow control for urban rail transit with the goal of maximizing passenger turnover is established based on the network topology of subway lines and the passenger flow demand matrix. Based on the AnyLogic software， the model reconstructs the urban rail transit system by constructing four types of agents： train， line， road network， and passenger， which realizes several passenger flow control methods， including boarding control， platform control， and gates control. In addition， a full-line coordinated passenger flow control strategy in multi stations of a single line in urban rail transit is proposed. Moreover， a passenger flow control experiment is conducted on a subway line with a scale of 23 stations and 122 933 passenger OD pairs， which increases the total passenger turnover by 1.01%.