基于多智能体仿真的交通诱导系统效率评价  

Operational Evaluation of Traffic Guidance Systems Based on Multi-Agent-Based Simulation

作　　者：唐克双[1] 张桁嘉 衣谢博闻[1] Tang Keshuang;Zhang Hengjia;Yixie Bowen(Key Laboratory of Road and Traffic Engineering of Ministry of Education, Tongji University, Shanghai 201804, China)

机构地区：[1]同济大学道路与交通工程教育部重点实验室,上海201804

出　　处：《系统仿真学报》2018年第7期2630-2639,共10页Journal of System Simulation

基　　金：国家自然科学基金(61673302)

摘　　要：为了研究基于可变情报板的交通诱导系统对交通资源利用效率的优化效果,利用多智能体仿真模型开发了仿真与评价平台。对3种典型交通诱导策略进行了仿真实验,采用平均行驶时间、平均巡泊时间和停车失败率的指标评价了诱导效果。结果表明:行车诱导和停车诱导分别对平均行驶时间和停车失败率有改善效果,并且交通诱导系统在不同交通需求水平下对交通运行状态和停车资源使用状况的影响存在差异性,在发布诱导信息时,需要综合考虑交通需求水平和诱导策略。To study the optimization effect of variable message sign-based traffic guidance system on the utilization efficiency of transportation resource, the multi-agent-based model is used to develop simulation and evaluation platform. The simulation experiments of three typical traffic guidance strategies are carried out. The effect of guidance is evaluated by the indexes as average driving time, average searching time for parking lot and parking failure rate. The result shows that the driving and parking guidance can help reduce the average driving time and the parking failure rate respectively. The influence of the traffic guidance system on the traffic operation status and the utilization of parking resource is different under different traffic demand levels, so the distribution of guidance information needs comprehensive consideration of traffic demand level and guiding strategy.

关 键 词：交通诱导系统 可变情报板 路径选择模型 停车选择模型 多智能体仿真 

分 类 号：U491.1[交通运输工程—交通运输规划与管理]