基于Agent的交叉口群微观交通仿真系统与应用

Agent-based microscopic traffic simulation system and its application at intersections group

为了改善车辆在交叉口群的运行效率,缩短车辆排队长度和减少车辆起停次数,在对交叉口群微观交通仿真中车辆、交叉口、车道、信号灯实体进行Agent建模基础上,采用AgentSpeak语言设计了车辆Agent的跟驰行为、交叉口避撞与左转待转行为、以及换道行为等模块,实现了基于Agent的交叉口群微观交通仿真系统。以西安市小寨商圈交叉口群为例,车辆Agent分别按照随机速度行驶和最大限速行驶意图模型进行仿真试验。研究结果表明:采用最大限速行驶策略时,车辆在交叉口群的平均延误时间可降低25.7%,平均起停次数可降低35.8%,这为改进现有交叉口群车辆通行控制策略和管理措施提供了思路及评价手段。

In order to improve the operational efficiency of vehicles,shorten the queue length and reduce the start-stop times of vehicles in the intersections group,Agent-based vehicle following model,intersection collision avoidance model waiting behaviors for turning left and lane change behaviors model were designed,and then a micro traffic simulation system in the intersections group was implemented by modeling the vehicle,intersection,lane and light in AgentSpeak language.In addition,taking the intersections group at Xiaozhai business area in Xi＇an an as example,two kinds of Agent intentions,random speed and maximum limit speed,were considered in the simulation.The results indicate that the vehicle with maximum limit speed intention can greatly decrease the average delay time by 25.7%and the start-stop times by35.8%in the intersections group,which can provide a guideline and evaluation tool for improving the existing intersection traffic control strategies and management measures.