基于Multi-Agent的动态路径诱导系统  

Dynamic Route Induction System Based on Multi-Agent

作　　者：朱晶晶[1] 罗志洁 

机构地区：[1]兰州交通大学交通运输学院,甘肃兰州730070 [2]浙江省水利水电勘测设计院,浙江杭州31002

出　　处：《西部交通科技》2015年第3期79-84,共6页Western China Communications Science & Technology

摘　　要：动态径路诱导是现代智能交通系统研究的一个重要领域。文章使用Multi-Agent技术思想,运用Java语言,结合动态路径诱导的实际需求开发动态路径诱导系统。基于各Agent间的实时交互信息,包括路段Agent提供的路况信息、路口Agent提供的交通信号信息、车辆Agent提供的车辆行驶及诱导需求信息,运用改造后的A*智能优化算法,实时计算诱导车辆当前位置下的最优路径。同时引入区域控制Agent与中心决策Agent对全局进行控制,避免局部最优及交通拥堵漂移现象的发生,从而达到整个路网的优化。完整地仿真了一个功能齐全的动态路径诱导系统,并且验证了该动态路径诱导模型的可行性及有效性。Dynamic touting induction is an important field of modern intelligent transportation systems research.This article developed the dynamic route guidance system by using the Multi-Agent technology concept and the Java language in combination with the actual needs of dynamic route guidance.Based on the real-time interaction information between various Agents,including the road traffic information provided by road Agent,traffic signal information provided by the intersection Agent,and the vehicles traveling and induction demand information provided by vehicle Agent,this article used the A*intelli-gent optimization algorithms after transformation for real-time computing the optimal path of induced vehicles under the current location.While the regional control Agent and central decision Agent were in-troduced for the overall control to avoid the occurrence of local optima and traffic congestion drift phenomenon,so as to achieve the optimization of entire road network.It completely simulated a fully func-tional dynamic route guidance system,and verified the feasibility and effectiveness of this dynamic route guidance model.

关 键 词：MULTI-AGENT 动态 路径诱导 A*算法 系统仿真 

分 类 号：U495[交通运输工程—交通运输规划与管理]