摘要
动态路径选择是城市交通流诱导系统的核心理论之一。为了兼顾个体出行者和路网系统管理者在路径选择过程中的利益,从分布式人工智能的角度出发,给出了一种基于多智能体协商的动态路径选择方法,将路网中的驾驶员、信息发布单元以及系统管理者分别看作不同的智能体进行建模,并给出智能体之间的路径选择协商模型。借助多智能体仿真软件Starlogo,对无信息无协商出行、有信息无协商出行和有信息有协商出行等三种不同的仿真方案进行模拟比较,仿真结果验证了协商方法在满足驾驶员出行需求以及提高路网整体效率方面的有效性和优越性。
Dynamic route choice is one of the key theories in the Urban Traffic Flow Guidance Systems （UTFGS）.With the purpose of giving consideration to the benefits of individual traveler and network system manager,a new dynamic route choice method,from the point of Distributed Artificial Intelligence （DAI）,was presented based on Multi-agent negotiation,regarding drivers in the network,the information center and the traffic system manager as different agents to modeling respectively,and then the route choice negotiation model among agents was given.By using Starlogo,the famous Multi-agent simulation software,three different simulation scenarios：none information without negotiation travel,information without negotiation travel,and information and negotiation travel,were simulated and compared.The results indicate the validity and superiority of the negotiation method in satisfying drivers＇ travel demands and improving the whole network performance.
出处
《系统仿真学报》
CAS
CSCD
北大核心
2010年第8期1890-1894,共5页
Journal of System Simulation
基金
国家自然科学基金(70673016)
哈尔滨工业大学优秀青年教师培养计划资助(HIT
2006
19)
关键词
交通流诱导
路径选择
多智能体
协商
traffic flow guidance
route choice
multi-agent
negotiation
