基于多Agent的交通信息智能预测系统的研究  

Research on intelligent prediction system of traffic information based on multi-Agent

作　　者：杨翠[1] 朱晶晶[1] 李欣娜[1] 

机构地区：[1]兰州交通大学交通运输学院,甘肃兰州730070

出　　处：《电子设计工程》2011年第21期31-33,共3页Electronic Design Engineering

基　　金：兰州交通大学大学生科技创新基金资助项目(DXS2011-026);兰州交通大学大学生创新性实验计划项目(201064)

摘　　要：城市交通控制和诱导系统是智能交通系统(Intelligent Transport System,ITS)的组成部分,而交通流预测特别是短时交通流预测是城市交通控制与交通诱导系统的基础。由于缺乏交通量预测所需历史数据以及判断影响交通量生成与增长因素的准确性不够,致使道路交通量的预测结果与实际值产生了较大差异。论述了灰色模型(Gray Model,GM)以及多Agent理论在交通量智能预测中的方法和应用技术。该方法利用多Agent间的相互通信、协作功能以及灰色模型的累加生成手段和微分方程描述,在一定预测时段内保证了预测数据良好的准确性和实用性。The traffic flow prediction, especially the prediction for short-term traffic flow, is the basis of urban traffic controlling and guidance system which are the components of ITS. Because the history data of traffic volume and the accuracy for justifying the factors that affect the generation and growth of tragic volume are scarce, the difference of road volume result between the forecasting and the true of is great. Consequently, this paper demonstrates the method and application of gray model and multi-Agent theory in traffic volume intelligence prediction. For a certain period, using the intercommunication and cooperation between the agents, and the summation generation method and partial equation description of gray model, it ensures the data＇s good accuracy and practicability.

关 键 词：智能交通 交通信息 灰色预测 多AGENT 协作 

分 类 号：U491[交通运输工程—交通运输规划与管理]