基于径向基函数神经网络的城市道路路段行程时间实时预测模型  

A Real-time Travel Time Prediction Model Based on RBF Neural Network

作　　者：刘江用[1] 云美萍[1] 闫亚文[1] 杨晓光[1] 

机构地区：[1]同济大学交通运输工程学院,上海201804

出　　处：《交通信息与安全》2011年第5期31-35,共5页Journal of Transport Information and Safety

基　　金：国家自然科学基金项目(批准号:70631002)资助

摘　　要：提出利用径向基函数(RBF)神经网络方法对城市道路路段行程时间进行建模预测,并结合线圈和视频实测数据进行仿真分析,以实际行程时间和模型输出的行程时间预测值比较验证了模型的合理性。并将RBF神经网络方法与BP神经网络方法进行比较,结果表明RBF神经网络相对于BP神经网络训练时间短,且预测精度更高。Travel time is a comprehensive index of traffic condition and the base of route guidance and traffic management. The prediction for travel time has been the key issue of ITS research. This paper provides a method to predict urban road travel time with Radial Basis Function Neural Network. Traffic data collected by loop and camera are used in simulation. The experiment results show the rationality of model. After comparison with BP neural network, the results prove that the RBF Neural network can predict the travel time in real time well and the adaptability and accuracy of RBF neural network are better than those of BP neural network.

关 键 词：行程时间 神经网络 径向基函数 

分 类 号：U491[交通运输工程—交通运输规划与管理]