基于广义径向基神经网络的城市道路运行状况评价  

Evaluation of city road running condition based on GRBF neural network

作　　者：王来军[1] 许晓楠[1,2] 张昕越 郭捷[3] 郭宏玉[1] 

机构地区：[1]长安大学汽车运输安全保障技术交通行业重点实验室,陕西西安710064 [2]长城汽车股份有限公司,河北保定071000 [3]陕西国防工业职业技术学院,陕西西安710300

出　　处：《长安大学学报（自然科学版）》2015年第4期103-111,共9页Journal of Chang’an University(Natural Science Edition)

基　　金：国家自然科学基金项目(60134010);陕西省自然科学基金项目(2012JM8036);陕西省'百人计划'项目;中央高校基本科研业务费专项资金项目(CHD2010JC032;CHD2011TD015;CHD2012JC015);陕西省协同创新计划项目(2015XT-29)

摘　　要：为缓解城市交通拥堵,提出了一种城市道路运行状况的评价方法并给出了具体的实现技术。以交通拥堵分析为途径来衡量城市道路运行状况,建立了道路拥堵度的评价指标体系,然后通过广义径向基函数（generalized radial basis function,GRBF）神经网络的优化算法建立城市道路交通拥堵度的计算模型。针对西安市典型局部路网,基于MATLAB 2010平台,计算该路网中各个样本路段的拥堵度,获得具体拥堵点的位置和发生时刻,并使用加权平均法对整个路网的交通运行状况进行评价,最后得到该路网的最优出行路径。研究结果表明：同时间段路网中南二环路东段最为拥堵,交通拥堵指数高达0.974 4;而对于样本路段翠华路,则晚高峰16：00~18：00最为拥堵,交通拥堵度最高达0.911 8,这些均与实际情况完全吻合。该方法能够准确得到量化的城市道路交通拥堵数值,从而完成对一定区域内的交通拥堵评价。To alleviate the urban traffic congestion, this paper proposed an evaluation method of city road operation and gave the specific implementation technology. The city road running condition was measured by means of the traffic congestion analysis, and an evaluation index system of road congestion degree was presented. Then the model of computing road traffic congestion degree was set up through the optimization algorithm of generalized radial basis function neural network. For a typical local road network of Xi＇an, the traffic congestion degrees of each sample road in the road network was computed based on MATLAB 2010, and the specific location and occurrence time of congestion were obtained. Then the traffic condition of the total road network was evaluated by using the weighted average method and one optimal travel path of the road network was gotten. The results show that the east of South Second Ring road is the most congested in the road network at the same time interval and the traffic congestion index arrives 0. 974 4; for the sample section on Cuihua road, it is congested at the evening peak 16.00~ 18：00 and the traffic congestion degree arrives 0. 911 8, which are consistent with the actual situation. Quantitative value of urban traffic congestion can be acquired by the method presented in this paper, and the evaluation of traffic congestion in certain area can be completed in this way. 10 tabs, 5 figs, 19 refs.

关 键 词：交通工程 交通拥堵 道路运行状况 广义径向基神经网络 系统评价 

分 类 号：U411[交通运输工程—道路与铁道工程] U491