基于遗传回归分析的无检测器交叉口流量预测  

Prediction of Traffic Volume at Non-detector Intersections Based on Genetic Algorithm and Stepwise Regression Analysis

作　　者：王媛[1] 杨兆升[1] 王志建[1] 王薇[1] 

机构地区：[1]吉林大学交通学院,长春130022

出　　处：《北京工业大学学报》2008年第10期1077-1083,共7页Journal of Beijing University of Technology

基　　金：国家'八六三'计划资助项目(2007AA12Z242;2007AA11Z218;2007AA11Z245).

摘　　要：对无检测器交叉口交通量预测方法进行了研究.介绍分析了逐步回归分析法和遗传算法,将2种方法结合起来提出一种基于遗传算法的无检测器交叉口流量预测方法,并用长春交叉口流量数据进行了验证,在与逐步回归分析法比较的基础上,证明了该方法的可行性且具有较高的预测精度.Volume predicting methods for intersections without detectors were studied. Stepwise regression analysis and genetic algorithm are introduced and combined to propose a voume predicting method for intersections without detectors based on genetic algorithm, which is validated by the data from the Changchun intersections. Compared with the predicted results of stepwise regression analysis, the feasibility and higher precision of this method are certificated.

关 键 词：交通运输工程 逐步回归分析 遗传算法 流量预测 

分 类 号：U491[交通运输工程—交通运输规划与管理]