基于SVM短时交通流量预测  

Short-term Traffic Flow Prediction Based on SVM

作　　者：蒋晓峰[1] 许伦辉[1] 朱悦[2] 

机构地区：[1]华南理工大学土木与交通学院,广东广州510461 [2]华南理工大学生物科学与工程学院,广东广州510006

出　　处：《广西师范大学学报（自然科学版）》2012年第4期13-17,共5页Journal of Guangxi Normal University:Natural Science Edition

基　　金：青年科学基金资助项目(51108191)

摘　　要：交通流量预测是智能交通系统中非常重要的研究领域,传统的预测方法在交通流量预测中有着非常广泛的应用。但是,在短时交通流量预测中,由于其影响因素错综复杂,传统的预测方法对于短时交通流量不能很好地进行预测。随着机器学习和数据挖掘各种理论的不断提出及完善,机器学习和数据挖掘与交通流量预测的结合是智能交通系统未来发展的一个重要方向。本文利用SVM(support vector machine)构建了短时交通流量预测模型,并利用遗传算法(genetic algorithm)对SVM的惩罚参数C和核参数σ进行优化,同时比较SVM中不同核函数,包括多项式核函数(polynomial kernel)和径向基核函数(RBF kernel)的预测效果。径向基SVM(RBF SVM)训练时间要比多项式SVM(polynomial SVM)短,预测准确率和精度也要比多项式SVM要好。从仿真结果上看,SVM非常适合应用于短时交通流量预测,能够取得很好的预测效果与精度。Traffic flow prediction is a very important area in intelligent transportation systems. Traditional prediction methods have a very wide range of applications in the traffic prediction. But traditional prediction methods does not work very well in short-term traffic flow prediction because of the complexity of the influencing factors. With the development of machine learning and data mining ,traffic flow prediction with a combination of machine learning and data mining has become more and more important as a research area. In this paper,SVM （Support Vector Machine） is used to build a short-term traffic flow prediction model,and Genetic Algorithm （GA） is used to optimize the SVM penalty factor C and kernel parameter a as well. The results of different kernel functions of SVM are compared,including polynomial kernel and RBF kernel. RBF SVM plays better than polynomial SVM with less training time and higher accuracy and SVM is very suitable for short-term traffic flow prediction.

关 键 词：SVM 交通流量 短时预测 遗传算法 

分 类 号：U491[交通运输工程—交通运输规划与管理]