交通运输系统工程与信息 ›› 2011, Vol. 11 ›› Issue (1): 157-162 .

• 系统工程理论与方法 • 上一篇    下一篇

基于混合AGO-SVM的高速公路短时交通量预测研究

张通;张骏*;杨霄   

  1. 西北工业大学 自动化学院, 西安 710072
  • 收稿日期:2010-11-02 修回日期:2010-12-09 出版日期:2011-02-25 发布日期:2012-12-20
  • 通讯作者: 张骏
  • 作者简介:张通(1984- ),男,陕西西安人,博士生.
  • 基金资助:

    西北工业大学创新基金(W016144)

Short-Term Highway Traffic Flow Prediction Based on Mixed AGO-SVM

ZHANG Tong;ZHANG Jun; YANG Xiao   

  1. College of Automation, Northwestern Polytechnic University, Xi’an 710072, China
  • Received:2010-11-02 Revised:2010-12-09 Online:2011-02-25 Published:2012-12-20
  • Contact: ZHANG Jun

摘要: 提出一种混合AGO-SVM高速公路交通量预测方法,原始交通量数据通过累加操作生成有规则的数据,预处理后的规则数据使用支持向量机法进行建模并预测,预测数据进行逆累加操作,获得下一时刻高速公路交通量的预测值,数据进行更新并保持样本序列不变从而进行高速公路交通量递推预测. 应用西宝高速交通量实际观测数据验证算法的有效性. 试验结果表明,在几种指标下该方法的预测精度比灰色模型法和支持向量机法的预测结果有所提高,是一种有效的高速公路交通流量预测方法.

关键词: 智能交通, AGO-SVM, 混合, 交通量预测, 高速公路

Abstract: Our goal is to present a mixed AGO-SVM highway traffic flow prediction method. Pretreatment traffic flow data following some rule is generated by accumulated generating operation and is modeled and predicted by support vector machine algorithm. Traffic flow data of the next moment in the highway is retrieved by inverse accumulated generating operation. Predicted data is kept updating, the sample sequence is maintained and the highway traffic flow is predicted through recursive prediction. The effectiveness of the algorithm is verified by Xi-Bao highway traffic flow data. The results show that the proposed method is more effective than gray model algorithm and support vector machine algorithm in prediction accuracy.

Key words: intelligent transportation, AGO-SVM, mix, traffic flow prediction, highway

中图分类号: