J4 ›› 2012, Vol. 25 ›› Issue (3): 23-28.doi: 10.3976/j.issn.1002-4026.2012.03.005

• 目录 • 上一篇    下一篇

基于RBF神经网络的城市快速路短时交通流预测研究

郑宣传1,韩宝明1,李得伟2   

  1. 1.北京交通大学交通运输学院,北京 100044;
    2.北京交通大学,轨道交通控制与安全国家重点实验室,北京 100044
  • 收稿日期:2012-03-27 出版日期:2012-06-20 发布日期:2012-06-20
  • 作者简介:作者简介:郑宣传(1989-),男,硕士研究生,研究方向为城市道路交通、城市轨道交通。Email:11121033@bjtu.edu.cn
  • 基金资助:

    国家科技支撑计划项目(2011BAG01B01);国家自然科学基金(61004105)

RBF neural network based urban expressway short-term traffic flow prediction

 ZHENG Xuan-Chuan1, HAN Bao-Ming1, LI De-Wei2   

  1. 1.School of Traffic and Transportation, Beijing Jiaotong University, Beijing 100044, China;
    2.State Key Laboratory of Rail Traffic Control and Safety, Beijing Jiaotong University, Beijing 100044, China
  • Received:2012-03-27 Online:2012-06-20 Published:2012-06-20

摘要:

      为检验灰色模型及径向神经网络模型用于短时交通流预测的可行性及适用性,本文分析和比较了灰色模型GM(1,1)和RBF径向神经网络模型对短时交通流的预测效果。仿真实例表明,灰色模型不适合用于短时交通流预测,而径向神经网络能够准确预测短期交通流的未来变化趋势,当径向基函数的分布密度值在0.8~1.0之间时能够取得较高的预测精度。

关键词: 交通流, 预测, 灰色模型, RBF神经网络

Abstract:

       We analyzed and compared the effects of gray model GM (1,1) and RBF neural network model on short-term traffic flow prediction to test their feasibility and applicability. Practical instances show that gray model is inapplicable to the short-term prediction of traffic flow, but RBF neural network model is applicable. Moreover, we can acquire higher prediction accuracy when the distribution density of the radial basis function is from 0.8 to 1.0.

Key words: traffic flow, prediction, gray model, RBF neural network

中图分类号: 

  • U491.1+12