计算机工程与应用 ›› 2009, Vol. 45 ›› Issue (28): 233-235.DOI: 10.3778/j.issn.1002-8331.2009.28.070

• 工程与应用 • 上一篇    下一篇

车辆跟驰投影寻踪回归模型

邢 丽,吴 芳,王晓原   

  1. 山东理工大学 交通与车辆工程学院 智能交通研究所,山东 淄博 255049
  • 收稿日期:2008-05-28 修回日期:2008-09-04 出版日期:2009-10-01 发布日期:2009-10-01
  • 通讯作者: 邢 丽

Projection pursuit regression car-following model

XING Li,WU Fang,WANG Xiao-yuan   

  1. Institute of Intelligent Transportation,School of Transportation and Vehicle Engineering,Shandong University of Technology,Zibo,Shandong 255049,China
  • Received:2008-05-28 Revised:2008-09-04 Online:2009-10-01 Published:2009-10-01
  • Contact: XING Li

摘要: 车辆跟驰模型是微观交通仿真的一个基本模型,基于非参数回归算法的跟车模型较好地解决了以往模型存在的典型问题,但随着样本维数增加,容易出现“维数祸根”现象。提出一种基于投影寻踪回归(PPR)技术的车辆跟驰模型,解决了“维数祸根”和高维数据间的非正态、非线性问题。PPR建模不需要对数据结构作任何假定,而只通过直接审视和分析数据进行建模,因此,该方法能充分地发掘数据中存在的信息,建立的模型符合客观实际,精度较高。经过实测数据验证,该算法用于车辆跟驰模型的研究是可行的。

关键词: 投影寻踪回归, 车辆跟驰, 交通仿真

Abstract: The car-following model is a basic model of microcosmic traffic simulation.Although the representative problems consisted in former models are solved preferably by the car-following model based on nonparametric regression arithmetic,the phenomenon of “dimension curse” has appeared following the increase of the sample dimension.A car-following model based on the projection pursuit regression which can solve “dimension curse” and non-normality among high-dimensions data is established.Supposition is not to be made and PPR model is established through scanning and analyzing the data structures directly.So,the information consisted in the data can be extracted adequately and the model which accords with actual situation has high precision.The algorithm is feasible through the field data test.

Key words: projection pursuit regression, car-following, traffic simulation

中图分类号: