Journal of Transportation Systems Engineering and Information Technology ›› 2003, Vol. 3 ›› Issue (4): 16-26 .

• ITS System and Technology • Previous Articles     Next Articles

ITS Data Archiving Strategy Based-on Sampling Approach

YU Lei1, QIAO Feng-xiang2, XU LI2, CHEN Xu-mei3
GENG Yan-bin3, WU Jia-qing3, YUAN Zhen-zhou3

  

  1. 1. Beijing Jiaotong University and Texas Southern University, Beijing 100044, china; 2. Texas Southern University, Houston, Texas 77004, U.S.A.; 3. Beijing Jiaotong University, Beijing 100044, China
  • Received:2003-09-16 Revised:1900-01-01 Online:2003-11-01 Published:2003-11-01

基于抽样技术的ITS数据存档策略

于雷1,乔凤翔2,徐力2 陈旭梅3,耿彦斌3,吴家庆3,袁振州3


  

  1. 1.北京交通大学,北京 100044;2.美国德克萨斯州南方大学;3.北京交通大学,北京 100044

Abstract: This paper presents an optimization-based sampling approach for data arching. This approach intends to identify the best representative samples of the raw ITS data based on either sum squire error(SSE) or cross validation(CV) while minimize the required storage size. The proposed approach is tested in the case study of TransGuide of San Antonio, Texas. After the proposed sampling approach is applied in the case study, only one tenth of the original data are needed to be stored, while the resulting optimal samples contain the maximum information of the raw data, which are able to meet the potential uses of various transportation purposes.

Key words: data archiving, sample, quality control, sum square error, cross validation

摘要: 介绍了ITS数据存储的最佳抽样方法,即对原始数据进行质量控制之后,运用误差平方和法(SSE)或互验法(CV)确定出原始ITS数据中最具有代表性的样本,从而降低所需的数据存储空间。通过开发相应的数据处理软件,针对美国德克萨斯州圣安东尼奥TransGuide交通管理中心的数据进行了测试,结果表明:对数据进行处理后,仅需存储十分之一的原始数据,所得到的最佳样本则包含了最多的原始数据信息,且该样本数据能够满足潜在的不同交通需求。

关键词: 数据存储, 抽样, 质量控制, 误差平方和法, 互验法