计算机工程与应用 ›› 2017, Vol. 53 ›› Issue (5): 266-270.DOI: 10.3778/j.issn.1002-8331.1506-0107

• 工程与应用 • 上一篇    

基于AF-SVR的城市快速路多源交通信息融合研究

丁宏飞1,2,秦  政1,刘  博1,李演洪1   

  1. 1.西南交通大学 交通运输与物流学院,成都 610031
    2.四川省交通运输厅 公路规划勘察设计研究院,成都 610041
  • 出版日期:2017-03-01 发布日期:2017-03-03

Study on urban expressways’ multi-source traffic data fusion technology based on AF-SVR model

DING Hongfei1,2, QIN Zheng1, LIU Bo1, LI Yanhong1   

  1. 1.School of Transportation and Logistics, Southwest Jiaotong University, Chengdu 610031, China
    2.Highway Planning, Survey, Design and Research Institute, Sichuan Provincial Transport Department, Chengdu 610041, China
  • Online:2017-03-01 Published:2017-03-03

摘要: 针对单一检测器所得到的交通数据不能够全面准确地反映实际的交通状态,提出一种基于AF-SVR模型的城市快速路多源交通信息融合的方法。首先通过将相同路段中不同检测器的速度数据作为学习样本输入到支持向量机回归模型(Support Vector Regression,SVR)中进行训练。然后利用鱼群算法(Artificial Fish,AF)对支持向量机回归模型中的参数进行优化,获得最优的信息融合模型,用于多源交通信息的融合,输出为能准确反映真实交通状态的速度数据,并用人工采集的速度数据作为真值进行验证。最后将此方法应用于成都市三环快速路路段上的多源交通信息融合,取得了令人满意的结果。

关键词: 多源交通信息, 信息融合, 支持向量机回归, 城市快速路, 鱼群算法

Abstract: Various detectors’ data can not fully reflect actual traffic conditions, each one of them contains limited traffic information, therefore a multi-source traffic data fusion technology of urban express way based on AF-SVR model is proposed in this paper. Firstly, th etraffic data collected by different detectors is used to train the Support Vector Regression(SVR) model. Secondly, the parameters of Support Vector Regression model are calibrated by Artificial Fish(AF) to build optimal model formulti-source traffic data fusion, the output can reflect actual traffic conditions, which is verified by field velocity data. Finally, this multi-source traffic data fusion technology is applied on 3rd Ring Road Expressway in Chengdu, the result of which turns out to be satisfactory.

Key words: multi-source traffic data, data fusion, Support Vector Regression(SVR), urban expressway, Artificial Fish(AF)