交通运输系统工程与信息 ›› 2018, Vol. 18 ›› Issue (4): 240-246.

• 案例分析 • 上一篇    

基于多维数据的高速公路服务水平实时判别方法

赵文忠 1,耿立艳 * 2,梁毅刚 2,张占福 3   

  1. 1. 河北曲港高速公路开发有限公司,河北 保定 071000;2. 石家庄铁道大学 经济管理学院,石家庄 050043; 3. 石家庄铁道大学 四方学院,石家庄 051132
  • 收稿日期:2018-04-18 修回日期:2018-05-27 出版日期:2018-08-25 发布日期:2018-08-27
  • 作者简介:赵文忠(1971-),男,河北鹿泉人,高级工程师.
  • 基金资助:

    国家自然科学基金(青年)/ National Natural Science Foundation for Young Scholars of China(61503261);河北省高等学校青年拔尖人才计划项目/Young Talents Program of Higher School of Hebei Province (BJ2014097).

A Real-time Identification Method of Highway Service Level Based on Multi-dimension Data

ZHAO Wen-zhong1, GENG Li-yan2, LIANG Yi-gang2, ZHANG Zhan-fu3   

  1. 1. Hebei Qu Gang Expressway Development Co., Ltd, Baoding 071000, Hebei, China; 2. School of Economics and Management, Shijiazhuang Tiedao University, Shijiazhuang 050043, China; 3. Sifang College, Shijiazhuang Tiedao University, Shijiazhuang 051132, China
  • Received:2018-04-18 Revised:2018-05-27 Online:2018-08-25 Published:2018-08-27

摘要:

随着高速公路的不断发展,交通信息逐渐呈现多元化,而将这些信息进行整合并对高速公路服务水平进行实时判别,对于后期的交通控制管理及实时高效提供道路信息至关重要.本文考虑了时间、天气、出行日期等多因素的影响,选用车速、密度作为高速公路服务水平评价指标,结合最小二乘支持向量机与聚类分析相关算法,提出了一种基于多维数据的高速公路服务水平实时判别模型,以实现对高速公路服务水平的参数预测及实时判别.选用河北省237 360组训练样本数据和6 048组预测样本数据验证了该模型的有效性.结果显示,该模型获得了较高的判别精度和较好的预测效果,是一种有效的高速公路服务水平实时判别方法.

关键词: 公路运输, 高速公路, 多维数据, 服务水平, 实时判别

Abstract:

With the continuous development of highway, traffic information of highway is diversified gradually. Integration of the traffic of highway and real-time identification of the service level of highway is very important for the later traffic controlling and management and provision of real-time and efficient road information. Combing least squares support vector machine with cluster analysis algorithm, this paper proposes a real-time identification model of highway service level based on multi-dimension data in order to achieve the parameter prediction and real-time identification of highway service level. The model includes many influence factors, such as time, weather, travel dates and so on. In addition, it selects vehicle speed and density to evaluate highway service level. The validity of the model is tested by using 237 360 training samples and 6 048 prediction samples of Hebei province. The results show that the model produces the higher identification accuracy and better prediction effect, which is a method for real-time identification of highway service level.

Key words: highway transportation, highway, multi-dimension data, service level, real-time identification

中图分类号: