计算机应用 ›› 2018, Vol. 38 ›› Issue (7): 1916-1922.DOI: 10.11772/j.issn.1001-9081.2018010090

• 数据科学与技术 • 上一篇    下一篇

基于车辆行驶数据的驾驶人行为谱分析方法

陈镜任1,2, 吴业福1,2, 吴冰1,2   

  1. 1. 武汉理工大学 计算机科学与技术学院, 武汉 430063;
    2. 交通物联网湖北省重点实验室(武汉理工大学), 武汉 430063
  • 收稿日期:2018-01-11 修回日期:2018-03-28 出版日期:2018-07-10 发布日期:2018-07-12
  • 通讯作者: 吴业福
  • 作者简介:陈镜任(1996-),女,江西南昌人,硕士研究生,主要研究方向:机器学习、大数据分析、交通安全;吴业福(1964-),男,湖北红安人,副教授,博士,主要研究方向:交通信息与安全、人工智能;吴冰(1992-),男,湖北武汉人,硕士,主要研究方向:大数据分析、交通安全。
  • 基金资助:
    国家科技支撑计划项目(2014BAG01B03-2)。

Driver behavior spectrum analysis method based on vehicle driving data

CHEN Jingren1,2, WU Yefu1,2, WU Bing1,2   

  1. 1. School of Computer Science and Technology, Wuhan University of Technology, Wuhan Hubei 430063, China;
    2. Hubei Key Laboratory of Transportation Internet of Things(Wuhan University of Technology), Wuhan Hubei 430063, China
  • Received:2018-01-11 Revised:2018-03-28 Online:2018-07-10 Published:2018-07-12
  • Supported by:
    This work is partially supported by the National Key Technology Research and Development Program (2014BAG01B03-2).

摘要: 针对我国驾驶人行为谱的研究尚不完善,专业领域内没有相应的行为谱分析工具的问题,提出了一套针对营运客车的完整的驾驶人驾驶行为谱体系并设计了一套分析工具。首先,设计并定义了驾驶人行为谱的特征指标和评价指标;其次,给出了驾驶人行为谱的特征指标分析、计算方法,采用基于马尔可夫链蒙特卡洛采样和离群点剔除的K-means算法对驾驶人的驾驶风格进行分析,采用回归学习对驾驶人的驾驶技能进行分析;然后,设计了基于车联网、大数据的驾驶人行为谱的基础数据采集和预处理方法;最后,采用Java语言、Spring MVC架构开发出驾驶人行为谱分析工具。将机器学习中的数据挖掘、数据分析算法与交通安全领域相结合,对完善我国驾驶人行为谱框架体系具有理论意义,为我国驾驶人行为谱的研究提供了一个科学、定量化分析的工具,对交管部门规范驾驶人驾驶行为、提高道路安全指数、制定合理的交通安全管理策略具有指导意义。

关键词: 驾驶人行为谱, 驾驶人行为谱分析工具, 车辆行驶数据, 车联网, 交通安全, 大数据

Abstract: Focusing on the issue that our country's driver behavior spectrum research is still not perfect, and there is no corresponding behavioral spectrum analysis tool in the professional field, a set of complete driver behavior spectrum system for commercial motor vehicle of passenger transport was proposed and an analyzing tool was designed. Firstly, the characteristic indexes and the evaluation indexes of driver behavior spectrum were designed and defined. Secondly, the characteristic indexes analysis method and algorithm of driver behavior spectrum were given, the improved K-means algorithm based on Markov chain Monte Carlo sampling and outlier removing was used to analyze driving styles of drivers, and regression learning was used to analyze driving skills of drivers. Then, the basic data acquisition scheme and preprocessing methods of driver behavior spectrum based on car networking and big data were designed and proposed. Finally, Java language and the Spring MVC (Model View Controller) architecture were used to develop the profiling tool of driver behavior spectrum. Data mining and data analysis methods in machine learning were combined with traffic safety, which has theoretical significance for perfecting the driver behavior spectrum framework. It provides a scientific and quantitative analysis tool for our country's driver behavior spectrum analysis work. It also provides guiding significance for traffic management department to standardize the driving behaviors of drivers, improves the road safety index and makes reasonable traffic safety management strategies.

Key words: driver behavior spectrum, driver behavior spectrum analysis tool, vehicle driving data, car networking, traffic safety, big data

中图分类号: