A graphical modeling method for individual driving behavior and its application in driving safety analysis using GPS data. (May 2019)

• Record Type: Journal Article
• Authors: Chen, Chen; Zhao, Xiaohua; Zhang, Yunlong; Rong, Jian; Liu, Xiaoming
• Is Part Of: Transportation research. Volume 63(2019)
• Page Start: 118
• Page End: 134
• Publication Date: 2019-05
• Subjects: Driving safety -- Individual behavior -- Graphical modeling -- GPS data; Automobile drivers -- Psychology -- Periodicals; Automobile driving -- Psychological aspects -- Periodicals; Transportation -- Psychological aspects -- Periodicals; 629.283019
• DOI: 10.1016/j.trf.2019.03.017
• Languages: English
• ISSNs: 1369-8478

Abstract

Highlights: A graphical method for intuitively illuminate fundamental driving habits of drivers. Describe the basic and internal driving patterns of drivers based on big data. A good application in the analysis of driving safety. Abstract: Due to differences in driving skills and personal characteristics among drivers, the behaviors of drivers when faced with various driving environments differ, causing different levels of driving safety concerns. In past research, the measurement of safety-related driving behavior mostly focused on classification, while few studies were concerned with individual driving behavior characteristics. However, it is important for drivers to recognize and correct their dangerous behaviors and optimize their driving. This paper presents a graphical method for modeling individual driving behaviors, and the results can be used in driving safety analysis. Based on the assumption that drivers have specific driving habits, typical driving patterns during driving are first detected and extracted. These typical driving patterns are then sorted according to their frequencies, forming a driving behavior graph that can directly illustrate each driver's behavior features. Furthermore, a quantitative analysis method for evaluating driving safety based on the behavior graph is provided. To verify the proposed method, a case study focusing on vehicles' longitudinal motion was conducted using GPS data collected from Beijing taxis. The results demonstrated that the graphical method can describe the individual features of a driver's longitudinal acceleration behavior and distinguish differences among drivers. The development of this method can help understand the individual features of driving behaviors and further support measures to optimize driving safety.