Driving Behavior Clustering for Hazardous Material Transportation Based on Genetic Fuzzy C-Means Algorithm

Abstract

In recent years, many accidents of hazardous material transportation have been caused by dangerous driving behavior such as excessive speed or rapid speed change. To solve the problem, in this paper a series of indicators for driving behavior evaluation are selected for quantitative analysis based on the massive data of the operating vehicle networked control system. After that, three main indicators are introduced in detail, that is, the acceleration deceleration behavior indicator, the over speed behavior indicator and the operation stability indicator. Then by using the GA-FCM clustering method, 40 drivers of hazardous material transportation are classified according to their behavior parameters. The results clearly show that the driving speed of the drivers with high-risk driving behavior, and most of the drivers with poor driving stability, are also at bad over speed level. Therefore, this paper creatively applies the method of combining factor analysis and GA-FCM clustering to the field of driving behavior of hazardous material transportation. It will facilitate enterprises and management departments to focus on controlling high-risk drivers, thereby reducing accidents.