In order to carry out classified management for different truck drivers, the differences in driving behavior of truck drivers are studied. In this paper, the vehicle driving data of 39 truck drivers in the natural driving state is obtained through the vehicle OBD equipment, and the characteristics of engine speed, vehicle lateral acceleration, driving speed and vehicle longitudinal acceleration in the vehicle travel information are extracted. In the data processing process, in order to eliminate the influence of the external environment such as road bumps and engine shake during the running of the vehicle on the data, the mean filtering method is used to smooth the acquired data. Adopt Gaussian mixture model establishes a driver's personal trait identification model, divide truck drivers into ordinary drivers, aggressive drivers and conservative drivers. One-way ANOVA was performed on the classification results, and there were significant differences between the variables and the difference was statistically significant. The classification results show that the driver's personal trait identification method based on Gaussian mixture model can effectively identify the driver type. Keywords: traffic safety, driving behavior, Gaussian mixture model, OBD data, mean filtering

CICTP