Research on prediction method of driver's lane changing behavior based on simulated driving

Abstract

The driver’s decision and behavior are important factors affecting traffic safety. This paper uses 22 participants to participate in a simulated driving experiment, integrated driver's driving behavior parameters and vehicle operating parameters for in-depth research as well as analysis, additionally statistical analysis on various indicators in lane change intention phase through independent sample T test, etc. The study has disclosed the effectively influence of the factors, which includes eye movement, head movement, and the vehicle related factors e.g.t vehicle speed, vehicle deviation angle, lateral acceleration, the distance to vehicle in front and the time to collision, etc. These factors change significantly between the different intention phases of lane keeping and lane change. Herein a steering prediction model is constructed based on logistic regression, which is verified to be highly effective. The accuracy of the prediction model in the experiment reached 94.4%, which means the effective complement of the prediction of lane change behavior.