A personalized curve driving model for intelligent vehicle

Although the intelligent vehicle can safely drive through the curve with traditional control methods, most of them are to limit the maximum speed according to the curvature of road, which cannot meet the comfortable requirement for passenger. In this paper, by analyzing the driver's curve driving data and considering the road construction, we divide the curve driving phase into three section: curve entry driving, mid curve driving and curve exit driving. In further, we present a novel Personalized Curve Driving model for Intelligent Vehicle, which includes an acceleration/deceleration model based on the GMM/GMR and a speed tracking algorithm. The model is trained with the human driving data and curve road parameters in Prescan simulation environment. The simulation results illustrate the model for Intelligent Vehicle can smoothly and stably pass through the curve like a human driver.