Control design of driver support system using multiple driver models

This paper proposes a control system design method concerned with a driver support system which assists driver's operation in an obstacle avoidance scene. In the procedure of the design method, multiple driver models regarding actual drivers are constructed by inverse models for vehicle dynamics. A controller of the support system is designed by model predictive control with constraint condition, in which a driver model for controller design is decided by switching the multiple models using vote form depending on an error signal between the actual driver and the driver model. By numerical experiments which simulate for avoiding the obstacle, it is shown that the proposed method has effect for a performance of avoidance task and switching to the driver model.