A Lane Keeping Assist Design: Adaptation to Driving Style Based on Aggressiveness

Abstract : The influence of driver style identification on the performance of adaptive driver assistance systems has been well explored. Typically, designed controllers focus on safety, fuel efficient performance, reduced emissions etc. In this work, a rule based approach is formulated to classify driver style based on levels of aggressiveness. Employing the identified driver style a adaptive robust lane keeping controller is formulated. With focus on lane keeping, a robust controller is designed using higher-order sliding mode for a driver-in-the-loop dynamic model developed with driver-style adaptation. The closed loop stability of the proposed design is established. The proposed cooperative control approach is verified through simulations over the Satory test track for different driving conditions and the results are presented.