Attack-Resilient Lateral Stability Control for Four-Wheel-Driven EVs Considering Changed Driver Behavior Under Cyber Threats

Abstract

This paper presents an attack-resilient lateral stability control design approach for four-wheel-driven electric vehicles (EVs) to enhance the cyber-physical security of the steering system. Specifically, this work examines cyber-attacks on the steering angle, which is considered one of the most safety-critical signals in vehicles. Firstly, a robust predictive controller is designed to mitigate the impact of cyber-attacks, which carries a low computational burden, making it applicable for real-time applications. Secondly, by developing a collaborative mechanism between the attack-resilient controller and the human driver’s action, the proposed method considers the altered driver behavior during cyber threats, e.g., enlarged driver’s neural response delay and muscle action delay, which to our knowledge has not been examined before. Thirdly, the lateral dynamics model and driver model are verified by experimental results under a production vehicle, based on which the effectiveness and feasibility of the proposed attack-resilient control methodology are demonstrated.