Event-Triggered Steering Control for Semiautonomous Vehicles with a Stochastic Driver Model

Abstract

This paper proposes an event-triggered steering control design for the lane-keeping of semiautonomous ground vehicles. In the proposed design method, a stochastic driver model accounting for the driver's mental states is proposed to characterize the steering behavior of human drivers. The actual driver mental states are hidden to the steering controller of the vehicles. A hidden Markov model is applied in the vehicle's steering controller design. By Lyapunov synthesis, a sufficient condition on the closed-loop stochastic stability with a disturbance attenuation level is provided for the driver-in-the-loop steering error dynamics of the vehicles. Simultaneously, a state-based event-triggered rule is determined by virtue of Lyapunov stability analysis. Finally, some simulation results are provided to support the proposed design method.