Event-triggered adaptive fuzzy control for automated vehicle steer-by-wire system with prescribed performance: Theoretical design and experiment implementation

Steer-by-wire (SbW) systems substitute the typical mechanical linkage between the steering wheel and the front wheel with electromechanical actuators, resulting in that the steering performance of SbW systems depends on the control of electromechanical actuators. To guarantee the transient and steady-state performance of SbW systems, this paper proposes an event-triggered adaptive fuzzy control for the SbW system subject to the unavailable steering angular velocity, the time-varying disturbance, and the limited communication resources. First, to solve the problem of uncertainty modeling and external interference, an interval type-2 fuzzy logic system (IT2 FLS) and a disturbance observer are used to approximate the unknown SbW system dynamics and disturbance, respectively. Second, considering the angular velocity of the front wheels is difficult to measure, an adaptive observer is developed to estimate the steering angular velocity. The observation error is proved by the Lyapunov theory to converge in finite time. Third, to guarantee the transient and steady-state performance of the SbW system, this paper develops an output feedback controller for the automated vehicles based on event-triggered control (ETC) and prescribed performance control (PPC). Theoretical analysis proves that the tracking error can converge to a preset range within the set time, and the communication data between the controller and the actuator is reduced. Finally, simulation and experiment verify the effectiveness and superiority of the proposed method. From the analysis of the estimation performance, the designed observer has strong adaptability, which can ensure better observation accuracy even if the system has uncertainty. From the analysis of control performance, the proposed controller can guarantee the output trajectory tracking performance attributes (maximum overshoot, convergence time, maximum steady-state error). More specifically, in the experiment, the tracking error can converge to within 0.075 rad in 5 s. Additionally, the RMSE of the experimental results of the designed method is 44.7% higher than that of the output-feedback adaptive neural controller. From the perspective of communication data, the control signal transmitted to the motor driver is effectively saved by 19.3 %.