Torque Vector Compensation Strategy with Adaptive Model Predictive Control Considering Steering Actuator Fault

This paper addresses the problem that the heavy-duty vehicles are more likely to suffer from rollover and drift over issues considering the steering actuator's fault. In order to isolate and estimate the faulty steering actuator, an integrated fault-tolerant control architecture is thus designed to cope with two types of actuators failure based on fault detection and diagnosis (FDD) and fault-tolerant (FT) control. Firstly, a fault isolation method based on plane projection theorem and an H-infinity fault estimator are designed to acquire the faulty signal and estimate the real-time fault coefficient. Then, considering the estimation error and uncertainty disturbance, an adaptive model predictive controller (MPC) is proposed to guarantee the controller's redundancy when the fault occurs on the actuators, while ensuring the rollover prevention and lateral stability. Finally, the proposed adaptive MPC architecture shows the effectiveness and superiority based on the simulation results in Matlab/CarSim full-vehicle model virtual experiment.