Fuzzy-Observer-Based Fault-Tolerant Steering Control for Autonomous Driving With Persistent Disturbances via Model Predictive Approach

The issue of fuzzy-observer-based (FOB) fault-tolerant (FT) steering control for autonomous driving with persistent disturbances via model predictive control approach is studied in this article, where velocity variation, potential faults, nonlinearity, and unmeasurable states are considered simultaneously. Due to the variable longitudinal velocity during cruising on diverse road conditions and various vehicle maneuvers, the Takagi–Sugeno fuzzy method is employed to handle parameter variations in the steering control system. For the sake of obtaining information on fault behavior and unmeasurable system states, this article constructs a fuzzy observer to estimate the fault signal and faulty system states. Improved conditions for designing the observer gains are proposed, such that the estimation error converges to a minimal robust positively invariant set. Subsequently, this article introduces a FOB FT model predictive controller, which is designed through the resolution of a Min-Max optimization issue. In the end, the benefits of the approach developed in this article are validated by utilizing the Carsim/Matlab joint simulation.