Output-Feedback Safe Tracking Control for Nonlinear Systems With Sensor Faults via Adaptive Critic Learning

Abstract

An output-feedback safe tracking control (OFSTC) scheme is investigated for nonlinear systems with sensor faults based on adaptive critic learning algorithm. By introducing a first-order filter, a mapping relationship between sensor faults and actuator faults is established, and an augmented system is constructed by integrating system state and filter output. Through the incorporation of robust adaptive terms, an output-based fault observer is developed to online identify sensor fault information, ensuring that observation errors converge asymptotically to zero. For optimal STC realization, an augmented tracking system is constructed by integrating the dynamics of tracking error, reference trajectory, and filter output. A modified cost function is designed to explicitly include sensor fault estimation and a discount factor based on the augmented tracking system. Then, the optimal STC strategy is derived by solving the Hamilton–Jacobi–Bellman equation using an adaptive critic structure with two tuned laws cooperatively. The application of the Lyapunov stability theorem demonstrates that the closed-loop system converges within a small neighborhood of the equilibrium point. Simulation results indicate the effectiveness of the proposed OFSTC method.