Optimized Backstepping Fault‐Tolerant Control for a Class of Nonlinear Strict‐Feedback Systems Subjected to Sensor Faults

ABSTRACT This paper proposes a fault‐tolerant optimized backstepping (OB) control scheme for a class of nonlinear strict‐feedback systems with sensor faults. For optimizing entire backstepping control, OB strategy needs to implement the reinforcement learning (RL) algorithm at every backstepping step. However, because of the presence of sensor bias faults, the RL training can only receive the damaged state information. To solve this problem, an adaptive algorithm of estimating the unknown bias parameters is integrated into RL. To make the integration smoothly, the RL is derived from a simple positive function that is equivalent to the Hamilton‐Jacobi‐Bellman (HJB) equation, and thus it can significantly mitigate the complexity of RL training algorithm. Finally, the effectiveness of the proposed control method is verified through Lyapunov stability analysis and simulation experiment.