Adaptive dynamic programming-based fault tolerant control for nonlinear time-delay systems

This study introduces a fault compensation control approach for a class of affine nonlinear time-delay systems using adaptive dynamic programming (ADP). By establishing an observer, the ADP-derived approach facilitates the estimation of potential actuator faults within the system. A novel value function is devised to incorporate the estimated fault and state delay. Subsequently, the control laws are derived from this innovative value function. The Hamilton-Jacobi-Bellman equation associated with this innovative value function is effectively solved using a critic neural network. Then, it is demonstrated through Lyapunov theory that the closed-loop systems achieve uniform ultimate boundedness. The efficacy of the current fault compensation strategy is validated through simulation instances. A notable contribution of the proposed method is its extension of the ADP technique to address the challenge of fault compensation in nonlinear time-delay systems.