Fault-Tolerant Control for Nonlinear System with Actuator Faults and Disturbances Via Adaptive Dynamic Programming

The fault-tolerant control problem of a class of non-linear affine systems with both actuator faults and disturbances is studied by means of adaptive dynamic programming in this paper. At first, we transform the optimal tracking problem into the optimal stabilization problem through system transformation. After that, an adaptive law is designed to estimate actuator faults, and the fault-tolerant control problem is equivalently transformed into an optimal controller solution problem through special design of performance index functions. The neural network is used to realize online policy iterative algorithm, and the approximately optimal controller can be obtained through the critic network. An disturbance compensation controller is also designed. The tracking error of the system is uniformly ultimately bounded is proved by the Lyapunov stability theorem, and the simulation verification is carried out as well.