An approach to nonlinear fault diagnosis based on neural network adaptive observer

A fault detection method based on neural networks on-line approximation structure for uncertain nonlinear system is presented in this paper. A neural networks observer is used to learning the nonlinear fault functions to monitor the abnormal behavior of dynamic system. When system faults occur, the on-line learning structure can approximate all possible unknown faults, then the faults are identified and accommodated. The uniformly ultimately bounded stability of closed-loop error system is guaranteed by Lyapunov stability theory and the weights is tuning without need of persistency of excitation.