Sensor Fault Diagnosis via Deterministic Learning

In this paper, an approach for the sensor faults detection of a class of nonlinear systems is proposed which can response sensor faults quickly. Firstly, through adaptive observer, high gain observer and deterministic learning (DL) theory, unknown sensor fault function can be approximated locally. The stability of the system and the convergence of the neural network are proved by using Lyapunov method and the lemma of uniformly completely observable. The states of system are stored in constant value neural network which can be reused by establish a set of dynamical estimators. If the residual norms of one fault estimator is less than the preset threshold in a finite time, the occurrence of the fault can be detected. The main contribution of this paper is that the convergence of the adaptive observer is proved without constructing a multi-delayed system. Simulation studies are included to demonstrate the effectiveness of this approach.