A Rapid Diagnosis Method of Small Faults Based on Adaptive Sliding Mode Observer

Abstract

In order to solve the problem of fast fault estimation involving large amplitude noise disturbance and small amplitude fault at the same time, a new sliding mode variable structure adaptive estimation method is designed. First, the original system is decoupled into two subsystems by constructing a suitable transformation matrix. For the subsystem containing only minor faults, a small fault fast estimation algorithm is proposed, which significantly improves the estimation performance of small faults. For the subsystem containing noise disturbances and small faults, a sliding mode observer is designed to eliminate the effects of noise and stabilize the integrated observer. Then the Lyapunov stability theory is used to prove the stability of the proposed integrated observer. Finally, the effectiveness of the method is verified by simulation experiments.