Adaptive Sliding Mode Recurrent Fuzzy Neural Network Estimator in Magnetic Bearing System

Abstract

This paper proposesd am adaptive sliding mode fuzzy neural network estimation (ASFNE) in the magnetic bearing system (MBS). The fuzzy neural network estimator has fuzzy rules base and neural network weights which the stability is proved by Lyapunov theorem in ASFNE. Therefore, ASFNE estimates system lump uncertainty to improve steady-state error and reduced chattering phenomenon. Finally, we compared ASFNE and sliding mode controller in MBS which ASFNE has better output responses.