Performance analysis of the sensorless adaptive sliding-mode neuro-fuzzy control of the induction motor drive with MRAS-type speed estimator

Abstract

This paper discusses a model reference adaptive sliding-mode control of the sensorless vector controlled induction motor drive in a wide speed range. The adaptive speed controller uses on-line trained fuzzy neural network, which enables very fast tracking of the changing speed reference signal. This adaptive sliding-mode neuro-fuzzy controller (ASNFC) is used as a speed controller in the direct rotor-field oriented control (DRFOC) of the induction motor (IM) drive structure. Connective weights of the controller are trained on-line according to the error between the actual speed of the drive and the reference model output signal. The rotor flux and speed of the vector controlled induction motor are estimated using the model reference adaptive system (MRAS) - type estimator. Presented simulation results are verified by experimental tests performed on the laboratory-rig with DSP controller.