Filter based torque ripple minimization of an adaptive neuro-fuzzy controller for induction motor

A novel filter based adaptive neuro-fuzzy (NF) speed controller coupled with an adaptive flux regulation for indirect field oriented induction motor (IM) drives is presented in this paper. A first order low pass digital infinite impulse response (IIR) filter is used in the input side of the proposed NF speed controller, which significantly softens the torque command generated by the adaptive neuro-fuzzy controller (NFC). Furthermore, the low pass filter based neuro-fuzzy controller (NFC) is able to eliminate the slight stable-state error of the speed, which the non-filter NFC based motor drive suffers from. In order to minimize the stable-state torque ripple of the induction motor, a speed error based reference flux regulation is employed in the proposed control scheme for induction motor drives. When the rotor speed is following the reference speed and kept in stable-state, the reference flux is moderately decreased, so that the torque ripple is minimized, and vice versa, the reference flux is increased up to the rated value when the speed error overruns. A complete simulation model for indirect field oriented control of IM incorporating the proposed filter based NFC and reference flux regulation is developed in Matlab/Simulink. The performance of the drive has been tested at different operating conditions. It has been proved that the proposed filter based neuro-fuzzy controller coupled with the torque ripple minimization method is suitable for application in high performance motor drive system