Neuro-fuzzy and fuzzy logic controllers based speed control of IPMSM drive — A torque ripple optimization approach

This paper presents a closed loop vector control for an interior permanent magnet synchronous motor (IPMSM) drive incorporating a neuro-fuzzy controller (NFC) and a fuzzy logic controller (FLC). The NFC serves as the speed controller of the drive. It is designed so as to generate the appropriate command q-axis current and ensure dynamic speed control. The back-propagation technique is used for the online tuning of this adaptive neuro-fuzzy inference system (ANFIS) parameters. Furthermore, a Mamdani type FLC is designed and incorporated to optimize the developed torque ripple by online adaptation of the hysteresis band limits of the PWM current controller. A performance comparison of the proposed NFC-FLC based IPMSM drive with conventional proportionalintegral (PI) controller based IPMSM drive having fixed hysteresis band limits is provided. Comparative simulation results demonstrate better torque response and dynamic speed performance of the proposed drive at different operating conditions.