Fuzzy logic based online adaptation of current and speed controllers for improved performance of IPMSM drive

Abstract

Precise torque and speed control of electric motors is a key issue in industries for variable speed drives (VSD). Over the years the induction motors have been widely utilized in industries for VSD applications. However, induction motor has some significant drawbacks like low efficiency, lagging power factor, asynchronous speed, low torque density etc. Nowadays the interior permanent magnet synchronous motor (IPMSM) is becoming popular for high performance variable speed drive (HPVSD) due to its high torque-current ratio, large power-weight ratio, high efficiency, high power factor, low noise and robustness as compared to conventional induction and other ac motors. Smooth torque response, fast and precise speed response, quick recovery of torque and speed from any disturbance and parameter insensitivity, robustness in variable speed domain and maintenance free operations are the main concerns for HPVSD. This work proposes a closed loop vector control of an IPMSM drive incorporating two separate fuzzy logic controllers (FLCs). Among them one FLC is designed. to minimize the developed torque ripple by varying online the hysteresis band of the PWM current controller. Another Sugeno type FLC is used to tune the gains of a proportional-integral (PI) controller where the PI controller actually serves as the primary speed controller. Thus, the limitations of traditional PI controllers will be avoided and the performance of the drive system can be improved. A flux controller is also incorporated in such a way that both torque and flux of the motor can be controlled while maintaining current and voltage constraints. The flux controller is designed based on maximum-torque- per-ampere (MTPA) operation below the rated speed and flux weakening operation above the rated speed. Thus, the proposed drive extends the operating speed limits for the motor and enables the effective use of the reluctance torque. In order to verify the performance of the proposed IPMSM drive, first a simulation model is developed using Matlab/Simulink. Then the complete IPMSM drive has been implemented in real-time using digital signal processor (DSP) controller board DS1104 for a laboratory 5 HP motor. The effectiveness of the proposed drive is verified both in simulation and experiment at different operating conditions. In this regard, a performance comparison of the proposed FLC based tuned PI and adapted hysteresis controllers based drive with the conventional PI and fixed bandwidth hysteresis controllers based drive is provided. These comparison results demonstrate the better dynamic response in torque and speed for the proposed IPMSM drive over a wide speed range.