Fuzzy Compensation of Model Predictive Control for PMSMs with External Disturbances

In some high-speed scenarios, encoders are not effective in providing real-time feedback. Besides, the encoder will increase the cost of Permanent Magnet Synchronous Motor (PMSM), both in terms of installation and maintenance. Therefore, designing sensorless algorithms in medium and high-speed scenarios will extend the practical application value of PMSM. Different from the angle estimation scheme for rotated coordinate system, model predictive control (MPC) will start from the stationary coordinate system to discuss the Lyapunov stability and directly obtain the speed and angle from the stability condition. The biggest advantage of this method is to avoid increasing the iteration error of observation angle. Meanwhile, it will properly simplify the conditions of stability. In addition, uncertain factors will adversely affect the performance of PMSMs. Then, different fuzzy schemes are introduced to compensate the uncertain factors in the sensorless mode, and the relation between the fuzzy error and the current error is proved strictly. Moreover, the realization condition of controller stability is discussed globally. Finally, the fuzzy compensations are proved to have good compensation for external disturbances in different experiments.