Sensorless Control of Linear Flux-Switching Permanent Magnet Motor Based on Improved MRAS

The sensorless control system of linear flux-switching permanent magnet (LFSPM) motor based on model reference adaptive system (MRAS) has the advantages of simple structure and good stability. However, it behaves poorly at low speed regions. In view of the problem of the conventional MRAS, this paper presents an improved MRAS based on artificial neural network (ANN). In this strategy, a two-layer ANN is selected as the adjustable model and the adaptive law with proportional-integral (PI) controller is replaced by the gradient descent algorithm. Besides, a sliding mode controller is designed for velocity loop to improve the rapidity and robustness of the system. Experimental results show that the proposed method has good dynamic performance and smaller speed estimation error at low speed regions.