Speed Sensorless Method of Bearingless Synchronous Reluctance Motor Based on WOA-Elman Neural Network

In the control of bearingless motors, it is essential to install sensors for detecting rotor position and speed signals. However, the conventional mechanical sensor installation increases motor size and costs, limiting its operational scope at high speeds. In order to realize the ultra-high speed, low cost, and practical operation of bearingless synchronous reluctance motor, a novel rotor position and speed estimation scheme base on the Elman neural network is proposed and designed. The Elman neural network's initial weights and thresholds are optimized using the whale optimization algorithm, which improved the prediction accuracy of the method. Subsequently, utilizing Matlab/Simulink software and an experiment platform, a bearingless synchronous reluctance motor control system is built to validate the proposed theory. Simulation and experimental results collectively demonstrate the correctness, anti-interference ability and feasibility of the proposed methodology.