Displacement Selfsensing Control of Outer Rotor Coreless Bearingless Permanent Magnet Synchronous Motor Based on Least Square Support Vector Machine Optimized by IMFO

To solve the problems of large size, high cost, and low reliability brought by mechanical displacement sensors to outer rotor coreless bearingless permanent magnet synchronous motor (ORC-BPMSM), a displacement self-sensing method based on the left inverse system of least squares support vector machine (LSSVM) is proposed. First, the working principle of ORC-BPMSM is introduced and the mathematical model of ORC-BPMSM is derived. Second, the left reversibility of the displacement subsystem is proved and the observation principle of the LSSVM left inverse system is explained. On this basis, the moth-flame optimization is improved by using dynamic inertia weighting strategy and Lévy flight, the parameters of LSSVM are optimized with the improved algorithm, and the displacement self-sensing system is constructed. Finally, simulations and experiments of steady-state suspension, speed variation, and anti-interference are conducted, and the results verify the feasibility and validity of the proposed displacement self-sensing method.