Acoustic Noise Based Permanent Magnet Flux Reduction Diagnosis and Current Compensation in PMSM

Diagnosis for permanent magnet (PM) flux reduction is necessary for determining motor efficiency. This paper proposes a real-time diagnosis model without invading the motor structure. In this method, acoustic signals acquired with a sound intensity meter under different motor operation conditions are analyzed and processed by a wavelet packet. Moreover, the Grey-Markov Chain model is applied to estimate the flux reduction which is fed to the current compensation algorithm in the vector control drive system. The proposed model is validated in a laboratory designed permanent magnet synchronous motor (PMSM) drive system under different speeds and temperatures.