Noninvasive Prognostic and Health Management for Permanent Magnet of PMSM Drive System Through Comprehensive Parameter Identification

Irreversible demagnetization of the rotor permanent magnets is prone to occur when the carrier operates under high temperature and heavy load conditions. Demagnetization fault not only induces parameter mismatches in the permanent magnet synchronous motors (PMSMs) model but also reduces the output power and torque capability of the motor. Given this, a permanent magnet fault prognostic and health management (PHM) scheme based on comprehensive parameter identification is proposed. To overcome parameter perturbation induced by the demagnetization effect and simultaneously compensate for the rank deficiency condition in the parameter estimation process of PMSMs, multigroup operating states of the motor are obtained by injecting the sinusoidal current signal into the d-axis. Furthermore, the inductance estimator based on the dynamic forgetting factor recursive least-squares (DFF-RLS) algorithm is designed to obtain the dq-axis inductance parameters in online by dynamically adjusting the forgetting factor. Additionally, a flux linkage monitor based on dual <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$H_{\infty } $ </tex-math></inline-formula> filter is designed according to the principle of minimizing the maximum estimation error. This monitor not only tracks the flux linkage parameter of the permanent magnets but also estimates the resistance parameter in real-time. Finally, the estimated results of the two monitors are cross-transmitted to eliminate the parameter mismatch effect caused by demagnetization fault, to achieve an accurate estimation of the average residual magnetic field strength of permanent magnets and demagnetization fault diagnosis. The experimental results show that the proposed scheme achieves high online estimation accuracy, and the relative estimation errors for comprehensive parameters on three PMSMs prototypes remain within 3%, which can realize the high-precision monitoring of magnet health.