An Adaptive Filter-Based Online Parameter Identification Method for Permanent Magnet Synchronous Motors

Abstract

The reliable and high-performance operation of permanent magnet synchronous motors (PMSMs) fundamentally depends on the accurate estimation of their electrical parameters. These parameters are affected by various operational factors like temperature and current, and their precise estimation is key to optimizing motor performance. In particular, the winding resistance and rotor flux in PMSMs vary with temperature, whereas stator inductances exhibit nonlinear behavior due to magnetic saturation influenced by the operating current. In this study, a novel online parameter estimation method is introduced, which leverages an adaptive filter based on the normalized least mean square (N-LMS) algorithm. The approach utilizes the voltage equations expressed in the d-q reference frame, which is commonly used for modeling PMSM dynamics. To facilitate the convergence of the parameter values, a low-value sinusoidal current is injected into the d-axis current. This excitation ensures that the system converges quickly and provides accurate parameter estimates. The results obtained under different load conditions and varying scenarios demonstrate the effectiveness of the proposed method, with high estimation accuracy achieved for motor parameters. The proposed N-LMS-based online estimation framework offers a substantial solution for online parameter estimation, enhancing the control and performance of PMSM drive systems.