Magnetic Flux Linkage Estimation of PMSM Based on Split Step Decoupling Strategy

The safety of permanent magnet synchronous motors relies heavily on the health status of their permanent magnets. Therefore, the magnetic flux linkage identification, as a key indicator of irreversible demagnetization, is crucial. Since the cross-coupling effect exacerbates the rank deficiency and coupling degree of permanent magnet synchronous motors, a magnetic flux linkage identification method based on split-step decoupling strategy is proposed to improve the identification accuracy. By analyzing the saliency consistency between the high frequency and fundamental frequency models, the identification model is separated from the cross-coupling effect in the first-step decouple. In the second-step decouple, the working point offset injection through the constant torque curve is used for the magnetic flux linkage identification. To enhance the identification accuracy, a composite filter is applied for the signal extraction and processing, and the dead time effect is also considered in the identification process. Experimental results confirm the effectiveness of the proposed method.