Unbiased High-Performance Multiparameter Identification Strategy for PMSM Based on Sampling Noise Excitation and Bias Compensation

Sampling noise can contribute positively to parameter identification, whereas all least-squares-based methods exhibit inherent bias when applied to permanent magnet synchronous motors (PMSMs). For the first time, this study establishes a relationship between estimation performance, variables, and sampling noise, while also highlighting the bias present in least squares (LSs) methods across different PMSM models. Building on these theoretical insights, a straightforward yet effective multiparameter identification strategy is proposed. This approach requires only a minor adjustment to the current controller to enhance information and tailors an unbiased iterative estimation algorithm capable of simultaneously identifying four parameters with high accuracy. In contrast to existing methods, it eliminates the need for external injection signals or complex sampling strategies, providing a practical solution that can be seamlessly integrated into existing drive systems. The theoretical findings and proposed strategy have been rigorously validated.