Design and Analysis of RLS-Based Torque Estimation for Control of EV Traction PMSMs

This paper addresses the critical need for accurate torque estimation in Permanent Magnet Synchronous Machines (PMSM), particularly relevant in the context of the growing elec-trification of automobiles. Focusing on both Surface Permanent Magnet Machines (SPMSM) and Interior Permanent Magnet Machines (IPMSM), the paper proposes a Recursive Least Squares (RLS) based estimator to identify machine parameters for precise torque estimation. This method requires minimal prior knowledge of machine parameters and is analytically shown to be effective for torque estimation with practical levels of parameter uncertainty present in the motor. Simulation studies are presented to verify the concept. Experimental validation on an in-house developed IPMSM supports the accuracy of the proposed method, highlighting its potential for enhancing electric vehicle control and safety.