Sensorless Full-Torque Startup of IPMSMs via Saturation-Embedded Inductance Regression

To enable full-torque sensorless startup for interior permanent magnet synchronous motors (IPMSMs), this letter presents a paradigm shift from passive saturation compensation to active exploitation of saturation-induced inductance. By encoding magnetic saturation characteristics into pulsating-voltage models and using a nonlinear observer to estimate the inductance regression, the method achieves accurate position estimation immune to load-induced nonlinearities. Experimental results show sustained full-torque production with high precision of the position estimation, eliminating torque derating. This addresses a key limitation in high-performance applications like EV traction and industrial servos.