Parameter identification for IPMSM position sensorless control based on unknown input observer

This paper proposes a new identification method for position sensorless control of Interior Permanent Magnet Synchronous Motors (IPMSMs) based on an unknown input observer. In position sensorless control, precise motor parameters are required to accurately estimate the rotor position. Thus, these parameters should be identified prior to the realization of position sensorless vector control. However, d-q axis model is not valid for parameter identification without position measurement because accurate position estimation is impossible. The authors have proposed a method for estimating rotor position using an unknown input observer that is possible to realize significantly robust to error in parameter identification. This paper proposes that the parameter identification method with the assistance of the robust unknown input observer in the position sensorless control.