Performance Improvement of IPMSM Position Sensorless Vector Control System by the Online Identification of Stator Resistance and Permanent Magnet Flux

Abstract

Some of the position sensorless vector control methods of IPMSM use the back electromotive force and require accurate values of motor parameter. Their stability is influenced by motor parameter errors. Particularly, it becomes a problem in the case of extreme temperature variation and operation at low speed. In this paper, the online identification of stator resistance and permanent magnet flux is proposed for performance improvement. The proposed method uses only an observer signal without any additional sensor or signal injection. Both parameters are identified simultaneously using an appropriate error function. The principle of the proposed method is presented in the experimental results.