Sensorless Backstepping Drive for a Five-Phase PMSM based on Unknown Input Observer

Abstract

A sensorless Backstepping drive was investigated in this work for the Five phase Permanent Magnet Synchronous Motor (PMSM), using an Unknown Input Observer for speed and load torque estimation. This control technique can solve the problem of the speed sensor fault, where the feedback speed information was deduced directly from the virtual sensor. Moreover, this method is able to ameliorate the performance of PMSM by the load torque estimation used for the real time compensation. The gains of Backstepping control and the Unknown Input Observer are carefully selected in aim to enhance estimation precision and stability dynamic performance. Simulation results prove both of efficiency and capability of the proposed approach to maintain a great operation and a continuity of the system operation.