AN ADVANCED RF-MRAS TECHNIQUE APPLYING TO SPEED SENSORLESS CONTROL

Abstract

Reference frame model reference adaptive system (RF_MRAS) is a typical method to generate an estimated speed signal for motor control without a speed encoder. The primary principle of the RF-MRAS method is to use two voltage and current structures to develop one kind of estimated rotor flux signal and then compare the difference between them for adapting control to an estimated speed signal. However, during operation, due to the increase in temperature, the stator resistance of the motor will change, affecting the speed control performance. This paper proposes an enhanced solution for the RF-MRAS technique based on the parallel integration of stator resistance estimation into the control model to increase motor performance. In this research, the stator voltage calculated from the DC-link voltage and switching pulses is proposed to be provided to the RF-MRAS model as the input. The paper’s objective is that the estimated speed should match the actual operating speed even under variable stator resistance conditions. The performance of the proposed model is investigated through various three case simulations, and the achievement results demonstrated its effectiveness under different operating conditions.