A Sensorless Predictive Torque Control for Induction Motor Using Ultra-local Model

In this paper, a sensorless predictive torque control (PTC) for induction motor using ultra-local model is studied. First, the speed and stator flux linkage are estimated online by the adaptive full-order observer (AFO). Second, in sensorless PTC, the stator current prediction model composed of the conventional prediction model (CPM) is sensitive to the motor parameters variation. In addition, the accuracy of the prediction model will be further decreased owing to the error of estimated speed and actual speed. In this paper, the ultra-local model (ULM) is adopted as the prediction model without considering any motor parameters. Thus, this method has the better robust performance. Finally, the scheme is proved by simulation.