Speed‐Sensorless Model‐Free Predictive Torque Control for Induction Motor Drive

ABSTRACT A speed sensor in a motor drive system decreases the reliability and increases the cost and hardware complexity of the system. In this paper, a speed‐sensorless model‐free predictive torque control (MF‐PTC) scheme has been proposed for an induction motor (IM) drive. An extended Kalman filter (EKF) is employed to estimate the motor speed and thus requirements of a speed sensor and pulse width modulation signals are avoided. The rotor flux is also estimated using the EKF based on the current model of the IM. An ultra‐local model is employed instead of the mathematical model of the IM for predicting the current and flux in the prediction step of the proposed MF‐PTC. The unknown state of the ultra‐local model is estimated using a linear extended state observer. Simulation and experimental results confirm that the proposed controller is robust against motor parameter variations. The speed estimator can estimate the motor speed accurately from the maximum speed down to a minimum of 1 Hz. The controller also demonstrates good performance in terms of torque response, stator flux response and stator current total harmonic distortion under various operating conditions.