On-line identification of induction motors: Experiments and results

Abstract

Induction motors are widely used in the industry. However, due to their involved mathematical models those depend on difficult to measure parameters such as leakage inductance. Therefore, simplified model approximations are usually used instead. In this paper, a system identification method based on auto regressive moving average (ARMA) models and hardware-in-the-loop (HIL) concept has been employed to identify and predict the behavior of a squirrel cage induction motor. The motor transfer function is identified online using an impulse input. Then, the identified model response is compared to the real system response with different input signals. Results show that the model used follows the real system response with good accuracy.