Parameter Estimation and Online Adaptation of Rotor Time Constant for Induction Motor Drive

Abstract

The performance of a sensor-based rotor flux-oriented vector control induction machine drive is highly influenced by the estimated value of the rotor time constant/rotor resistance which varies continuously with temperature when it is driven with the variable loads. Therefore, to achieve accurate torque control, online tunning of the rotor time constant is necessary. This article proposes a novel and simple adaption technique to estimate the rotor time constant. In this proposed method, the reference model of the rotor flux vector is derived using the voltage model of the induction machine, which is independent of the rotor time constant. The current controller uses the model of the rotor flux vector that is derived using the current model of the induction machine. This current model is dependent on the rotor time constant. The error between the rotor flux vectors derived using the current model and the voltage model is minimized using the least-squares method. In this process, the rotor time constant is always tuned to its actual value. The article also presents a simple approach to estimate the parameters of induction machine like the stator resistance, the net leakage inductance, and the initial value of the rotor time constant in an offline mode at a standstill condition. To validate the proposed algorithm, two experimental prototypes are developed for two different power ratings of 3.7 and 67 kW. Experimental results from these prototypes confirm the effectiveness, reliability, and stability of the proposed algorithm both in the motoring and regenerative modes of operation. This method of the adaptation of the rotor time constant is simple and, therefore, useful for practicing engineers.