Rotor Time Constant Identification Method for Induction Motors Based on an Improved Stator-Flux Model

In induction motor drive systems employing indirect field-oriented control (IFOC), variations in the rotor time constant lead to inaccurate estimation of the rotor field phase, thereby preventing accurate decoupling of torque and flux. To solve this problem, this paper designs an online rotor time constant identification scheme based on an improved stator-flux model. Considering that the stator-flux model is susceptible to DC offset, a stator-flux observer with DC-offset compensation is designed. Based on the improved stator-flux model, a rotor-flux voltage model is established as the reference model of the model reference adaptive system (MRAS), while the rotor-flux current model is adopted as the adjustable model. The corresponding adaptation law is derived based on Popov’s hyperstability theory to ensure the stability of the identification system. Simulation results show that the proposed observer reduces the relative flux estimation root-mean-square error from 37.38% to 2.05% in the DC-offset test. In the parameter identification test, the estimated inverse rotor time constant enters a ±2% error band in approximately 0.43 s and achieves a relative root-mean-square error of 0.51% over the interval from 1.5 to 2 s.