Parameter Identification of Induction Motors Based on Third-Order Generalized Integrators

This paper proposes a novel algorithm for the parameter identification of induction motors based on third-order generalized integrators (TOGIs), where both the electrical and mechanical parameters are estimated. First, AC voltages of 0.7 V at 200 Hz and 1.2 V at 5 Hz superimposed on a DC bias component are injected into the induction motor by a PWM inverter, from which the amplitude of the stator current and the phase angle between stator voltage and stator current are calculated by the TOGI. These signals are used to estimate the stator transient inductance, rotor resistance, and rotor time constant precisely. On the other hand, a dynamic test is performed to identify the mechanical parameter. A sinusoidal speed reference is commanded to the induction motor, and the corresponding motor torque signal is then utilized to estimate the moment of inertia with the help of the TOGI. Since the TOGI behaves as an adaptive filter, the adverse effects of DC-offset and measurement noises are mitigated. The effectiveness of the proposed identification method has been verified by experimental results, showing errors of +2.6% for the stator transient inductance, +2.8% for the rotor resistance, and -1.8% for the rotor time constant compared to those obtained following IEEE Std. 112; and 1.3% for the moment of inertia compared to the value in the technical datasheet.