A reduced-order model based induction machine self-commissioning method

Abstract

A new method which uses the voltage source PWM inverter (VSI) to identify the induction machine (IM) parameters is proposed based on the reduced-order models of induction machine. All the variables for parameter extraction are obtained by the inverter and converted to the dq frame using the FOC software modules. The variables are DC values and low-pass filter may be used to reduce the harmonic and noise effects. The least-square and curve fitting methods are employed to remove the nonlinear effects. The proposed method provides high accuracy for induction machine parameter characterization and can be used to measure the saturated magnetizing inductance as a function of frequency and magnetizing current. It can be integrated with the FOC software as the self-commissioning function without much cost of additional software.