Determination of the induction machine parameters from starting transient measurements

Abstract

This Thesis study the parameters estimation of the induction machine with the single-cage and doublecage models, using the stator currents, voltages and mechanical speed from a starting transient. The dynamic equations of the induction machine are expressed in the synchronous reference frame because in this reference the values of the direct and quadrature currents have a slower variation than those in the other reference frames. This reference frame is important to facilitate the data filtering. The proposed method for the single-cage parameters estimation introduces an approximation of the rotor flux using the stator currents and voltages. It has been demonstrated in the thesis that this approximation improves the accuracy of the calculated parameters significantly. Two different methods to calculate the parameters have been developed. The first is based on the linear least-square method and the second on the resolving of a nonlinear system of equations. A detailed error study of the parameters and predicted behavior have been made in the thesis. An interesting result of the error study is that it is very important that the speed of the points be used in the calculations. It is very important to use the points with the maximum torque speed and points near the synchronous speed. The thesis presents the two first methods developed for parameters estimation of the induction machine with double-cage model using data from a starting transient. The other methods known in the literature use steady state data. Both methods use data from the point of maximum torque to the synchronous speed to calculate the parameters of a single-cage model, because the torque- and current-speed curves of this single-cage fit very well with the torque- and current-speed curves of the double-cage. Later, the first method uses a point with a speed near to zero speed and a point in the synchronous speed and other single-cage is calculated. With these two singlecages the impedances of three points are calculated and they are used to estimate the six parameters of the double-cage model solving a nonlinear system of equations. The second method, called instantaneous power method, uses the fact that the averaged transient torque has similar values to the steady-state torque in the mechanical transient region. Then, several impedance values are calculated for points with different speeds in this region, where the effects of the double-cage are more important. Finally, with impedance values in the synchronous speed and the maximum torque speed obtained with the regression method and
 impedance values obtained with the instantaneous power method, a set of non linear equations are defined. Their solution is the estimated parameters of the induction machine double-cage model.