Estimation of coefficients of multivariable power series approximating magnetic nonlinearity of AC machines

Abstract

Abstract:Energy based approach was used in the study to formulate a set of functions approximating the magnetic flux linkages versus independent currents. The simplest power series that approximates field co-energy and linked fluxes for a two winding core of an induction machine are described by a set of common unknown coefficients. The authors tested three algorithms for the coefficient estimation using Weighted Least-Squared Method for two different positions of the coils. The comparison of the approximation accuracy was accomplished in the specified area of the currents. All proposed algorithms of the coefficient estimation have been found to be effective. The algorithm based solely on the magnetic field co-energy values is significantly simpler than the method based on the magnetic flux linkages estimation concept. The algorithm based on the field coenergy and linked fluxes seems to be the most suitable for determining simultaneously the coefficients of power series approximating linked fluxes and field co-energy.

Model of the double-rotor induction motor in terms of electromagnetic differential

“…The mathematical model of the double rotor machines was formulated using the Lagrange energy method (Lagrange equations). A detailed description of the methodology for modelling induction machines using Lagrange equations is given in [8][9]. The first step, to develop a machine circuit model, is to divide it into lumped conservative and dissipative elements, both electrical and mechanical.…”