Parameter Estimation of a Two-Section Transformer Winding Model using Pseudo-Random Impulse Sequence Perturbation

Abstract

This paper proposes the use of the Pseudo-Random Impulse Sequence as perturbation signal for estimating the parameters of a lumped parameter circuit model commonly used in the wideband modelling of distributed transformer windings. Impedance and voltage ratio transfer functions are derived analytically for the model and the associated frequency responses are determined. These results are used to validate the frequency responses simulated using a Pseudo-Random Impulse Sequence perturbation signal. A constrained Globalsearch optimisation algorithm is used to estimate the model parameters, where the cost function is defined in terms of the frequency responses of various transfer functions. A total of five case studies are investigated. The results show that the proposed method is suitable for estimating the parameters of the target system with a fair degree of accuracy. Inclusion of a midpoint voltage in the parameter estimation process allows for most parameters to be estimated with accuracy of the order of 1%.