Validation of the Orthogonal Fitting Results for Short-Circuit Impedance Identification

Abstract

Background: Accurate transformer models are essential for electromagnetic transient (EMT) simulations, particularly for evaluating Transient Recovery Voltage (TRV) during switching operations. However, many transformer models used in simulation software rely on simplified parameter identification methods developed before the widespread availability of modern digital measurement techniques. A common approach combines low-signal frequency response measurements with Vector Fitting, producing pole-residue models without providing information on parameter uncertainty. This study investigates an alternative identification method based on orthogonal regression. Methods: The proposed approach applies orthogonal regression based on Koopmans–Levin method to identify the transformer short-circuit impedance from measurement data. The method is validated by comparing the identified model with reference measurements and assessing its suitability for TRV-oriented transient simulations. Results: The proposed identification procedure provides an accurate representation of the transformer short-circuit impedance while enabling the estimation of parameter uncertainty. Compared with conventional frequency-response-based identification, the method offers improved confidence in model parameters and better reflects measurement variability. Conclusions: Orthogonal regression is a promising alternative for transformer parameter identification in EMT studies. By incorporating parameter uncertainty into the identification process, the method enhances the reliability of transformer models used for TRV analysis and other transient simulations.