Valuable tips for transformer ladder network parameters estimation

Estimation of the parameters of a transformer ladder network model based on the measured impedance frequency response (IFR) faces major challenges. Nonlinear nature of the system, mutual interaction, frequency dependency of the elements and the number of parameters are those that make the accurate and valid model estimation task involved. In this respect, based on the transformer physical insight, certain tips are derived which contract the search space, suggest the number of the ladder network model sections, provide appropriate initial conditions those improving the performance of the genetic nonlinear optimization algorithm. For the transformer model to be accurate for transient behavior analysis, a 19-section laddernetwork model with frequency dependent elements (except capacitors) is considered. The model is modified by adding in-between sections capacitors. The estimation results indicate that while both models succeed in simulating precisely the dominant resonances, the non dominant resonances are better captured by the modified ladder network model.