Parameter identification of extended Debye model based on Kramers-Kronig transform and frequency domain differentiation

It is of great significance to identify the parameters of the oil-paper insulation extended Derbyshire model, however, the identification accuracy and model reliability of most methods are not high. To address this kind of problem, this paper adopts the method based on the frequency domain decoupling of the Kramers-Kronig transform to achieve the accurate identification of the extended Derbyshire model. Firstly, the frequency-domain dielectric spectrum is decoupled by Kramers-Kronig transform to accurately distinguish the insulation resistance spectral lines, geometrical capacitance spectral lines, and multiple relaxation polarisation spectral lines in the extended Debye model, and the parameters of insulation resistance and geometrical capacitance are identified by the least-squares method. Subsequently, the real part of the polarised complex capacitance spectral lines are decomposed into a number of real part of the micromolecular spectral lines with peak points, and the accurate parameters of each polarisation branch are solved by the corresponding peak point features. Finally, the accuracy of the proposed identification method is verified by combining simulation and examples, which provides an accurate model reference for future research on the internal fundamentals of oil-paper insulation.