The simulation analysis of transformer recovery voltage by field and circuit method based on PSO algorithm

Oil immersed distribution transformers are widely used in the distribution network. The internal insulation system is composed of oil-paper composite material. The distribution transformer has short-time overload operation, so deterioration of the insulation performance of internal oil-paper system will take place. It is a convenient and effective method to evaluate the insulation status of oil-paper insulation system to a certain extent. The typical oil-paper insulation system model of the distribution transformer is established in the ANSYS simulation environment. The voltage waveform during the recovery voltage measurement is applied to the model, the characteristics of the E-field distribution at critical time points are analyzed. The principle of recovery voltage is analyzed using lumped circuit. The multi-parameter extended Debye model is resolved by the theoretical formula. On the MATLAB/Simulink platform, the Debye parameter model is extended to the lumped parameter circuit. The Simulink-PSO algorithm is used to realize the nonlinear fitting the Debye parameters, and the influence principle of the lumped parameters on the polarization spectrum is analyzed quantitatively. The study found that the field-circuit simulation analysis shows the transient process exists in the recovery voltage measurement. The charge transfer at the interface of oil-paper composite insulation is the main reason for the transient process. MATLAB/Simulink can be used to construct the multi-branch extended Debye model, and carried out numerical simulation. The PSO algorithm can effectively search the lumped parameter of Debye model. In this paper, the electric field and circuit method are used to analyze the recovery voltage of oil-paper composite insulation, the lumped parameter inversion algorithm of Debye model is proposed, it provides the theoretical guidance for the insulation condition monitoring of oil immersed distribution transformer.