Intelligent Evaluation Of Vegetable Oil Paper Insulation Condition Based On Frequency Domain Dielectric Spectroscopy

As an important part of traditional transformers,insulating oil has always been a research focus.At present,most traditional transformers use mineral oil,but with the continuous development of clean energy,the safety and environmental problems caused by mineral oil are becoming more and more serious.It is imperative to find a clean alternative.Vegetable oil has received extensive attention in recent years due to its high ignition point,environmental protection,and renewable advantages.The main component of vegetable insulating oil is triglyceride,which is different from mineral oil,resulting in different physical and chemical properties such as flash point and ignition point.Therefore,for the wide application of vegetable insulating oil,it is an indispensable long-term subject to in-depth study of the insulation state of the entire system from various aspects.The main work done in this paper is as follows:1)In order to study the aging law of vegetable oil-impregnated paper insulation system and insulation diagnosis method,this paper introduces the concepts of fractional element and fractional calculus,and proposes a method for parameter identification of vegetable oil-impregnated paper insulation fraction Maxwell model based on particle swarm optimization.Through experiments,the real part and imaginary part of the complex permittivity measured by the FDS method are used to obtain the curve,and the characteristic quantity related to the aging state in the Maxwell model of the score is obtained through these experimental data.2)Verify the accuracy of the Maxwell model.Select the characteristic quantity relaxation time and shape parameters obtained in the Maxwell model,and use mathematical fitting to obtain the parameter polymerization degree DP that characterizes the aging state of the vegetable oil-impregnated paper insulation test sample.There is a better exponential function relationship with the relaxation time τ,which is related to the parameters.Good linear relationship.It is proved that Maxwell model can characterize the aging process of vegetable oil-paper insulation system.3)Based on the FDS data measured in the laboratory and the transformer FDS data measured on site,the insulation status of the transformer oil-paper insulation system is analyzed using the method of intelligent diagnosis.First,the data is fuzzy and reduced to form a rule set;second,the neural network is trained through the rule set to obtain the evaluation rule table of the transformer oil-paper insulation system;finally,an example verifies the effectiveness of this rule table.