Research On Aging Performance Of Oil-Paper Insulation Equipment Based On Polarization Spectrum And Characteristic Parameters

The subject derives from the National Natural Science Foundation Project of China that is the diagnosis method of transformer oil-paper insulation based on circuit analysis and recovery voltage response(item number:61174117). This paper takes oil-paper insulation transformer as the research object, and further researches the relationship that polarization spectrum and characteristics parameters of return voltage with the oil-paper insulation condition, based on the computer model and a simulation platform. Through many simulations and experiments, the recovery voltage characteristics of oil paper insulation system are discussed. The return voltage diagnosis mentalities combine with the field test results, and it is feasible that using the recovery voltage characteristic parameters to diagnose insulation condition. Therefore, the research not only has the important research value, but also has a good extensive practical application prospect.The main works of this thesis are listed as following:1. Based on the extend Debye dielectric response model, it is studying on how to accurately identify model parameters. This paper introduces a hierarchical concept in the parameter identification for the first time, and makes Hierarchical Particle Swarm Optimization algorithm.At last it achieves equivalent parameter identification based on the recovery voltage test data. By comparison with the Particle Swarm Optimization algorithm, the new algorithm is more effective, flexible and accurate.2. The internal relationships between insulation state, polarization spectrum, recovery voltage characteristic parameters and equivalent model parameters are studied by the simulation. Then it is proposed that the correspondence between the model number of branches and the insulating state, namely the insulation condition worse, the model branches’number should be increased. The non-standard polarization spectrum is successfully simulated through the equivalent model. It reveals the internal influence factors of non-standard polarization spectrum and proposes model identification research’s ideas on non-standard polarization spectrum.3. According to the insulation structure of actual oil-paper insulation transformer, a Ⅱ type analog insulation system is made. Based on simulation experiment platform, the paper studies the charging voltage and the ratio of charging to discharging time’s influence on test data via experiments. Return voltage characteristics are analyzed from the moisture content, the aging, the insulating paper layer and the air gap. Nowadays, study on the use of return voltage method to detect insulation system with air gaps is little, so the research’s direction has some value.4. The article uses insulation diagnosis mentality to analyze the field test result, and proposes the synthesis diagnosis method that is based on the polarization spectrum, the peak value time and the initial slope of RV, such as:a) Through on-site test of the same voltage level and different capacity oil-paper insulation transformers, it can be got that near capacity of the transformers are comparable, b) The method can be used to accurately identify fault phase from the different polarization spectrum and characteristic parameters, c) According to field test, it indicates that the numbers of model branches indeed are correlated with the insulating state, d) Non-standard polarization spectrum obtained from the actual test can be used to establish equivalent model, and the water is one of the non-standardization’s reason.