Research On Parameter Identification Of Oil-paper Insulation Equipment Model And State Association On FDS

Oil-paper insulation equipment is an important part of power system, in the power grid which serves for energy conversion, insulation, protection and measurement, etc. Operation experiences show that the insulation performance degradation is the main reason of oil-paper insulation equipment failure. Therefore, accurate diagnosis of their insulation condition is significant to ensure the safe operation of the power grid. Frequency domain spectroscopy method based on frequency domain dielectric response is a noninvasive and non-destructive diagnosis technology, which has the advantages of abundant information and strong anti- interference ability and hence it has been widely used. But current research is mainly limited in analyzing the impact on frequency domain dielectric response curve from oil-paper insulation condition, and few studies focus on the state of the dielectric response curves, and they rarely identify parameters from frequency domain dielectric response or establish quantitative relationship with the insulating state.First of all, in this paper, the extended Debye model was employed to describe the relaxation polarization process of the dielectric response. Expression of FDS based on extended Debye model was derived, and through changing the parameters of the model, the change of insulation state was simulated. The impact on FDS curve from parameters changes in the model was analyzed. Simulated and test results and be verified and supplementary to each other, which is helpful to interpret the experimental results of FDS.Secondly, identification of the extended Debye model parameters based on FDS test results was implemented, using combined genetic algorithm and Levenberg Marquardt algorithm. The determination of optimization target and polarization branch number was demonstrated. Whereafter, FDS was measured on two transformer bushings in which damping process from outs ide to inside were simulated. The proposed algorithm was verified based on the FDS test results. The identified parameters of two bushings are analyzed, and it confirmed the feasibility of extended Debye model parameters to characterize the insulation stat e. Based on the model identification parameters, the polarization curves of the recovery voltage maximum value are derived, and compared with the test curve, it is proved that the calculated results are in good agreement with the test results.Finally, the FDS test results of different aging degree, with different moisture content and different testing temperature of oil paper insulation sample were analyzed, and the impact on FDS test results from three state factors was summarized; the FDS tests results of the different state were used to identify the extended Debye model parameters by the above optimization algorithm. The influence of insulation state on model parameters was analyzed, the relationship between parameters and state parameters was quantified, and quantitative correlation model is established. And according to the relational model, the method based on correlation between the reconstruction curve and the test curve of calculating oil-paper insulation was proposed, and the oil-paper insulation o f samples was taken to verify the effectiveness of the diagnosis method.