| Transformer is the key equipment for voltage converting,power transmission and distribution in power system.For a long time,accurately assessing the aging condition of oil–paper insulation,developing an effective maintenance and replacement strategy and extending the service life of the transformer as long as possible has become an imperative task in the power industry.The polarization and depolarization current(PDC)and frequency domain spectroscopy(FDS)based on the dielectric relaxation theory,which are not only non-destructive diagnostic technique,but also make up the drawbacks of traditional physicochemical and electrical diagnosis methods,are widely used in the field of condition assessment of oil-impregnated electrical equipment,especially power transformers.However,at present,dielectric response(DR)technique are mostly limited to the evaluation of moisture content,rarely used in the aging assessment of field transformers.In addition,the influences of aging state are usually ignored in the assessment of moisture.Therefore,it is of great theoretical significance and practical application value to investigate a new DR method,which can accurately diagnose the aging and moisture condition of oil-paper insulation of transformers.In this thesis,the equivalent transformation between the DR of transformer main insulation and that of the oil-impregnated pressboard was first analyzed by measuring the time-frequency domain DR of oil-paper composite insulation model and oil-impregnated insulating pressboard.Secondly,in the laboratory,a large number of oil-impregnated pressboard samples with different aging and moisture condition were prepared by accelerated thermal aging and moisture absorption experiments,and their time-frequency DR were measured at different temperatures,then followed by the investigation of their time-frequency DR properties and activation energies under different aging and moisture condition.Afterwards,the dielectric fingerprint characteristics sensitive to the change of aging and moisture condition of the oil-impregnated pressboard were extracted.Finally,a method used for the assessment of aging and moisture condition of transformer main insulation was proposed,and its effectiveness was preliminary verified by practical cases.The main innovative achievements of this thesis are as following:(1)The feasibility of XY model theoretical equation in realizing the equivalence conversion between the FDS of transformer main insulation and that of oil-impregnated pressboard was systematically verified by experimental study.On this basis,Fourier transform and the extended Debye model were introduced to realize the mutual conversion between time and frequency domain spectroscopy of the oil-impregnated insulating pressboard,and which effectively solves the problem that PDC of transformer oil-impregnated pressboard can not be obtained by traditional time-domain DR test.(2)The influence of the combined effects of aging and moisture on DR and activation energy of oil-impregnated pressboard was systematically studied.The study shows that if the combined effects of these two are not considered,it will result in an error of 34% for moisture assessment or 53% for aging assessment of oil-impregnated pressboard.The two-factor analysis of variance indicates that aging and moisture condition do not present a significant effect on the activation energy of oil-impregnated pressboard,and the average activation energy(0.95eV)can be used to improve the applicability of the main curve technique in eliminating temperature influence on FDS.(3)The dielectric fingerprint characteristics which can characterize the aging and moisture condition of oil-impregnated pressboard were constructed for the first time.According to the PDC curve and the modified Cole–Cole model parameters,the time-frequency domain dielectric characteristics sensitive to aging and moisture condition of oil-impregnated pressboard were extracted,and considering the combined effects of aging and moisture on the DR,physicochemical characteristics were introduced to assist in condition distinction,the dielectric fingerprint characteristics formed after further characteristic extraction by the principal component analysis can characterize the aging and moisture condition.(4)It is innovative to introduce the dielectric fingerprint recognition method to assess aging and moisture condition of transformer main insulation.In view of the indistinguishable effects of aging and moisture on DR,it is suggested to group the aging and moisture condion together into composite state,and to identify them by extracted dielectric fingerprints.This method has combined XY model and time-frequency domain spectroscopy transform technique,and can be used for the assessment of aging and moisture of transformer main insulation with different insulation structures.Its effectiveness was preliminary verified in assessment results of five field transformers. |
PDF Full Text Request