Research On Dissolved Gas Analysis And Fault Diagnosis Method In Vinsoil Insulating Oil Transformer

Natural ester insulating oil transformers have increasingly become excellent in environmental protection and fire safety,and have been used in many fields.At present,there is a lack of experimental data for natural ester insulating oil transformers in China,and the research on the mechanism of gas production under faults is still lacking.The fault diagnosis methods suitable for mineral oil are not suitable for natural ester insulating oil.Therefore,it is imperative to study fault diagnosis methods suitable for natural ester insulating oil transformers.Vinsoil insulating oil was selected as the research object from natural ester insulating oil.The dissolved gas characteristics of Vinsoil insulating oil under electric and thermal faults were explored,and the differences of gas production rules under electric heating faults were compared.Meanwhile,a fault diagnosis method suitable for Vinsoil insulated oil transformer was proposed Based on dissolved gas analysis technology and artificial intelligence technology.A fault model of Vinsoil insulated oil transformer was developed and the dissolved gas content in oil was measured.The distributed model was selected to simulate the electrical fault and thermal fault respectively.The volume and shell material of the model were determined.A new thermal fault model and heating source were designed,and the existing electrical fault model was improved.A total of 306 groups of oil and gas data were measured by gas chromatography under six typical transformer faults(partial discharge,low energy discharge,high energy discharge,low temperature overheating,medium temperature overheating and high temperature overheating).The contents of dissolved gas in oil under different faults were analyzed.The variation rule of the content of gases(hydrogen,methane,ethane,acetylene,ethylene,carbon monoxide,carbon dioxide)in seven types of oil under six typical transformer faults was explored over time and so were the corresponding characteristic gases under different faults.The differences of dissolved gases in oil under electrical faults and thermal faults were compared in terms of gas concentration and gas types.The influence of fault severity on decomposition reaction and product of insulating oil was analyzed.In order to avoid the subjectivity of model parameter selection,AWOA algorithm Based on adaptive weight and search strategy was used to optimize the total random forest number and random forest number of cascade forest.Then,dissolved gas in seven kinds of Vinsoil insulating oil was used as the model characteristics.Finally,a fault diagnosis model of Vinsoil insulated oil transformer Based on AWOA optimized cascade forest was formed.In order to verify the validity of the model,firstly,the fault characteristics were compared with IEC three-ratio method,Roger ratio and common DGA data applicable to mineral oil.Secondly,it was compared with common optimization algorithms such as particle swarm optimization and whale optimization.Finally,it was compared with random forest,cascade forest,particle swarm optimization cascade forest and other fault diagnosis models.