Transformer Condition Assessment And Partial Discharge Fault Diagnosis

Transformer is the core hub in power system,so the safe and stable operation of power system is the basis of normal work.Our country transformer has been put into operation for a long time,a lot of transformer are to a certain extent,the insulation aging,so the transformer faults caused by electricity accidents emerge endlessly in recent years,it's greatly threatens the safe and stable operation of power grid.Therefore,the state evaluation and fault diagnosis of transformer accurately predicted the development trend of it,to staff in a timely manner to understand the operation of the transformer condition,adopt the correct maintenance strategy,to ensure the safe and stable operation of power grid has very important significance.Thesis collected a large number of data,the field test data,procedures and expert advice,and research the data analysis method,to establish the system of state evaluation method,the support vector machine(SVM)algorithm,the partial discharge fault characteristics,for the use of big data method for transformer condition assessment is studied,based on the partial discharge point on the fault diagnosis method has made certain progress.The main research results are as follows:1)setting up of a combination of evaluation index of the core and non-core index combining multidimensional transformer condition assessment system.The core index mainly includes oil chromato graphic data,oil test data,electrical test information data,non-core indicators mainly include environmental factors,meteorology factors,attachment information and shipment inspection record.2)when the transformer condition assessment,make full use of the core evaluation index on the basis of status evaluation,digging the correlation between the evaluation index data,considering the influence of the core indicators of transformer running state.Set up a state of big data based on information entropy evaluation method and evaluation method of combining the conventional transformer comprehensive evaluation system.3)using the factor analysis method to simplify the partial discharge characteristic factor from 24 to eight,as the subsequent extreme learning machine fault diagnosis threshold.And using the rapid perception of artificial fish around a lot of things,the advantages of the extreme learning machine is optimized,avoid it in the search for the optimal solution in the defect of local minima.