Transformer Fault Diagnosis And Prediction Based On SVM

As the hub equipment of the power system,the operation status of the power transformer is directly related to whether the power system can operate stably.Therefore,it is necessary to make the transformer fault to accurately judge and predict in advance.Usually,the dissolved gas analysis(DGA)in oil is used as the fault diagnosis method,and the DGA can accurately and effectively find the hidden dangers of the transformer and its degree of development.The type and rate of gas produced when the transformer fails are different according to the cause of its failure.Therefore,it is of great significance to make timely fault diagnosis and predict the occurrence of transformer.This paper will use SVM method,and apply it to the research of transformer fault diagnosis and prediction problems,and will use SVM classification and regression algorithm to build two models of fault diagnosis model and fault prediction model respectively,specifically as follows:(1)The choice of support vector machine hyperparameters will affect the accuracy of the fault diagnosis model.In order to eliminate this impact,this paper preliminarily selects the strong distributed PSO and GWO with strong group search ability,and uses two methods to optimize the kernel parameters and penalty factors to verify the simulation effect.In order to further improve the accuracy of the model,this paper combines the advantages of the two algorithms,takes into account the accuracy and convergence speed of the algorithm,and finally adopts a PSO-based improved GWO to optimize the fault diagnosis and prediction model.Finally,the model was applied to the transformer.Through a large number of experimental data analysis,the test set accuracy of the model was finally improved to 95.9184%,which further shows that the fault diagnosis model proposed in this paper has a better effect and has certain practical significance.(2)A regression model is used to analyze the prediction of the gas concentration data in the transformer oil to realize the prediction of the transformer failure.To verify the prediction accuracy of the regression model,the combined fault prediction model,PSOSVR and GWO-SVR are constructed.The prediction effect of PSO-SVR and GWO-SVR models is compared with that of 5-fold cross model.After many simulation experiments,the results show that PSO-SVR and GWO-SVR models have high prediction accuracy and good convergence.In order to further improve the prediction power of the model,a transformer fault prediction method based on PSO-GWO-SVR model was constructed,and the prediction effect of this model was compared with the prediction of PSO-SVR,GWO-SVR and discount discount cross method.After many simulation experiments,the results show that the PSO-GWO-SVR model is the best effective,and the combined prediction model proposed in this paper has higher accuracy.Figure [20] Table [23] Reference [80]...