Improved Svm and Ann in Incipient Fault Diagnosis of Power Transformers Using Clonal Selection Algorithms

Based on statistical learning theory (SLT), the support vector machine (SVM) is well recognized as a powerful computational tool for problems with nonlinearity having high dimensionalities. Solving the problem of feature and kernel parameter selection is a difficult task in machine learning and of high practical relevance in blurred fault diagnosis. We explored the feasibility of applying an artificial neural network (ANN) and multi-layer SVM with feature and radial basis function (RBF) kernel parameter selection to diagnose incipient fault in power transformers by combining a clonal selection algorithm (CSA). Experimental results of practical data demonstrate the effectiveness and improved efficiency of the proposed approach, quickens operations, and also increases the accuracy of the classification.