Application of S-transform modular matrix and LS-SVM in identification of lightning and switching overvoltages
Bin Dai · Dianli zidonghua shebei · 2012
S-transform modular matrix and LS-SVM(Least Square-Support Vector Machine) are applied to identify the lightning overvoltage and switching overvoltage.The singular value decomposition decomposes the S-transform modular matrix of zero sequence overvoltage signals into different time-frequency characteristic subspaces,from which five statistical features are extracted and used as the input vectors of the LS-SVM classifiers to identify the lightning and switching overvoltages.Results of the test with measured overvoltage data indicate that,the characteristic dimension is small,the features are immune to electromagnetic noises,the training times is low,and the recognition rate is high.The proposed method can be well applied in identification of lightning and switching overvoltages.