Classification for Power Quality Disturbances Based on Cubic B-Spline Wavelet and Decision Tree

An efficient method for power quality disturbances (PQD) classification is put forward which combines with cubic B-spline wavelet and C4.5 algorithm in decision tree. At first the mathematic models are established. Then multi-scale decomposition for PQD is carried out using cubic B-spline wavelet. A three dimensional eigenvector is built by selecting the mean, standard deviation and energy entropy of the wavelet coefficients. To improve the classification accuracy, de-noising solution is implemented before extracting features. Finally PQD classification is carried out based on C4.5 algorithm. The simulation verifies its validity to classify PQD.