Abstract:Firstly, db4-wavelet is used to analyze PQD signals with multi-resolution decomposition of 10 layers, and energy differences of every level between PQD disturbance signal and standard signal as feature vectors;Then, it used PCA to analyze feature vectors and takes the front six dimensions are extracted as the eigenvectors of classification.The method of cross-validation is used for the training set to select the optimal parameters adaptively and constructs the training set model;finally, the testing set is substituted into the training set model for classification.The test results show that this method can realize the classification of PQD, and has high resolution, strong resistance to noise, fast classification speed, and thus is suitable for the classification of PQD.