Power Quality Dectection and Classification in Active Distribution Networks Based on Improved Empiricial Wavelet Transform and Disperson Entropy

In this paper, a model including wind power generation, photovoltaic power generation and electric vehicle for high permeability active distribution network (ADN) is established. The power quality (PQ) disturbance signals in the high permeability are extracted, and the characteristics of disturbance signals are analyzed in the situation of grid connection, interruption and islanding. The multi-scale fluctuation dispersion entropy (MFDE) initialized by the improved empirical wavelet transform (IEWT) is utilized to detect and classify the disturbance signals in the high permeability ADN. First, the eigenvectors of the disturbance signals are obtained by using the multi-scale fluctuation dispersion entropy initiated by the IEWT, and then the reduced eigenvectors are put into the support vector machine to classify the PQ disturbances caused by the access of the different distributed generators accessed. The classification results are compared with that in the traditional methods and other similar ways; the effectiveness of the IEWT-MFDE system is verified.