Complex Power Quality Disturbance Classification Based on Fuzzy Transfer Field and DPN-17

Disturbances of power quality in modern power systems tend to be complicated due to the diversification of electrical equipment and the complexity of equipment working conditions. Traditional classification methods are difficult to meet the classification requirement of multiple power quality disturbances (PQDs). In this paper, a hybrid power quality disturbance classification method based on fuzzy transfer field and DPN-17 is proposed. Fuzzy transfer field is a conversion scheme proposed to solve the confusion problem in the process of amplitude-dependent signal transformation in the existing mainstream dimension conversion methods. Fuzzy idea is used to blur the specific coordinates of signal points, and precisely encode the timing relationship and distribution of signal points, so that it can effectively encode the timing relationship and distribution information of signals. It can retain the features well in the process of downsampling; DPN-17 is implemented by referring to the shallow residual network model and using dual-path neural network modular fast combination stack. Simulation and experiments show that the proposed method can effectively solve the confusion problem of dimension conversion schemes, and can achieve accurate and rapid identification of complex power quality disturbances with an accuracy rate of 97.40%.