Identification of Voltage Sag Sources Based on Wavelet Transform Sample Entropy and Extreme Learning Machine

Accurate identification of voltage sag sources is the basis for the allocation of responsibilities and governance decisions for voltage sags. In order to solve the problem of voltage sag sources identification, this paper proposes a voltage sag sources identification method based on the combination of wavelet transform sample entropy and extreme learning machine. First, wavelet transform is used to decompose the voltage signal, and the sample entropy is extracted from its approximate coefficients and detail coefficients to form a feature vector; then based on the extreme learning machine theory, a voltage sag sources identification model is established, and the feature vector is input into the identification model to identify the type of voltage sag source. Through simulation analysis, the performance of the proposed method is evaluated by the confusion matrix and its evaluation index, which proves the correctness and feasibility of the proposed method.