DWT-Based Islanding Detection Method for DFIG-typed WTs

Unintentional grid disconnection of wind turbines (WTs) from the existing power grid can cause serious failures for the power system. WTs operating in island mode must be disconnected from the network within a maximum of 2 seconds because the grid is suddenly disconnected. In this study, a medium-power WT model with a double-fed asynchronous generator (DFIG) was proposed and a discrete wavelet transform (DWT)-based island mode detection method was presented. The coefficients of the approximation for island mode detection from a single voltage signal received from the point of common coupling in WT have been obtained in the proposed method by taking advantage of the ability of DWT to analyze both frequency and time domain. The obtained DWT approach coefficients were determined in island mode study using experimentally determined threshold values under different grid condition. The non-detection zone (NDZ) problem was eliminated and island mode detection was performed in less than 0.2 seconds with the proposed method. Detection of island mode using a single signal is also advantageous in terms of applicability compared to existing islanding detection methods.