A comparative study of discrete wavelet transforms in islanding detection of single-phase grid-connected photovoltaic system

Single-phase photovoltaic (PV) systems are the major population of distributed generation adopted by domestic and small commercial consumers. The research contributions on islanding detection for PV systems are mostly for three-phase systems. Passive islanding detection from local signals also suffers from non-detection. Here a single-phase grid-connected PV system is simulated in MATLAB/Simulink. Different discrete wavelet transforms (DWT) signal processing technique is implemented on signals from the model. The wavelet coefficients are analyzed to select a feature. The cases of load variation and irradiation changes are studied. The wavelet type and decomposition level are comparatively studied for different DWT methods. The study concludes with an appropriate feature for the time localization of the islanding condition which is a computationally fast method, reduces non-detection and is suitable in real-time.