Wavelet-based fault detection in grid-connected photovoltaic systems

This paper proposes a real-time wavelet-based fault detection as well as a real-time analysis of voltage sag/swell due to faults in a grid-connected photovoltaic system, which is the first step for a wavelet-based islanding protection. The spectral energy of the voltage is decomposed in scaling coefficient energy and wavelet coefficient energy. The real-time analysis of the scaling coefficient energy provides the undervoltage and overvoltage protection, whereas the wavelet coefficient energy provides a high-speed fault detection. The wavelet-based fault detection was assessed in an experimental grid-connected photovoltaic system.