Study on islanding detection of grid-connected distributed generations with random properties

Abstract

With the development of renewable energy, distributed generations such as photovoltaic generation, wind power systems need to be connected to a utility power grid for the systems to work effectively as energy sources. Islanding detection is one of the most important techniques to develop for the distributed generation systems. A novel active islanding detection algorithm based in pseudo-random signal is proposed in this paper. The modulation waveform will be modified by a binary pseudo-random signal (M-sequence signal), then the voltage of PCC (the point of common coupling) would have random properties. When in grid-connected scenario, the voltage of PCC is dominated by voltage of power grid and the variance has no remarkable peak; however, once the islanding event takes place, the variance of PCC has a remarkable peak. Therefore the islanding can be detected by monitoring a peak of the variance. Doing simulation and experiment based on photovoltaic grid-connected generation systems, the experimental results demonstrate effectiveness and being easy to realize the proposed algorithm.