Fault Identification Algorithm for Grid Connected Photovoltaic Systems using Machine Learning Techniques

The motivation and background behind the fault detection for grid connected solar power plant is presented in this paper. The major issues encountered when integrating a PV system to the grid include multi-peak phenomenon due to partial shading, regulation of circulating currents, the impact of grid impedances on PV system stability, Fault Ride-Through (FRT) Capability, and anti-islanding detection. Hence, fault detection and condition monitoring system are necessary for smooth operation. In this paper, a fault classification technique for single-phase grid connected PV systems is developed. Wavelet Transform and Neural network approaches are used for developing the fault classification algorithm. The results depicted that the developed fault detection algorithm shows a significant improvement in the classification accuracy with 98.4%.