Application of ANN for forecasting of PV plant output power – Case study Oryahovo

Nowadays, the prediction of energy production of grid connected photovoltaic (PV) power plant is important for ensuring the quality and continuity of power supply. In this context, this paper analyzes different approaches for prediction of PV power output, using Artificial Neural Network (ANN). The Multi-Layer Perceptron (MLP) ANN method is widespread in the scientific literature for foresting of the solar irradiance and PV power output. The ANN is modeled using five years of solar data, measured or estimated using previous research and the measured power output at each 10 minutes. The accuracy of the optimal configuration is around 8% for the RRMSE.