Artificial neural network predictor for grid-connected solar photovoltaic installations at atmospheric temperature

An increase in environmental awareness, renewable energy usage and concern for energy security have resulted in the advent of Solar Photovoltaic (PV) systems as a sustainable form of alternative energy. Lack of area-specific forecasts for the power output of grid-connected photovoltaic system hinders in tapping the full potential of abundant solar power. The objective of this paper is to estimate the profile of power output of a grid connected 20kWp solar power plant in a reputed manufacturing industry located in Tiruchirappalli, India [10° 44' 42.3816" N, 78° 47' 9.4524" E] using artificial intelligence techniques. An Artificial Neural Network (ANN) based model is proposed as a prediction model in this paper. An experimental database comprising of each day's solar power output and atmospheric temperature (from 31^st May 2014 to 31^st July 2014) has been used for training the ANN. The regression mapping of the neural network was carried out with the Neural Network Fitting Toolbox of MATLAB and simulated with Neuro Solutions development environment. Statistical error analysis in terms of Mean Squared Error (MSE) was calculated on the Day-Ahead Forecasting results and was found to be in the range of 0.019 to 0.025, signifying good accuracy and efficiency. Reliable area-specific solar power production map can provide better utilization of solar energy resource and help in power system management.