Short-term Power Prediction using ANN

An accurate prediction of solar energy becomes imperative for the planning and optimization of solar-based energy systems. The present research involves the implementation of Artificial Neural Network (ANN) models employing a cascade forward backpropagation algorithm for predicting short-term PV power using meteorological parameters based on distinct weather conditions. Prediction of solar energy during clear weather is easily done; however, the challenge lies in prediction under cloudy weather conditions. Therefore, the present work involves the prediction of power in solar PV systems for clear, hazy, partly and fully cloudy weather in composite climatic zone. Models are developed by simulating in MATLAB platform and for validating the accuracy of the results, statistical evaluation indices are used. The model can be used easily for predicting power for the preliminary design of solar-based applications.