A Back Propagation Neural Network Model for Predicting Solar Irradiation

A Back Propagation Neural Network Model for Predicting Solar Irradiation

Prabhat Kumar Mishra1, Prof. Sanmati Jain2

Abstract— Artificial intelligence and machine learning have made its presence felt ubiquitously in different avenues of research and technology wherein the data is large and complex. In the proposed work, to forecast solar irradiation energy; whose structure uses the back-propagation algorithm is used. The averaging  window approach is also been used so as to increase the accuracy of prediction. The performance metrics which have been evaluated are the gradient, combination co-efficients, mean squared error and mean absolute percentage error. The system attains a MAPE of 1.77%. Hence the accuracy attained is 98.23%. The mean square error has been chosen as the performance function for the proposed algorithm.

Keywords—Solar Energy Prediction, Back Propagation, Mean Absolute Percentage Error (MAPE), Mean Squared Error (MSE), Regression.