Modelling the Solar Energy Potential in Kerala using Artificial Neural Networks

Abstract: In this work, an artificial neural network (ANN) based model for prediction of solar energy potential in Palakkad, a state in Kerala, was developed. Standard multilayered, feed-forward, back-propagation neural networks with different architecture were designed using python deep learning tool ‘keras’. Meteorological data of Palakkad for period of 5 years (2015 june–2021 may) from NASA Geo-satellite database were used for the training and testing the network. Meteorological data (mean temperature, wind speed, solar irradiation and relative Humidity) were used as inputs to the network, while the solar radiation intensity was used as the output of the network. The results show that the correlation coefficients between the ANN predictions and actual Mean monthly global solar radiation intensities for training and testing datasets were higher than 90%, thus suggesting a high reliability of the model for evaluation of solar radiation in Palakkad. The predicted solar radiation values from the model were given in form of daily maps. The daily mean solar radiation potential in Palakkad ranged from 7.01–5.62 to 5.43–3.54 kW h/m2 day, respectively. A tabulated daily data for future 1 year (2021 may-2022 may) were generated according with the date. The model can be used easily for estimation of solar radiation for preliminary design of solar applications in Palakkad region.