English

Abstract

Renewable sources are essential in fulfilling affordable, clean, and sustainable energy requirements. Consistent wind speed forecasting and accurate solar radiation prediction are required to minimize economic losses and enhance the security of power usage. Predicting solar radiation and wind speed is difficult due to their uncertain behaviour. Motivated by these aspects, this study implemented an Internet of Things (IoT) and artificial neural network (ANN) based system for predicting solar radiation and wind speed. Initially, IoT and long-range (LoRa) enabled hardware was developed for obtaining real-time data. In this study, customized hardware was developed with the amalgamation of LoRa, WI-Fi and many sensors. Solar grid and windmill sensor nodes are customized hardware deployed to collect real-time data to create datasets that predict solar radiation and wind speed using wind speed, solar radiation, humidity, temperature, voltage, and current parameters. For the prediction, in this study, we have employed the ANN techniques, i.e., Levenberg Marquardt (LM) and Bayesian regularization (BR) methods. We have found that the proposed ANN framework using the LM algorithm provides accuracy in the prediction compared to the BR algorithm with less root mean square error (RMSE), mean square error (MSE), fast convergence speed, and more significant Correlation coefficient (R).