Design of Bacterial Foraging Optimization model with Deep Support Vector Machine for Solar Radiation Prediction using Weather Forecasting Data

Recently, the utilization of solar resources is significantly increased owing to the exponential rise in energy consumption over the globe. In order to proficiently manage the solar resources, a generalized solar radiation (SR) prediction model is necessary to compute the efficiency of the solar system. At the same time, inaccurate SR prediction outcomesmight estimate the load wrongly and result in insufficient energy supply. The recent developments of deep learning (DL) and metaheuristics paves a way for effective SR prediction model. In this aspect, this paper devises a new bacterial foraging optimization (BFO) with deep support vector machine (DSVM) based regression, named BFO-DSVM model for SR prediction. The proposed BFO-DSVM technique involves a three-stage process namely preprocessing, DSVM based prediction, and BFO based parameter optimization.The proposed BFO-DSVM technique involves the design of DSVM model to predict the SR using the weather data. Besides, in order to improve the performance of the DSVM model, the BFO algorithm is used to optimally tune the parameters involved in it. A wide range of simulations takes place to ensure the betterment of the BFO-DSVM model. The simulation results demonstrate the promising performance of the BFO-DSVM model over the recent state of art SR prediction approaches.