Size Optimization of Grid-Tied Hybrid Energy System by Employing Forecasted Meteorological Data

Abstract

Embracing hybrid energy systems (HES) to ensure access to clean, reliable, and cost-effective energy is necessary for nations that are striving for sustainable development. By leveraging precise meteorological data from forecasts, the HES can be rendered more accurate. Thus, firstly, the research presented here employed four machine learning approaches, such as Gaussian process regression (GPR), support vector regression, extreme gradient boosting, and decision trees, to carry out hourly forecasting of meteorological data over a year. The results obtained revealed that the GPR outperformed the other three forecasting models. For this reason, the forecasted meteorological data acquired from GPR is employed in the sizing of the HES. Tunicate swarm algorithm (TSA), a recently developed method, is applied to perform the size optimization of HES capable of meeting the energy necessities at remote sites in the Indian province of Uttar Pradesh. Following a comparative study of TSA, particle swarm optimization, and harmony search, TSA proved to yield a better outcome. Additionally, the simulation result showed a 0.33% cut in the per-unit cost of energy when forecasted data becomes the basis for the optimization of system size.