optimal selection and efficient utilization of particle swarm optimization methods for designing renewable energy microgrids

In recent years, renewable energy sources have gained significant attention. optimizing small-scale renewable energy systems plays a crucial role in the effective and economical use of these resources. particle swarm optimization (pso) is a popular stochastic optimization method widely applied in various fields. however, standard pso techniques face challenges, including high computational complexity and rapid convergence rates. this study presents a modified pso, comprehensive learning particle swarm optimization (clpso), and generalized pso (gepso) techniques to optimize the capacity sizing of hybrid power generation systems. these systems include photovoltaic (pv), wind, and battery units to supply power to an information and communication technology (ict) center. the research evaluates two scenarios: a standalone system with pv, wind, and battery units, and a grid-connected system with pv and wind units. results demonstrate that the clpso technique significantly reduces overall investment costs compared to standard pso, mpso, and gepso algorithms, by 53.34% and 27.28% for standalone and grid-connected systems, respectively. furthermore, clpso reduces computation time by 57.9% in grid-connected systems and improves energy procurement efficiency, decreasing the required energy purchased from the grid by up to 11.84%. ultimately, clpso outperforms other pso techniques in terms of both precision and efficiency, making it the most suitable method for solving optimization problems in renewable microgrid design.