A comparison of different metaheuristic optimization algorithms on hydrogen storage-based microgrid sizing

Abstract

Microgrids (MGs) with a high penetration of renewable energy are becoming increasingly popular, mainly due to the need for a sustainable and environmentally friendly power system. However, the stochastic characteristic of renewable energy sources makes it a considerable challenge when designing a microgrid. Appropriate installation of energy storage systems (ESSs) such as battery and hydrogen storage systems are needed to counter the intermittent nature of energy sources. This study presents a comparison and evaluation of eight different metaheuristic approaches for optimizing the size of a hydrogen storage-based microgrid, with the aims of minimizing the microgrid’s cost and ensuring the ability to regulate the energy flow within the system. In addition, the optimization algorithm considers the power of the photovoltaic (PV) system, electrolyzer, fuel cell, and the capacity of the battery and hydrogen tank as decision variables. Results of numerical simulations proved that, under the above problem framework, the particle swarm optimization algorithm outperforms the rest. The algorithm is able to produce an optimized microgrid with a 25.3% lower annual system cost compared to the worst-performing algorithm. Its ability to escape the local optimum solution is also showcased.