Optimized control of industrial microgrid energy storage using particle swarm techniques

Renewable energy sources have rapidly developed over the past few years. The stochastic nature of the generated energy in photovoltaic systems (PV) and wind power plants is causing more interest in energy storage systems (ESS). In commercial installations, deterministic methods are used to control the power of the storage, which is not efficient. Developing algorithms that optimize economic and technical aspects is necessary. Methods based on computational intelligence (CI) can be a solution. The paper presents a novel CI algorithm for optimizing power flow in microgrids using the particle swarm optimization method. The economic and technical efficiency of control is achieved by combining multi-ple criteria in the objective function. The solution is universal, scalable, and can be applied to any industrial or residential microgrid. The method uses short-term forecasts of local generation and load and specifications of ESS, ensuring that technological constraints are maintained. Analyses were conducted for a whole year for a real industrial microgrid. The paper presents the selected results of the study. The efficiency of the proposed algorithm is compared with the results obtained by a deterministic algorithm aimed at maximizing au-toconsumption. Using the PSO algorithm resulted in an economic effect of €6 635 with 461 full discharge cycles, compared to €2 287 and 110 cycles for the deterministic approach, meaning an increase of more than 2.5 times. However, such storage operation requires more intensive work, affecting its lifetime. Further research can develop objective functions that, without compromising economic effects, support microgrid operation: improving power quality, minimizing voltage fluctuations.