Performance analysis of Hybridized energy storage systems using metaheuristic offline tuning approaches

Designers must evaluate control complexity, energy, precision, robustness, system efficiency, and dependability when choosing a power system digital controller. The examined system is part of isolated DC-Microgrids (DC-MGs) with tiny power-scall clustered hybridized energy storage system (HESS) based Lithium-ion (Li-ion) and Supercapacitor (SC) banks supporting photovoltaic (PV) renewable energy resources (RERs). This research examines HESS performance under pulsing and disturbing load and uncertain PV fluctuations, examining the effects of classical linear PI and nonlinear Super-Twisting-Sliding-Mode-Controller (STSMC) in Dual-closed-loop design. Frequency-decoupling-based current sharing control solved sluggish Li-ion reactions and reduced HESS and load split stress. Trail-and-error tuning of PI and STSMC improves converter response characteristics. Genetic-based algorithm (GA) was applied to improve controllers' responses. Root-Mean-Square-Error (RMSE), Mean-Absolute-Error (MAE), and Integral-Square-Error (ISE) metrics were used to compare controller performance and balance users' requirements with stability objectives. Matlab Simulink simulations showed adequate PI performance despite low performance and tracking. GA-tuned-PI and GA-tuned-STSMC had the fastest rising time of 0.8(sec) compared to primary PI. GA-tuned-STSMC outperformed with super-smooth transients, extra-minimized steady-state chattering of 0.93(%), slight overshoot of 0.38(%) corresponding to 97(%) improvement of primary PI, and the lowest transient peak of 1.0045. (p.u). Thus, GA-PI, STSMC, and GA-STSMC decreased power losses by 34, 36, and 75%, respectively.