Adaptive Multi-objective Real-Time Hierarchical Control for Isolated Microgrid Clusters Utilizing an Enhanced Particle Swarm Optimization Strategy to Optimize Costs and Emissions

Abstract

This paper introduces an adaptive hierarchical control for an isolated microgrid cluster (IMGC) leveraging a realtime multi-objective particle swarm optimization (MOPSO) algorithm. It simultaneously considers CO2 emissions minimization as a tertiary control objective and total losses minimization as a primary control objective, integrating grid-supporting and grid-feeding inverters for MG interconnection. The effectiveness of the MOPSObased hierarchical control is demonstrated across multiple scenarios. Compared to a hierarchical control based on proportional power distribution relative to the rated inverter capacities of the MGs, the proposed method shows a 27.21% reduction in total losses and a 7.66% reduction in CO2 emissions. When compared with an optimization based on the fmincon solver, the proposed approach achieves a 22.92% reduction in losses and a 3.5% decrease in emissions. Additionally, centralized secondary control improves MRE indices by 100.09%, ITAE by 28.5%, ITSE by 43.78%, IAE by 30.61%, and ITSE by 47.72%, compared to the primary control strategy based on proportional approach. The MOPSO approach demonstrates robustness and flexibility, maintaining stable frequency and voltage within set thresholds during MG failures and sudden demand changes. Finally, the practical feasibility of the proposed approach is verified in a hardware-in-the-loop experimental setup using an OPAL-RT4512 unit and a dSPACE MicroLabBox. The experimental results, utilizing a time step of 50 µs, are consistent with the simulation outcomes, ensuring voltage and frequency control as its rated references.