Bi-Level Optimization Framework for Energy Management in Networked Microgrid

The global energy sector is transforming significantly, driven by the need for an efficient and sustainable generation with the goal of net-zero carbon emissions. Microgrids (MGs) have therefore risen to the challenge by facilitating the decentralization of power supply and consumption. This paper proposes a Bi-Level Optimization approach to promote optimal decision-making in Networked Microgrids (NMGs) by leveraging the real-time data processing capabilities of edge-fog devices. Furthermore, the study seeks to optimize both independent MG and NMG, in a bid to achieve cost optimization. To accomplish this, the framework utilizes two optimization strategies: an internal Mixed-Integer Nonlinear Programming (MINLP) optimizer within the microgrids to minimize costs and optimize energy source utilization at the MG level, and an external MINLP optimizer at the fog level to handle energy transactions between the MGs. The simulation analysis on Australian grid data demonstrates that this framework effectively decreases operational expenses by efficiently managing energy use inside and between MGs, improving cost effectiveness by 19.5%.