Microgrid optimization & real-time control via a digital architecture for dynamic orchestration

The integrated management of modern microgrids presents significant challenges arising from the heterogeneous nature of distributed energy resources, the need for real-time adaptive control, and the growing requirement for regulatory compliance, challenges that existing energy management frameworks address only in isolation. This paper presents a novel digital architecture that synergizes cloud computing, a full alternating current optimal power flow formulation, and real-time edge control within a vertical, iterative, and distributed paradigm. The alternating current optimal power flow explicitly incorporates electric vehicle charging via the constant impedance-constant current-constant power load model and heating, ventilation, and air conditioning thermal dynamics, alongside integration of electric vehicles and controllable demand. This allows for the realization of high-level objectives, low-level real-time control and protection strategies, and the dynamic adaptation of control signals to real-time network dynamics. A dynamic re-triggering mechanism ensures the optimization adapts to real-time network events within one 5-minute cycle, while edge-level protection and control functions respond in real-time independently of cloud communication status. Validation through a Hardware-in-the-Loop setup at the Power Networks Demonstration Centre across seven operational scenarios demonstrates 288 consecutive alternating current optimal power flow runs with zero failed solves, a maximum voltage deviation of 2. 3% from nominal, and 100% market dispatch accuracy. The proposed architecture provides a cost-effective alternative to conventional solutions, with edge hardware below 1000 per unit compared to industry-standard costs of 155, 000–470, 000 per MW, underscoring the practical applicability and real-world readiness of the proposed framework.