Network-physical coupling technology applicable to distributed energy microgrids and emergency recovery systems in disaster-prone areas

Island communities connected to mainland grids face severe risks from natural disasters that damage undersea and overhead transmission lines, resulting in prolonged power outages. This study proposes a network-physical coupling framework for distributed energy microgrids to improve resilience and operational efficiency in disaster-prone regions. Multi-source datasets, including IRENA economic data, CAIDA communication metrics, EM-DAT disaster records, and World Bank regional indicators, are integrated, cleaned, normalized, and aligned with load and renewable generation patterns. HOMER Pro optimization is applied to determine optimal photovoltaic and battery configurations by minimizing Net Present Cost (NPC) and Levelized Cost of Energy (LCOE). A graph-based network-physical model represents electrical and communication infrastructures, enabling topology evaluation, fault propagation analysis, and energy management. MQTT and SCADA protocols simulate real-time communication performance under latency and packet loss conditions, while AC/DC power flow analysis ensures grid stability. Agent-based simulations using AnyLogic evaluate system behavior during normal and disaster scenarios. Results demonstrate improved cost-resilience balance, enhanced reliability, and faster recovery of island microgrids.