A Closed-Loop Multi-Timescale Energy Management System for V2G-Enabled Commercial Building Microgrids

Abstract

Vehicle-to-grid (V2G) integration in commercial building microgrids (CBMGs) offers a promising path for grid support, economic arbitrage, and resilience enhancement. However, practical implementation is hindered by the optimization–execution gap, where high-level aggregated commands fail to match low-level physical charger capacities and individual battery boundaries, as well as by the lack of sociotechnical coupling under extreme weather events, where vehicle owner range anxiety dominates. To address these challenges, a closed-loop multi-timescale energy management system for V2G-enabled CBMGs under exogenous meteorological conditions is proposed. The framework features an integrated four-layer cyber–physical control architecture connecting macroscopic day-ahead scheduling, receding-horizon model predictive control (MPC), discrete real-time parking slot allocation with hardware safety boundary constraints, and equipment-level power flow execution. To handle extreme events, an exogenous meteorological stress index is formulated to quantify ambient structural hazards and temperature deviations, which are then mapped to owner range anxiety and loss-aversion behaviors using prospect theory. Rather than relying on heuristic rule-switching, the optimizer executes a smooth and continuous transition from normal economic peak-shaving to active pre-disaster energy reservation and load demand survival. The cyber–physical system is validated using high-fidelity simulations under typical summer and winter blizzard scenarios. The results demonstrate that the proposed hierarchical architecture successfully eliminates optimization–execution mismatches and guarantees zero load shedding. Furthermore, sensitivity analyses establish the optimal system configuration with the critical defense tolerance of 0.6 and the baseline anxiety ratio of 4, which successfully resolves the trade-off between premature defensive actions and insufficient energy reserves while considering human behavioral uncertainty.