Decision-dependent robust optimization for smart microgrid with hydrogen and EV integration

Modern energy systems face escalating complexity from renewable intermittency, hydrogen demand volatility, and bidirectional vehicle-grid interactions. This study proposes a hierarchical robust optimization framework (TRO-DDU) to coordinate microgrids, energy hubs, hydrogen refueling stations, and electric/fuel cell vehicles under decision-dependent uncertainty, including those arising from dynamic integration and varying fleet sizes of vehicles. Unlike conventional methods, TRO-DDU dynamically adjusts uncertainty sets based on operational decisions, enabling proactive risk mitigation across four interdependent layers: hydrogen refueling, vehicle-to-grid coordination, hub-level energy conversion, and grid-wide resilience. Leveraging enhanced Benders decomposition with mixed-integer linear and quadratic programming, the framework resolves multi-stage trade-offs between cost efficiency and worst-case robustness. Case studies on 69- and 118-bus renewable microgrids demonstrate significant cost reductions (up to 9.7%) and resilience improvements (12–15%) over existing approaches, with 92% hydrogen tank utilization and 87% renewable energy penetration under extreme storage constraints. The model’s scalability and adaptability highlight its potential for designing resilient, low-carbon energy networks amid accelerating electrification and hydrogen mobility transitions.