Resilient Constraint Energy Management for Microgrids: Integrating Wasserstein DRO and CVaR-Constrained MPC Under Renewable Uncertainty

The integration of high levels of renewable energy into microgrids introduces significant volatility. This demands energy management systems (EMS) that balance economic efficiency with resilience. This work proposes a Resilient Constraint Energy Management System (RCEMS) that combines Wasserstein Distributionally Robust Optimization (DRO) and Conditional Value-at-Risk (CVaR)-constrained Model Predictive Control (MPC). The proposed framework helps communities achieve energy independence, reduce carbon footprints, and enhance climate resilience. The proposed two-layer architecture handles non-Gaussian uncertainties in renewable generation and load demand. The day-ahead layer uses Wasserstein DRO for scheduling, while the real-time layer employs risk-aware MPC with CVaR constraints to ensure probabilistic resilience. The proposed RCEMS is validate on a modified IEEE 34-node microgrid with solar PV, wind, and storage. Results show that the proposed method reduces daily operational costs by 12–18% compared to reinforcement learning (RL) and stochastic EMS methods. It also maintains voltage stability within 0.95–1.05 pu during extreme conditions. Key resilience metrics include a 98% islanding success rate which is 8–13% higher than benchmarks and only 2% state-of-charge violations compared to 28% in RL-based approaches. The system demonstrates Pareto-optimal performance, balancing costs i.e., $500–550/day and resilience with 95–98% success across 100 Monte Carlo trials. Tests under hurricanes, cyberattacks, and multi-contingency outages confirm the framework’s adaptability to spatio-temporal uncertainties and its ability to take rapid corrective actions using Phasor Measurement Unit (PMU) feedback. The results highlight RCEMS as a scalable solution for microgrids with high renewable penetration, offering robust trade-offs between cost efficiency and survivability. The framework does not rely on restrictive parametric assumptions, thus providing a foundation for climate-resilient and sustainable energy infrastructures.