Multi-Scenario Optimization of Microgrid Energy Management Using GA and PSO

Residential microgrids face growing challenges in balancing energy cost, renewable integration, and battery health under dynamic operating conditions. To address this, a multi-objective energy management strategy is developed using Genetic Algorithm (GA) and Particle Swarm Optimization (PSO) to optimize power dispatch across photovoltaic (PV), battery, and grid resources. Both algorithms are evaluated under four realistic scenarios: normal operation, PV generation drop, load surge, and grid outage. GA consistently outperforms PSO in resilience-focused conditions, achieving up to 10% lower operational cost, 42.9% less battery usage, and 51.1% reduction in unmet load during outages. PSO, while faster, exhibits higher battery cycling and reduced reliability under constraints. These results highlight key trade-offs between robustness and convergence speed, guiding the design of adaptive, scenario-aware EMS strategies for intelligent microgrids.