Comparative Analysis of Ga and Nsga-Ii for Multi-Objective Energy Management: Cost, Battery Degradation, and Renewable Utilization

This research presents a comparative evaluation of Genetic Algorithm (GA) and Non-dominated Sorting Genetic Algorithm II (NSGA-II) for multi-objective optimization in an energy management system (EMS) that incorporates a photovoltaic (PV) system and battery energy storage. The EMS is designed to minimize total operational cost, reduce battery degradation, and maximize the use of renewable energy. A simulation framework was developed in Python using real-world data for load demand and PV generation, along with a dynamic pricing structure for electricity imported from the grid. GA was applied as a single-objective optimization method using a weighted fitness function, while NSGA-II was implemented as a multi-objective evolutionary algorithm capable of producing a set of Pareto optimal solutions. The algorithms were evaluated across several realistic operational scenarios to assess their effectiveness. Results indicate that NSGA-II consistently achieves better performance in balancing objectives, particularly in reducing cost and improving energy distribution, with only a modest increase in computational complexity. This study demonstrates the potential of multi-objective evolutionary algorithms to improve the efficiency and flexibility of modern EMS applications.