A multi-objective optimization approach based on the Non-Dominated Sorting Genetic Algorithm II for power coordination in battery energy storage systems for DC distribution network applications

Advancements in power electronics and renewable energy have transformed traditional distribution networks into active systems with distributed energy resources (DERs). DC networks provide an ideal platform for efficient DER integration, specifically Energy Storage Systems (ESS), underscoring the need for optimized coordination strategies. This study addresses the energy coordination problem in ESS within active DC distribution networks. In such networks, ESS must be efficiently managed due to daily demand variability and solar generation fluctuations, with proactive strategies essential to reduce operational costs and energy losses. The proposed methodology uses a multi-objective nonlinear programming approach, resolved through a master-slave framework. In the master stage, the non-dominated genetic algorithm (NSGA-II) is employed to determine ESS charge and discharge actions, while in the slave stage, an hourly power flow method based on successive approximations assesses the objective functions. The methodology was validated on a 33-node DC test system adapted to the generation and demand conditions of Medellin, Colombia. MATLAB simulations demonstrate that NSGA-II outperforms other methods in Pareto front quality, best solution, average response, and standard deviation. Results show that NSGA-II is the optimal choice for minimizing operational costs and losses in DC networks, confirming its applicability and effectiveness within the studied context. This methodology represents an advancement in ESS management for distribution networks, particularly in applications where anticipating demand and generation variability is critical.