Multi-objective optimization of energy storage configuration considering uncertainties using an improved non-dominated sorting genetic algorithm II

To address the coordinated planning problem of renewable energy (RE) and energy storage system (ESS) under high-penetration RE power systems, this work proposes a multi-objective optimization framework based on an improved NSGA-II algorithm (INSGA-II). Which bolsters convergence and preserves population diversity through four enhancement strategies: adaptive parameter adjustment, improved crossover and mutation operators, enhanced polynomial variation, and an elite retention strategy. A multi-objective optimization model is developed to minimize the annualized total cost, RE curtailment rate, and loss of load probability. RE generation uncertainty is modeled via probabilistic distributions from historical data. The performance of the proposed method is evaluated using standard metrics, including hypervolume (HV), inverted generational distance (IGD), and diversity. Comparative case studies demonstrate that INSGA-II achieves improved convergence characteristics and better solution distribution compared with NSGA-II, NSGA-III, multi-objective particle swarm optimization (MOPSO), and strength Pareto evolutionary algorithm 2 (SPEA2). In terms of economic performance, the annualized cost obtained by INSGA-II is 3.45% and 8.26% lower than those of MOPSO and SPEA2, respectively. Moreover, INSGA-II achieves the lowest RE curtailment rate among all compared methods. For a representative configuration (163 MW/274 MWh), INSGA-II can effectively identify configurations with a balanced trade-off among economic performance, RE utilization, and system reliability, providing multiple feasible solutions for decision-makers.