Day Ahead Optimal Scheduling of Smart Community Microgrid Based on Improved NSGA-III

Abstract

Urban Microgrids have complex topologies and operating environments. This leads to extreme volatility and instability when high percentage of renewable energy is integrated into them. To address these problems, we propose a day-ahead optimal scheduling model for the smart community microgrid on the basis of the improved non-dominated sorting genetic algorithm III (NSGA-III). It minimizes the day-ahead operating cost of the microgrid while balancing electricity load. Firstly, we design a wind/solar/hydrogen storage self-consistent energy system inside the community microgrid to realize the efficient utilization of renewable energy. Secondly, we introduce an adaptive mutation-crossover strategy, and propose a fuzzy inference-based NSGA-III algorithm, to solve the slow convergence speed and poor convergence stability of the current multi-objective optimization algorithms. Finally, we design an optimal microgrid scheduling strategy based on typical day. In the day ahead operation cycle of the community microgrid, it maximizes the absorption of wind/solar unit outputs. This improves the operational efficiency of the microgrid and effectively reducing system operating costs. Extensive experiments demonstrate that the proposed model achieves superior performance compared with the original NSGA-III methods. The total daily operating costs of the proposed microgrid is reduced by 4.5%, and the renewable energy consumption rates is increased by 8.1%.