Optimized energy management in Grid-Connected microgrids leveraging K-means clustering algorithm and Artificial Neural network models

Abstract

The increasing integration of renewable energy sources (RESs) in grid-connected microgrids necessitates advanced energy management strategies to enhance efficiency, reliability, and sustainability. This study proposes an optimized energy management framework leveraging the One-to-One-Based Optimizer (OOBO) for microgrid scheduling, combined with K-means clustering and Artificial Neural Networks (ANNs) for load forecasting. The proposed method dynamically schedules distributed energy resources (DERs), battery energy storage systems (BESS), and diesel generators while minimizing operational costs and carbon emissions. Simulation results demonstrate that the OOBO-based optimization achieves a 20–48% reduction in operational costs and a 25–38% decrease in carbon emissions, outperforming conventional methods such as Particle Swarm Optimization (PSO), Genetic Algorithm (GA), and Differential Evolution (DE). The comparative analysis highlights the superior convergence speed of OOBO, reducing computational time by 30–45%, making it suitable for real-time applications. Furthermore, the study evaluates three scenarios: reliance solely on a diesel generator, optimization without BESS, and optimization with BESS, where BESS integration led to a 38% reduction in emissions compared to diesel generator-only configurations. The novelty of this work lies in the synergistic integration of OOBO, AI-driven forecasting models, and adaptive resource scheduling, ensuring optimal cost savings and energy efficiency. The results confirm the scalability and robustness of the proposed framework, making it a promising solution for future multi-microgrid and multi-energy system applications. These findings provide a strong foundation for sustainable energy transitions, reducing dependence on fossil fuels and enhancing grid stability.