An Intelligent Artificial Neural Network-Adaptive Particle Swarm Optimization Framework for Optimal Energy Management of Grid-Connected Renewable Energy Microgrids

The increasing integration of renewable energy resources into electrical power systems has intensified the need for intelligent energy management strategies capable of addressing the variability of photovoltaic generation and electricity demand. This paper presents a hybrid Artificial Neural Network–Adaptive Particle Swarm Optimization (ANN–APSO) framework for optimal energy management of a grid-connected renewable energy microgrid. Historical household electricity demand data were preprocessed and transformed into temporal and statistical features for short-term load forecasting using an Artificial Neural Network. The forecasted demand profile was subsequently utilized by an Adaptive Particle Swarm Optimization algorithm to optimize battery charging and discharging schedules while satisfying power balance and battery State-of-Charge constraints. The proposed framework was implemented in Python and evaluated using forecasting accuracy, operating cost, renewable energy utilization, and grid energy import. The selected ANN architecture achieved superior forecasting performance compared with the persistence benchmark, while the APSO algorithm outperformed conventional Particle Swarm Optimization and a rule-based energy management strategy by reducing grid energy import, improving renewable energy utilization, and producing the lowest average operating cost under the simulated operating conditions. These results demonstrate that integrating predictive load forecasting with adaptive optimization provides an effective and practical approach for intelligent energy management of grid-connected renewable energy microgrids. The proposed framework offers a scalable solution for future smart-grid applications and distributed renewable energy systems.