Enhanced battery integrated microgrid energy management system using an optimized approach for robust handling of feeble power and dynamic load variations

Battery-integrated microgrid energy management systems aim to enhance reliability and stability by regulating voltage, balancing renewable generation with dynamic load variations, and mitigating low power generation conditions in hybrid environments. However, uncertainties in renewable energy forecasting can reduce efficiency and increase operational costs. To address these issues, this manuscript proposes an enhanced energy management framework based on a Non-bridged Isolated Positive Cuk converter integrated with a hybrid Granger Causality-Inspired Graph Neural Network and Electric Eel Foraging Optimization (GCGNN-EEFO) approach. The GCGNN model performs accurate short-term forecasting of renewable generation and load demand, while EEFO optimally schedules battery operation and converter switching. The proposed technique was executed and compared with other existing approaches on the MATLAB platform. Results establish that the proposed method achieves an efficiency of 98.78% and reduces operational costs to 0.776$, indicating improved performance and reliable microgrid operation under variable conditions.