Optimizing Energy Management in Renewable DC Micro-Grids through Distributed Charge/Discharge Control Using FLO-MSNN

This work proposes a novel control technique to ensure stable operation of direct current (DC) microgrids powered by renewable energy in various operating modes. The proposed control framework combines Frilled Lizard Optimization (FLO) with Mix Style Neural Network (MSNN) for adaptive control of charge/discharge cycle and energy availability prediction. FLO algorithm ensures fast and adaptive response by optimizing power flow under nonlinear and unstable operating conditions. Simultaneously, MSNN improves robustness in predicting learning relationship between voltage, current and irradiance patterns. The aim of the proposed framework is to minimize the operational cost and improve system efficiency. The proposed method is simulated utilizing MATLAB and its performance is benchmarked against hybrid models such as cuckoo search–recurrent neural network (CS-RNN) and Ant Colony Optimization-Feed forward Neural Network (ACO-FNN) with cost reduction upto 28.57% and efficiency improvement upto 8.89% in comparison to these hybrid methods. The proposed framework also had the lowest voltage deviation of 0.64%. These findings highlight the reliability, adaptiveness and cost-effectiveness of the proposed energy management system in renewable-based DC microgrids.