A NARMA-L2 ANN-controlled hybrid storage-based energy management framework for enhanced microgrid operation

This paper presents an advanced energy management system (EMS) for a hybrid microgrid comprising photovoltaic (PV), Fuel cell (FC), battery, supercapacitor (SC), and AC grid sources. The proposed architecture ensures efficient power distribution and system stability through intelligent coordination of multiple energy resources under varying irradiance and load conditions. Two multiport converter topologies are adopted: one connects the PV and FC to a common DC link, and the other integrates battery and SC storage units, enabling flexible and optimized power flow. A NARMA-L2 Artificial Neural Network (ANN)-based controller is implemented to manage power exchange intelligently by predicting and adapting to real-time variations in demand and supply. The EMS operates efficiently in both power surplus and deficit modes, dynamically engaging sources and storage units to ensure continuous load support. Simulation results demonstrate that the system maintains a constant load power of 6 kW despite fluctuations in PV and FC outputs, with the battery and SC compensating during transitions. The coordinated control ensures voltage stability, minimizes conversion stages, and enables quick response to transients. This integrated and intelligent control approach offers a robust and sustainable solution for microgrids and renewable-based EV charging infrastructures.