Neural-Network Based Energy Management System for Battery-Supercapacitor Hybrid Storage

As the need for reducing carbon dioxide (CO₂) emissions, clean energy solutions such as standalone photovoltaic (PV) system paired with energy storage system offer a solution. Batteries are usually used to store surplus energy due to their high-energy density that can lead to increased stress and reduced lifespan when subjected to sudden changes in irradiation and load. Combining with supercapacitor that has a high-power density can reduce the battery stress and extend the battery lifespan. This paper suggests the implementation of a hybrid energy storage system that integrates battery and supercapacitor managed through a neural network to optimize energy management. Additionally, the paper considers variations in the state of charge (SOC) of both energy storage components. The research findings demonstrate the effectiveness of neural networks in optimizing energy distribution, thus enhancing the efficiency and reliability of hybrid energy storage systems. Furthermore, this approach efficiently regulates the distribution of power between both storage units in a very short timeframe compared to optimization method, typically less than one second.