Akademik Veri Yönetim Sistemi
International Journal of Hydrogen Energy, cilt.144, ss.593-610, 2025 (SCI-Expanded)
The production, storage, and conversion of hydrogen into energy, as well as its use in areas such as green ammonia production for agriculture or the catalysis of natural gas, are of significant interest due to their stable structure and source diversity. For this purpose, energy islands (EIs) have been established near consumption points, and renewable energy (RE) obtained from photovoltaic (PV) panels and wind turbines (WTs) is used for green hydrogen production (GHP). In these EIs, hydrogen production from renewable sources shows significant growth, contributing to the power-to-X (P2X) system in terms of storage and flexibility. GHP through renewables will likely become a prominent solution for EIs within the next ten years. One of the bottlenecks here is not to reflect the adverse effects of the variable nature of renewables while transferring sustainable energy. Herein, to enhance operation sustainability, stability, and reliability of renewable-based distributed energy resources (DERs), machine learning (ML)-based techniques can be neat and auxiliary solutions. Power electronic converters (PECs) have such a duty as being the backbone of transferring energy in renewables. The crucial matter is here to keep the inputs and outputs of the converters as stable as possible. In this context, this paper outlines an ML approach to reduce the computational burden and enhance reliability. Therefore, this paper proposes the utilization of an EI to strengthen the stability and reliability of the general scheme. This work is a preliminary attempt and effective solution to establish these EIs, including GHP, considering feasibility criteria and improving reliability. Furthermore, this study examines the essential components, design criteria, challenges, and future issues for system establishment. It aims to facilitate the work of researchers in this field and further enhance the development of EIs.