Solar Power Estimation Methods Using ANN and CA-ANN Models for Hydrogen Production Potential in Mediterranean Region

Abstract

Among renewable energy sources, investments in solar and hydrogen energy as eco-friendly energy sources are continuously expanding. The energy generated by solar photovoltaic (PV) panels, varies during the day. In order to ensure energy continuity, hybrid systems are being developed by integrating solar energy systems with different energy systems. Solar-assisted hydrogen production energy systems have a number of advantages, such as having zero carbon emissions, being able to be used as fuel in fuel cells, and having a high energy density of hydrogen. The sustainable energy obtained from solar energy is used to meet the energy demand in numerous applications due to the high solar energy potential in Mediterranean Region. In this article, solar PV panel output power prediction based on Artificial Neural Network (ANN) and Cultural Algorithm-Artificial Neural Network (CA-ANN) for a solar-assisted hydrogen production energy system have been performed. In this study, the real time data of a 4 MW solar PV power plant is used to estimate solar PV output power. Predicting the output power of PV power plants with high accuracy is of great importance in the realization of energy management, hydrogen production and storage strategies. The numerical results obtained at the end of the study are presented. The results obtained with the ANN and CA-ANN methods are given in detail. In addition, the hydrogen production amount in the Mediterranean region is calculated using MATLAB/Simulink software. The analysis results show that the Mediterranean region has high efficiency hydrogen production potential.

KEYWORDS:

• Artificial neural network
• Cultural algorithm
• Hydrogen energy
• Renewable energy
• solar energy

DISCLOSURE STATEMENT

No potential conflict of interest was reported by the author(s).

Additional information

Notes on contributors

Kübra Tümay Ateş

Kübra Tümay Ateş, after successfully completing her master's degree in Çukurova University, Department of Industrial Engineering, she was enrolled in the PhD program. The dissertation topic she worked on was supported by TUBITAK and she served as a scholar in this project. She still continues to work at Çukurova University. Her current research interests include artificial neural networks, decision making and artificial intelligence. Corresponding author. Email: [email protected]