1. MMC
  2. All Volumes and Issues
  3. Volume 12, Number 4, 2025
  4. Analysis and Prediction of Hydrogen Yield Using Artificial Neural Networks

Analysis and Prediction of Hydrogen Yield Using Artificial Neural Networks

MMC.
2025;
Volume 12, Number 4
: pp. 1313–1319
Received: May 05, 2025
Revised: December 05, 2025
Accepted: December 06, 2025

Kouraa N., Abderafi S., Abdelouahed L.  Analysis and Prediction of Hydrogen Yield Using Artificial Neural Networks.  Mathematical Modeling and Computing. Vol. 12, No. 4, pp. 1313–1319 (2025)

Authors:
  1. N. Kouraa
  2. S. Abderafi
  3. L. Abdelouahed
1
Mohammadia School of Engineers, Mohammed V University in Rabat, Rabat, Morocco
2
Mohammadia School of Engineers, Mohammed V University in Rabat, Rabat, Morocco
3
Normandie Univ, INSA Rouen, UNIROUEN, Rouen, France

This study aims to design a reliable prediction model for H$_2$ production by steam methane reforming using artificial neural networks (ANN).  To achieve this, experimental data related to this process were first analyzed and then used to develop an ANN model implemented in Python.  Model construction was carried out in three stages: network architecture design, training, and performance evaluation.  Data analysis revealed that temperature has the dominant influence on the H$_2$ yield, followed by space time.  The resulting ANN model, with an optimal (2–5–1) architecture, is capable of predicting hydrogen yield with a mean relative error (MRE) of 8.91%.  This model provides satisfactory prediction accuracy and can be used to control the Ni/Al$_2$O$_3$-catalyzed steam methane reforming process for hydrogen production.

data analysis
artificial neural network
steam methane reforming
hydrogen yield
temperature
space time
  1. Pereira V. G. F., Rodrigues C. P., Toniolo F. S.  Ni/Al$_2$O$_3$ supported on cordierite monoliths for methane steam reforming: Influence of catalyst coating methodology.  Catalysis Communic.  183, 106759 (2023).
  2. Omata K., Endo Y., Ishii H., Masuda A., Yamada M.  Effective additives of Ni/$\alpha$-Al$_2$O$_3$ catalyst at low methane conversion of oxidative reforming for syngas formation.  Applied Catalysis A: General.  351 (1), 54–58 (2008).
  3. Khan F., Khan O., Parvez M., Almujibah H., Pachauri P., Yahya Z., Ahamad T., Yadav A. K., Ağbulut Ü.  Innovative hydrogen production from waste bio-oil via steam methane reforming: An advanced ANN-AHP-k-means modelling approach using extreme machine learning weighted clustering.  International Journal of Hydrogen Energy.  105, 1080–1091 (2025).
  4. Pafili A., Charisiou N. D., Douvartzides S. L., et al.  Recent Progress in the Steam Reforming of Bio-Oil for Hydrogen Production: A Review of Operating Parameters, Catalytic Systems and Technological Innovations.  Catalysts.  11 (12), 1526 (2021).
  5. Lu C., Li S., Lu Z.  Building Energy Prediction Using Artificial Neural Networks: A Literature Survey.  Energy & Buildings.  262, 111718 (2022).
  6. Saaidi A., Bichri A., Abderafi S.  Efficient machine learning model to predict dynamic viscosity in phosphoric acid production.  Results in Engineering.  18, 101024 (2023).
  7. Yu X., Shen Y., Guan Z., Zhang D., Tang Z., Li W.  Multi-Objective Optimization of ANN-Based PSA Model for Hydrogen Purification from Steam-Methane Reforming Gas.  International Journal of Hydrogen Energy.  46 (21), 11740–11755 (2021).
  8. Belmajdoub F., Abderafi S.  Efficient machine learning model to predict fineness, in a vertical raw meal of Morocco cement plant.  Results in Engineering.  17, 100833 (2023).
  9. Mehrabian M., Mahmoudimehr J.  A correlation for optimal steam-to-fuel ratio in a biogas-fueled solid oxide fuel cell with internal steam reforming by using Artificial Neural Networks.  Renewable Energy.  219, 119397 (2023).
  10. Jafarizadeh A., Panjepour M., Emami M. D.  Advanced modelling and optimization of steam methane reforming: From CFD simulation to machine learning – Driven optimization.  International Journal of Hydrogen Energy.  96, 1262–1280 (2024).
  11. Ko Y., Kang J. H., Do H., Kum J., Lee C. H.  Hybrid process using cryogenic and pressure swing adsorption process for CO$_2$ capture and extra  H$_2$ production from a tail gas in a steam methane reforming plant.  Energy Conversion and Management.  328, 119561 (2025).
  12. Szablowski L., Wojcik M., Dybinski O.  Review of steam methane reforming as a method of hydrogen production.  Energy.  316, 134540 (2025).
  13. Sanyal A., Malalasekera W., Bandulasena H., Wijayantha K. G. U.  Review of the production of turquoise hydrogen from methane catalytic decomposition: Optimising reactors for Sustainable Hydrogen production.  International Journal of Hydrogen Energy.  72, 694–715 (2024).
  14. De Cata D., Mazzeo L., Piemonte V., Giaconia A.  Electrified steam methane reforming as efficient pathway for sustainable hydrogen production and industrial decarbonization: A critical review.  International Journal of Hydrogen Energy.  105, 31–44 (2025).
  15. Arregi A., Amutio M., Lopez G., Bilbao J., Olazar M.  Evaluation of thermochemical routes for hydrogen production from biomass: A review.  Energy Conversion and Management.  165, 696–719 (2018).
  16. Arregi A., Lopez G., Amutio M., Barbarias I., Bilbao J., Olazar M.  Hydrogen production from biomass by continuous fast pyrolysis and in-line steam reforming.  RSC Advances.  6, 25975–25985 (2016).
  17. Sooria M., Arezoob B., Dastres R.  Artificial neural networks in supply chain management, a review.  Journal of Economy and Technology.  1, 179–196 (2023).
  18. Jbari Y., Abderafi S.  Liquid Density Prediction of Ethanol/Water, Using Artificial Neural Network.  Biointerface Research in Applied Chemistry.   12 (4), 5625–5637 (2022).
  19. Jbari Y., Abderafi S.  Design and optimization of reverse osmosis unit for vinasse purification of Moroccan distillery.  Computers & Chemical Engineering.  175, 108280 (2023).
  20. Maniatopoulos A., Mitianoudis N.  Learnable Leaky ReLU (LeLeLU): An Alternative Accuracy-Optimized Activation Function.  Information.  12 (12), 513 (2021).
  21. Olden J. D., Joy M. K., Death R. G.  An accurate comparison of methods for quantifying variable importance in artificial neural networks using simulated data.  Ecological Modelling.  178 (3–4), 389–397 (2004).
  22. Xu M., Wong T. C., Chin K. S.  Modeling daily patient arrivals at Emergency Department and quantifying the relative importance of contributing variables using artificial neural network.  Decision Support Systems.  54 (3), 1488–1498 (2013).
  23. Garson G. D.  Interpreting neural-network connection weights.  AI Expert.  6 (4), 47–51 (1991).
  24. Olden J. D., Jackson D. A.  Illuminating the "black box": a randomization approach for understanding variable contributions in artificial neural networks.  Ecological Modelling.  154, 135–150 (2002).