Research Article

Using LSTM and GRU-Based Deep Learning Approaches: Sustainable Energy Optimization in the Steel Industry

Volume: 2 Number: 1 May 31, 2026
EN TR

Using LSTM and GRU-Based Deep Learning Approaches: Sustainable Energy Optimization in the Steel Industry

Abstract

This study develops and investigates predictive models for energy consumption using deep learning techniques, focusing on a smart small-scale steel manufacturing facility in South Korea. It aims to predict energy consumption (Usage_kWh) to achieve sustainable energy optimization in the steel industry. The dataset used in this study was compiled from daily, monthly, and annual data obtained from the Korea Electric Power Corporation’s website and consists of 35,040 observations and 11 features. Energy consumption data are acquired through IoT-based monitoring systems and utilized for predictive modeling. The dataset incorporates key industrial variables, including lagging and leading current reactive power, lagging and leading current power factor, carbon dioxide (CO₂) emissions, and load type classifications. Various industrial variables were used as the basis for developing different deep learning (Long Short-Term Memory - LSTM / Gated Recurrent Unit - GRU) and machine learning architectures. The models were evaluated comparatively. The LSTM model achieved the highest mean absolute error (MAE) score of 5.75 while the AdaBoostRegressor model achieved an R² score of 0.90%. The results obtained in this study contribute directly to a sustainable green industry by predicting energy consumption in steel industry facilities and ensure the optimization of energy efficiency. This model can be leveraged to support the design of energy-efficient systems, facilitating the optimization of energy consumption and informing policy development in smart city frameworks.

Keywords

References

  1. United Nations, "Transforming our world: The 2030 Agenda for Sustainable Development," UN Publishing, 2015. [Online]. Available: https://sdgs.un.org/goals. [Accessed: April 9, 2026].
  2. European Commission, "The European green deal," 2019. [Online].
  3. International Energy Agency, "Iron and Steel Technology Roadmap," IEA, 2020. [Online]. Available: https://www.iea.org. [Accessed: April 9, 2026].
  4. E. Gholamzadeh and A. Ghaemi, "Optimizing electricity consumption in direct reduction iron processes using RSM, MLP, and RBF models," Scientific Reports, vol. 15, no. 1, p. 34947, 2025.
  5. H. Bai, Y. Chen, H. Bai, et al., "Energy Optimization and Efficiency Improvement Model for Enterprise Production Process Based on Deep Learning Under the Background of Carbon Peak and Carbon Neutrality," International Journal of Computational Intelligence Systems ,vol. 18, p. 169, 2025. doi:10.1007/s44196-025-00901-9
  6. J. Chen, Q. Gong, Z. Cao, M. Liu, M. Xie, and G. Zhao, "Three-layer design and optimization of CO2 emission reduction in the iron and steel industry based on ‘BRL’ industrial metabolism," Energy, vol. 315, p. 134387, 2025. doi:10.1016/j.energy.2025.134387
  7. K. Karthick, R. Dharmaprakash, and S. Sathya, "Predictive modeling of energy consumption in the steel industry using CatBoost regression: A data-driven approach for sustainable energy management," International Journal of Robotics and Control Systems, vol. 4, no. 1, pp. 33-49, 2024.
  8. K. Kerdprasop, N. Kerdprasop, and P. Chuaybamroong, "Deep learning and machine learning models to predict energy consumption in steel industry," International Journal of Machine Learning, vol. 13, no. 4, pp. 142-145, 2023.

Details

Primary Language

English

Subjects

Satisfiability and Optimisation

Journal Section

Research Article

Publication Date

May 31, 2026

Submission Date

April 16, 2026

Acceptance Date

May 3, 2026

Published in Issue

Year 2026 Volume: 2 Number: 1

APA
Özer, Z., & Doğan, A. (2026). Using LSTM and GRU-Based Deep Learning Approaches: Sustainable Energy Optimization in the Steel Industry. Innovative Artificial Intelligence, 2(1), 1-9. https://izlik.org/JA53RB83LT
AMA
1.Özer Z, Doğan A. Using LSTM and GRU-Based Deep Learning Approaches: Sustainable Energy Optimization in the Steel Industry. INNAI. 2026;2(1):1-9. https://izlik.org/JA53RB83LT
Chicago
Özer, Zeynep, and Alican Doğan. 2026. “Using LSTM and GRU-Based Deep Learning Approaches: Sustainable Energy Optimization in the Steel Industry”. Innovative Artificial Intelligence 2 (1): 1-9. https://izlik.org/JA53RB83LT.
EndNote
Özer Z, Doğan A (May 1, 2026) Using LSTM and GRU-Based Deep Learning Approaches: Sustainable Energy Optimization in the Steel Industry. Innovative Artificial Intelligence 2 1 1–9.
IEEE
[1]Z. Özer and A. Doğan, “Using LSTM and GRU-Based Deep Learning Approaches: Sustainable Energy Optimization in the Steel Industry”, INNAI, vol. 2, no. 1, pp. 1–9, May 2026, [Online]. Available: https://izlik.org/JA53RB83LT
ISNAD
Özer, Zeynep - Doğan, Alican. “Using LSTM and GRU-Based Deep Learning Approaches: Sustainable Energy Optimization in the Steel Industry”. Innovative Artificial Intelligence 2/1 (May 1, 2026): 1-9. https://izlik.org/JA53RB83LT.
JAMA
1.Özer Z, Doğan A. Using LSTM and GRU-Based Deep Learning Approaches: Sustainable Energy Optimization in the Steel Industry. INNAI. 2026;2:1–9.
MLA
Özer, Zeynep, and Alican Doğan. “Using LSTM and GRU-Based Deep Learning Approaches: Sustainable Energy Optimization in the Steel Industry”. Innovative Artificial Intelligence, vol. 2, no. 1, May 2026, pp. 1-9, https://izlik.org/JA53RB83LT.
Vancouver
1.Zeynep Özer, Alican Doğan. Using LSTM and GRU-Based Deep Learning Approaches: Sustainable Energy Optimization in the Steel Industry. INNAI [Internet]. 2026 May 1;2(1):1-9. Available from: https://izlik.org/JA53RB83LT