Data-Driven Digital Twin for Gas Turbine Power Output Estimation and Performance Deviation Monitoring: Data-Driven Digital Twin for Gas Turbine

Asaad Saber Ameen; Ibrahim Ismail Hamarash

doi:10.21070/jeeeu.v10i1.1738

Authors

Asaad Saber Ameen Salahaddin University, Engineering College
II Hamarash Department of Electrical Engineering, College of Engineering, Salahaddin University-Erbil, Kurdistan Region, Iraq https://orcid.org/0000-0002-3673-5653

DOI:

https://doi.org/10.21070/jeeeu.v10i1.1738

Keywords:

Digital Twin (DT), Gas Turbine (GT), Machine Learning, Power Output Estimation, Industry 4.0, anomaly detection., Digital Twin (DT), Gas Turbine (GT), Machine Learning, Power Output Estimation, Industry 4.0, anomaly detection

Abstract

Digital twin which creates virtual representation of physical systems using real-time data are increasingly employed for performance monitoring, operational optimization, and decision support in the energy sector in power generation , the electrical power output in gas turbines varies significantly with environmental and operational conditions. Accurate estimation of the expected power output (MW) is indispensable for benchmarking and degradation detection. This study proposes a data-driven digital twin framework for predicting power output and monitoring performance deviation for an operational GE Frame 9E gas turbine power plant. Real-time data is acquired from a Mark VIe control system and archived in the AVEVA PI historian (PI ProcessBook).Four machine learning algorithms, Support Vector Regression (SVR), Random Forest regression, Extreme Gradient Boosting, and a feed-forward Artificial Neural Network (ANN) were implemented and evaluated to find the most suitable model for the digital twin, The XGBoost-based digital twin achieved the best overall accuracy (RMSE = 1.4056 MW and R² = 0.9125) and generated the most stable deviation behavior, supporting performance benchmarking in industrial operation.. The digital twin calculates the deviation (DMW) between actual and expected power output, enabling continuous non-intrusive monitoring of performance shifts. This study provides a framewok for integrating real-time data with machine learning models to deploy the digital twins for condition monitoring in gas turbine operations.

Author Biography

II Hamarash, Department of Electrical Engineering, College of Engineering, Salahaddin University-Erbil, Kurdistan Region, Iraq

Dr. Ibrahim Ismael Hamad (Ibrahim Hamarash) is a Professor of Electrical and Computer Engineering at Salahaddin University-Erbil. He holds a BSc in Electronics and Communication Engineering from Salahaddin University, an MSc in Control Systems Engineering from Mosul University, an MRes in Computer Science from Staffordshire University (UK), and a PhD in Control Systems Engineering from The University of Technology. Additionally, he obtained an ICT project management certificate from UCSD, USA. In 2007, he was a visiting postdoctoral researcher at Monash University, Australia, where he was awarded the Endeavor Medal by the Australian Government for his postdoctoral research in Control Systems. He has also won research grants from NSF (USA), IRIS (USA), delPHE (UK), DAAD and GIZ (Germany), UNESCO, and IEEE.

Dr. Ibrahim is the co-founder of the Mobile Education Centre for Regeneration Energy and Automation (MECREA) in collaboration with Karlsruhe University for Applied Science, Germany. He also established the Erbil Centre for Sustainable Development alongside five professors from international universities to support the KRG. He founded and led the Software Engineering Department at SU from 2000 to 2004 and served as the Head of the Electrical Engineering Department at SU from 2007 to 2011. Furthermore, he was the Vice President for Scientific Affairs and Postgraduate Studies from 2011 to 2017 and the Dean of the Engineering College at SU from 2017 to 2018.

As an author, Dr. Ibrahim wrote the first Computer Dictionary (English-Kurdish) in 1999 and a textbook titled Computers for Year 7 and 8 students in the Kurdistan Region in 2001. He has published over 40 scientific articles and has served as a general chair and keynote speaker at numerous international conferences in Kurdistan and abroad. He designed and supervised numerous software, electrical, and computer projects from 1993 to 2011 in the industrial sector. Dr. Ibrahim has supervised over 40 MSc students and 7 PhD students in Electrical and Computer Engineering and Information Technology. He is a senior member of IEEE.

References

[1] D. P. Francis, S. Lazarova-Molnar, and N. Mohamed, "Towards data-driven digital twins for smart manufacturing," in International Conference On Systems Engineering, 2020, pp. 445-454: Springer.

[2] S. J. A. T. E. Kim, "Prediction-focused machine learning-based visualization of compressor flow in a gas turbine engine digital twin," p. 127154, 2025.

[3] Z. Liu and I. A. J. E. Karimi, "Gas turbine performance prediction via machine learning," vol. 192, p. 116627, 2020.

[4] M. P. Boyce, Gas turbine engineering handbook. Elsevier, 2011.

[5] A. S. H. Ameen, I. I. J. Z. J. o. P. Hamarash, and A. Sciences, "A Machine Learning Models for Predicting Power Output in Gas Turbines: A Case Study from Perdawood Power Station, Erbil, Iraqi Kurdistan," vol. 37, no. 5, pp. 115-124, 2025.

[6] H. Lasi, P. Fettke, H.-G. Kemper, T. Feld, M. J. B. Hoffmann, and i. s. engineering, "Industry 4.0," vol. 6, no. 4, pp. 239-242, 2014.

[7] J. Ullrich, A. G. Voyiatzis, and E. R. Weippl, "Secure cyber-physical production systems: Solid steps towards realization," in 2016 1st International workshop on cyber-physical production systems (CPPS), 2016, pp. 1-4: IEEE.

[8] T. Khan, U. Khan, A. Khan, C. Mollan, I. Morkvenaite-Vilkonciene, and V. J. M. Pandey, "Data-Driven Digital Twin Framework for Predictive Maintenance of Smart Manufacturing Systems," vol. 13, no. 6, p. 481, 2025.

[9] T. H.-J. Uhlemann, C. Lehmann, and R. J. P. C. Steinhilper, "The digital twin: Realizing the cyber-physical production system for industry 4.0," vol. 61, pp. 335-340, 2017.

[10] S. Alketbi, A. B. Nassif, M. A. Eddin, I. Shahin, and A. Elnagar, "Predicting the power of a combined cycle power plant using machine learning methods," in 2020 International conference on communications, computing, cybersecurity, and informatics (CCCI), 2020, pp. 1-5: IEEE.

[11] C. A. Saleel, "Forecasting the energy output from a combined cycle thermal power plant using deep learning models," Case Studies in Thermal Engineering, vol. 28, p. 101693, 2021/12/01/ 2021.

[12] Y. Tunckaya and E. J. J. o. t. E. I. Koklukaya, "Comparative analysis and prediction study for effluent gas emissions in a coal-fired thermal power plant using artificial intelligence and statistical tools," Journal of the Energy Institute, vol. 88, no. 2, pp. 118-125, 2015.

[13] M. A. Saeed et al., "Electrical power output prediction of combined cycle power plants using a recurrent neural network optimized by waterwheel plant algorithm," Frontiers in Energy Research, vol. 11, p. 1234624, 2023.

[14] Q. Yi, H. Xiong, and D. J. E. Wang, "Predicting Power Generation from a Combined Cycle Power Plant Using Transformer Encoders with DNN," vol. 12, no. 11, p. 2431, 2023.

[15] N. Pachauri and C. W. J. I. A. Ahn, "Electrical energy prediction of combined cycle power plant using gradient boosted generalized additive model," vol. 10, pp. 24566-24577, 2022.

[16] A. Asghar, T. A. H. Ratlamwala, K. Kamal, M. Alkahtani, E. Mohammad, and S. J. H. Mathavan, "Sustainable operations of a combined cycle power plant using artificial intelligence based power prediction," vol. 9, no. 9, 2023.

[17] J. L. Lobo, I. Ballesteros, I. Oregi, and J. J. a. p. a. Del Ser, "Real-time Electrical Power Prediction in a Combined Cycle Power Plant," 2019.

[18] N. S. Santarisi and S. S. J. E.-E. J. o. E. T. Faouri, "Prediction of combined cycle power plant electrical output power using machine learning regression algorithms," vol. 6, no. 8, p. 114, 2021.

[19] B. Xu et al., "A case study of digital-twin-modelling analysis on power-plant-performance optimizations," vol. 3, no. 3, pp. 227-234, 2019.

[20] J. Friederich, D. P. Francis, S. Lazarova-Molnar, and N. J. C. i. I. Mohamed, "A framework for data-driven digital twins of smart manufacturing systems," vol. 136, p. 103586, 2022.

[21] J. Lim et al., "The EPRI Gas Turbine Digital Twin–a Platform for Operator Focused Integrated Diagnostics and Performance Forecasting," in Turbo Expo: Power for Land, Sea, and Air, 2021, vol. 84966, p. V004T09A009: American Society of Mechanical Engineers.

[22] Y. Ma, X. Zhu, J. Lu, P. Yang, and J. J. S. Sun, "Construction of data-driven performance digital twin for a real-world gas turbine anomaly detection considering uncertainty," vol. 23, no. 15, p. 6660, 2023.

[23] D. Ariansyah et al., "Digital twin development: a step by step guideline," 2020.

[24] A. Redelinghuys, A. Basson, and K. Kruger, "A six-layer digital twin architecture for a manufacturing cell," in International Workshop on Service Orientation in Holonic and Multi-Agent Manufacturing, 2018, pp. 412-423: Springer.

[25] J. T. Lim, A. Habibullah, and E. Y. K. J. E. Ng, "Towards a digital twin for gas turbines: Thermodynamic modeling, critical parameter estimation, and performance optimization using PINN and PSO," vol. 18, no. 14, p. 3721, 2025.

[26] K. Rawal and A. Ahmad, "A Comparative Analysis of Supervised Machine Learning Algorithms for Electricity Demand Forecasting," in 2022 Second International Conference on Power, Control and Computing Technologies (ICPC2T), 2022, pp. 1-6.

[27] W.-C. Hong, "Hybrid evolutionary algorithms in a SVR-based electric load forecasting model," International Journal of Electrical Power & Energy Systems, vol. 31, no. 7, pp. 409-417, 2009/09/01/ 2009.

[28] L. A. Abad, S. M. Sarabia, J. M. Yuzon, and M. C. Pacis, "A Short-Term Load Forecasting Algorithm Using Support Vector Regression & Artificial Neural Network Method (SVR-ANN)," in 2020 11th IEEE Control and System Graduate Research Colloquium (ICSGRC), 2020, vol. 11th, pp. 138-143: IEEE.

[29] B. Mallala, A. I. U. Ahmed, S. V. Pamidi, M. O. Faruque, and R. J. R. i. E. Reddy, "Forecasting global sustainable energy from renewable sources using random forest algorithm," vol. 25, p. 103789, 2025.

[30] L. Bandić, M. Hasičić, and J. Kevrić, "Prediction of power output for combined cycle power plant using random decision tree algorithms and ANFIS," in International Symposium on Innovative and Interdisciplinary Applications of Advanced Technologies, 2019, pp. 406-416: Springer.

[31] M. J. T. E. P. o. S. T. E. Cosar and Mathematics, "Utilizing the eXtreme gradient boosting algorithm for artificial intelligence-supported learning analytics application," vol. 28, pp. 277-285, 2024.

[32] T. M. Singh and K. Lippert, "Digital Twins in Human Activity Prediction on Gait Using Extreme Gradient Boosting Local Binary Pattern: Healthcare 6.0," in Exploring the Advancements and Future Directions of Digital Twins in Healthcare 6.0: IGI Global, 2024, pp. 240-261.

[33] Z. A. Ali, Z. H. Abduljabbar, H. A. Tahir, A. B. Sallow, and S. M. J. A. J. o. N. U. Almufti, "eXtreme gradient boosting algorithm with machine learning: A review," vol. 12, no. 2, pp. 320-334, 2023.

[34] A. Wankhede and V. Ghate, "Prediction of power of a combine gas and steam turbine using artificial neural network," in 2018 3rd IEEE International Conference on Recent Trends in Electronics, Information & Communication Technology (RTEICT), 2018, pp. 1103-1108: IEEE.

Data-Driven Digital Twin for Gas Turbine Power Output Estimation and Performance Deviation Monitoring