INTELLIGENT SYSTEM FOR PREDICTIVE MAINTENANCE OF A COMBINED-CYCLE GAS TURBINE PLANTS BASED ON A UNIVERSAL ALGORITHM OF DIAGNOSTICS AND FORECASTING

Gulyamov Shukhrat Mannapovich; Doshanova Malika Yuldashovna; Shimbergenova Anora Zhengisbay kizi; Avezzhanov Bobur Zhamalovich

Authors

Gulyamov Shukhrat Mannapovich TSTU named after Islam Karimov
Doshanova Malika Yuldashovna TUIT named after Muhammad al-Khwarizmi
Shimbergenova Anora Zhengisbay kizi TUIT named after Muhammad al-Khwarizmi
Avezzhanov Bobur Zhamalovich TUIT named after Muhammad al-Khwarizmi

Keywords:

algorithm, machine learning, combined cycle power plant, condition diagnostics, predictive maintenance, residual life forecasting, repair optimization, intelligent diagnostics

Abstract

This paper examines the problem of intelligent diagnostics, technical condition prediction, and optimization of repair measures for a combined-cycle gas turbine power plant as a complex power engineering system. Combined-cycle gas turbines are highly efficient and widely used in modern energy systems. However, their operation places significant demands on equipment reliability and timely maintenance. Traditional maintenance methods based on scheduled maintenance intervals do not always effectively account for the actual condition of the equipment and its degradation dynamics.
This paper proposes a universal diagnostics and forecasting algorithm based on the analysis of time series of operating parameters, the identification of diagnostic features, and the use of data mining methods. The algorithm includes the stages of data preprocessing, the formation of a feature space, technical condition classification, the prediction of the remaining equipment life, and the optimization of repair time.
The architecture of an intelligent diagnostics and forecasting system, integrated into the monitoring and control system of a combined-cycle gas turbine, has been developed. The system processes monitoring data, evaluates the equipment's technical condition index, predicts the probability of failure, and generates maintenance recommendations.
To test the feasibility of the proposed approach, a computational experiment was conducted on time series of operating parameters of a combined cycle gas turbine. Various equipment degradation scenarios were considered, including increased bearing vibration, changes in temperature conditions, and decreased heat exchange efficiency. The simulation results showed that the proposed algorithm effectively identifies early signs of technical deterioration, predicts the remaining service life of equipment, and optimizes repair time.
The proposed approach can be used in intelligent predictive maintenance systems for power equipment and contributes to improved reliability, safety, and cost-effectiveness of combined cycle gas turbine operation.
Keywords: algorithm, machine learning, combined cycle gas turbine, time series, technical condition diagnostics, predictive maintenance, residual service life prediction, repair optimization, intelligent diagnostics.

References

1. Anping Wan, Fu Gu, Jianhong Chen, Lingjuan Zheng, Philip Hall, Yangjian Ji, Xinjian Gu, Prognostics of gas turbine: A condition-based maintenance approach based on multi-environmental time similarity, Mechanical Systems and Signal Processing, Volume 109, 2018, Pages 150-165, https://doi.org/10.1016/j.ymssp.2018.02.027.

2. Mohammadreza Tahan, Elias Tsoutsanis, Masdi Muhammad, Z.A. Abdul Karim, Performance-based health monitoring, diagnostics and prognostics for condition-based maintenance of gas turbines: A review, Applied Energy, Volume 198, 2017, Pages 122-144, https://doi.org/10.1016/j.apenergy.2017.04.048.

3. Martha A. Zaidan, Rishi Relan, Andrew R. Mills, Robert F. Harrison, Prognostics of gas turbine engine: An integrated approach, Expert Systems with Applications, Volume 42, Issue 22, 2015, Pages 8472-8483, https://doi.org/10.1016/j.eswa.2015.07.003.

4. Juan José Montero Jimenez, Sébastien Schwartz, Rob Vingerhoeds, Bernard Grabot, Michel Salaün, Towards multi-model approaches to predictive maintenance: A systematic literature survey

on diagnostics and prognostics, Journal of Manufacturing Systems, Volume 56, 2020, Pages 539-557, https://doi.org/10.1016/j.jmsy.2020.07.008.

5. De Castro-Cros, M.; Velasco, M.; Angulo, C. Machine-Learning-Based Condition Assessment of Gas Turbines - A Review. Energies 2021, 14, 8468. https://doi.org/10.3390/en14248468.

6. Savolainen, J., Urbani, M. Maintenance optimization for a multi-unit system with digital twin simulation. J Intell Manuf 32, 1953–1973 (2021). https://doi.org/10.1007/s10845-021-01740-z

7. Fu, Y., Zhu, G., Zhu, M. et al. Digital Twin for Integration of Design-Manufacturing-Maintenance: An Overview. Chin. J. Mech. Eng. 35, 80 (2022). https://doi.org/10.1186/s10033-022-00760-x.

8. Dong Zhong, Zhelei Xia, Yian Zhu, Junhua Duan, Overview of predictive maintenance based on digital twin technology, Heliyon, Volume 9, Issue 4, 2023, e14534, https://doi.org/10.1016/j.heliyon.2023.e14534.

9. Xiaofeng LIU, Yingjie CHEN, Liuqi XIONG, Jianhua WANG, Chenshuang LUO, Liming ZHANG, Kehuan WANG, Intelligent fault diagnosis methods toward gas turbine: A review, Chinese Journal of Aeronautics, Volume 37, Issue 4, 2024, Pages 93-120, https://doi.org/10.1016/j.cja.2023.09.024.

10. Anping Wan, D.U. Chenyu, Chen Peng, Khalil AL-Bukhaiti, Predictive modeling of combined cycle power plant performance using a digital twin-based neural ODE approach, Journal of Building Engineering, Volume 96, 2024, 110390, https://doi.org/10.1016/j.jobe.2024.110390.

11. Wei Cheng, Hassaan Ahmad, Lin Gao, Ji Xing, Zelin Nie, Xuefeng Chen, Zhao Xu, Rongyong Zhang, Diagnostics and Prognostics in Power Plants: A systematic review, Reliability Engineering & System Safety, Volume 255, 2025, 110663, https://doi.org/10.1016/j.ress.2024.110663.