Adaptive PSO-Optimised Random Forest for Power Output Prediction in Combined Cycle Power Plants
DOI:
https://doi.org/10.61467/2007.1558.2027.v18i1.1473Keywords:
Combined cycle power plant, power output prediction, random forest, adaptive particle swarm optimisation, hyperparameter optimisationAbstract
A hybrid model (RF and PSO) is proposed to predict power generation in combined-cycle power plants. The study was conducted using 9,568 operational records collected over six years, with ambient temperature, atmospheric pressure, relative humidity, and exhaust vacuum as input conditions. Upon examining the dataset for physical consistency and identifying ambient temperature as the primary predictive variable in terms of power output, a correlation of -0.9485 was also obtained. During the training process, the baseline RF with grid search optimization achieved an RMSE of 3.334528 MW. Adaptive-PSO was evaluated by applying linear, exponential, and oscillatory update schemes. The results show that the oscillatory variant had the best cross-validation RMSE of 3.329416 MW, while the linear and exponential variants exhibited the lowest test RMSE of 3.330879 MW. A statistically significant difference was found between the oscillatory PSO variant and the reference model according to the Wilcoxon signed-rank test.
Spanish-language metadata / Metadatos en español
Título en español:
Random Forest optimizado mediante PSO adaptativo para la predicción de la potencia de salida en centrales eléctricas de ciclo combinado
Resumen:
Se propone un modelo híbrido (RF y PSO) para predecir la generación de energía en centrales eléctricas de ciclo combinado. El estudio se llevó a cabo utilizando 9.568 registros operativos recopilados durante seis años, con la temperatura ambiente, la presión atmosférica, la humedad relativa y el vacío de escape como condiciones de entrada. Tras examinar la consistencia física del conjunto de datos e identificar la temperatura ambiente como la principal variable predictora de la potencia de salida, también se obtuvo una correlación de −0,9485. Durante el proceso de entrenamiento, el RF de referencia con optimización mediante búsqueda en cuadrícula alcanzó un RMSE de 3,334528 MW. El PSO adaptativo se evaluó mediante la aplicación de esquemas de actualización lineal, exponencial y oscilatorio. Los resultados muestran que la variante oscilatoria obtuvo el mejor RMSE de validación cruzada, con un valor de 3,329416 MW, mientras que las variantes lineal y exponencial presentaron el menor RMSE de prueba, de 3,330879 MW. De acuerdo con la prueba de rangos con signo de Wilcoxon, se encontró una diferencia estadísticamente significativa entre la variante oscilatoria de PSO y el modelo de referencia.
Palabras Claves:
Central eléctrica de ciclo combinado, predicción de la potencia de salida, Random Forest, optimización adaptativa por enjambre de partículas, optimización de hiperparámetros
Smart citations:
References
Alketbi, S., Nassif, A. B., Eddin, M. A., Shahin, I., & Elnagar, A. (2020). Predicting the power of a combined cycle power plant using machine learning methods. In 2020 International Conference on Communications, Computing, Cybersecurity, and Informatics (CCCI) (pp. 1–5). IEEE. https://doi.org/10.1109/CCCI49893.2020.9256742
Ameen, A. S. H., & Hamarash, I. I. (2025). A machine learning models for predicting power output in gas turbines: A case study from Perdawood Power Station, Erbil, Iraqi Kurdistan. Zanco Journal of Pure and Applied Sciences, 37(5), 115–124. https://doi.org/10.21271/ZJPAS.37.5.9
Arakelyan, E. K., Boldyrev, I. A., Evseev, K. V., & Gorban, Y. A. (2021). Application of machine learning methods for optimizing the technical and economic performance of generating systems. IOP Conference Series: Materials Science and Engineering, 1035(1), 012011. https://doi.org/10.1088/1757-899X/1035/1/012011
Asghar, A., Ratlamwala, T. A. H., Kamal, K., Alkahtani, M., Mohammad, E., & Mathavan, S. (2023). Sustainable operations of a combined cycle power plant using artificial intelligence based power prediction. Heliyon, 9(9), e19562. https://doi.org/10.1016/j.heliyon.2023.e19562
Bishop, C. M. (1995). Neural networks for pattern recognition. Oxford University Press.
Breiman, L. (2001). Random forests. Machine Learning, 45(1), 5–32. https://doi.org/10.1023/A:1010933404324
Dikra, S. H., Gawanmeh, A., & Rabia, K. (2026). Data-driven analysis and optimization of combined cycle power plants using machine learning models. Journal of Engineering and Technological Sciences, 58(2), 202–216. https://doi.org/10.5614/j.eng.technol.sci.2026.58.2.4
Huang, G.-B., Zhu, Q.-Y., & Siew, C.-K. (2006). Extreme learning machine: Theory and applications. Neurocomputing, 70(1–3), 489–501. https://doi.org/10.1016/j.neucom.2005.12.126
Kurker, F. (2026). PSO–fmincon optimization-based explainable ensemble learning approach for reliable estimation of electrical power output in combined cycle power plants. Electric Power Systems Research, 257, 112969. https://doi.org/10.1016/j.epsr.2026.112969
Lorencin, I., Anđelić, N., Mrzljak, V., & Car, Z. (2019). Genetic algorithm approach to design of multi-layer perceptron for combined cycle power plant electrical power output estimation. Energies, 12(22), 4352. https://doi.org/10.3390/en12224352
Rouibah, B., Labdaoui, A., Aggoun, H., & Güler, İ. (2025). Power forecasting using ANN and ELM: A comparative evaluation of machine learning approaches. Mathematical Modelling of Engineering Problems, 12(1), 1–8. https://doi.org/10.18280/mmep.120101
Saleel, C. A. (2021). Forecasting the energy output from a combined cycle thermal power plant using deep learning models. Case Studies in Thermal Engineering, 28, 101693. https://doi.org/10.1016/j.csite.2021.101693
Santarisi, N. S., & Faouri, S. S. (2021). Prediction of combined cycle power plant electrical output power using machine learning regression algorithms. Eastern-European Journal of Enterprise Technologies, 6(8 [114]), 16–26. https://doi.org/10.15587/1729-4061.2021.245663
Sharma, H., Marinovici, L., Adetola, V., & Schaef, H. T. (2023). Data-driven modeling of power generation for a coal power plant under cycling. Energy and AI, 11, 100214. https://doi.org/10.1016/j.egyai.2022.100214
Song, Y., Park, J., Suh, M.-S., & Kim, C. (2024). Prediction of full-load electrical power output of combined cycle power plant using a super learner ensemble. Applied Sciences, 14(24), 11638. https://doi.org/10.3390/app142411638
Sun, L., Liu, T., Xie, Y., Zhang, D., & Xia, X. (2021). Real-time power prediction approach for turbine using deep learning techniques. Energy, 233, 121130. https://doi.org/10.1016/j.energy.2021.121130
Tüfekci, P., & Kaya, H. (2014). Combined cycle power plant [Data set]. UCI Machine Learning Repository. https://doi.org/10.24432/C5002N
Wang, J., Lu, S., Wang, S.-H., & Zhang, Y.-D. (2022). A review on extreme learning machine. Multimedia Tools and Applications, 81(29), 41611–41660. https://doi.org/10.1007/s11042-021-11007-7
Zhan, Z.-H., Zhang, J., Li, Y., & Chung, H. S.-H. (2009). Adaptive particle swarm optimization. IEEE Transactions on Systems, Man, and Cybernetics, Part B (Cybernetics), 39(6), 1362–1381. https://doi.org/10.1109/TSMCB.2009.2015956
Zhao, Y., & Foong, L. K. (2022). Predicting electrical power output of combined cycle power plants using a novel artificial neural network optimized by electrostatic discharge algorithm. Measurement, 198, 111405. https://doi.org/10.1016/j.measurement.2022.111405
Downloads
Published
How to Cite
Issue
Section
License
Copyright (c) 2026 International Journal of Combinatorial Optimization Problems and Informatics

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.