Artificial Intelligence and Optimization Based Output Power Forecasting in Power Plants

Öz

The demand for energy has been rapidly increasing on a global scale due to industrialization and population growth. This increase necessitates making energy production processes more efficient, sustainable, and predictable. Therefore, forecasting models based on artificial intelligence and heuristic optimization techniques have become a crucial component of decision support systems in the energy sector. In this study, a forecasting model based on Particle Swarm Optimization (PSO) was developed, and the hyperparameters of the Long Short-Term Memory (LSTM) model used for forecasting were optimized using PSO. During the training and testing stages, a dataset consisting of operational data from a power plant was utilized. The model's performance was evaluated using statistical error metrics such as the coefficient of determination (R²), root mean square error (RMSE), mean squared error (MSE), and mean absolute error (MAE). The results demonstrate that the proposed PSO-based optimization approach provides high accuracy in energy production forecasting and offers a significant alternative to traditional methods.

Anahtar Kelimeler

• LSTM
• PSO
• POWER PLANTS
• POWER FORECASTING

Enerji Santrallerinde Yapay Zekâ ve Optimizasyon Temelli Çıkış Gücü Kestirimi

Öz

Enerji talebi, sanayileşme ve nüfus artışına bağlı olarak küresel ölçekte hızla artmaktadır. Bu artış, enerji üretim süreçlerinin daha verimli, sürdürülebilir ve öngörülebilir hale getirilmesini zorunlu kılmaktadır. Bu nedenle, özellikle yapay zekâ ve sezgisel optimizasyon temelli tahmin modelleri, enerji alanında karar destek mekanizmalarının önemli bir bileşeni haline gelmiştir.Bu çalışmada, Parçacık Sürü Optimizasyonu (PSO) algoritmasına dayalı bir tahmin modeli geliştirilmiş; ve tahmin için kullanılan Uzun Kısa Süreli Bellek (LSTM) modelinin hiperparametreleri PSO kullanılarak optimize edilmiştir. Modelin eğitim ve test süreçlerinde, bir enerji santraline ait operasyonel verilerden oluşan bir veri seti kullanılmıştır. Modelin performansı; belirleme katsayısı (R²), kök ortalama kare hata (RMSE), ortalama kare hata (MSE) ve ortalama mutlak hata (MAE) gibi istatistiksel hata metrikleri ile değerlendirilmiştir. Sonuçlar, önerilen PSO tabanlı optimizasyon yaklaşımının enerji üretim tahmininde yüksek doğruluk sunduğunu ve geleneksel yöntemlere kıyasla önemli bir alternatif oluşturduğunu göstermektedir.

Anahtar Kelimeler

• LSTM
• PSO
• ENERJİ SANTRALLERİ
• GÜÇ KESTİRİMİ

Kaynakça

1. P. Tüfekci, “Prediction of full load electrical power output of a base load operated combined cycle power plant using machine learning methods,” International Journal of Electrical Power and Energy Systems, vol. 60, pp. 126–140, 2014.
2. Y. S. H. Najjar and A. Abu-Shamleh, “Performance evaluation of a large-scale thermal power plant based on the best industrial practices,” Sci Rep, vol. 10, no. 1, Dec. 2020.
3. N. Isık Delibalta, “Zaman Serileri Tahmininde Melez Bir Yaklaşim,” 2023.
4. C. Korkmaz and İ. Kacar, “Zaman Serisinin Kestirimi İçin Uzun-Kısa Süreli Bellek Ağı Yaklaşımı,” Çukurova Üniversitesi Mühendislik Fakültesi Dergisi, vol. 39, no. 4, pp. 1053–1066, Dec. 2024.
5. Y. M. Zhu, L. Pan, J. Shen, and J. Zhang, “Multi-Step Prediction of Short-Term Electrical Load Using The Multi-Network Model Based on PSO-LSTM Algorithm,” Proceedings - 2022 4th International Conference on Electrical Engineering and Control Technologies, CEECT 2022, pp. 118–123, 2022.
6. A. Pranolo et al., “Exploring LSTM-based Attention Mechanisms with PSO and Grid Search under Different Normalization Techniques for Energy demands Time Series Forecasting,” Knowledge Engineering and Data Science, vol. 7, no. 1, p. 1, Apr. 2024.
7. A. Sharma, A. Sharma, J. K. Pandey, and M. Ram, “Swarm intelligence: Foundation, principles, and engineering applications,” Swarm Intelligence: Foundation, Principles, and Engineering Applications, pp. 1–140, Feb. 2022.
8. V. A. G. Raju, J. Nayak, P. B. Dash, and M. Mishra, “Short-term solar irradiance forecasting model based on hyper-parameter tuned LSTM via chaotic particle swarm optimization algorithm,” Case Studies in Thermal Engineering, vol. 69, p. 105999, May 2025.

9. L. P. Dai, “Performance analysis of deep learning-based electric load forecasting model with particle swarm optimization,” Heliyon, vol. 10, no. 16, p. e35273, Aug. 2024.
10. N. S. Santarisi and S. S. F. Doctor, “Predıctıon Of Combıned Cycle Power Plant Electrıcal Output Power Usıng Machıne Learnıng Regressıon Algorıthms,” Eastern-European Journal of Enterprise Technologies, vol. 6, no. 8(114), pp. 16–26, 2021.
11. M. A. Saeed et al., “Electrical power output prediction of combined cycle power plants using a recurrent neural network optimized by waterwheel plant algorithm,” Front Energy Res, vol. 11, 2023.
12. M. Chen, P. Xu, Z. Liu, F. Liu, H. Zhang, and S. Miao, “Air pollution prediction based on optimized deep learning neural networks: PSO-LSTM,” Atmos Pollut Res, vol. 16, no. 3, p. 102413, Mar. 2025.
13. Y. Yao, L. Han, and J. Wang, “LSTM-PSO: Long Short-Term Memory Ship Motion Prediction Based on Particle Swarm Optimization,” chine, Aug. 2018, pp. 1–5.
14. V. A. Truong, N. S. Dinh, T. L. Duong, N. T. Le, C. D. Truong, and L. T. Nguyen, “Hybrid LSTM-PSO optimization techniques for enhancing wind power bidding efficiency in electricity markets,” Ain Shams Engineering Journal, vol. 16, no. 2, p. 103285, Feb. 2025.
15. A. Bazi, F. Basim Ismail, A. Al-Bazi, R. Hikmat Al-Hadeethi, and D. Singh, “Application of Intelligent Computational Techniques in Power Plants: A Review,” Advances In Industrial Engineering And Management, vol. 10, no. 1, pp. 10–21, Oct. 2021.
16. M. B. Ozdemir, T. Menlik, H. İ. Variyenli, and L. Sevin, “Politeknik Dergisi Journal of Polytechnic,” pp. 971–978, 2017.
17. Y. Song, J. Park, M. S. Suh, and C. Kim, “Prediction of Full-Load Electrical Power Output of Combined Cycle Power Plant Using a Super Learner Ensemble,” Applied Sciences (Switzerland), vol. 14, no. 24, Dec. 2024.
18. M. A. Saeed et al., “Electrical power output prediction of combined cycle power plants using a recurrent neural network optimized by waterwheel plant algorithm,” Front Energy Res, vol. 11, 2023.
19. H. Kaya and P. Tufekci, “Local and GlobalLearning Methods for Predicting Power of a Combined Gas & Steam Turbine,” Mar. 2012.
20. R. Eberhart and J. Kennedy, “A New Optimizer Using Particle Swarm Theory,” 1995.
21. J. Kennedy and R. Eberhart, “Particle swarm optimization,” Proceedings of ICNN’95 - International Conference on Neural Networks, vol. 4, pp. 1942–1948, 1995.
22. M. Jacob, C. Neves, and D. Vukadinović Greetham, “Short Term Load Forecasting,” pp. 15–37, 2020.
23. G. Memarzadeh and F. Keynia, “Short-term electricity load and price forecasting by a new optimal LSTM-NN based prediction algorithm,” Electric Power Systems Research, vol. 192, p. 106995, Mar. 2021.
24. R. Zhou and X. Zhang, “Short-term power load forecasting based on ARIMA-LSTM,” J Phys Conf Ser, vol. 2803, no. 1, 2024.
25. M. Milanova et al., “Automatic Evaluation of Neural Network Training Results,” Computers 2023, Vol. 12, Page 26, vol. 12, no. 2, p. 26, Jan. 2023.