Predicting Full Load, Partial Load Efficiency of a Combined Cycle Power Plant Using Machine Learning Methods

Kaewprapha, Phisan; Prempaneerach, Pradya; Singh, Vishakha; Tinikul, Trilot; Intarangsi, Nutty; Kijkanjanarat, Taweesak

doi:10.1109/icccs55155.2022.9846118

Cited by 3 publications

(1 citation statement)

References 4 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Studi ini menyimpulkan bahwa integrasi machine learning dapat membantu pengelola pembangkit dalam mengidentifikasi pola-pola kompleks dalam data lingkungan, sehingga memungkinkan pengoptimalan operasional yang lebih cepat. dalam penelitian [9] menyoroti peran penting machine learning dalam prediksi efisiensi energi dan pengoperasian optimal pada pembangkit listrik.…”

Section: Pendahuluanunclassified

Prediksi Daya Listrik Pada Pembangkit Listrik Siklus Gabungan Berdasarkan Kondisi Lingkungan Menggunakan Metode Machine Learning

Hendra

2023

EJI

View full text Add to dashboard Cite

The utilization of machine learning methods in energy simulation enables the optimization of energy use and improves energy efficiency. In this research, the modeling of predicting power output was conducted under full load conditions in a Combined Cycle Power Plant (CCPP) based on the surrounding environmental conditions. Historical data of CCPP operation were used to model and predict power output under various environmental conditions. In this study, four machine learning algorithms, namely Linear Regression (LR), Decision Tree (DT), Random Forest (RF), and Artificial Neural Network (ANN), were compared and evaluated for their performance. The evaluation metrics used to measure the model performance were Mean Absolute Error (MAE), Root Mean Square Error (RMSE), and R-Squared. The research results indicate that the Random Forest (RF) model achieved the best performance compared to other models with MAE of 2.314, RMSE of 3.372, and R-squared of 0.961. Additionally, the RF model also performed the best compared to other models in external testing with new data, where RF obtained values of MAE 2.579, RMSE 3.315, and R-squared 0.957. These results are consistent with the previous testing, indicating that RF has stable and reliable performance in predicting larger and more diverse datasets. This research contributes to understanding the potential application of machine learning in the power generation industry, especially in CCPP.

show abstract

Section: Pendahuluanunclassified