A benchmark study of supervised learning methods for predicting the live steam production of thermal power plants
- Power plant operators increasingly rely on predictive models to diagnose and monitor their systems. Data-driven prediction models are generally simple and can have high precision, making them superior to physics-based or knowledge-based models, especially for complex systems like thermal power plants. However, the accuracy of data-driven predictions depends on (1) the quality of the dataset, (2) a suitable selection of sensor signals, and (3) an appropriate selection of the training period. In some instances, redundancies and irrelevant sensors may even reduce the prediction quality. We investigate ideal configurations for predicting the live steam production of a solid fuel-burning thermal power plant in the pulp and paper industry for different modes of operation. To this end, we benchmark four machine learning algorithms on two feature sets and two training sets to predict steam production. Our results indicate that with the best possible configuration, a coefficient of determination of R^2 = 0.95 and a mean absolute error of MAE=1.2 t/h with an average steam production of 35.1 t/h is reached. On average, using a dynamic dataset for training lowers MAE by 32% compared to a static dataset for training. A feature set based on expert knowledge lowers MAE by an additional 32 %, compared to a simple feature set representing the fuel inputs. We can conclude that based on the static training set and the basic feature set, machine learning algorithms can identify long-term changes. When using a dynamic dataset the performance parameters of thermal power plants are predicted with high accuracy and allow for detecting short-term problems.
Author: | Gleb Prokhorskii, Elias Eder, Souman Rudra, Markus PreißingerORCiD |
---|---|
DOI: | https://doi.org/10.5281/zenodo.10245218 |
ISBN: | 978-1-912669-63-9 |
Parent Title (English): | 10th Heat Powered Cycles Conference Proceedings. 3-6 September, 2023. Edinburgh, Scotland, UK. |
Publisher: | The University of Edinburgh |
Place of publication: | Edinburgh, Scotland |
Editor: | Roger Riehl, Giulio Santori, Markus Preißinger |
Document Type: | Conference Proceeding |
Language: | English |
Year of publication: | 2023 |
Release Date: | 2024/01/31 |
Tag: | Live steam prediction; Monitoring system; Supervised learning; Thermal power plant |
Number of pages: | 12 |
First Page: | 313 |
Last Page: | 324 |
Organisationseinheit: | Forschung / Forschungszentrum Energie |
Forschung / Josef Ressel Zentrum für Intelligente Thermische Energiesysteme | |
DDC classes: | 600 Technik, Medizin, angewandte Wissenschaften / 600 Technik |
Open Access?: | ja |
Peer review: | wiss. Beitrag, peer-reviewed |
Publicationlist: | Preißinger, Markus |
Eder, Elias | |
Prokhorskii, Gleb | |
Licence (German): | Creative Commons - CC BY - International - Attribution- Namensnennung 4.0 |