Mehnatkesh, Hossein, Gordon, David et Koch, Charles Robert.
2025.
« Temporal Kolmogorov-Arnold networks for control-oriented modeling of hydrogen/diesel dual-fuel engines ».
In Proceedings of the CSME-CFDSC-CSR 2025 International Congress (Montreal, QC, Canada, May 25-28, 2025)
Coll. « Progress in Canadian Mechanical Engineering », vol. 8.
Prévisualisation |
PDF
332 - Temporal Kolmogorov-Arnold network.pdf - Version publiée Licence d'utilisation : Tous les droits réservés aux détenteurs du droit d'auteur. Télécharger (1MB) | Prévisualisation |
Résumé
Hydrogen/Diesel Dual-Fuel (HDDF) engines are being investigated as a promising solution for reducing carbon dioxide and providing a transition to zero carbon fuels. However, modeling these engines presents challenges due to their nonlinear dynamics and complex interactions among the two fuels. Computational fluid dynamics modeling is often inaccurate across various operating conditions, conventional data-driven approaches frequently face issues related to model size and lack of interpretability, and based on the complexity of the system, there is a lack of accurate physics-based modeling. This study explores Temporal Kolmogorov-Arnold Networks (T-KANs) as an alternative to conventional Machine Learning (ML) models to capture the behavior of HDDF engines. T-KANs offer a structured framework for function approximation that can efficiently learn the underlying dependencies with historical data while ensuring interpretability. A T-KAN model is developed and trained using experimental HDDF engine data, demonstrating its ability to predict the performance metrics of indicated mean effective pressure and oxides of nitrogen emissions. A comparative analysis of seven different models of well-known ML methods highlights the advantages of T-KANs in terms of accuracy, generalization, and computational efficiency. Given its computational efficiency and a coefficient of determination of 0.956, the T-KAN network with a 10-lookback period is suitable for a model-based controller. This configuration effectively utilizes historical data while functioning without reliance on sensor feedback.
| Type de document: | Compte rendu de conférence |
|---|---|
| Éditeurs: | Éditeurs ORCID Hof, Lucas A. NON SPÉCIFIÉ Di Labbio, Giuseppe NON SPÉCIFIÉ Tahan, Antoine NON SPÉCIFIÉ Sanjosé, Marlène NON SPÉCIFIÉ Lalonde, Sébastien NON SPÉCIFIÉ Demarquette, Nicole R. NON SPÉCIFIÉ |
| Date de dépôt: | 18 déc. 2025 15:16 |
| Dernière modification: | 18 déc. 2025 15:16 |
| URI: | https://espace2.etsmtl.ca/id/eprint/32457 |
Actions (Authentification requise)
 |
Dernière vérification avant le dépôt |