Manev, Nikola and Dimitrovski, Dame and Nikolov, Elenior (2023) Diesel Engine Modelling with the Use of Artificial Neural Networks to Decrease Simulation Processing Requirements. In: 26th Conference on Process Integration, Modelling and Optimisation for Energy Saving and Pollution Reduction, 8-11 Oct 2023, Thessaloniki, Greece.
![[thumbnail of 0197.pdf]](/style/images/fileicons/text.png) |
Text
0197.pdf - Published Version Download (562kB) |
Abstract
As the complexities of the internal combustion engine increase, to optimize engine performance and reduce emissions levels, automotive industries look to shorten engine calibration and modeling time. The following paper proposes a diesel engine model that makes use of a pre-built thermodynamic mean-value engine model (MVEM) that integrates novel regression methods and foremost Artificial Neural Networks (ANNs) to capture the inter-dependence of engine control with the output engine operating characteristics and output emission, making it particularly useful in new engine calibration. The primary benefits of such an integration are maintaining sufficient accuracy characteristic of more complex engine models (thermodynamic, wave-action, or even computational fluid dynamics models) while significantly reducing the simulation processing requirements, which adds to the streamlining of the final stages of the new engine design process. The results from this paper point to the fact that ANNs provide the best that fit for engine data compared to older, better-accepted regression methods such as radial basis function (RBF) or polynomial regression while increasing model response time. Since ANNs can be successfully integrated with existing thermodynamic models, this opens the major potential to simplify engine modeling by reducing modeling time and building more efficient and environmentally friendly diesel engines.
| Item Type: | Conference or Workshop Item (Paper) |
|---|---|
| Subjects: | Engineering and Technology > Mechanical engineering Engineering and Technology > Other engineering and technologies |
| Divisions: | Military Academy |
| Depositing User: | Nikola Manev |
| Date Deposited: | 11 Dec 2024 11:15 |
| Last Modified: | 11 Dec 2024 11:15 |
| URI: | https://eprints.ugd.edu.mk/id/eprint/35215 |
Actions (login required)
 |
View Item |