TU Darmstadt / ULB / TUprints

Deep feature learning of in-cylinder flow fields to analyze cycle-to-cycle variations in an SI engine

Dreher, Daniel ; Schmidt, Marius ; Welch, Cooper ; Ourza, Sara ; Zündorf, Samuel ; Maucher, Johannes ; Peters, Steven ; Dreizler, Andreas ; Böhm, Benjamin ; Hanuschkin, Alexander (2023)
Deep feature learning of in-cylinder flow fields to analyze cycle-to-cycle variations in an SI engine.
In: International Journal of Engine Research, 2021, 22 (11)
doi: 10.26083/tuprints-00020179
Article, Secondary publication, Publisher's Version

[img] Text
Copyright Information: In Copyright.

Download (4MB)
Item Type: Article
Type of entry: Secondary publication
Title: Deep feature learning of in-cylinder flow fields to analyze cycle-to-cycle variations in an SI engine
Language: English
Date: 28 November 2023
Place of Publication: Darmstadt
Year of primary publication: November 2021
Place of primary publication: London
Publisher: SAGE Publications
Journal or Publication Title: International Journal of Engine Research
Volume of the journal: 22
Issue Number: 11
DOI: 10.26083/tuprints-00020179
Corresponding Links:
Origin: Secondary publication DeepGreen

Machine learning (ML) models based on a large data set of in-cylinder flow fields of an IC engine obtained by high-speed particle image velocimetry allow the identification of relevant flow structures underlying cycle-to-cycle variations of engine performance. To this end, deep feature learning is employed to train ML models that predict cycles of high and low in-cylinder maximum pressure. Deep convolutional autoencoders are self-supervised-trained to encode flow field features in low dimensional latent space. Without the limitations ascribable to manual feature engineering, ML models based on these learned features are able to classify high energy cycles already from the flow field during late intake and the compression stroke as early as 290 crank angle degrees before top dead center (-290° CA) with a mean accuracy above chance level. The prediction accuracy from -290° CA to -10° CA is comparable to baseline ML approaches utilizing an extensive set of engineered features. Relevant flow structures in the compression stroke are revealed by feature analysis of ML models and are interpreted using conditional averaged flow quantities. This analysis unveils the importance of the horizontal velocity component of in-cylinder flows in predicting engine performance. Combining deep learning and conventional flow analysis techniques promises to be a powerful tool for ultimately revealing high-level flow features relevant to the prediction of cycle-to-cycle variations and further engine optimization.

Uncontrolled Keywords: Deep learning, machine learning, feature analysis, particle image velocimetry, in-cylinder flow, cycle-to-cycle variations, IC engine
Status: Publisher's Version
URN: urn:nbn:de:tuda-tuprints-201798
Classification DDC: 000 Generalities, computers, information > 004 Computer science
600 Technology, medicine, applied sciences > 620 Engineering and machine engineering
Divisions: 16 Department of Mechanical Engineering > Institute of Reactive Flows and Diagnostics (RSM)
Date Deposited: 28 Nov 2023 10:36
Last Modified: 01 Dec 2023 11:02
SWORD Depositor: Deep Green
URI: https://tuprints.ulb.tu-darmstadt.de/id/eprint/20179
PPN: 513585923
Actions (login required)
View Item View Item