Multiobjective model-based optimization of diesel injection rate profile by machine learning methods

Eero Immonen; Mika Lauren; Lassi Roininen; Simo Särkkä

doi:10.1109/SysCon47679.2020.9349028

Multiobjective model-based optimization of diesel injection rate profile by machine learning methods

Eero Immonen, Mika Lauren, Lassi Roininen, Simo Särkkä

Research output: Chapter in Book/Report/Conference proceeding › Conference article in proceedings › Scientific › peer-review

2 Citations (Scopus)

Abstract

The contribution of this article is to present a model-based machine learning methodology for automatic and simultaneous optimization of the power output and exhaust emissions of diesel internal combustion (IC) engines. We carry out parametric optimization of the rate profile at which fuel is injected into the cylinder for producing minimal nitrogen oxide (NOx) emissions and maximal cylinder power (nIMEP) output, on a computational simulation model of an Agco Power 44 AWI engine calibrated by measurements. Our results display the tradeoffs in reaching these two contradictory optimization objectives on the Pareto frontiers. We show that the so-called boot injection profile, which is commonly used in practice, also emerges through mathematical optimization as a reasonable compromise of the objectives.

Original language	English
Title of host publication	Proceedings of 14th Annual IEEE International Systems Conference, SYSCON 2020
Publisher	IEEE
Number of pages	6
ISBN (Electronic)	978-1-7281-5365-0
DOIs	https://doi.org/10.1109/SysCon47679.2020.9349028
Publication status	Published - 9 Feb 2021
MoE publication type	A4 Conference publication
Event	IEEE International Systems Conference - Virtual, Montreal, Canada Duration: 24 Aug 2020 → 27 Aug 2020 Conference number: 14

Publication series

Name	Annual IEEE Systems Conference
ISSN (Electronic)	2472-9647

Conference

Conference	IEEE International Systems Conference
Abbreviated title	SYSCON
Country/Territory	Canada
City	Montreal
Period	24/08/2020 → 27/08/2020

Keywords

Diesel engine
Fuel injection
Machine learning
Modeling and simulation
Multiobjective optimization
NOx emissions

Access to Document

10.1109/SysCon47679.2020.9349028

Cite this

@inproceedings{10d9336d13ba4b39995ba41443d5ec5b,

title = "Multiobjective model-based optimization of diesel injection rate profile by machine learning methods",

abstract = "The contribution of this article is to present a model-based machine learning methodology for automatic and simultaneous optimization of the power output and exhaust emissions of diesel internal combustion (IC) engines. We carry out parametric optimization of the rate profile at which fuel is injected into the cylinder for producing minimal nitrogen oxide (NOx) emissions and maximal cylinder power (nIMEP) output, on a computational simulation model of an Agco Power 44 AWI engine calibrated by measurements. Our results display the tradeoffs in reaching these two contradictory optimization objectives on the Pareto frontiers. We show that the so-called boot injection profile, which is commonly used in practice, also emerges through mathematical optimization as a reasonable compromise of the objectives.",

keywords = "Diesel engine, Fuel injection, Machine learning, Modeling and simulation, Multiobjective optimization, NOx emissions",

author = "Eero Immonen and Mika Lauren and Lassi Roininen and Simo S{\"a}rkk{\"a}",

note = "Publisher Copyright: {\textcopyright} 2020 IEEE.; IEEE International Systems Conference, SYSCON ; Conference date: 24-08-2020 Through 27-08-2020",

year = "2021",

month = feb,

day = "9",

doi = "10.1109/SysCon47679.2020.9349028",

language = "English",

series = "Annual IEEE Systems Conference",

publisher = "IEEE",

booktitle = "Proceedings of 14th Annual IEEE International Systems Conference, SYSCON 2020",

address = "United States",

}

Immonen, E, Lauren, M, Roininen, L & Särkkä, S 2021, Multiobjective model-based optimization of diesel injection rate profile by machine learning methods. in Proceedings of 14th Annual IEEE International Systems Conference, SYSCON 2020., 09349028, Annual IEEE Systems Conference, IEEE, IEEE International Systems Conference, Montreal, Canada, 24/08/2020. https://doi.org/10.1109/SysCon47679.2020.9349028

Multiobjective model-based optimization of diesel injection rate profile by machine learning methods. / Immonen, Eero; Lauren, Mika; Roininen, Lassi et al.
Proceedings of 14th Annual IEEE International Systems Conference, SYSCON 2020. IEEE, 2021. 09349028 (Annual IEEE Systems Conference).

Research output: Chapter in Book/Report/Conference proceeding › Conference article in proceedings › Scientific › peer-review

TY - GEN

T1 - Multiobjective model-based optimization of diesel injection rate profile by machine learning methods

AU - Immonen, Eero

AU - Lauren, Mika

AU - Roininen, Lassi

AU - Särkkä, Simo

N1 - Conference code: 14

PY - 2021/2/9

Y1 - 2021/2/9

N2 - The contribution of this article is to present a model-based machine learning methodology for automatic and simultaneous optimization of the power output and exhaust emissions of diesel internal combustion (IC) engines. We carry out parametric optimization of the rate profile at which fuel is injected into the cylinder for producing minimal nitrogen oxide (NOx) emissions and maximal cylinder power (nIMEP) output, on a computational simulation model of an Agco Power 44 AWI engine calibrated by measurements. Our results display the tradeoffs in reaching these two contradictory optimization objectives on the Pareto frontiers. We show that the so-called boot injection profile, which is commonly used in practice, also emerges through mathematical optimization as a reasonable compromise of the objectives.

AB - The contribution of this article is to present a model-based machine learning methodology for automatic and simultaneous optimization of the power output and exhaust emissions of diesel internal combustion (IC) engines. We carry out parametric optimization of the rate profile at which fuel is injected into the cylinder for producing minimal nitrogen oxide (NOx) emissions and maximal cylinder power (nIMEP) output, on a computational simulation model of an Agco Power 44 AWI engine calibrated by measurements. Our results display the tradeoffs in reaching these two contradictory optimization objectives on the Pareto frontiers. We show that the so-called boot injection profile, which is commonly used in practice, also emerges through mathematical optimization as a reasonable compromise of the objectives.

KW - Diesel engine

KW - Fuel injection

KW - Machine learning

KW - Modeling and simulation

KW - Multiobjective optimization

KW - NOx emissions

UR - http://www.scopus.com/inward/record.url?scp=85093366109&partnerID=8YFLogxK

U2 - 10.1109/SysCon47679.2020.9349028

DO - 10.1109/SysCon47679.2020.9349028

M3 - Conference article in proceedings

AN - SCOPUS:85093366109

T3 - Annual IEEE Systems Conference

BT - Proceedings of 14th Annual IEEE International Systems Conference, SYSCON 2020

PB - IEEE

T2 - IEEE International Systems Conference

Y2 - 24 August 2020 through 27 August 2020

ER -

Multiobjective model-based optimization of diesel injection rate profile by machine learning methods

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this