Comparison and combination of NLPQL and MOGA algorithms for a marine medium-speed diesel engine optimisation

Nao Hu, Peilin Zhou, Jianguo Yang

Research output: Contribution to journalArticle

7 Citations (Scopus)

Abstract

Seven engine design parameters were investigated by use of NLPQL algorithm and MOGA separately and together. Detailed comparisons were made on NOx, soot, SFOC, and also on the design parameters. Results indicate that NLPQL algorithm failed to approach optimal designs while MOGA offered more and better feasible Pareto designs. Then, an optimal design obtained by MOGA which has the trade-off between NOx and soot was set as the starting point of NLPQL algorithm. In this situation, an even better design with lower NOx and soot was approached. Combustion processes of the optimal designs were also disclosed and compared in detail. Late injection and small swirl were reckoned to be the main reasons for reducing NOx. In the end, RSM contour maps were applied in order to gain a better understanding of the sensitivity of import parameters on NOx, soot and SFOC.
LanguageEnglish
Pages138-152
Number of pages15
JournalEnergy Conversion and Management
Volume133
Early online date16 Dec 2016
DOIs
Publication statusPublished - 1 Feb 2017

Fingerprint

Soot
Diesel engines
Engines
Optimal design

Keywords

  • comparison
  • combination
  • NLPQL
  • MOGA
  • RSM

Cite this

@article{683822ac73554eac846c8710b908bc48,
title = "Comparison and combination of NLPQL and MOGA algorithms for a marine medium-speed diesel engine optimisation",
abstract = "Seven engine design parameters were investigated by use of NLPQL algorithm and MOGA separately and together. Detailed comparisons were made on NOx, soot, SFOC, and also on the design parameters. Results indicate that NLPQL algorithm failed to approach optimal designs while MOGA offered more and better feasible Pareto designs. Then, an optimal design obtained by MOGA which has the trade-off between NOx and soot was set as the starting point of NLPQL algorithm. In this situation, an even better design with lower NOx and soot was approached. Combustion processes of the optimal designs were also disclosed and compared in detail. Late injection and small swirl were reckoned to be the main reasons for reducing NOx. In the end, RSM contour maps were applied in order to gain a better understanding of the sensitivity of import parameters on NOx, soot and SFOC.",
keywords = "comparison, combination, NLPQL, MOGA, RSM",
author = "Nao Hu and Peilin Zhou and Jianguo Yang",
year = "2017",
month = "2",
day = "1",
doi = "10.1016/j.enconman.2016.11.066",
language = "English",
volume = "133",
pages = "138--152",
journal = "Energy Conversion and Management",
issn = "0196-8904",

}

Comparison and combination of NLPQL and MOGA algorithms for a marine medium-speed diesel engine optimisation. / Hu, Nao; Zhou, Peilin; Yang, Jianguo.

In: Energy Conversion and Management, Vol. 133, 01.02.2017, p. 138-152.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Comparison and combination of NLPQL and MOGA algorithms for a marine medium-speed diesel engine optimisation

AU - Hu, Nao

AU - Zhou, Peilin

AU - Yang, Jianguo

PY - 2017/2/1

Y1 - 2017/2/1

N2 - Seven engine design parameters were investigated by use of NLPQL algorithm and MOGA separately and together. Detailed comparisons were made on NOx, soot, SFOC, and also on the design parameters. Results indicate that NLPQL algorithm failed to approach optimal designs while MOGA offered more and better feasible Pareto designs. Then, an optimal design obtained by MOGA which has the trade-off between NOx and soot was set as the starting point of NLPQL algorithm. In this situation, an even better design with lower NOx and soot was approached. Combustion processes of the optimal designs were also disclosed and compared in detail. Late injection and small swirl were reckoned to be the main reasons for reducing NOx. In the end, RSM contour maps were applied in order to gain a better understanding of the sensitivity of import parameters on NOx, soot and SFOC.

AB - Seven engine design parameters were investigated by use of NLPQL algorithm and MOGA separately and together. Detailed comparisons were made on NOx, soot, SFOC, and also on the design parameters. Results indicate that NLPQL algorithm failed to approach optimal designs while MOGA offered more and better feasible Pareto designs. Then, an optimal design obtained by MOGA which has the trade-off between NOx and soot was set as the starting point of NLPQL algorithm. In this situation, an even better design with lower NOx and soot was approached. Combustion processes of the optimal designs were also disclosed and compared in detail. Late injection and small swirl were reckoned to be the main reasons for reducing NOx. In the end, RSM contour maps were applied in order to gain a better understanding of the sensitivity of import parameters on NOx, soot and SFOC.

KW - comparison

KW - combination

KW - NLPQL

KW - MOGA

KW - RSM

UR - http://www.sciencedirect.com/science/journal/01968904

U2 - 10.1016/j.enconman.2016.11.066

DO - 10.1016/j.enconman.2016.11.066

M3 - Article

VL - 133

SP - 138

EP - 152

JO - Energy Conversion and Management

T2 - Energy Conversion and Management

JF - Energy Conversion and Management

SN - 0196-8904

ER -