hh.sePublications
Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
Optimisation of the Hybrid Feature Learning Algorithm for RUL estimation
Halmstad University, School of Information Technology.
Halmstad University, School of Information Technology.
Responsible organisation
2024 (English)Independent thesis Advanced level (degree of Master (Two Years)), 20 credits / 30 HE creditsStudent thesis
Abstract [en]

A B S T R A C TIn recent years, machine learning (ML) algorithms have been used tominimize maintenance costs and identify problems early in the auto-motive sector. The breakdown of a component or equipment impactsthe performance and cost, and hence it is considered a crucial stepin various domains. To fulfill this purpose, different approaches canbe considered to improve the application in real-world problems. Thedetermination of an asset’s residual useful life of a component at aspecific time is known as "remaining useful life" (RUL). The exten-sive evolution of data makes it challenging to analyze and interprethigh-level and valuable features from the data. The issue arises inall disciplines, and the automotive industry is no exception, giventhe large number of sensors to consider. Existing RUL research hasnot given much thought to the influence of high dimensionality dataon component maintenance and deterioration. The fundamental pur-pose of feature selection (FS) is to select a subset of features from thedata without compromising model performance. Lately, the trend ofresearch in feature selection has been based on bio-inspired methods.This work proposes a hybrid approach to the FS problem that com-bines Ant Colony Optimization (ACO) and Particle Swarm Optimiza-tion (PSO). When tested on public datasets, our results demonstrate arise in regression accuracy and a reduction in the number of selectedfeatures. ACO-PSO applies advantage from ACO to handle directlywith nominal attributes and uses advantage from PSO to handle con-tinuous features, utilizing the "the best of both worlds" in a singlealgorithm. This approach is combined with the random forest classi-fier for choosing the relevant and appropriate features. To gain betterresults from the algorithm, the performance of these approaches iscompared and the feature set with the best accuracy has been usedfor regression tasks. The results are evaluated in 5 public domaindatasets as well, and results show an increase in accuracy for the re-gression and the reduction of selected features number.

Place, publisher, year, edition, pages
2024.
Keywords [en]
Remaining Useful Life, Feature Selection, Ant colony optimization, particle swarm optimization, large datasets.
National Category
Engineering and Technology
Identifiers
URN: urn:nbn:se:hh:diva-55148OAI: oai:DiVA.org:hh-55148DiVA, id: diva2:1923400
Educational program
Master's Programme in Embedded and Intelligent Systems, 120 credits
Available from: 2025-01-07 Created: 2024-12-23 Last updated: 2025-10-01

Open Access in DiVA

fulltext(1709 kB)84 downloads
File information
File name FULLTEXT02.pdfFile size 1709 kBChecksum SHA-512
0454174f22c118e373f98b87981b70f77002c5a04b9527b0c6a3789e1edac6febda9f7e09219b8ba5c69f1f7b6d8a43b2b3b6d10beffa47219ad488cf7235a20
Type fulltextMimetype application/pdf

By organisation
School of Information Technology
Engineering and Technology

Search outside of DiVA

GoogleGoogle Scholar
Total: 101 downloads
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

urn-nbn

Altmetric score

urn-nbn
Total: 380 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf