Welding parameter optimization based on Gaussian process regression Bayesian optimization algorithm

Dillon Sterling, Tyler Sterling, Yuming Zhang, Heping Chen

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

18 Scopus citations

Abstract

In welding processes, welding parameters have a significant impact on weld quality and mechanical properties of welded joints. For example, if the welding current is not tuned properly, the welding arc becomes unstable which will cause an unacceptable weld. Therefore welding parameters must be optimized in order to achieve best weld quality. However current methods have many limitations in exploring optimal welding parameters. In this paper, Gaussian Process Regression is applied to model the relationship between the welding performance indices and welding parameters. Bayesian Optimization Algorithm is adopted to balance the modeling and optimization processes and optimize welding parameters. Experiments were performed for the Gas tungsten arc welding (GTAW) process and the results demonstrate the effectiveness of the proposed algorithm. Compared to the existing methods, the proposed method greatly improves the welding parameter optimization process; moreover it can be applied with fewer experiments compared with existing methods which will reduce the testing cost and effort.

Original languageEnglish
Title of host publication2015 IEEE Conference on Automation Science and Engineering
Subtitle of host publicationAutomation for a Sustainable Future, CASE 2015
Pages1490-1496
Number of pages7
ISBN (Electronic)9781467381833
DOIs
StatePublished - Oct 7 2015
Event11th IEEE International Conference on Automation Science and Engineering, CASE 2015 - Gothenburg, Sweden
Duration: Aug 24 2015Aug 28 2015

Publication series

NameIEEE International Conference on Automation Science and Engineering
Volume2015-October
ISSN (Print)2161-8070
ISSN (Electronic)2161-8089

Conference

Conference11th IEEE International Conference on Automation Science and Engineering, CASE 2015
Country/TerritorySweden
CityGothenburg
Period8/24/158/28/15

ASJC Scopus subject areas

  • Control and Systems Engineering
  • Electrical and Electronic Engineering

Fingerprint

Dive into the research topics of 'Welding parameter optimization based on Gaussian process regression Bayesian optimization algorithm'. Together they form a unique fingerprint.

Cite this