A material stack-up combination identification method for resistance spot welding based on dynamic resistance. (August 2020)
- Record Type:
- Journal Article
- Title:
- A material stack-up combination identification method for resistance spot welding based on dynamic resistance. (August 2020)
- Main Title:
- A material stack-up combination identification method for resistance spot welding based on dynamic resistance
- Authors:
- Zhou, Lei
Zheng, Wenjia
Li, Tianjian
Zhang, Tianyi
Zhang, Zhongdian
Zhang, Ye
Wu, Zhicheng
Lei, Zhenglong
Wu, Laijun
Zhu, Shiliang - Abstract:
- Highlights: A feasible scheme is proposed to improve the generalization ability of adaptive control model. Material combination identification method based on dynamic resistance was proposed for the first time in this article. A feasible data processing process is proposed, including signal acquisition, dynamic resistance calculation, filtering, and dimensionality reduction. The classification performance of three supervised classification models(support vector machines, logical regression, and random forest) is analyzed in detail. Abstract: Adaptive control of the resistance spot welding process has always been a hot issue in the field of resistance spot welding. Due to the significant difference in weldability of different materials, the basis of adaptive control is the identification of material stack-up (including material types, layer number of metal sheets, the thickness of metal sheet). The material stack-up identification method based on dynamic resistance was proposed for the first time in this article. Ten different types of material stack-up commonly used in the manufacture of automobile body-in-white were selected in the experiment, and a total of 550 dynamic resistance samples were collected. The dynamic resistance value was used as the feature, and the supervised classification algorithms including support vector machines, logical regression, and random forest were used to classify the dynamic resistance, to realize the recognition of material stack-up. TheHighlights: A feasible scheme is proposed to improve the generalization ability of adaptive control model. Material combination identification method based on dynamic resistance was proposed for the first time in this article. A feasible data processing process is proposed, including signal acquisition, dynamic resistance calculation, filtering, and dimensionality reduction. The classification performance of three supervised classification models(support vector machines, logical regression, and random forest) is analyzed in detail. Abstract: Adaptive control of the resistance spot welding process has always been a hot issue in the field of resistance spot welding. Due to the significant difference in weldability of different materials, the basis of adaptive control is the identification of material stack-up (including material types, layer number of metal sheets, the thickness of metal sheet). The material stack-up identification method based on dynamic resistance was proposed for the first time in this article. Ten different types of material stack-up commonly used in the manufacture of automobile body-in-white were selected in the experiment, and a total of 550 dynamic resistance samples were collected. The dynamic resistance value was used as the feature, and the supervised classification algorithms including support vector machines, logical regression, and random forest were used to classify the dynamic resistance, to realize the recognition of material stack-up. The data preprocessing part, include signal acquisition, dynamic resistance calculation, filtering, and dimensionality reduction was introduced in detail. The results show that before dimensionality reduction, the classification accuracy of the random forest is the highest, reaching 93.9%. After dimensionality reduction, the classification performance of logistic regression is the best, and the accuracy is 96.97%. The requirement of adaptive control for the accuracy of material stack-up recognition can be satisfied. … (more)
- Is Part Of:
- Journal of manufacturing processes. Volume 56:Part A(2020)
- Journal:
- Journal of manufacturing processes
- Issue:
- Volume 56:Part A(2020)
- Issue Display:
- Volume 56, Issue 2020 (2020)
- Year:
- 2020
- Volume:
- 56
- Issue:
- 2020
- Issue Sort Value:
- 2020-0056-2020-0000
- Page Start:
- 796
- Page End:
- 805
- Publication Date:
- 2020-08
- Subjects:
- Resistance spot welding -- Material stack-up -- Machine learning -- dynamic resistance
Production management -- Data processing -- Periodicals
Manufacturing processes -- Periodicals
Procestechnologie
Productietechniek
Production -- Gestion -- Informatique -- Périodiques
Fabrication -- Périodiques
Manufacturing processes
Production management -- Data processing
Periodicals
670.5 - Journal URLs:
- http://www.sciencedirect.com/science/journal/15266125 ↗
http://www.elsevier.com/journals ↗ - DOI:
- 10.1016/j.jmapro.2020.04.051 ↗
- Languages:
- English
- ISSNs:
- 1526-6125
- Deposit Type:
- Legaldeposit
- View Content:
- Available online (eLD content is only available in our Reading Rooms) ↗
- Physical Locations:
- British Library DSC - 5011.640000
British Library DSC - BLDSS-3PM
British Library HMNTS - ELD Digital store - Ingest File:
- 13685.xml