Comparative Study on Online Prediction of TP2 Rolled Copper Tube Wall Thickness Based on Different Proxy Models

doi:10.3390/ma17235685

IMR OpenIR

	Comparative Study on Online Prediction of TP2 Rolled Copper Tube Wall Thickness Based on Different Proxy Models
	Yue, Fengli 1; Sha, Zhuo 1; Sun, Hongyun 1; Liu, Huan 1; Chen, Dayong 2,3; Liu, Jinsong 1,2,3; Chen, Chuanlai 3
通讯作者	Yue, Fengli(flyue@163.com)
	2024-12-01
发表期刊	MATERIALS
卷号	17 期号:23 页码:24
摘要	The wall thickness of the TP2 copper tube casting billet is not uniform after a three-roll planetary rotational rolling, which affects the wall thickness uniformity of the copper tube in the subsequent process. In order to study the influence of wall thickness at different positions of copper pipe after rolling on the wall thickness of copper pipe after joint drawing, an online ultrasonic test platform was used to measure the wall thickness of copper pipe after tying, and based on the test data, a finite element model of copper pipe billet was established, and the numerical simulation of joint drawing wall thickness was conducted. Based on the results of the ultrasonic testing experiment and finite element simulation, different neural network models were used to predict the joint tensile wall thickness with the data of the ultrasonic testing experiment as input and the results of finite element simulation as output. The prediction effect of different neural network models was compared, and the results showed that the prediction and fitting effect of the SVM model was better, but overfitting occurred during the fitting process. Furthermore, particle swarm optimization is used to optimize the penalty parameter C and the kernel parameter g in the SVM model. Compared with the traditional SVM model, the PSO-SVM model is more suitable for the prediction of joint tensile wall thickness, which can better guide the production to solve this problem.
关键词	joint drawing ultrasonic testing numerical simulation neural network
资助者	Basic Research Projects of Liaoning Provincial Department of Education ; Foundation Project - Science and Technology Plan Project of Chang Zhou
DOI	10.3390/ma17235685
收录类别	SCI
语种	英语
资助项目	Basic Research Projects of Liaoning Provincial Department of Education ; Foundation Project - Science and Technology Plan Project of Chang Zhou[CQ20220057] ; [LJKMZ20220591]
WOS研究方向	Chemistry ; Materials Science ; Metallurgy & Metallurgical Engineering ; Physics
WOS类目	Chemistry, Physical ; Materials Science, Multidisciplinary ; Metallurgy & Metallurgical Engineering ; Physics, Applied ; Physics, Condensed Matter
WOS记录号	WOS:001376441900001
出版者	MDPI
引用统计
文献类型	期刊论文
条目标识符	http://ir.imr.ac.cn/handle/321006/181421
专题	中国科学院金属研究所
通讯作者	Yue, Fengli
作者单位	1.Shenyang Ligong Univ, Automot & Transportat Coll, Shenyang 110159, Peoples R China 2.Chinese Acad Sci, Inst Met Res, Shi Changxu Mat Innovat Ctr, Shenyang 110016, Peoples R China 3.Changzhou Enreach Copper Co Ltd, Changzhou 213149, Peoples R China
推荐引用方式 GB/T 7714	Yue, Fengli,Sha, Zhuo,Sun, Hongyun,et al. Comparative Study on Online Prediction of TP2 Rolled Copper Tube Wall Thickness Based on Different Proxy Models[J]. MATERIALS,2024,17(23):24.
APA	Yue, Fengli.,Sha, Zhuo.,Sun, Hongyun.,Liu, Huan.,Chen, Dayong.,...&Chen, Chuanlai.(2024).Comparative Study on Online Prediction of TP2 Rolled Copper Tube Wall Thickness Based on Different Proxy Models.MATERIALS,17(23),24.
MLA	Yue, Fengli,et al."Comparative Study on Online Prediction of TP2 Rolled Copper Tube Wall Thickness Based on Different Proxy Models".MATERIALS 17.23(2024):24.