中国科学院机构知识库网格
Chinese Academy of Sciences Institutional Repositories Grid
首页 机构 成果 学者
  
  1. CAS IR Grid
  2. 合肥物质科学研究院
  3. 中国科学院合肥物质科学研究院
Application of physics-guided deep learning model in tool wear monitoring of high-speed milling

文献类型：期刊论文

作者Li, Shenshen1; Li, Jun1; Zhu, Kunpeng1,2
刊名MECHANICAL SYSTEMS AND SIGNAL PROCESSING
出版日期2025
卷号224
ISSN号0888-3270
DOI10.1016/j.ymssp.2024.111949
通讯作者Zhu, Kunpeng(zhukp@iamt.ac.cn)
英文摘要As an important part of intelligent machining system, tool wear monitoring plays a crucial role in ensuring workpiece quality and process safety. At present, the models based on tool wear monitoring mainly include data-driven models and physics-based models. However, data-driven models are limited by physical inconsistency, and physics-based models usually lack accurate description of the machining process for process control. To solve these issues, this study proposes a physics-guided deep learning model for tool wear monitoring. Firstly, a dual scale time series model is established, and physical constraints are added to the model according to the degradation characteristics of tool wear. Secondly, the physics-based loss function is introduced through the physical model of tool wear to constrain the training process of the model. Finally, a model agnostic meta learning algorithm is used to train a pre-weight for the model, so that the model can be quickly applied to different processing conditions. The experimental results show that the physics-guided deep learning model proposed has good accuracy and physical consistency. In the high-speed milling tests, the average MAPE of tool wear prediction is 0.039, and the physical consistency index is 0.035 mu m.
WOS关键词NEURAL-NETWORK ; PROGNOSTICS
资助项目National Natural Science Foundation of China[52175528] ; National Key Research and Development Program of China ; Chinese Ministry of Science and Technology[2018YFB1703200]
WOS研究方向Engineering
语种英语
WOS记录号WOS:001324716100001
出版者ACADEMIC PRESS LTD- ELSEVIER SCIENCE LTD
资助机构National Natural Science Foundation of China ; National Key Research and Development Program of China ; Chinese Ministry of Science and Technology
源URL[http://ir.hfcas.ac.cn:8080/handle/334002/135625]  
专题中国科学院合肥物质科学研究院
通讯作者Zhu, Kunpeng
作者单位1.Wuhan Univ Sci & Technol, Inst Precis Mfg, Sch Machinery & Automat, Wuhan 430081, Peoples R China
2.Chinese Acad Sci, Inst Intelligent Machines, Hefei Inst Phys Sci, Changzhou 213164, Peoples R China
推荐引用方式
GB/T 7714		 Li, Shenshen,Li, Jun,Zhu, Kunpeng. Application of physics-guided deep learning model in tool wear monitoring of high-speed milling[J]. MECHANICAL SYSTEMS AND SIGNAL PROCESSING,2025,224.
APA Li, Shenshen,Li, Jun,&Zhu, Kunpeng.(2025).Application of physics-guided deep learning model in tool wear monitoring of high-speed milling.MECHANICAL SYSTEMS AND SIGNAL PROCESSING,224.
MLA Li, Shenshen,et al."Application of physics-guided deep learning model in tool wear monitoring of high-speed milling".MECHANICAL SYSTEMS AND SIGNAL PROCESSING 224(2025).

入库方式： OAI收割

来源：合肥物质科学研究院

浏览0
下载0
收藏0
其他版本

除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。