A Study of Tool Wear Prediction Based on Digital Twins

In the context of the global intelligent transformation of manufacturing, digital twin technology, through the deep integration of physical entities and virtual models, provides an innovative path for the implementation of smart manufacturing. Taking the VMC-C50 five-axis CNC machine tool milling ti...

Full description

Saved in:
Bibliographic Details
Main Authors: LIU Minghao, MAO Xinhui, XIA Wei, YUE Caixu, LIU Xianli
Format: Article
Language:zho
Published: Harbin University of Science and Technology Publications 2025-02-01
Series:Journal of Harbin University of Science and Technology
Subjects:
Online Access:https://hlgxb.hrbust.edu.cn/#/digest?ArticleID=2400
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849426579265945600
author LIU Minghao
MAO Xinhui
XIA Wei
YUE Caixu
LIU Xianli
author_facet LIU Minghao
MAO Xinhui
XIA Wei
YUE Caixu
LIU Xianli
author_sort LIU Minghao
collection DOAJ
description In the context of the global intelligent transformation of manufacturing, digital twin technology, through the deep integration of physical entities and virtual models, provides an innovative path for the implementation of smart manufacturing. Taking the VMC-C50 five-axis CNC machine tool milling titanium alloy as the research object, a digital twin architecture-based milling tool wear monitoring system was constructed based on the technical route of “ virtual-real interaction, data-driven”. By integrating the physical perception layer, virtual modeling layer, data interconnection layer, and intelligent service layer, a bidirectional communication mechanism between the physical machine tool and the virtual model was established, achieving full-factor mapping and dynamic optimization of the machining process. With tool wear prediction as the application scenario, a deep learning model based on the fusion of multi-scale convolutional neural network, residual network, bidirectional long short-term memory network, and gated recurrent unit (MSCNN-ResNet-BiLSTM-GRU) was proposed. This model can deeply extract spatial features and dynamic temporal features, significantly improving prediction accuracy compared to conventional models. Through virtual-real interaction and data fusion mechanisms, it provides an engineering solution for the dynamic perception of tool wear during the milling process.
format Article
id doaj-art-e9f459ff845a41e6bb866cbf814f5177
institution Kabale University
issn 1007-2683
language zho
publishDate 2025-02-01
publisher Harbin University of Science and Technology Publications
record_format Article
series Journal of Harbin University of Science and Technology
spelling doaj-art-e9f459ff845a41e6bb866cbf814f51772025-08-20T03:29:19ZzhoHarbin University of Science and Technology PublicationsJournal of Harbin University of Science and Technology1007-26832025-02-013001708110.15938/j.jhust.2025.01.008A Study of Tool Wear Prediction Based on Digital TwinsLIU Minghao0MAO Xinhui1XIA Wei2YUE Caixu3LIU Xianli4Key Laboratory of Advanced Manufacturing and Intelligent Technology, Ministry of Education, Harbin University of Science and Technology, Harbin 150080, ChinaKey Laboratory of Advanced Manufacturing and Intelligent Technology, Ministry of Education, Harbin University of Science and Technology, Harbin 150080, ChinaKey Laboratory of Advanced Manufacturing and Intelligent Technology, Ministry of Education, Harbin University of Science and Technology, Harbin 150080, ChinaKey Laboratory of Advanced Manufacturing and Intelligent Technology, Ministry of Education, Harbin University of Science and Technology, Harbin 150080, ChinaKey Laboratory of Advanced Manufacturing and Intelligent Technology, Ministry of Education, Harbin University of Science and Technology, Harbin 150080, ChinaIn the context of the global intelligent transformation of manufacturing, digital twin technology, through the deep integration of physical entities and virtual models, provides an innovative path for the implementation of smart manufacturing. Taking the VMC-C50 five-axis CNC machine tool milling titanium alloy as the research object, a digital twin architecture-based milling tool wear monitoring system was constructed based on the technical route of “ virtual-real interaction, data-driven”. By integrating the physical perception layer, virtual modeling layer, data interconnection layer, and intelligent service layer, a bidirectional communication mechanism between the physical machine tool and the virtual model was established, achieving full-factor mapping and dynamic optimization of the machining process. With tool wear prediction as the application scenario, a deep learning model based on the fusion of multi-scale convolutional neural network, residual network, bidirectional long short-term memory network, and gated recurrent unit (MSCNN-ResNet-BiLSTM-GRU) was proposed. This model can deeply extract spatial features and dynamic temporal features, significantly improving prediction accuracy compared to conventional models. Through virtual-real interaction and data fusion mechanisms, it provides an engineering solution for the dynamic perception of tool wear during the milling process.https://hlgxb.hrbust.edu.cn/#/digest?ArticleID=2400digital twinscnc machine toolstool wear predictiondeep learningconvolutional neural networksbidirectional long short-term memory networks
spellingShingle LIU Minghao
MAO Xinhui
XIA Wei
YUE Caixu
LIU Xianli
A Study of Tool Wear Prediction Based on Digital Twins
Journal of Harbin University of Science and Technology
digital twins
cnc machine tools
tool wear prediction
deep learning
convolutional neural networks
bidirectional long short-term memory networks
title A Study of Tool Wear Prediction Based on Digital Twins
title_full A Study of Tool Wear Prediction Based on Digital Twins
title_fullStr A Study of Tool Wear Prediction Based on Digital Twins
title_full_unstemmed A Study of Tool Wear Prediction Based on Digital Twins
title_short A Study of Tool Wear Prediction Based on Digital Twins
title_sort study of tool wear prediction based on digital twins
topic digital twins
cnc machine tools
tool wear prediction
deep learning
convolutional neural networks
bidirectional long short-term memory networks
url https://hlgxb.hrbust.edu.cn/#/digest?ArticleID=2400
work_keys_str_mv AT liuminghao astudyoftoolwearpredictionbasedondigitaltwins
AT maoxinhui astudyoftoolwearpredictionbasedondigitaltwins
AT xiawei astudyoftoolwearpredictionbasedondigitaltwins
AT yuecaixu astudyoftoolwearpredictionbasedondigitaltwins
AT liuxianli astudyoftoolwearpredictionbasedondigitaltwins
AT liuminghao studyoftoolwearpredictionbasedondigitaltwins
AT maoxinhui studyoftoolwearpredictionbasedondigitaltwins
AT xiawei studyoftoolwearpredictionbasedondigitaltwins
AT yuecaixu studyoftoolwearpredictionbasedondigitaltwins
AT liuxianli studyoftoolwearpredictionbasedondigitaltwins