A Novel Tool Wear Identification Method Based on a Semi-Supervised LSTM

Machine learning models have been widely used in the field of cutting tool wear identification, achieving favorable results. However, in actual industrial scenarios, obtaining sufficient labeled samples is time consuming and costly, while unlabeled samples are abundant and easy to collect. This situ...

Full description

Saved in:
Bibliographic Details
Main Authors: Xin He, Meipeng Zhong, Chengcheng He, Jinhao Wu, Haiyang Yang, Zhigao Zhao, Wei Yang, Cong Jing, Yanlin Li, Chen Gao
Format: Article
Language:English
Published: MDPI AG 2025-02-01
Series:Lubricants
Subjects:
Online Access:https://www.mdpi.com/2075-4442/13/2/72
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850081778414387200
author Xin He
Meipeng Zhong
Chengcheng He
Jinhao Wu
Haiyang Yang
Zhigao Zhao
Wei Yang
Cong Jing
Yanlin Li
Chen Gao
author_facet Xin He
Meipeng Zhong
Chengcheng He
Jinhao Wu
Haiyang Yang
Zhigao Zhao
Wei Yang
Cong Jing
Yanlin Li
Chen Gao
author_sort Xin He
collection DOAJ
description Machine learning models have been widely used in the field of cutting tool wear identification, achieving favorable results. However, in actual industrial scenarios, obtaining sufficient labeled samples is time consuming and costly, while unlabeled samples are abundant and easy to collect. This situation significantly affects the model’s performance. To address this challenge, a novel semi-supervised method, based on long short-term memory (LSTM) networks, is provided. The proposed method leverages both small labeled and abundant unlabeled data to improve tool wear identification performance. The proposed method trains an initial tool wear regression model using LSTM, using a small amount of labeled samples. It then uses manifold regularization to generate pseudo-labels for the unlabeled samples. These pseudo-labeled samples are combined with the original labeled samples to retrain the MR–LSTM model iteratively to improve its performance. This process continues until a termination condition is met. The method considers the correlation between sample labels and feature structures, as well as the correlation between global and local sample labels. Experiments involving milling tool wear identification demonstrate that the proposed method significantly outperforms support vector regression (SVR) and recurrent neural network (RNN)-based methods, when a small amount of labeled samples and abundant unlabeled samples are available. The average <i>R<sup>2</sup></i> values in terms of the proposed method’s predicted results can reach above 0.95. The proposed method is a potential technique for low-cost tool wear identification, without the need to collect a large number of labeled samples.
format Article
id doaj-art-c94de329148d4c728fb7b27579cd8d8b
institution DOAJ
issn 2075-4442
language English
publishDate 2025-02-01
publisher MDPI AG
record_format Article
series Lubricants
spelling doaj-art-c94de329148d4c728fb7b27579cd8d8b2025-08-20T02:44:39ZengMDPI AGLubricants2075-44422025-02-011327210.3390/lubricants13020072A Novel Tool Wear Identification Method Based on a Semi-Supervised LSTMXin He0Meipeng Zhong1Chengcheng He2Jinhao Wu3Haiyang Yang4Zhigao Zhao5Wei Yang6Cong Jing7Yanlin Li8Chen Gao9College of Mechanical and Electrical Engineering, Jiaxing Nanhu University, Jiaxing 314001, ChinaCollege of Mechanical and Electrical Engineering, Jiaxing Nanhu University, Jiaxing 314001, ChinaZhejiang XCC Group Co., Ltd., Shaoxing 312500, ChinaCollege of Mechanical and Electrical Engineering, Jiaxing Nanhu University, Jiaxing 314001, ChinaCollege of Mechanical and Electrical Engineering, Jiaxing Nanhu University, Jiaxing 314001, ChinaZhejiang XCC Group Co., Ltd., Shaoxing 312500, ChinaZhejiang XCC Group Co., Ltd., Shaoxing 312500, ChinaCollege of Mechanical and Electrical Engineering, Jiaxing Nanhu University, Jiaxing 314001, ChinaCollege of Mechanical and Electrical Engineering, Jiaxing Nanhu University, Jiaxing 314001, ChinaSchool of Mechatronics and Transportation, Jiaxing Nanyang Polytechnic Institute, Jiaxing 314003, ChinaMachine learning models have been widely used in the field of cutting tool wear identification, achieving favorable results. However, in actual industrial scenarios, obtaining sufficient labeled samples is time consuming and costly, while unlabeled samples are abundant and easy to collect. This situation significantly affects the model’s performance. To address this challenge, a novel semi-supervised method, based on long short-term memory (LSTM) networks, is provided. The proposed method leverages both small labeled and abundant unlabeled data to improve tool wear identification performance. The proposed method trains an initial tool wear regression model using LSTM, using a small amount of labeled samples. It then uses manifold regularization to generate pseudo-labels for the unlabeled samples. These pseudo-labeled samples are combined with the original labeled samples to retrain the MR–LSTM model iteratively to improve its performance. This process continues until a termination condition is met. The method considers the correlation between sample labels and feature structures, as well as the correlation between global and local sample labels. Experiments involving milling tool wear identification demonstrate that the proposed method significantly outperforms support vector regression (SVR) and recurrent neural network (RNN)-based methods, when a small amount of labeled samples and abundant unlabeled samples are available. The average <i>R<sup>2</sup></i> values in terms of the proposed method’s predicted results can reach above 0.95. The proposed method is a potential technique for low-cost tool wear identification, without the need to collect a large number of labeled samples.https://www.mdpi.com/2075-4442/13/2/72tool wear identificationlong short-term memorymanifold regularizationsemi-supervised learning
spellingShingle Xin He
Meipeng Zhong
Chengcheng He
Jinhao Wu
Haiyang Yang
Zhigao Zhao
Wei Yang
Cong Jing
Yanlin Li
Chen Gao
A Novel Tool Wear Identification Method Based on a Semi-Supervised LSTM
Lubricants
tool wear identification
long short-term memory
manifold regularization
semi-supervised learning
title A Novel Tool Wear Identification Method Based on a Semi-Supervised LSTM
title_full A Novel Tool Wear Identification Method Based on a Semi-Supervised LSTM
title_fullStr A Novel Tool Wear Identification Method Based on a Semi-Supervised LSTM
title_full_unstemmed A Novel Tool Wear Identification Method Based on a Semi-Supervised LSTM
title_short A Novel Tool Wear Identification Method Based on a Semi-Supervised LSTM
title_sort novel tool wear identification method based on a semi supervised lstm
topic tool wear identification
long short-term memory
manifold regularization
semi-supervised learning
url https://www.mdpi.com/2075-4442/13/2/72
work_keys_str_mv AT xinhe anoveltoolwearidentificationmethodbasedonasemisupervisedlstm
AT meipengzhong anoveltoolwearidentificationmethodbasedonasemisupervisedlstm
AT chengchenghe anoveltoolwearidentificationmethodbasedonasemisupervisedlstm
AT jinhaowu anoveltoolwearidentificationmethodbasedonasemisupervisedlstm
AT haiyangyang anoveltoolwearidentificationmethodbasedonasemisupervisedlstm
AT zhigaozhao anoveltoolwearidentificationmethodbasedonasemisupervisedlstm
AT weiyang anoveltoolwearidentificationmethodbasedonasemisupervisedlstm
AT congjing anoveltoolwearidentificationmethodbasedonasemisupervisedlstm
AT yanlinli anoveltoolwearidentificationmethodbasedonasemisupervisedlstm
AT chengao anoveltoolwearidentificationmethodbasedonasemisupervisedlstm
AT xinhe noveltoolwearidentificationmethodbasedonasemisupervisedlstm
AT meipengzhong noveltoolwearidentificationmethodbasedonasemisupervisedlstm
AT chengchenghe noveltoolwearidentificationmethodbasedonasemisupervisedlstm
AT jinhaowu noveltoolwearidentificationmethodbasedonasemisupervisedlstm
AT haiyangyang noveltoolwearidentificationmethodbasedonasemisupervisedlstm
AT zhigaozhao noveltoolwearidentificationmethodbasedonasemisupervisedlstm
AT weiyang noveltoolwearidentificationmethodbasedonasemisupervisedlstm
AT congjing noveltoolwearidentificationmethodbasedonasemisupervisedlstm
AT yanlinli noveltoolwearidentificationmethodbasedonasemisupervisedlstm
AT chengao noveltoolwearidentificationmethodbasedonasemisupervisedlstm