Due to the open-chain, multilink structure and weak rigidity characteristics of the industrial robots, the stability of their milling process is poor, significantly exacerbating tool wear and breakage. Therefore, this article proposes a tool wear condition monitoring method based on variational autoencoders (VAE) and deep learning neural networks. This method combines unsupervised and supervised learning to analyze the mapping relationship between multidomain features of monitoring signals and tool wear, achieving the monitoring and identification of tool wear conditions. Firstly, a multidomain feature extraction method is proposed to obtain wear characteristics from milling signals. Then, based on the proposed method, VAE is combined with long short-term memory neural networks (LSTM), Transformer encoders integrated with LSTM layers, and bidirectional LSTM neural networks (BiLSTM) to construct three tool wear prediction models: VAE-LSTM, VAE-LSTM-Transformer, and VAE-BiLSTM. These models achieve deep integration of multidomain features and map them to tool wear values. Finally, experiments were conducted to validate the tool wear conditions for both robot milling and traditional machine tool milling. The results show that this method has high prediction accuracy, strong adaptability to changes in milling parameters, and strong compatibility with different milling equipment.
TaozhangWYuWYufeiW, et al.Application of deep learning in robot milling chattering identification. Mech Sci Technol Aerosp Eng2021; 40: 188–192.
2.
ZeyuR. Research status and development trend of tool wear monitoring technology. China Plant Eng2019; 02: 76–77.
3.
PeibinTBaojinWJintaoS. Research status and prospect of intelligent monitoring technology for cutter wear state. Forest Mach Woodwork Equip2021; 49: 4–8. +20.
4.
ChunlongPErhuaWYiZ. Research on monitoring method of wear state of micro-milling tool. Comp Meas Contr2021; 29: 22–28. +40.
YulongZChaofeiZJianT, et al.Intelligent fault diagnosis using image representation of multi-domain features. J Intell Fuzzy Syst2022; 42: 1317–1329.
7.
MikołajczykTNowickiKKłodowskiA, et al.Neural network approach for automatic image analysis of cutting edge wear. Mech Syst Signal Process2017; 88: 100–110.
8.
LinW-JChenJ-WJhuangJ-P, et al.Publisher correction: integrating object detection and image segmentation for detecting the tool wear area on stitched image. Sci Rep2021; 11: 21729.
9.
Fernández-RoblesLAzzopardiGAlegreE, et al.Machine-vision-based identification of broken inserts in edge profile milling heads. Robot Comp Integ Manuf2017; 44: 276–283. DOI:https://doi.org/10.1016/j.rcim.2016.10.004.
10.
ZhanqiangLXingA. Wear characteristics of cutting tools in high speed machining. Tribology2002; 22: 4.
11.
CorneRNathCMansoriEM, et al.Study of spindle power data with neural network for predicting real-time tool wear/breakage during inconel drilling. J Manuf Syst2017; 43: 287–295. DOI:https://doi.org/10.1016/j.jmsy.2017.01.004.
12.
PeiyuanSWeiweiYChenghanW, et al.Surface roughness prediction method based on convolutional autoencoder feature extraction. Mach Des Res2022; 38: 145–149. +154.
13.
GuoLLiNJiaF, et al.A recurrent neural network based health indicator for remaining useful life prediction of bearings. Neurocomputing2017; 240: 98–109.
14.
GomesMCBritoLCda SilvaMB, et al.Tool wear monitoring in micromilling using Support Vector Machine with vibration and sound sensors. Prec Eng2021; 67: 137–151.
15.
UpaseRAmbhoreN. Experimental investigation of tool wear using vibration signals: an ANN approach. Mater Today Proc2020; 24: 1365–1375.
16.
BhatNNDuttaSVashisthT, et al.Tool condition monitoring by SVM classification of machined surface images in turning. Int J Adv Manuf Technol2016; 83: 1487–1502.
17.
BijunHNiCLiangL, et al.Research on drilling tool state recognition based on wavelet packet analysis and LS-SVM. Tool Eng2019; 53: 3–9.
18.
BenkedjouhTZerhouniNRechakS. Tool wear condition monitoring based on continuous wavelet transform and blind source separation. Int J Adv Manuf Technol2018; 97: 3311–3323.
19.
HesserFDMarkertB. Tool wear monitoring of a retrofitted CNC milling machine using artificial neural networks. Manuf Lett2018; 19: 1–4.
20.
LiWLiuT. Time varying and condition adaptive hidden Markov model for tool wear state estimation and remaining useful life prediction in micro-milling. Mech Syst Signal Process2019; 131: 689–702.
21.
HalgamugeNMDamindaENirmalathasA. Best optimizer selection for predicting bushfire occurrences using deep learning. Nat Hazards2020; 103: 1–16.
22.
VahidNSinaDKareemA, et al.Tool wear monitoring by ensemble learning and sensor fusion using power, sound, vibration, and AE signals. Manuf Lett2021; 30: 32–38.
23.
ZhaoRYanRChenZ, et al.Deep learning and its applications to machine health monitoring. Mech Syst Signal Process2019; 115: 213–237.
24.
ChaobiaoZYanmingS. Tool wear prognostic based on feature engineering and deep feedforward network. Mach Des Manuf2020; 06: 190–193.
25.
XianliLXuebingLMingnaD, et al.Intelligent management and control technology of cutting tool life-cycle for intelligent manufacturing. J Mech Eng2021; 57: 196–219.
26.
HuiyongLSongZFengjianL, et al.Tool wear detection based on improved CNN-BiLSTM model. China Mech Eng2022; 33: 1940–1947. +1956.
27.
AnQTaoZXuX, et al.A data-driven model for milling tool remaining useful life prediction with convolutional and stacked LSTM network. Measurement (Mahwah N J)2020; 154: 107461–107461.
28.
YuWKimYIMechefskeC. Analysis of different RNN autoencoder variants for time series classification and machine prognostics. Mech Syst Signal Process2021; 149: 107322–107322. DOI:https://doi.org/10.1016/j.ymssp.2020.107322.
29.
HuangZZhuJLeiJ, et al.Tool wear predicting based on multi-domain feature fusion by deep convolutional neural network in milling operations. J Intell Manuf2019; 31: 1–14.
30.
XianliLShaoyangLXuebingL, et al.Intelligent tool wear monitoring based on parallel residual and stacked bidirectional long short-term memory network. J Manuf Syst2021; 60: 608–619.
31.
LiJHuangRLiW. Intelligent fault diagnosis for bearing dataset using adversarial transfer learning based on stacked auto-encoder. Proc Manuf2020; 49: 75–80.
32.
BaiYXiangSChengF, et al.A dynamic-inner LSTM prediction method for key alarm variables forecasting in chemical process. Chin J Chem Eng2023; 55: 266–276.
33.
ChenpengZWeijunWZhichenH, et al.Milling cutter wear prediction based on feature extraction and long short-term memory neural networks. Contr Info Technol2021; 4: 59–65.
34.
FanZDechenYShengzhuoY, et al.Lifetime prediction of bearings based on the Transformer—LSTM network. J Vibr Shock2024; 43: 320–328.
35.
KaixingZDelongZChang’anZ, et al.Tool wear monitoring using a novel parallel BiLSTM model with multi-domain features for robotic milling Al7050-T7451 workpiece. Int J Adv Manuf Technol2023; 129: 1883–1899.
36.
ZhangYGuoKSunJ, etal. Method of Postures Selection for Industrial Robot Joint Stiffness Identification2021-01-01; 9: 62583–62592.
37.
AppelmanHCastleJCenL, et al.A method for mode coupling chatter detection and suppression in robotic milling. J Manuf Sci Eng2018; 140: 081015–081015.
38.
Chang’anZKaiGJieS. Sound singularity analysis for milling tool condition monitoring towards sustainable manufacturing. Mech Syst Signal Process2021; 157: 107738–107738.
39.
PoongadanSLineeshMC. Non-linear time series prediction using improved CEEMDAN, SVD and LSTM. Neural Process Lett2024; 56: 164–164.
40.
TaoWChanghuaLYiningS, et al.Automatic ECG classification using continuous wavelet transform and convolutional neural network. Entropy2021; 23: 119–119.
41.
XiangjieLYuanyanLXiaobingK, et al.Deep neural network for forecasting of photovoltaic power based on wavelet packet decomposition with similar day analysis. Energy2023; 271: 126963–126963.
42.
ChengCLiJLiuY, etal. Deep convolutional neural network-based in-process tool condition monitoring in abrasive belt grinding. Computers in Industry2019-04; 106: 1–13.
43.
FatemehAAntoineTMarcT. Tool condition monitoring using spectral subtraction and convolutional neural networks in milling process. Int J Adv Manuf Technol2018; 98: 3217–3227.
44.
PatelDThakkerHKiranMB, et al.Surface roughness prediction of machined components using gray level co-occurrence matrix and Bagging Tree. FME Transact2020; 48: 468–475.
45.
FalessiDHuangJNarayanaL, et al.On the need of preserving order of data when validating within-project defect classifiers. Empir Softw Eng2020; 25: 4805–4830.
46.
WangJJYanJLiC, et al.Deep heterogeneous GRU model for predictive analytics in smart manufacturing: application to tool wear prediction. Comput Ind2019; 111: 1–14.
47.
ShahMBoradeHSanghaviV, et al.Enhancing tool wear prediction accuracy using Walsh–Hadamard transform, DCGAN and dragonfly algorithm-based feature selection. Sensors2023; 23: 3833.
48.
KarlFPielokTMoosbauerJ, et al.Multi-objective hyperparameter optimization in machine learning—an overview. ACM Trans Evol Learn2023; 3: 1–50.
49.
JinjiangWYilinLRuiZ, et al.Physics guided neural network for machining tool wear prediction. J Manuf Syst2020; 57: 298–310.
50.
ZhangYLZhangCTanJ, et al.Intelligent fault diagnosis using image representation of multi-domain features. J Intell Fuzzy Syst2022; 42: 1317–1329.
51.
XueweiZTianbiaoYPengfeiX, et al.In-process stochastic tool wear identification and its application to the improved cutting force modeling of micro milling. Mech Syst Signal Process2022; 164: 108233–108233.
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
