An innovative strategy for real-time tool anomaly detection in CNC milling processes using time series monitoring

Abstract

Tool anomalies in computer numerical control (CNC) milling processes are unpredictable, hindering the promotion of fully automated machining. Traditional detection systems often struggle with stability due to the irregular and non-parametric nature of signals generated during dynamic milling. This study proposes an improved probability and statistics-based model for constructing tool anomaly thresholds in the time domain, with the decision-making strategy seamlessly integrated into CNC milling systems. A robust data acquisition and preprocessing framework was developed to improve the accuracy and reliability of real-time monitoring data. A Gaussian process model was employed to construct an anomaly detection threshold data set from irregular signals. Anomalies were identified when monitoring indicators surpassed the established threshold. The proposed method was validated through milling experiments of a turbine blisk, demonstrating an overall anomaly detection accuracy of 91.45%, which exceeds those of four other typical anomaly detection methods. These results confirm the effectiveness of the proposed method and its potential applicability in industry.

Graphical abstract

Keywords

anomaly detection / cutting tool / cutting force / threshold / machining system

References

[1] Fragapane G, Eleftheriadis R, Powell D, Antony J. A global survey on the current state of practice in zero defect manufacturing and its impact on production performance. Computers in Industry, 2023, 148: 103879

[2] Pfirrmann D, Baumann J, Krebs E, Biermann D, Wiederkehr P. Material defects detection based on in-process measurements in milling of Ti6246 alloy. Procedia CIRP, 2021, 99: 165–170

[3] Zhang X, Gao Y, Guo Z C, Zhang W, Yin J, Zhao W H. Physical model-based tool wear and breakage monitoring in milling process. Mechanical Systems and Signal Processing, 2023, 184: 109641

[4] Zheng P, Wang H H, Sang Z Q, Zhong R Y, Liu Y K, Liu C, Mubarok K, Yu S Q, Xu X. Smart manufacturing systems for Industry 4.0: conceptual framework, scenarios, and future perspectives. Frontiers of Mechanical Engineering, 2018, 13(2): 137–150

[5] Liu B F, Zhang Y F, Lv J X, Majeed A, Chen C H, Zhang D. A cost-effective manufacturing process recognition approach based on deep transfer learning for CPS enabled shop-floor. Robotics and Computer-Integrated Manufacturing, 2021, 70: 102128

[6] Ying Z Z, Shu L M, Kizaki T, Iwama M, Sugita N. Hybrid approach for onsite monitoring and anomaly detection of cutting tool life. Procedia CIRP, 2021, 104: 1541–1546

[7] Bai L L, Liu H G, Zhang J, Zhao W H. Real-time tool breakage monitoring based on dimensionless indicators under time-varying cutting conditions. Robotics and Computer-Integrated Manufacturing, 2023, 81: 102502

[8] Mahmood J, Mustafa G, Ali M. Accurate estimation of tool wear levels during milling, drilling and turning operations by designing novel hyperparameter tuned models based on LightGBM and stacking. Measurement, 2022, 190: 110722

[9] CaronEngineering. Tool Monitoring System for CNC Machining. 2024, available at the website of www.caroneng.com/products/tmac

[10] Peco Chacón A M, Segovia Ramirez I, García Márquez F P. False alarm detection in wind turbine by classification models. Advances in Engineering Software, 2023, 177: 103409

[11] Atluru S, Huang S H, Snyder J P. A smart machine supervisory system framework. The International Journal of Advanced Manufacturing Technology, 2012, 58(5–8): 563–572

[12] Wang S, Liang Y C, Li W D, Cai X T. Big data enabled intelligent immune system for energy efficient manufacturing management. Journal of Cleaner Production, 2018, 195: 507–520

[13] Garghetti F, Grasso M, Pacella M, Fogliazza G, Colosimo B M. Multi-stream big data mining for industry 4.0 in machining: novel application of a Gated Recurrent Unit Network. Procedia CIRP, 2023, 118: 431–436

[14] Quatrini E, Costantino F, Di Gravio G, Patriarca R. Machine learning for anomaly detection and process phase classification to improve safety and maintenance activities. Journal of Manufacturing Systems, 2020, 56: 117–132

[15] Liu Z P, Lang Z Q, Gui Y F, Zhu Y P, Laalej H. Digital twin-based anomaly detection for real-time tool condition monitoring in machining. Journal of Manufacturing Systems, 2024, 75: 163–173

[16] Wang S H, Lei Y G, Lu N, Li X, Yang B. A multi-sensor relation model for recognizing and localizing faults of machines based on network analysis. Frontiers of Mechanical Engineering, 2023, 18(2): 20

[17] Sun S X, Liu Y C, Hu X F, Zhang W J. A semisupervised autoencoder-based method for anomaly detection in cutting tools. Journal of Manufacturing Processes, 2023, 93: 315–327

[18] Bernini L, Albertelli P, Monno M. Mill condition monitoring based on instantaneous identification of specific force coefficients under variable cutting conditions. Mechanical Systems and Signal Processing, 2023, 185: 109820

[19] Du H Z, Duan Z Y. Finder: A novel approach of change point detection for multivariate time series. Applied Intelligence, 2022, 52(3): 2496–2509

[20] Hu G Z, Li H F, Xia Y Q, Luo L X. A deep Boltzmann machine and multi-grained scanning forest ensemble collaborative method and its application to industrial fault diagnosis. Computers in Industry, 2018, 100: 287–296

[21] Liu Y Q, Yu D P, Zhao W, Zhang K. Segmentation-assisted classification model with convolutional neural network for weld defect detection. Advances in Engineering Software, 2024, 198: 103788

[22] Tian Y, Fu M Y, Wu F. Steel plates fault diagnosis on the basis of support vector machines. Neurocomputing, 2015, 151: 296–303

[23] Kumar P, Raouf I, Kim H S. Transfer learning for servomotor bearing fault detection in the industrial robot. Advances in Engineering Software, 2024, 194: 103672

[24] Wang Y H, Zheng L Y, Wang Y W. Event-driven tool condition monitoring methodology considering tool life prediction based on industrial internet. Journal of Manufacturing Systems, 2021, 58: 205–222

[25] Kong D D, Chen Y J, Li N. Gaussian process regression for tool wear prediction. Mechanical Systems and Signal Processing, 2018, 104: 556–574

[26] Fan J, Zhang Q R, Zhu J L, Zhang M, Yang Z, Cao H X. Robust deep auto-encoding Gaussian process regression for unsupervised anomaly detection. Neurocomputing, 2020, 376: 180–190

[27] Qiang B Y, Shi K N, Liu N, Ren J X, Shi Y Y. Integrating physics-informed recurrent Gaussian process regression into instance transfer for predicting tool wear in milling process. Journal of Manufacturing Systems, 2023, 68: 42–55

[28] Lauro C H, Brandão L C, Baldo D, Reis R A, Davim J P. Monitoring and processing signal applied in machining processes – a review. Measurement, 2014, 58: 73–86

[29] Peng D F, Li H K, Dai Y B, Wang Z D, Ou J Y. Prediction of milling force based on spindle current signal by neural networks. Measurement, 2022, 205: 112153

[30] Lévesque M, Tipper D. A survey of clock synchronization over packet-switched networks. IEEE Communications Surveys and Tutorials, 2016, 18(4): 2926–2947

[31] Du Z H, Chen X F, Zhang H, Zi Y Y, Yan R Q. Multiple fault separation and detection by joint subspace learning for the health assessment of wind turbine gearboxes. Frontiers of Mechanical Engineering, 2017, 12(3): 333–347

[32] Li Y, Zhao Z C, Wang K, Qian N, Fu Y C, Cao S F. Active-passive hybrid feed rate control systems in CNC machining: mitigating force fluctuations and enhancing tool life. Journal of Manufacturing Systems, 2024, 77: 184–195

[33] Jiang J C, Zhang R H, Wu Y T, Chang C, Jiang Y. A fault diagnosis method for electric vehicle power lithium battery based on wavelet packet decomposition. Journal of Energy Storage, 2022, 56: 105909

[34] InternationalOrganization for Standardization. Tool life testing in milling. Part 2: End milling. ISO 8688-2:1989. 1989

RIGHTS & PERMISSIONS

Higher Education Press