Quality analysis for condition monitoring of a manufacturing process integrating acoustic processing based on three distinct ML algorithms
-
Robin Bäßler
Robin Bäßler studied mechanical engineering at Aalen University. After completing his research master’s degree in Advanced Materials and Manufacturing, he started working as an academic assistant at the Institute of Driveline Technology Aalen (IAA).
,
Akash Mangaluru Ramananda
Akash Mangaluru Ramananda has a Bachelor’s degree in Mechanical Engineering from Visvesvaraya Technological University and around six years of professional experience in R&D at Infosys Mysore, where he focused on mechanical design for autonomous vehicles and robotics automation platforms. Working as a research assistant at the Institute of Driveline Technology Aalen (IAA) whilst also completing a research master’s degree in Advanced Materials and Manufacturing.
and
Markus Kley
Markus Kley studied mechanical engineering and completed his PhD (Dr.-Ing.) at University of Stuttgart. He worked for Voith Turbo in Crailsheim in various of positions after receiving his degree. He was the head of engine component technologies in his previous role. At Aalen University, he was later named Professor of Design in General Mechanical Engineering. He administers the “Cooperative Doctoral Program PROMISE 4.0” at Aalen University of Applied Sciences, which includes Stuttgart University, Esslingen University, Heilbronn University, and Aalen University. He also coordinates the System Integration/Optimization-Methodology Testing research area in the Aalen University research building. In addition, he is the head of the Steinbeis Transfer Platform Industry 4.0 (TPBW 4.0) and the Steinbeis Transfer Center for Innovative Drive Technology and Waste Heat Utilization.
Abstract
Over the years, it has been proven that all machines that run continuously or even for a short period of time experiences breakdowns, reduced efficiency and damage to the parts. Considering one such manufacturing process is welding process where strip welding is studied for condition monitoring and prediction of online quality of the weld. The proposed embodiment discusses a predictive condition monitoring of a welding process, using acoustic sensors and processing their data. Converting the processed data to a predictive model, comparing different Machine Learning (ML) algorithms like Decision Tree (DT), Artificial Neural Networks (ANN) and Convolutional Neural Networks (CNN) based on the validation and training results, selecting the suitable algorithm for online or real-time weld quality prediction is the objective of the proposed embodiment. According to the research, effectively executing the proposed solution can improve its life, efficiency, and utilization. The training accuracy is observed to be 97 % for DT, 50 % for ANN and 67 % for CNN. DT, on the other hand, has a validation accuracy of 94.7 %, ANN has 47.65 % and CNN has 65.8 %. DT produced the best results in the study, initially DT prediction model is used in offline testing of the new data-set for individual weld class and the system produced 97 % accuracy. Furthermore the DT prediction model is implemented in online weld quality prediction in three classes: Bad weld, OK weld and Good weld where the system produced 90 % accuracy in real-time prediction.
Zusammenfassung
Im Laufe der Jahre hat sich gezeigt, dass bei allen Maschinen, die kontinuierlich oder auch nur für kurze Zeit laufen, Ausfälle, verminderte Effizienz und Schäden an den Teilen auftreten. Ein solcher Fertigungsprozess ist das Schweißen, bei dem ein Ensdlosschweißverfahren zur Zustandsüberwachung und Vorhersage der Online-Qualität der Schweißnaht untersucht wird. In der vorgeschlagenen Ausführungsform wird eine prädiktive Zustandsüberwachung eines Schweißprozesses unter Verwendung akustischer Sensoren und der Verarbeitung ihrer Daten erörtert. Die Umwandlung der verarbeiteten Daten in ein Vorhersagemodell, der Vergleich verschiedener Algorithmen des Maschine Learning (ML) wie Decision Tree (DT), Artificial Neural Networks (ANN) und Convolutional Neural Networks (CNN) auf der Grundlage der Validierungs- und Trainingsergebnisse und die Auswahl des geeigneten Algorithmus für die Online- oder Echtzeit-Vorhersage der Schweißqualität ist das Ziel der vorgeschlagenen Ausführungsform. Die Untersuchung hat ergeben, dass eine effektive Ausführung der vorgeschlagenen Lösung die Lebensdauer, Effizienz und Auslastung verbessern kann. Die Trainingsgenauigkeit liegt bei 97 % für DT, 50 % für ANN und 67 % für CNN. DT hingegen hat eine Validierungsgenauigkeit von 94,7 %, ANN von 47,65 % und CNN von 65,8 %. DT lieferte die besten Ergebnisse in der Studie. Zunächst wurde das DT-Vorhersagemodell bei Offline-Tests des neuen Datensatzes für einzelne Schweißnahtklassen verwendet, und das System erreichte eine Genauigkeit von 97 %. Darüber hinaus wird das DT-Vorhersagemodell zur Online-Vorhersage der Schweißnahtqualität in drei Klassen eingesetzt: Schlechte Schweißnaht, OK-Schweißnaht und gute Schweißnaht, wobei das System eine Genauigkeit von 90 % bei der Echtzeitvorhersage erreichte.
Funding source: AiF Projekt
Award Identifier / Grant number: ZF4724201GR9
Funding statement: This work is supported by AiF Projekt GmbH as the project management organisation of the Federal Ministry for Economic Affairs and Energy within the context of a ZIM cooperation project under the funding code ZF4724201GR9.
About the authors
Robin Bäßler studied mechanical engineering at Aalen University. After completing his research master’s degree in Advanced Materials and Manufacturing, he started working as an academic assistant at the Institute of Driveline Technology Aalen (IAA).
Akash Mangaluru Ramananda has a Bachelor’s degree in Mechanical Engineering from Visvesvaraya Technological University and around six years of professional experience in R&D at Infosys Mysore, where he focused on mechanical design for autonomous vehicles and robotics automation platforms. Working as a research assistant at the Institute of Driveline Technology Aalen (IAA) whilst also completing a research master’s degree in Advanced Materials and Manufacturing.
Markus Kley studied mechanical engineering and completed his PhD (Dr.-Ing.) at University of Stuttgart. He worked for Voith Turbo in Crailsheim in various of positions after receiving his degree. He was the head of engine component technologies in his previous role. At Aalen University, he was later named Professor of Design in General Mechanical Engineering. He administers the “Cooperative Doctoral Program PROMISE 4.0” at Aalen University of Applied Sciences, which includes Stuttgart University, Esslingen University, Heilbronn University, and Aalen University. He also coordinates the System Integration/Optimization-Methodology Testing research area in the Aalen University research building. In addition, he is the head of the Steinbeis Transfer Platform Industry 4.0 (TPBW 4.0) and the Steinbeis Transfer Center for Innovative Drive Technology and Waste Heat Utilization.
References
1. R. Zurawski, The Industrial Information Technology Handbook, CRC, Boca Raton, FL, London (2004), ISBN: 0-8493-1985-4.10.1201/9781420036336Search in Google Scholar
2. S. Lacey, An overview of bearing vibration analysis, Maintenance and Asset Management Journal (2008), 32–42.Search in Google Scholar
3. M. Bauer, F. Wagner, M. Kley, Optimierung der Sensorpositionierung bei schwingungsbasierter Wälzlagerzustandsüberwachung unter Einbezug von Systemeigenmoden, tm-Technisches Messen (2021), DOI: 10.1515/teme-2021-0045.Search in Google Scholar
4. M. Bauer, M. Hoffmann, M. Kley, Method for detecting the influence of external vibration excitations in rolling bearing condition monitoring, in: Vibrations, 343–353, VDI Verlag, Dusseldorf (2019), DOI: 10.51202/9783181023662-343.Search in Google Scholar
5. J. C. Kabugo, S.-L. Jämsä-Jounela, R. Schiemann et al., Industry 4.0 based process data analytics platform: A waste-to-energy plant case study, International Journal of Electrical Power & Energy Systems 115 (2020), 343–353, DOI: 10.1016/j.ijepes.2019.105508.Search in Google Scholar
6. F. Souza, R. Araújo, J. Mendes, Review of soft sensor methods for regression applications, Chemometrics and Intelligent Laboratory Systems 152 (2016), 69–79, DOI: 10.1016/j.chemolab.2015.12.011.Search in Google Scholar
7. J.-J. Wang, Y.-H. Zheng, L.-B. Zhang, L.-X. Duan, R. Zhao, Virtual sensing for gearbox condition monitoring based on kernel factor analysis, Petroleum Science 14 (2017), 539–548, DOI: 10.1007/s12182-017-0163-4.Search in Google Scholar
8. A. Albers, J.-M. Veith, B. Krüger, M. Behrendt, Schätzung von Drehmomenten in Fahrzeugantriebsträngen aus BUS-Signalen mithilfe künstlicher neuronaler Netze, IPEK, KITopen-ID: 1000085490.Search in Google Scholar
9. M. Acosta, S. Kanarachos, M. Fitzpatrick, A virtual sensor for integral tire force estimation using tire model-less approaches and adaptive unscented Kalman filter, in: ICINCO (2017), 386–397, DOI: 10.5220/0006394103860397.Search in Google Scholar
10. E. Regolin, A. Alatorre, M. Zambelli, A. Victorino, A. Charara, A. Ferrara, A sliding-mode virtual sensor for wheel forces estimation with accuracy enhancement via EKF, IEEE Transactions on Vehicular Technology 68(4) (2019), 3457–3471, DOI: 10.1109/TVT.2019.2903598.Search in Google Scholar
11. T.Bäßler, R. Bäßler, M. Kley, Augmented mel-spectrogram VGG-16 model for axial and radial load classification at wire-race bearings, tm-Technisches Messen (2022), DOI: 10.1515/teme-teme.2022.0039.Search in Google Scholar
12. R.Bäßler, T. Bäßler, M. Kley, Classification of load and rotational speed at wire-race bearings using Convolutional Neural Networks with vibration spectrograms, tm-Technisches Messen (2022), DOI: 10.1515/teme-2021-0143.Search in Google Scholar
13. M. Bauer, R. Bäßler, M. Kley, Lastdetektierung durch intelligente Schwingungsanalyse in Leichtbaulagern, in: Stuttgarter Symposium für Produktentwicklung, SSP (2019), 73–82, DOI: 10.18419/opus-10394.Search in Google Scholar
14. R. Bäßler, M. Bauer, M. Kley, Einsatz eines virtuellen Sensors an einem Drahtwälzlager, in: Aalener Kolloquium antriebstechnische Anwendungen, AKAA (2020), 112–117, ISBN: 978-3-8440-7529-8.Search in Google Scholar
15. L. Hafner, R. Bäßler, S. Schwarzer, F. Dohnal, M. Kley, Bubble size detection by process ancillaries, Chemie Ingenieur Technik 94(8) (2022), 1096–1104, DOI: 10.1002/cite.202100175.Search in Google Scholar
16. A. Sumesh, K. Rameshkumar, K. Mohandas et al., Use of machine learning algorithms for weld quality monitoring using acoustic signature, Procedia Computer Science 50 (2015), 316–322, DOI: 10.1016/j.procs.2015.04.042.Search in Google Scholar
17. R. B. Randall, Vibration-based Condition Monitoring: Industrial, Aerospace and Automotive Applications, John Wiley & Sons Inc. (2021), ISBN: 9781119477556.10.1002/9781119477631Search in Google Scholar
18. A. Brandt, Noise and Vibration Analysis: Signal Analysis and Experimental Procedures, Wiley-Blackwell, Oxford (2011), ISBN: 9780470746448.10.1002/9780470978160Search in Google Scholar
19. J. Blough, Improving the Analysis of Operating Data on Rotating Automotive Components, A Bell & Howell Information Company, USA (1998), UMI: 9918136.Search in Google Scholar
20. J. W. Cooley, J. W. Tukey, An Algorithm for the Machine Calculation of Complex Fourier Series, Mathematics of Computation, vol. 19 (1965).10.1090/S0025-5718-1965-0178586-1Search in Google Scholar
21. A. Bouzida, O. Touhami, R. Ibtiouen, A. Belouchrani, M. Fadel, A. Rezzoug, Fault diagnosis in industrial induction machines through discrete wavelet transform, IEEE Trans. Ind. Electron. 58 (2015), 4385–4395, DOI: 10.1109/TIE.2010.2095391.Search in Google Scholar
22. A. Baccigalupi, A. Liccardo, The Huang Hilbert transform for evaluating the instantaneous frequency evolution of transient signals in non-linear systems, Measurement 86 (2016), 1–13, DOI: 10.1016/j.measurement.2016.02.018.Search in Google Scholar
23. Jagath Sri Lal Senanayaka, Online Condition Monitoring of Electric Powertrains using Machine Learning and Data Fusion, Doctoral dissertations at the University of Agder 270, Media 07 Kristiansand (2020), ISBN: 978-82-7117-971-7.Search in Google Scholar
24. Z. Ge, Z. Song, S. X. Ding, B. Huang, Data mining and analytics in the process industry: the role of machine learning, IEEE Access 5 (2017), 20590–20616, DOI: 10.1109/ACCESS.2017.2756872.Search in Google Scholar
© 2022 Walter de Gruyter GmbH, Berlin/Boston
Articles in the same Issue
- Frontmatter
- Research Articles
- Verzeichnungskorrektur einer omnidirektionalen Kamera durch 3D-Referenzobjekte
- The influence of workpiece surface texture on visual measurement of roughness
- Deploying the high-power pulsed lasers in precision force metrology – Towards SI traceable and practical force quantization by photon momentum
- Quality analysis for condition monitoring of a manufacturing process integrating acoustic processing based on three distinct ML algorithms
- Three-dimensional finite element analysis and research of leakage magnetic field of flake, pit, through-hole defect in petroleum pipelines
- Generation of reproducible and scalable local damage to rolling bearings for the development of condition monitoring systems through the use of laser technology
Articles in the same Issue
- Frontmatter
- Research Articles
- Verzeichnungskorrektur einer omnidirektionalen Kamera durch 3D-Referenzobjekte
- The influence of workpiece surface texture on visual measurement of roughness
- Deploying the high-power pulsed lasers in precision force metrology – Towards SI traceable and practical force quantization by photon momentum
- Quality analysis for condition monitoring of a manufacturing process integrating acoustic processing based on three distinct ML algorithms
- Three-dimensional finite element analysis and research of leakage magnetic field of flake, pit, through-hole defect in petroleum pipelines
- Generation of reproducible and scalable local damage to rolling bearings for the development of condition monitoring systems through the use of laser technology
