Machine condition monitoring in FDM based on electret microphone, SVM, and neural networks

被引:3
作者
Lopes, Thiago Glissoi [1 ]
Aguiar, Paulo Roberto [1 ]
de Carvalho Monson, Paulo Monteiro [1 ]
D'Addona, Doriana Marilena [2 ]
Conceicao Junior, Pedro de Oliveira [3 ]
de Oliveira Junior, Reinaldo Gotz [1 ,4 ]
机构
[1] Sao Paulo State Univ UNESP, Dept Elect Engn, Fac Engn, Bauru, SP, Brazil
[2] Univ Naples Federico II, Dept Chem Mat & Ind Prod Engn, Naples, Italy
[3] Univ Sao Paulo, Sao Carlos Sch Engn, Dept Elect & Comp Engn, Sao Carlos, SP, Brazil
[4] Sao Paulo State Technol Coll, Biomed Syst, Bauru, SP, Brazil
关键词
3D printing; Monitoring; Electret microphone; Signal processing; Neural networks; SUPPORT VECTOR MACHINE; POWER SPECTRAL DENSITY; ACOUSTIC-EMISSION; FAULT-DIAGNOSIS; SEPARATION; VIBRATION; SIGNALS;
D O I
10.1007/s00170-023-12375-0
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
The fused deposition modeling (FDM) process, also known as 3D printing, deals with the manufacture of parts by adding layers of fused filament. Research on manufacturing process monitoring is on the rise, with an emphasis on investigating low-cost transducers as substitutes for the traditional, pricier options. The present study addresses a critical gap in the literature concerning the monitoring of the FDM process using acoustic signals from an electret microphone attached to the extruder. By employing an extensive signal processing and feature extraction analysis, including RMS values, ratio of power (ROP), and count statistics, this research uncovers distinguishable patterns in raw signals that relate to different machine conditions such as normal operation, extruder clogging, and filament shortages. Additionally, machine learning algorithms, specifically neural networks and support vector machine (SVM), are utilized to classify these machine conditions. Notably, signal filtering is found to significantly improve the classification models. The spectral analysis further contributes to characterizing the printing process, especially in identifying frequency values associated with defects. In conclusion, the methodology developed in this study holds promise for real-time monitoring systems, as it showcases high accuracy in classifying machine conditions and offers the potential to ensure quality and detect anomalies early in the printing process. Future research is encouraged to refine the methodology and explore its scalability across different FDM systems and materials.
引用
收藏
页码:1769 / 1786
页数:18
相关论文
共 69 条
[31]   Low-cost automatic identification of nozzle clogging in material extrusion 3D printers [J].
Lambos, Nikolaos ;
Vosniakos, George-Christopher ;
Papazetis, George .
30TH INTERNATIONAL CONFERENCE ON FLEXIBLE AUTOMATION AND INTELLIGENT MANUFACTURING (FAIM2021), 2020, 51 :274-279
[32]   Real-time distortion monitoring during fused deposition modeling via acoustic emission [J].
Li, Feng ;
Yu, Zhonghua ;
Yang, Zhensheng ;
Shen, Xuanwei .
STRUCTURAL HEALTH MONITORING-AN INTERNATIONAL JOURNAL, 2020, 19 (02) :412-423
[33]   Real-time polymer flow state monitoring during fused filament fabrication based on acoustic emission [J].
Li, Hao ;
Yu, Zhonghua ;
Li, Feng ;
Kong, Qingshun ;
Tang, Jie .
JOURNAL OF MANUFACTURING SYSTEMS, 2022, 62 :628-635
[34]   An improved fault diagnosis approach for FDM process with acoustic emission [J].
Liu, Jie ;
Hu, Youmin ;
Wu, Bo ;
Wang, Yan .
JOURNAL OF MANUFACTURING PROCESSES, 2018, 35 :570-579
[35]  
Lopes Thiago Glissoi, 2019, 2019 7th International Engineering, Sciences and Technology Conference (IESTEC), P544, DOI 10.1109/IESTEC46403.2019.00103
[36]  
Lopes T.G., 2022, Eng. Proc, V27, DOI [10.3390/ecsa-9-13285, DOI 10.3390/ECSA-9-13285]
[37]   Study of the Use of Piezoelectric Diaphragm as a Low-Cost Alternative to the Acoustic Emission Sensor in Dressing Operation of Aluminum Oxide Wheels [J].
Lopes, Wenderson N. ;
Aguiar, Paulo R. ;
Conceicao Junior, Pedro O. ;
Dotto, Fabio R. L. ;
Fernandez, Breno O. ;
Bianchi, Eduardo C. .
IEEE SENSORS JOURNAL, 2021, 21 (16) :18055-18062
[38]   Digital signal processing of acoustic emission signals using power spectral density and counts statistic applied to single-point dressing operation [J].
Lopes, Wenderson Nascimento ;
Ferreira, Fabio Isaac ;
Alexandre, Felipe Aparecido ;
Santos Ribeiro, Danilo Marcus ;
Conceicao Junior, Pedro de Oliveira ;
de Aguiar, Paulo Roberto ;
Bianchi, Eduardo Carlos .
IET SCIENCE MEASUREMENT & TECHNOLOGY, 2017, 11 (05) :631-636
[39]  
Lyu J., 2019, 39 COMP INF ENG C, DOI [10.1115/DETC2019-97963, DOI 10.1115/DETC2019-97963]
[40]   Tool Condition Monitoring of Single-Point Dresser Using Acoustic Emission and Neural Networks Models [J].
Martins, Cesar H. R. ;
Aguiar, Paulo R. ;
Frech, Arminio, Jr. ;
Bianchi, Eduardo Carlos .
IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT, 2014, 63 (03) :667-679