Fault Diagnosis and Prediction System for Metal Wire Feeding Additive Manufacturing

• Published in: Sensors (Basel, Switzerland) Vol. 24; no. 13; p. 4277
• Main Authors: Xie, Meng, Shi, Zhuoyong, Yue, Xixi, Ding, Moyan, Qiu, Yujiang, Jia, Yetao, Li, Bobo, Li, Nan
• Format: Journal Article
• Language: English
• Published: Switzerland MDPI AG 01.07.2024
• Subjects: Accuracy; Additive manufacturing; Data processing; Deep learning; Design; Fault diagnosis; Humidity; LabVIEW; LSTM neural network; Neural networks; Nuclear power plants; Sensors; Software; Unmanned aerial vehicles; additive manufacturing; fault diagnosis; LSTM neural network; LabVIEW
• ISSN: 1424-8220; 1424-8220
• DOI: 10.3390/s24134277

In the process of metal wire and additive manufacturing, due to changes in temperature, humidity, current, voltage, and other parameters, as well as the failure of machinery and equipment, a failure may occur in the manufacturing process that seriously affects the current situation of production efficiency and product quality. Based on the demand for monitoring of the key impact parameters of additive manufacturing, this paper develops a parameter monitoring and prediction system for the additive manufacturing feeding process to provide a basis for future fault diagnosis. The fault diagnosis and prediction system for metal wire supply and additive manufacturing utilizes STM 32 as its core, enabling the capture and transmission of temperature, humidity, current, and voltage data. The upper computer system, designed on the LabVIEW 2019 virtual instrument platform, incorporates an LSTM neural network model and facilitates a connection between LabVIEW and MATLAB 2019 to achieve the prediction function. The monitoring and prediction system established in this study is intended to provide basic research assistance in the field of fault diagnosis.