Exploring the Processing Paradigm of Input Data for End-to-End Deep Learning in Tool Condition Monitoring

• Published in: Sensors (Basel, Switzerland) Vol. 24; no. 16; p. 5300
• Main Authors: Wang, Chengguan, Wang, Guangping, Wang, Tao, Xiong, Xiyao, Ouyang, Zhongchuan, Gong, Tao
• Format: Journal Article
• Language: English
• Published: Switzerland MDPI AG 15-08-2024; MDPI
• Subjects: Accuracy; Algorithms; bidirectional long short-term memory; convolutional neural network; Deep learning; end-to-end; Engineering; input data; Manufacturing; Methods; Neural networks; Sensors; Signal processing; Time series; tool condition monitoring; end-to-end; deep learning; bidirectional long short-term memory; convolutional neural network; input data; tool condition monitoring
• ISSN: 1424-8220; 1424-8220
• DOI: 10.3390/s24165300

Tool condition monitoring technology is an indispensable part of intelligent manufacturing. Most current research focuses on complex signal processing techniques or advanced deep learning algorithms to improve prediction performance without fully leveraging the end-to-end advantages of deep learning. The challenge lies in transforming multi-sensor raw data into input data suitable for direct model feeding, all while minimizing data scale and preserving sufficient temporal interpretation of tool wear. However, there is no clear reference standard for this so far. In light of this, this paper innovatively explores the processing methods that transform raw data into input data for deep learning models, a process known as an input paradigm. This paper introduces three new input paradigms: the downsampling paradigm, the periodic paradigm, and the subsequence paradigm. Then an improved hybrid model that combines a convolutional neural network (CNN) and bidirectional long short-term memory (BiLSTM) was employed to validate the model's performance. The subsequence paradigm demonstrated considerable superiority in prediction results based on the PHM2010 dataset, as the newly generated time series maintained the integrity of the raw data. Further investigation revealed that, with 120 subsequences and the temporal indicator being the maximum value, the model's mean absolute error (MAE) and root mean square error (RMSE) were the lowest after threefold cross-validation, outperforming several classical and contemporary methods. The methods explored in this paper provide references for designing input data for deep learning models, helping to enhance the end-to-end potential of deep learning models, and promoting the industrial deployment and practical application of tool condition monitoring systems.