Process data based estimation of tool wear on punching machines using TCN-Autoencoder from raw time-series information
Tóm tắt
Tracking the wear states of tools on punching machines is necessary to reduce scrap rates. In this paper, we propose a method to estimate wear state of punches using Temporal Convolutional Network Autoencoder (TCN-Autoencoder), one of the deep learning techniques for learning time-series information with convolutional architecture. Approach involves inputting raw time-series information, such as sensor, vibration and audio data, into TCN-Autoencoder, and calculating the reconstruction error between the output and the input data. The reconstruction error is used as “anomaly score” and indicates the distance from the normal state. By training TCN-Autoencoder only with data annotated as “normal” state, the reconstruction error becomes larger when inputting abnormal state data, which corresponds the wear state of the punch. Performance is evaluated on experimental measurement data that spans various wear states of the punch. The results showed our model can estimate anomalies faster than the conventional machine-learning-based anomaly estimation method, while maintaining the high estimation accuracy. This is due to TCN-Autoencoder being able to learn from both frequency and time domain.
Từ khóa
Tài liệu tham khảo
De Silva, 2007
Byrne, 1995, Tool condition monitoring(TCM)—the status of research and industrial application, CIRP Ann-Manuf Techn, 44, 541, 10.1016/S0007-8506(07)60503-4
Jin, 2000, Diagnostic feature extraction from stamping tonnage signals based on design of experiments, J Manuf Sci E-T ASME, 122, 360, 10.1115/1.538926
Li, 2008, Transient dynamical analysis of strain signals in sheet metal stamping processes, Int J Mach Tool Manu, 48, 576, 10.1016/j.ijmachtools.2007.06.010
Doege, 2001, 172
Klingenberg, 2004, 169
Lee, 1997, Automatic supervision of blanking tool wear using pattern recognition analysis, Int J Mach Tool Manu, 37, 1079, 10.1016/S0890-6955(97)88104-7
Breitling, 1997, Process control in blanking, J Mater Process Tech, 71, 187, 10.1016/S0924-0136(97)00167-2
Ge, 2002, Feature extraction from energy distribution of stamping process using wavelet transform, J Vib Control, 8, 1023, 10.1177/107754602029577
Sari, 2017, Preliminary study for online monitoring during the punching process, Int J Adv Manuf Technol, 88, 2275, 10.1007/s00170-016-8956-y
Zhang, 2002, Bispectral analysis for on-line monitoring of stamping operation, Eng Appl Artif Intel, 15, 97, 10.1016/S0952-1976(02)00007-6
Kim, 1983, Punch press monitoring with acoustic emission (AE) part I: signal characterization and stock hardness effects, J Eng Mater-TASME, 105, 295, 10.1115/1.3225661
Wadi, 1997, Using neural networks to model the blanking process, J Mater Process Tech, 91, 52, 10.1016/S0924-0136(98)00426-9
Schuler, 1998
Ubhayaratne, 2015
Mardapittas, 1992, Expert system for tool wear monitoring in blanking intelligent fault diagnosis-part 1
Baldi, 2012, 37
Lea, 2017, 156
Ide, 2015, Anomaly detection and change detection
Cover, 1967, Nearest neighbor pattern classification, IEEE transactions on information theory, 13, 21, 10.1109/TIT.1967.1053964
Khreich, 2017, An anomaly detection system based on variable N-gram features and one-class SVM, Information and Software Technology, 91, 186, 10.1016/j.infsof.2017.07.009
Abadi, 2016, 265