Machine auscultation: enabling machine diagnostics using convolutional neural networks and large-scale machine audio data

Abstract

Acoustic signals play an essential role in machine state monitoring. Efficient processing of real-time machine acoustic signals improves production quality. However, generating semantically useful information from sound signals is an ill-defined problem that exhibits a highly non-linear relationship between sound and subjective perceptions. This paper outlines two neural network models to analyze and classify acoustic signals emanating from machines: (i) a backpropagation neural network (BP-NN); and (ii) a convolutional neural network (CNN). Microphones are used to collect acoustic data for training models from a computer numeric control (CNC) lathe. Numerical experiments demonstrate that CNN performs better than the BP-NN.

Keywords

Acoustic signal processing / Machine performance / Backpropagation neural network (BP-NN) / Convolutional neural network (CNN)

References

[1] Huang HB, Huang XR, Li RX, et al. Sound quality prediction of vehicle interior noise using deep belief networks. Appl Acoust, 2016, 113: 149-161.

[2] Sharan RV, Moir TJ. An overview of applications and advancements in automatic sound recognition. Neurocomputing, 2016, 200: 22-34.

[3] Kumon M, Yoshihiro ITO, Nakashima T, et al. Sound source classification using support vector machine. IFAC Proc Vol, 2007, 40(13): 465-470.

[4] Thaler T, Potočnik P, Bric I, et al. Chatter detection in band sawing based on discriminant analysis of sound features. Appl Acoust, 2014, 77: 114-121.

[5] Dombovari Z, Barton DAW, Wilson RE, et al. On the global dynamics of chatter in the orthogonal cuttingmodel. Int J Non-lin Mech, 2011, 46(1): 330-338.

[6] Pan G, Xu H, Kwan CM, et al. Modeling and intellligent chatter control strategies for a lathe machine. Control Eng Pract, 1996, 4(12): 1647-1658.

[7] https://github.com/ruoyuyang1991/machine-auscultation-classification. Accessed 10 April 2019

[8] Singh KK, Singh R, Kartik V. Comparative study of chatter detection methods for high-speed micromilling of Ti6Al4V. Proc Manufacturing, 2015, 1(1): 593-606.

[9] Li XQ, Wong YS, Nee AYC. Tool wear and chatter detection using the coherence function of two crossed accelerations. Int J Mach Tools Manuf, 1997, 37(4): 425-435.

[10] Kuljanic E, Sortino M, Totis G. Multisensor approaches for chatter detection in milling. J Sound Vib, 2008, 312(4–5): 672-693.

[11] Toh CK. Vibration analysis in high speed rough and finish milling hardened steel. J Sound Vib, 2004, 278(1): 101-115.

[12] Soliman E, Ismail F. Chatter detection by monitoring spindle drive current. Int J Adv Manuf Technol, 1997, 13(1): 27-34.

[13] Delio T, Tlusty J, Smith S. Use of audio signals for chatter detection and control. J Manuf Sci Eng, 1992, 114(2): 146-157.

[14] Li XQ, Wong YS, Nee AYC. A comprehensive identification of tool failure and chatter using a parallel multi-art2 neural network. J Manuf Sci Eng, 1988, 120(2): 433-442.

[15] Jiang AY, Zhang C (2006) Hybrid HMM/SVM method for predicting cutting chatter. In: Proceedings of SPIE–the international society for optical engineering 6280:62801Q-8

[16] Yao ZH, Mei DQ, Chen ZC. On-line chatter detection and identification based on wavelet and support vector machine. J Mater Process Technol, 2010, 210(5): 713-719.

[17] Bediaga I, Munoa J, Hernández J, et al. An automatic spindle speed selection strategy to obtain stability in high-speed milling. Int J Mach Tools Manuf, 2009, 49(5): 384-394.

[18] Zhang CL, Yue X, Jiang YT. A hybrid approach of ann and hmm for cutting chatter monitoring. Adv Mater Res, 2010, 97: 3225-3232.

[19] Cao HR, Lei YG, He ZJ. Chatter identification in end milling process using wavelet packets and hilbert–huang transform. Int J Mach Tools Manuf, 2013, 69: 11-19.

[20] Kondo E, Ota H, Kawai T. A new method to detect regenerative chatter using spectral analysis, part 1: Basic study on criteria for detection of chatter. J Manuf Sci Eng, 1997, 119(4A): 461-466.

[21] Tansel IN, Wang X, Chen P, et al. Transformations in machining. Part 2: evaluation of machining quality and detection of chatter in turning by using s-transformation. Int J Mach Tools Manuf, 2006, 46(1): 43-50.

[22] Cho DW, Eman KF. Pattern recognition for on- line chatter detection. Mech Syst Signal Process, 1988, 2(3): 279-290.

[23] Grabec I, Gradišek J, Govekar E. A new method for chatter detection in turning. CIRP Ann Manuf Technol, 1999, 48(1): 29-32.

[24] Berger B, Belai C, Anand D. Chatter identification with mutual information. J Sound Vib, 2003, 267(1): 178-186.

[25] Choi T, Shin YC. On-line chatter detection using wavelet-based parameter estimation. Trans-Am Soc Mech Eng J Manuf Sci Eng, 2003, 125(1): 21-28.

[26] Lécun Y, Bottou L, Bengio Y, et al. Gradient-based learning applied to document recognition. Proc IEEE, 1998, 86(11): 2278-2324.

[27] Krizhevsky A, Sutskever I, Hinton GE. Imagenet classification with deep convolutional neural networks. Adv Neural Inf Process Syst, 2012, 1: 1097-1105.

[28] Simonyan K, Zisserman A (2014) Very deep convo- lutional networks for large-scale image recognition. arXiv preprint arXiv:1409–1556

[29] Szegedy C, Liu W, Jia YQ er al (2015) Going deeper with convolutions. In: Proceedings of the IEEE conference on computer vision and pattern recognition, pp 1–9

[30] Girshick R, Donahue J, Darrell T (2014) Rich feature hierarchies for accurate object detection and semantic segmentation. In: Proceedings of the IEEE conference on computer vision and pattern recognition, pp 580–587

[31] Girshick R (2015) Fast R-CNN. In: Proceedings of the IEEE international conference on computer vision, pp 1440–1448

[32] Ren SQ, He KM, Girshick R, et al. Faster R-CNN: Towards real-time object detection with region proposal networks. Adv Neural Inform Process Syst, 2015, 1: 91-99.

[33] LeCun Y, Bengio Y, Hinton G. Deep learning. Nature, 2015, 521(7553): 436-444.

[34] Piczak KJ (2015) Environmental sound classification with convolutional neural networks. In: IEEE 25th International workshop on machine learning for signal processing (MLSP), 2015, pp 1–6

[35] Boddapatia V, Petef A, Rasmusson J, et al. Classifying environmental sounds using image recognition networks. Proc Comput Sci, 2017, 112: 2048-2056.

[36] Fu Y, Zhang Y, Gao Y, et al. Machining vibration states monitoring based on image representation using convolutional neural networks. Eng Appl Artif Intell, 2017, 65: 240-251.

[37] Loh WL. On latin hypercube sampling. Ann Stat, 1996, 24(5): 2058-2080.

[38] Vaseghi SV. Advanced digital signal processing and noise reduction, 2008, New York: Wiley.

[39] Gupta CN, Palaniappan R, Swaminathan S, et al. Neural network classification of homomorphic segmented heart sounds. Appl Soft Comput, 2007, 7(1): 286-297.

[40] Kotani M, Katsura M, Ozawa S. Detection of gas leakage sound using modular neural networks for unknown environments. Neurocomputing, 2004, 62: 427-440.

[41] Potočnik P, Thaler T, Govekar E. Multisensory chatter detection in band sawing. Proc CIRP, 2013, 8: 469-474.

[42] Liang HY, Nartimo I (1998) A feature extraction algorithm based on wavelet packet decomposition for heart sound signals. In: Proceedings of the IEEE-SP international symposium, pp 93–96

[43] Babaei S, Geranmayeh A. Heart sound reproduction based on neural network classification of cardiac valve disorders using wavelet transforms of pcg signals. Comput Biol Med, 2009, 39(1): 8-15.

[44] O’shaughnessy D. Speech communication: human and machine, 1987, Hyderabad: Universities Press.

[45] Krizhevsky A, Hinton G. Learning multiple layers of features from tiny images. Technical report, University of Toronto, 2009, 1(4): 7

[46] Srivastava N, Hinton G, Krizhevsky A, et al. Dropout: A simple way to prevent neural networks from overfitting. J Mach Learn Res, 2014, 15(1): 1929-1958.