Iranian Journal of Manufacturing Engineering

Iranian Journal of Manufacturing Engineering

Fault diagnosis radial bearing mounted on shaft using classification with audio spectrogram and convolutional neural network

Document Type : Original Article

Authors
Mehdi Ustad
Amir Rasti
Armin Rezaei
Department of Mechanical Engineering, Tarbiat Modares University, Tehran, Iran
10.22034/ijme.2025.521365.2077
Abstract
Bearing failures are a primary cause of breakdowns in rotating machinery. As such, there is an increasing demand for effective bearing fault detection methods to prevent machinery failures. Previous studies have explored temperature measurement and vibration monitoring for fault detection, but these approaches face limitations due to noise interference. Consequently, researchers have turned to sound signal monitoring. Among modern techniques, Mel Frequency Cepstral Coefficients (MFCCs), audio spectrograms, and two-dimensional Convolutional Neural Networks (CNNs) have attracted significant interest. However, existing MFCC methods require high sampling rates and wide frequency bands. In this study, we propose a spectrogram-based method emphasizing frequency features and noise reduction using filters, integrated with an optimized CNN. The spectrogram analyzes a narrow frequency band and generates low-resolution images, which are then processed by a CNN designed with convolutional and fully connected layers. Experimental results demonstrate that the proposed system achieves 99. 88% accuracy with reduced complexity. The optimized CNN has 622.77 kB of parameters and 1.53×106 FLOPs. This fault diagnosis system proves effective even under varying rotational frequencies and low sampling rates.
Keywords
Fault diagnosis
Classification
Machine learning
Convolutional neural network
Radial Bearing
[1] Grebenik J, Zhang Y, Bingham C, Srivastava S. Roller element bearing acoustic fault detection using smartphone and consumer microphones comparing with vibration techniques. In2016 17th International Conference on Mechatronics-Mechatronika (ME) 2016 Dec 7 (pp. 1-7). IEEE.
[2] Rasti A, Jafari D, Rafierad I. Design and performance analysis of integrated roller bearing and roller screw mechanism for precision linear motion. Iranian Journal of Manufacturing Engineering. 2025 Apr 21;12(2):59-69. [in Persian] doi: 10.22034/ijme.2025.495859.2038
[3] Ramezani NM, Rasti A, Sadeghi MH, Pour BJ, Hajideh MR. Experimental study of tool wear and surface roughness on high speed helical milling in D2 steel. Modares Mech Eng. 2016;15(20):198-202. [in Persian]
[4] Hassanpour H, Rasti A, Sadeghi MH, Khosrowshahi JH. Investigation of roughness, topography, microhardness, white layer and surface chemical composition in high speed milling of Ti-6Al-4V using minimum quantity lubrication. Machining Science and Technology. 2020 Sep 2;24(5):719-38. doi: 10.1080/10910344.2020.1752237
[5] Zeinolabedin-Beygi A, Naeini HM, Talebi-Ghadikolaee H, Rabiee AH, Hajiahmadi S. Predictive modeling of spring-back in pre-punched sheet roll forming using machine learning. The Journal of Strain Analysis for Engineering Design. 2024 Oct;59(7):463-74. doi: 10.1177/03093247241263685
[6] Hajiahmadi S, Moslemi Naeini H, Talebi-Ghadikolaee H, Safdarian R, Zeinolabedin-Beygi A. Insights into spring-back prediction: a comparative analysis of constitutive models for perforated U-shaped roll-formed steel profiles. The International Journal of Advanced Manufacturing Technology. 2024 Sep;134(3):1915-33. doi: 10.1007/s00170-024-14211-5
[7] Zhang S, Zhang S, Wang B, Habetler TG. Deep learning algorithms for bearing fault diagnostics—A comprehensive review. IEEE access. 2020 Feb 10;8:29857-81. doi: 10.1109/ACCESS.2020.2972859
[8] Abdenebi R, Fethi D, Abdelkrim N, Badis D. Gearbox fault diagnosis using the short-time cepstral features. In2022 2nd International Conference on Advanced Electrical Engineering (ICAEE) 2022 Oct 29 (pp. 1-8). IEEE. doi: 10.1109/ICAEE53772.2022.9962139
[9] Kumar A, Vashishtha G, Gandhi CP, Zhou Y, Glowacz A, Xiang J. Novel convolutional neural network (NCNN) for the diagnosis of bearing defects in rotary machinery. IEEE Transactions on Instrumentation and Measurement. 2021 Feb 1;70:1-0. doi: 10.1109/TIM.2021.3055802
[10] Azeez AA, Alkhedher M, Gadala MS. Thermal imaging fault detection for rolling element bearings. In2020 Advances in science and engineering technology international conferences (ASET) 2020 Feb 4 (pp. 1-5). IEEE. doi: 10.1109/ASET48392.2020.9118361
[11] Liu H, Li L, Ma J. Rolling bearing fault diagnosis based on STFT‐deep learning and sound signals. Shock and Vibration. 2016;2016(1):6127479. doi: 10.1155/2016/6127479
[12] Sun D, Fan Y, Wang G. Gradient-oriented prioritization in meta-learning for enhanced few-shot fault diagnosis in industrial systems. Applied Sciences. 2023 Dec 25;14(1):181. doi: 10.3390/app14010181
[13] Shan S, Liu J, Wu S, Shao Y, Li H. A motor bearing fault voiceprint recognition method based on Mel-CNN model. Measurement. 2023 Feb 15;207:112408. doi: 10.1016/j.measurement.2022.112408
[14] Elhami S, Razfar MR, Farahnakian M, Rasti A. Application of GONNS to predict constrained optimum surface roughness in face milling of high-silicon austenitic stainless steel. The International Journal of Advanced Manufacturing Technology. 2013 May;66(5):975-86. doi: 10.1007/s00170-012-4382-y
[15] Li D, Li B, Wang C, Cheng P, Jiao B. A fault diagnosis equipment of motor bearing based on sound signal and CNN. InJournal of Physics: Conference Series 2021 Sep 1 (Vol. 2010, No. 1, p. 012159). IOP Publishing. doi: 10.1088/1742-6596/2010/1/012159
[16] Wang H, Liu Z, Peng D, Cheng Z. Attention-guided joint learning CNN with noise robustness for bearing fault diagnosis and vibration signal denoising. ISA transactions. 2022 Sep 1;128:470-84. doi: 10.1016/j.isatra.2021.11.028
[17] Wang X, Mao D, Li X. Bearing fault diagnosis based on vibro-acoustic data fusion and 1D-CNN network. Measurement. 2021 Mar 1;173:108518. doi: 10.1016/j.measurement.2020.108518
[18] Chen J, Jiang J, Guo X, Tan L. An Efficient CNN with Tunable Input-Size for Bearing Fault Diagnosis. Int. J. Comput. Intell. Syst.. 2021 Jan 1;14(1):625-34. doi: 10.2991/ijcis.d.210113.001
[19] Xia M, Li T, Xu L, Liu L, De Silva CW. Fault diagnosis for rotating machinery using multiple sensors and convolutional neural networks. IEEE/ASME transactions on mechatronics. 2017 Jul 17;23(1):101-10. doi: 10.1109/TMECH.2017.2728371
[20] Akpudo UE, Hur JW. A cost-efficient MFCC-based fault detection and isolation technology for electromagnetic pumps. Electronics. 2021 Feb 10;10(4):439. doi: 10.3390/electronics10040439
[21] Zhao Y, Zhang N, Zhang Z, Xu X. Bearing fault diagnosis based on mel frequency cepstrum coefficient and deformable space-frequency attention network. IEEe Access. 2023 Apr 3;11:34407-20. doi: 10.1109/ACCESS.2023.3264276
[22] Pham MT, Kim JM, Kim CH. Deep learning-based bearing fault diagnosis method for embedded systems. Sensors. 2020 Dec 2;20(23):6886. doi: 10.3390/s20236886
[23] Mohamad TH. Nonlinear Dynamics-Based Machine Learning Algorithms for Condition-Based Maintenance. Villanova University; 2021.
[24] Sehri M, Dumond P, Bouchard M. University of Ottawa constant load and speed rolling-element bearing vibration and acoustic fault signature datasets. Data in Brief. 2023 Aug 1;49:109327. doi: 10.1016/j.dib.2023.109327