Fault Detection in Spur Gears through Vibration Signal Analysis and Machine Learning Techniques

Abstract:

This research uses a multi-domain technique to give a thorough analysis of mechanical gears health evaluation that includes time, frequency, and time-frequency signal analysis. The research seeks to discover patterns indicative of healthy, partially damaged, or fully damaged gear states using a variety of graphical representations, including time and frequency plots, the Short-Time Fourier Transform (STFT), and scalograms which are visual representations of the wavelet transform of a signal. Advanced machine learning models are used to improve diagnostic accuracy when manual identification of these trends becomes difficult. The goal is to achieve a validation accuracy greater than 70% a threshold selected based on prior studies indicating that this level ensures reliable fault detection for industrial applications while balancing computational constraints. The reliability and effectiveness of gear monitoring systems can be increased by integrating contemporary signal processing and machine learning approaches, as demonstrated by this research, which also advances the identification of gear faults. Among the conclusions are the outcomes of tests done to identify gear problems in which authors were able to train a model with more than 72% accuracy and able to propose Artificial Intelligence model for classification of faults in gears.

• Periodical: Engineering Innovations (Volume 15)
• Pages: 21-32
• DOI: https://doi.org/10.4028/p-O9sSAg
• Online since: November 2025
• Authors: Muhammad Talha Asif, Shahzad Ahmad*, Asif Khan, Sohail Malik
• Keywords: Fault Diagnosis, Gears Condition Monitoring, Machine Learning, Scalograms, Short-Time Fourier Transform

References

[1] Nazarova Nafisa Ilhomovna., "The Importance of role of Gears in Mechanical Engineering," in Proceedings of the Thirteenth EuroSys Conference, Porto Portugal: ACM, Apr. 2018, p.1–15

DOI: 10.1145/3190508.3190551

[2] J. A. Faria, E. Nunes, and M. A. Matos, "Cost and quality of service analysis of production systems based on the cumulative downtime," Int. J. Prod. Res., vol. 48, no. 6, p.1653–1684, Mar. 2010.

DOI: 10.1080/00207540802068631

[3] E. P. Carden and P. Fanning, "Vibration based condition monitoring: A review," 2004.

DOI: 10.1177/1475921704047500

[4] N. H. Chandra and A. S. Sekhar, "Fault detection in rotor bearing systems using time frequency techniques," Mech Syst Signal Process, vol. 72–73, p.105–133, May 2016.

DOI: 10.1016/j.ymssp.2015.11.013

[5] T. Praveenkumar, M. Saimurugan, P. Krishnakumar, and K. I. Ramachandran, "Fault diagnosis of automobile gearbox based on machine learning techniques," in Procedia Engineering, Elsevier Ltd, 2014, p.2092–2098.

DOI: 10.1016/j.proeng.2014.12.452

[6] Y. Bella, A. Oulmane, and M. Mostefai, "Industrial Bearing Fault Detection Using Time-Frequency Analysis," 2018. [Online]. Available: www.etasr.com

DOI: 10.48084/etasr.2135

[7] P. Ong, A. John Koshy, K. H. Lai, C. K. Sia, and M. Ismon, "A deep learning approach for health monitoring in rotating machineries using vibrations and thermal features," Decision Analytics Journal, vol. 10, Mar. 2024.

DOI: 10.1016/j.dajour.2024.100399

[8] M. Sohaib et al., "Gearbox fault diagnosis using improved feature representation and multitask learning," Front Energy Res, vol. 10, Sep. 2022.

DOI: 10.3389/fenrg.2022.998760

[9] H. Sindi, M. Nour, M. Rawa, Ş. Öztürk, and K. Polat, "A novel hybrid deep learning approach including combination of 1D power signals and 2D signal images for power quality disturbance classification," Expert Syst Appl, vol. 174, Jul. 2021.

DOI: 10.1016/j.eswa.2021.114785

[10] F. Tan, "Improving Feature Selection Techniques for Machine Learning Improving Feature Selection Techniques for Machine Learning Feng Tan".

[11] F. Tan, "Improving Feature Selection Techniques for Machine Learning Improving Feature Selection Techniques for Machine Learning Feng Tan".

[12] M. A. Khamis, W. Gomaa, and W. F. Ahmed, "Machine learning in computational docking," Mar. 01, 2015, Elsevier.

DOI: 10.1016/j.artmed.2015.02.002

[13] F. P. García Márquez, A. M. Tobias, J. M. Pinar Pérez, and M. Papaelias, "Condition monitoring of wind turbines: Techniques and methods," Renew Energy, vol. 46, p.169–178, Oct. 2012.

DOI: 10.1016/j.renene.2012.03.003

[14] P. Nunes, J. Santos, and E. Rocha, "Challenges in predictive maintenance – A review," Feb. 01, 2023, Elsevier Ltd.

DOI: 10.1016/j.cirpj.2022.11.004

[15] H. Bhavsar and A. Ganatra, "A Comparative Study of Training Algorithms for Supervised Machine Learning," vol. 2, no. 4, 2012.

[16] L. Lévesque, "Revisiting the Nyquist criterion and aliasing in data analysis," Eur. J. Phys., vol. 22, no. 2, p.127–132, Mar. 2001.

DOI: 10.1088/0143-0807/22/2/304

[17] Y. Yu et al., "Dynamic control flow in large-scale machine learning," in Proceedings of the Thirteenth EuroSys Conference, Porto Portugal: ACM, Apr. 2018, p.1–15.

DOI: 10.1145/3190508.3190551

[18] I.O. Muraina, "Ideal Dataset Splitting Ratios in Machine Learning Algorithms: General Concerns for Data Scientists and Data Analysts".

[19] C. Schaffer and S. A. H. C. Edu, "Selecting a Classification Method by Cross-Validation," 1993.

[20] B. Al‐Najjar, "Impact of real‐time measurements of operating conditions on effectiveness and accuracy of vibration‐based maintenance policy ‐ A case study in paper mill," J. Qual. Maint. Eng., vol. 6, no. 4, p.275–287, Dec. 2000.

DOI: 10.1108/13552510010346815

[21] A. A. Rasheed, "Improving prediction efficiency by revolutionary machine learning models," Mater. Today Proc., vol. 81, p.577–583, 2023.

DOI: 10.1016/j.matpr.2021.04.014