Modelling of improved LSTM +1D convolution neural network methods for the diagnosis of SKF bearings

  • Authors

    • Signe Feudjeu Josias Éric University of Ngaoundere
    • Nzié Wolfgang University of Ngaoundere
    • Ngnassi Djami Aslain Brisco Dr
    • Ntsama Eloundou Pascal University of Bertoua
    • Chassem Priva University of Dschang
    2024-06-18
    https://doi.org/10.14419/pf65qa49
  • Modeling; Diagnostics; Convolutional Neural Networks; LSTM.

  • The ability to accurately detect and predict faults in automotive bearings is essential for diagnostic applications in the maintenance process. Although previous methods can accurately identify the various faults on bearings, they mostly produce erroneous results in the presence of certain mechanical factors when classifying the data. We propose a new diagnostic framework based on one-dimensional convolutional neural network (CONV1D) modelling and improved long short-term memory (LSTM), together with confusion matrices to evaluate data classification using the Deep Learning algorithm. Our framework classifies the data by taking into account the mechanical factors of the bearings (sudden load, rotation speed, operating temperature, etc.). Our results improve the training accuracy of the model to over 96.6%, with a percentage error of 23.29% for 50 iterations (repetitions). This percentage of training accuracy could be closer to 100% and that of the error margins to 0% if we increase the number of iterations. These results underline the promise of our method across our model and indicate how future expansion of the model by combining three methods can lead to further improvements in training accuracy with fewer errors and fewer iterations.

  • References

    1. Gu, K., Zhang, Y., Liu, X., Li, H., & Ren, M. (2021). DWTLSTM-Based Fault Diagnosis of Rolling Bearings with Multi-Sensors. Elec-tronics, 10 (7), 2076. https://doi.org/10.3390/electronics10172076.
    2. Oh, S., Han, S., & Jeong, J. (2021). Multi-scale convolutional recurrent neural network for bearing fault detection in noisy manufactur-ing environments. Applied Sciences, 11(9), 3963. https://doi.org/10.3390/app11093963.
    3. Zhao, Z., Xu, Q., & Jia, M. (2016). Improved shuffled frog leaping algorithm-based BP neural network and its application in bearing early fault diagnosis. Neural Computing and Applications, 27, 375-385. https://doi.org/10.1007/s00521-015-1850-y.
    4. Khorram, A., Khalooei, M., & Rezghi, M. (2021). End-to-end CNN+ LSTM deep learning approach for bearing fault diagnosis. Ap-plied Intelligence, 51(2), 736-751. https://doi.org/10.1007/s10489-020-01859-1.
    5. Qiao, M., Yan, S., Tang, X., & Xu, C. (2020). Deep convolutional and LSTM recurrent neural networks for rolling bearing fault diag-nosis under strong noises and variable loads. Ieee Access, 8, 66257-66269. https://doi.org/10.1109/ACCESS.2020.2985617.
    6. Slavič, J., Brković, A., & Boltežar, M. (2011). Typical bearing-fault rating using force measurements: application to real data. Journal of Vibration and Control, 17(14), 2164-2174. https://doi.org/10.1177/1077546311399949.
    7. Amar Chiter. (2001). Detection and diagnosis of bearing faults: contribution to the maintenance of rotating machines. Master’s thesis in optics and accuracy mechanics». UFAS.
    8. Schmidhuber, J. (2015). Deep learning in neural networks: An overview. Neural networks, 61, 85-117. https://doi.org/10.1016/j.neunet.2014.09.003.
    9. Pan, H., He, X., Tang, S., & Meng, F. (2018). An improved bearing fault diagnosis method using one-dimensional CNN and LSTM. Journal of Mechanical Engineering/Strojniški Vestnik, 64.
    10. Fukushima, K. (1980). Neocognitron: A self-organizing neural network model for a mechanism of pattern recognition unaffected by shift in position. Biological cybernetics, 36(4), 193-202. https://doi.org/10.1007/BF00344251.
    11. Wen, L., Li, X., Gao, L., & Zhang, Y. (2017). A new convolutional neural network-based data-driven fault diagnosis method. IEEE Transactions on Industrial Electronics, 65(7), 5990-5998. https://doi.org/10.1109/TIE.2017.2774777.
    12. Hochreiter, S., & Schmidhuber, J. (1997). Long short-term memory. Neural computation, 9(8), 1735-1780. https://doi.org/10.1162/neco.1997.9.8.1735.
    13. Huang, H., & Baddour, N. (2018). Bearing vibration data collected under time-varying rotational speed conditions. Data in brief, 21, 1745-1749. https://doi.org/10.1016/j.dib.2018.11.019.

  • Downloads

  • How to Cite

    Feudjeu Josias Éric , S. ., Wolfgang, N. ., Djami Aslain Brisco , N. ., Eloundou Pascal , N. ., & Priva , C. (2024). Modelling of improved LSTM +1D convolution neural network methods for the diagnosis of SKF bearings. International Journal of Engineering & Technology, 13(2), 194-203. https://doi.org/10.14419/pf65qa49