Aiming at the problem of how to recognize the difficulty of identifying industrial equipment faults in the field of industrial production, this paper proposes an image recognition method for industrial equipment faults based on GWO-GoogleNet and transfer learning. First, the original signals of different types of faults of industrial equipment are time-frequency converted into images to generate corresponding data sets. Secondly, the deep learning network with the same parameter level is selected for noise resistance test, and the GoogleNet network with the best relative noise resistance is selected. Then, with the final error of the network as the optimization goal, the parameters of batch sampling number and learning rate are optimized, and the optimized network framework is completed. Finally, based on this network framework, its trained parameters are transferred to the target domain, and a small amount of target domain data is used to fine-tune the network. The method is verified on the bearing dataset of Western Reserve University, and the results show that the method has good recognition ability on bearing fault images, and still has good fault detection ability under migration conditions.


    Zugriff

    Zugriff prüfen

    Verfügbarkeit in meiner Bibliothek prüfen

    Bestellung bei Subito €


    Exportieren, teilen und zitieren



    Titel :

    Fault image recognition method of industrial equipment based on GWO-GoogleNet and transfer learning


    Beteiligte:
    Sun, Zecheng (Autor:in)


    Erscheinungsdatum :

    11.10.2023


    Format / Umfang :

    3929526 byte




    Medientyp :

    Aufsatz (Konferenz)


    Format :

    Elektronische Ressource


    Sprache :

    Englisch



    Deep learning algorithm for autonomous driving using GoogLeNet

    Al-Qizwini, Mohammed / Barjasteh, Iman / Al-Qassab, Hothaifa et al. | IEEE | 2017


    Deep Learning Algorithm for Autonomous Driving Using GoogLeNet

    Al-Qizwini, Mohammed / Barjasteh, Iman / Al-Qassab, Hothaifa et al. | British Library Conference Proceedings | 2017