| Abstrakt: |
Magnetic flux leakage (MFL) testing is an important way of non-destructive testing, preventing some major accidents of hoist equipment by identifying the damage of wire ropes, whereas, in some working conditions such as mines and oil wells, the inevitable vibration and harsh environment will generate noise and interfere with the MFL signal, which makes it difficult to identify the damage. As a classification network, convolutional neural network is positive in recognition accuracy and noise resistance, but it is hardly used in the damage identification of wire rope. To improve the recognition accuracy of damage identification under strong noise background, we propose a method of wire rope damage identification via Light-EfficientNetV2 and MFL image. First, the MFL signal is segmented and rearranged to form the MFL image, and then, the image is classified by Light-EfficientNetV2. Then, to improve the efficiency, we reduce the layers of EfficientNetV2 to make it lighter. Finally, the availability of this method is proved by the validation set. Within five neural networks, the recognition accuracy of Light-EfficientNetV2 is the highest. Moreover, as the noise intensity increases, the recognition accuracy of Light-EfficientNetV2 is higher than EfficientNetV2, which has important value in the application of the wire rope damage identification under strong noise background. [ABSTRACT FROM AUTHOR] |
Full text from SpringerLink