Identification of buried depth of steel bars in concrete based on GAF-CNN

Reinforced concrete is widely used in the transmission of distribution networks due to its low cost and high reliability. When the internal steel structure is unreasonable or the steel quality does not meet the standards, the various properties of reinforced concrete will be greatly reduced, causing huge safety hazards. This paper takes concrete poles as the research object. Aiming at the current status of measuring the thickness of the steel protection layer in concrete, this paper proposes an off-line non-destructive detection method for the state of steel bars in concrete based on deep learning and pulsed eddy current. The time pattern of the received signal is explicitly encoded into different types of spatial images, namely Gramian Angular Fields (GAF), so that the convolutional neural network (CNN) is used to learn and classify the characteristics of the steel protection layer thickness measurement. A pulsed eddy current detection reinforced concrete model is built to collect the induced voltage changes due to the change of the buried depth of the steel bar, so as to verify the method in this paper, and it is concluded that GAF-CNN has good recognition accuracy in steel bar buried depth recognition.