The Laminated Pipeline Material Delamination Identification Based on Artificial Intelligence Schemes

Electrical Potential Change (EPC)-based faults detection is a promising methods in new ways. In this paper, BFRP composite pipeline system is used for analysis by using Electrical Capacitance Sensor (ECS) for delamination monitoring inside layers of pipe. The dielectric properties of the pipe material are gauged for intact and delamination stats. The array and capacitance value of the electrical contact, the change of capacitance and the changes in the distribution of node potential are tested. The results show that these diagrams can reflect the delamination information of the pipeline according to the author’s previous work. However, because the model is 3D and has different bending and twisting effects, these diagrams may not be a good choice to accurately detect delamination information. So, a novel Convolutional Neural Network (CNN) algorithm is employed to train and test the EPC diagrams to detect the delamination. The current technique have high values of accuracy rate (P%), recall rate (R%), and F-score (F%) equal to 95.2%, 93.7%, and 90.9% respectively, this indicates that the current technique show the recognition efficiency and accuracy.