Hıdır Selçuk NOĞAY
Energy transmission lines are exposed to an extra force due to external factors such as snow, ice, wind, rain, sun etc. Therefore, conductors are chosen to withstand such adverse effects. Conductors used in overhead transmission lines can be classified as copper conductor, all aluminum conductor (AAC), all aluminum alloy conductor (AAAC) and aluminium conductor steel reinforced (ACSR) according to the material used. Copper conductor is not preferred much in overhead transmission lines due to reasons such as weight and cost. As the voltage increases in overhead lines and the diameter of the conductor decreases, the electric field intensity will increase, so a corona event occurs. Multi-bundle conductors are used because the corona phenomenon becomes important at voltages above 220 kV. It is important to know the class of the conductor used in any malfunction that may occur in overhead lines. However, the classification process should be performed here based on the overhead transmission line cable images obtained from afar or close-up based on the image. Because it is important to be able to automatically determine the class of the cable used in a transmission line in any intervention, in order to be able to detect early and continue the repair or replacement process with a cable of the same class. In this respect, in order to solve this problem to some extent, an image-based triple automatic classification process has been carried out in this study. For this, according to the material used first, overhead transmission line conductors were accepted as two classes as aluminum and steel-cored-aluminum, and multi-bundle conductors were added to these two classes as a third class. In this way triple classification according to both the material used and the multi-bundle preferences was performed by applying an image-based deep learning model. As a result of the study, using the convolutional neural network model, overhead transmission line cables can be classified based on images with good accuracy. ORCID NO: 0000-0001-9105-508X

Anahtar Kelimeler: CNN, TL, Transmission Lines, AAC, AAAC, ACSR