Deep Learning Based Arc Detection in Pantograph-Catenary Systems

Abstract

Pantograph-catenary systems are the most important parts of electric trains. Faults that occur in pantograph-catenary systems seriously affect railway transportation. Arcs are the most important reporters of pantograph-catenary systems. Detection of arcs that give early signal of these faults is very important. In this paper, an approach using deep learning is proposed for the detection of arcs in pantograph-catenary systems. Arc detection is performed using CNN (Convolutional Neural Network). Deep learning have gained great importance in recent years. In this study, experimental results show that the proposed method is quite successful in detecting the arc.