In the electric railway system, catenary support components (CSCs) are critical elements that sustain the catenary system’s contact line and messenger. Precise location and identification of component types are vital to subsequent troubleshooting. There are many types of CSCs with muti-scales, and previous work could not effectively deal with them. In this study, we present a novel positioning strategy for accurately locating CSCs. Firstly, we use a transformer framework to extract detailed component features. Secondly, we use FPN to fuse multiscale features on the feature map. Following that, we trained a detector with several detection heads, gradually increasing the threshold of the detection heads to improve detection accuracy. In the experiment, the adopted ST has a higher detection accuracy for CSCs when compared with several current competitive methods. Finally, we discussed the current status and prospects of deep learning for CSCs positioning.
Swin Transformer-Based Positioning Methodology for Catenary Support Components in High-Speed Railway
Lect. Notes Electrical Eng.
International Conference on Electrical and Information Technologies for Rail Transportation ; 2021 October 21, 2021 - October 23, 2021
Proceedings of the 5th International Conference on Electrical Engineering and Information Technologies for Rail Transportation (EITRT) 2021 ; Kapitel : 71 ; 627-636
22.02.2022
10 pages
Aufsatz/Kapitel (Buch)
Elektronische Ressource
Englisch
Swin Transformer-Based Positioning Methodology for Catenary Support Components in High-Speed Railway
British Library Conference Proceedings | 2022
|Catenary Support Components and Their Characteristics in High-Speed Railways
Springer Verlag | 2023
|High-speed railway catenary dropper calibration special operation vehicle
Europäisches Patentamt | 2021
|