Automatic Processing Analysis Of Infrared Images For Monitoring Pantograph Catenary Interactions
Price
Free (open access)
Transaction
Volume
46
Pages
10
Published
2007
Size
982 kb
Paper DOI
10.2495/CMEM070801
Copyright
WIT Press
Author(s)
A. Balestrino, O. Bruno, A. Landi & L. Sani
Abstract
This paper shows a new sensor based on infrared images for monitoring pantograph catenary interaction. Due to the displacement of the contact point with respect to the reference position an image processing analysis of each frame is performed by using the Hough transformation in order to detect automatically monitoring variables of interest (e.g. temperature at the contact point and the position of the support towers along the railway line). Keywords: infrared image, pantograph-catenary interaction, Hough transformation, segment detection. 1 Introduction In order to improve the maintenance activities, a relevant objective for railway companies is the development of new sensors for a continuous monitoring of the quality of the current transmission between the overhead line and the collector strips of the pantograph. This problem is grown crucial with the introduction of the European Standard 96/48/CE about the interoperability. It is well-known that a poor electric contact between is the origin of arcing between the overhead wire and the collector strips of the pantograph. Several investigations have shown that the main damages of the overhead contact line installations are caused by the short term thermal effect of these arcing. The process which leads to the deterioration of the contact wire is associated with a localised recrystallization of the copper (i.e. a transaction to a stable crystalline microstructure with loss of all physical characteristics typical of the cold-drawn) and the formation of pits and dents on the surface. Unfortunately high speeds worsen the problem and a monitoring system has to be set up and tested to plan maintenance activity.
Keywords
infrared image, pantograph-catenary interaction, Hough transformation, segment detection.