Fraunhofer-Gesellschaft

Publica

Hier finden Sie wissenschaftliche Publikationen aus den Fraunhofer-Instituten.

Support of state-dependent maintenance by advanced information technology

 

World Congress on Railway Research 1997. Proceedings. Vol. B
Florenz, 1997
pp.315-321
World Congress on Railway Research (WCRR) <1997, Florenz>
English
Conference Paper
Fraunhofer IITB ( IOSB) ()
diagnosis; maintenance; model application; pattern recognition; signal analysis; state determination

Abstract
Reduction of maintenance costs without reduction of safety and reliability is a need for all railway companies. One commonly taken approach to reach this goal is the transition from the concept of maintenance activities at fixed time intervals to a state-dependent maintenance concept. Basically such a concept requires continuous availability of information about the actual status of the maintenance objects. Regarding respective boundary conditions and available measuring principles this requirement can only be partly met. Certain inspections can only be executed in running sheds or by test runs of dynamometer cars, others are completed by detectors at certain points along the track lines. A current trend is to monitor components crucial for safe and reliable railway traffic continuously. Thus investigations are undertaken to prove the feasibility and profitability of concepts to equip e. g. signalling devices and rolling stock with appropriate sensors and diagnostic systems. Beyond tha t it must be explored if at least an overall quasi-continuous inspection of the track state can be done by installing robust low-cost diagnostic systems in a larger number of standard cars for regular service. In all these cases advanced methods of information technology are powerful tools for extracting the relevant information from the huge amount of sensor data. At the same time they may help to reduce the amount of additional equipment. Examples for the application of signal analysis, pattern recognition and model-based diagnosis are given in this paper.

: http://publica.fraunhofer.de/documents/PX-35357.html