Title :
Abnormality Detection in Event Data and Condition Counters on Regina Trains
Author :
Holst, Anders ; Ekman, Jan ; Larsen, Stefan
Author_Institution :
Swedish Inst. of Comput. Sci.
Abstract :
The Regina trains, manufactured by Bombardier Transportation, contain software and hardware to generate as well condition data as event data that can be used to monitor the condition of the trains. In this paper we present the necessary equations for abnormality detection of both event data and condition counters in a general setting. The use of the equations is illustrated on authentic data from Regina trains
Keywords :
Bayes methods; condition monitoring; fault diagnosis; locomotives; railway engineering; statistical analysis; Bombardier Transportation; Regina trains; abnormality detection; condition counters; condition data; condition monitoring; event data; Bayesian statistics; Condition monitoring; Statistical anomaly detection;
Conference_Titel :
Railway Condition Monitoring, 2006. The Institution of Engineering and Technology International Conference on
Conference_Location :
Birmingham
Print_ISBN :
0-86341-732-9
