You are here: Home :: Area NEM :: Engineering :: Automation

Alexander Geistler, Frank Böhringer

Stochastic, Recursive Turnout Classification for Track-selective Location of Rail Vehicles

Keywords: on-board train location, eddy current sensors, railway turnouts, Bayes classification, HMM

On-board train location systems are mainly used for customer services and logistics but there is also great need to extend these systems to security-related tasks, which always depend on a track-selective location. Since eddy current sensors can be applied to detect and classify turnouts, a track-selective location is possible. To estimate the probability of a classified turnout, a Bayes classifier has been used successfully. The track assignment uses statistical methods, like Markov models, which are based on recursive methods. First field tests indicate the robustness of the location methods presented.

at – Automatisierungstechnik, Oldenbourg Wissenschaftsverlag

Print ISSN: 0178-2312
Volume: 54, 03/2006
Pages: 113 - 122

Show full article (external site)

Show all available items of this journal

 

Area NEM