The importance of railway transportation has been increasing in the world.Considering the current and future estimates of high cargo and passengertransportation volume in railways, prevention or reduction of delays due to anyfailure is becoming ever more crucial. Railway turnout systems are one of themost critical pieces of equipment in railway infrastructure. When incipientfailures occur, they mostly progress slowly from the fault free to the failurestate. Although studies focusing on the identification of possible failures inrailway turnout systems exist in the literature, neither the detection norforecasting of failure progression has been reported. This paper presents asimple state-based prognostic method that aims to detect and forecast failureprogression in electro-mechanical systems. The method is compared with HiddenMarkov Model based methods on real data collected from a railway turnout system.Obtaining statistically sufficient failure progression samples is difficultconsidering that the natural progression of failures in electro-mechanicalsystems may take years. In addition, validating the classification model isdifficult when the degradation is not observable. Data collection and modelvalidation strategies for failure progression are also presented.
展开▼
