A DFT Modeling Approach for Infrastructure Reliability Analysis of Railway Station Areas

Matthias Volk, Norman Weik, Joost Pieter Katoen, Nils Nießen

Publikation: Beitrag in Buch/Bericht/KonferenzbandKonferenzbeitragBegutachtung

4 Zitate (Scopus)

Abstract

Infrastructure failures—in particular in station and junction areas—are one of the most important causes for train delays in railway systems. Individually, subsystems, such as track circuits or radio communication, are well understood and have been analyzed using formal methods. However, verification of the capability of station areas to fulfill operational design specifications as a whole remains widely open. In this paper, we present a fully automatic translation from station area infrastructure to dynamic fault trees (DFT) with special emphasis on field elements including switches, signals and track occupation detection systems. Reliability is assessed in terms of train routability, where feasible train routes consist of the set of train paths projected in the interlocking system including their requirements w.r.t. the state of field elements. Analysing the DFTs by probabilistic model checking techniques allows for new performance metrics based on, e.g., conditional events or the sequence of failures, which can serve to provide additional insights into the criticality of field elements. We demonstrate the feasibility of the DFT-based analysis based on data for railway stations in Germany where the generated DFTs consist of hundreds of elements.

OriginalspracheEnglisch
TitelFormal Methods for Industrial Critical Systems - 24th International Conference, FMICS 2019, Proceedings
Redakteure/-innenKim Guldstrand Larsen, Tim Willemse
Herausgeber (Verlag)Springer Verlag
Seiten40-58
Seitenumfang19
ISBN (Print)9783030270070
DOIs
PublikationsstatusVeröffentlicht - 2019
Extern publiziertJa
Veranstaltung24th International Conference on Formal Methods for Industrial Critical Systems, FMICS 2019 - Amsterdam, Niederlande
Dauer: 30 Aug. 201931 Aug. 2019

Publikationsreihe

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Band11687 LNCS
ISSN (Print)0302-9743
ISSN (elektronisch)1611-3349

Konferenz

Konferenz24th International Conference on Formal Methods for Industrial Critical Systems, FMICS 2019
Land/GebietNiederlande
OrtAmsterdam
Zeitraum30/08/1931/08/19

Fingerprint

Untersuchen Sie die Forschungsthemen von „A DFT Modeling Approach for Infrastructure Reliability Analysis of Railway Station Areas“. Zusammen bilden sie einen einzigartigen Fingerprint.

Dieses zitieren