How to Be Sure a Faulty System Does Not Always Appear Healthy?
- 204 Downloads
Fault diagnosis is a crucial and challenging task in the automatic control of complex systems, whose efficiency depends on the diagnosability property of a system. Diagnosability describes the system property allowing one to determine with certainty whether a given fault has effectively occurred based on the available observations. However, this is a quite strong property that generally requires a high number of sensors. Consequently, it is not rare that developing a diagnosable system is too expensive. In this paper, we analyze a new discrete event system property called manifestability, that represents the weakest requirement on observations for having a chance to identify on line fault occurrences and can be verified at design stage. Intuitively, this property makes sure that a faulty system cannot always appear healthy, i.e., has at least one future behavior after fault occurrence observably distinguishable from all normal behaviors. Then, we prove that manifestability is a weaker property than diagnosability before proposing an algorithm with PSPACE complexity to automatically verify both properties. Furthermore, we prove that the problem of manifestability verification itself is PSPACE-complete. The experimental results show the feasibility of our algorithm from a practical point of view. Finally, we compare our approach with related work.
KeywordsFault Occurrence Discrete Event Systems (DESs) Stochastic Diagnosability Critical Pairs Inﬁnite Words
- 1.Agarwal, A., Madalinski, A., Haar, S.: Effective verification of weak diagnosability. In: Proceedings of the 8th IFAC Symposium on Fault Detection, Supervision and Safety for Technical Processes (SAFEPROCESS 2012), pp. 636–641. IFAC (2012)Google Scholar
- 3.Bertrand, N., Haddad, S., Lefaucheux, E.: Foundation of diagnosis and predictability in probabilistic systems. In: 34th International Conference on Foundation of Software Technology and Theoretical Computer Science, FSTTCS 2014, 15–17 December 2014, New Delhi, India, pp. 417–429 (2014)Google Scholar
- 4.Bertrand, N., Haddad, S., Lefaucheux, E.: Diagnosis in infinite-state probabilistic systems. In: 27th International Conference on Concurrency Theory, CONCUR 2016, 23–26 August 2016, Québec City, Canada, pp. 37:1–37:15 (2016)Google Scholar
- 5.Bonchi, F., Pous, D.: Checking NFA Equivalence with Bisimulations up to Congruence. In: Proceedings of 40th ACM SIGPLAN-SIGACT Symposium on Principles of Programming Languages (POPL-2013), pp. 457–468. ACM (2013)Google Scholar
- 11.Papineau, D.: Philosophical Naturalism. Blackwell Publishers, Hoboken (1993)Google Scholar
- 12.Pencolé, Y.: Diagnosability analysis of distributed discrete event systems. In: Proceedings of the 16th European Conference on Articifial Intelligent (ECAI 2004), pp. 43–47. IOS Press, Nieuwe Hemweg (2004)Google Scholar
- 14.Schumann, A., Huang, J.: A scalable jointree algorithm for diagnosability. In: Proceedings of the 23rd American National Conference on Artificial Intelligence (AAAI 2008), pp. 535–540. AAAI Press, Menlo Park (2008)Google Scholar
- 15.Schumann, A., Pencolé, Y.: Scalable diagnosability checking of event-driven system. In: Proceedings of the Twentieth International Joint Conference on Artificial Intelligence (IJCAI 2007), pp. 575–580. International Joint Conferences on Artificial Intelligence Inc., Menlo Park (2007)Google Scholar
- 20.Ye, L., Dague, P.: Diagnosability analysis of discrete event systems with autonomous components. In: Proceedings of the 19th European Conference on Artificial Intelligence (ECAI 2010), pp. 105–110. IOS Press, Nieuwe Hemweg (2010)Google Scholar
- 21.Ye, L., Dague, P., Longuet, D., Briones, L.B., Madalinski, A.: Fault manifestability verification for discrete event systems. In: Proceedings of the 22nd European Conference on Artificial Intelligence (ECAI 2016), pp. 1718–1719. IOS Press (2016)Google Scholar