Electrical and Electronic Engineering
false
http://link.springer.com/10.1007/978-1-4471-4276-8_5
2019-04-15T13:26
chapters
2013
Diagnosis and Automata
85-106
Fault diagnosis and state estimation are two central and typical problems one may face in the monitoring of discrete-event systems. This chapter examines these two problems in the simple setting of automata. It is first explained that diagnosis and state estimation are two related problems. Then one describes the construction of an observer (resp. a diagnoser) both for standard and for probabilistic automata. A section is dedicated to diagnosability issues, that is the ability to detect the occurrence of an unobservable fault event after a bounded number of observations following that fault. The chapter then proposes an opening to the case of distributed systems, made of several interacting components, but still assuming a sequential semantics (i.e. ignoring the possible parallelism of some events). One first presents a modularity property on observers and diagnosers of distributed systems, in a rather specific case. The general case is then examined, and a distributed procedure is described to recover the runs of a distributed system that can explain a set of distributed observations collected in this system. Finally, the chapter closes on a discussion about the interest of true concurrency semantics for the monitoring of large distributed systems.
https://scigraph.springernature.com/explorer/license/
2013-01-01
en
chapter
doi
10.1007/978-1-4471-4276-8_5
978-1-4471-4275-1
Control of Discrete-Event Systems
978-1-4471-4276-8
Springer Nature - SN SciGraph project
London
Springer London
Seatzu
Carla
University of Rennes 1
INRIA Rennes Bretagne Atlantique, Campus de Beaulieu, 35042 Rennes cedex, France
Manuel
Silva
readcube_id
358bc59eb5da39079d003219e3a5f765d0227f4a0343fe45f215acc97d4fe8c0
Eric
Fabre
dimensions_id
pub.1011238631
Jan H.
van Schuppen
Engineering