Optimizing Sensor Activation in a Language Domain for Fault Diagnosis

Sensors are activated/deactivated according to a sensor activation policy (SAP), which is a decision function whose domain is the language of the system. An SAP is minimal for diagnosability if it is at the point where the system is diagnosable, but any additional deactivation prevents correct syste...

Full description

Saved in:
Bibliographic Details
Published inIEEE transactions on automatic control Vol. 64; no. 2; pp. 743 - 750
Main Authors Wang, Weilin, Gong, Chaohui, Wang, Di
Format Journal Article
LanguageEnglish
Published New York IEEE 01.02.2019
The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Subjects
Activation
Automata
Chaotic communication
Complexity theory
Deactivation
Delays
discrete event systems
Fault diagnosis
Observers
Optimization
sensor activation
Sensors
Trajectory
Online AccessGet full text

Cover

More Information
Summary:Sensors are activated/deactivated according to a sensor activation policy (SAP), which is a decision function whose domain is the language of the system. An SAP is minimal for diagnosability if it is at the point where the system is diagnosable, but any additional deactivation prevents correct system diagnosis. Given that the system is diagnosable when all sensors are always activated, we find in this paper that minimal SAP for diagnosability does not always exist. We therefore develop algorithms to compute a minimal SAP under which the system is pseudodiagnosable. Adhering to such a policy extends the window of opportunity for proper system diagnosis indefinitely, but with the compromise that later policy amendments are allowed.
ISSN:0018-9286
1558-2523
DOI:10.1109/TAC.2018.2844409