Generalized Sensor Mapping for Fault Diagnosis of Discrete Event Systems
IEEE Transactions on Automation Science and Engineering · 2026 · 23 · 8710-8721
DOI: 10.1109/tase.2026.3686409 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Transactions on Automation Science and Engineering, 2026. Verfasst von Victor Hugo Cruz und 2 weiteren Autor:innen. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 37 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Coordenação de Aperfeiçoamento de Pessoal de Nível Superior, National Council for Scientific and Technological Development, Carlos Chagas Filho Foundation.
Abstract (Originalsprache)
The first step in the design of fault diagnosis systems for industrial plants is to obtain a model that describes both normal and fault behaviors. In the context of fault diagnosis of discrete event systems (DES), Sensor Mapping (SM) is the main tool used to obtain an automaton model. The SM works by converting a Moore automaton, whose sensor readings are the state outputs, into a Mealy automaton, whose events associated with the sensor readings are integrated into the system model. However, SM has the following drawbacks: 1) it is limited to states associated with a single sensor reading, and 2) since the information associated with commands and sensor readings is both grouped in a single event, it can lead to an incorrect estimation in real time of the states of the monitored system. In this paper, we generalize the SM modeling tool and propose the Generalized Sensor Mapping (GSM) method that does not have the aforementioned restrictions of the SM. In this regard, in the GSM proposed here, the events associated with sensors and actuators are separate. In addition, the GSM can also be used in Moore automata that have states whose outputs are associated with more than one reading. The efficiency of the proposed method is illustrated by means of a practical example. Note to Practitioners—The construction of discrete event models for industrial systems, for example, a process control systems with upper-layer supervision, requires transitions to be labeled by events associated with actions (commands) or change in the sensor readings. However, before actually obtaining a discrete event model with this property, it is necessary to create an initial model where the information regarding the system dynamics is in the states. It is very common that two or more sensor readings be associated with the same state; for example, the readings of a level sensor used in a chemical tank to inform if the level is rising or going down may display two readings depending on the action imposed by the controller. The existing methodology combine control actions and sensor readings to label the transitions of the discrete event model. Such an approach may lead the fault diagnosis system designed based on this model to issue an incorrect decision regarding the fault occurrence. Another drawback is that it can only be applied when the states are associated with a single sensor reading. In this paper, we present a new methodology for the construction of discrete event models which is completely practical-oriented, straightforward to use and provides a model whose transitions are either command or sensor reading events. In addition, the resulting model can be applied not only to fault diagnosis but also to all applications of discrete event systems.
Förderung
- Coordenação de Aperfeiçoamento de Pessoal de Nível Superior
- National Council for Scientific and Technological Development
- Carlos Chagas Filho Foundation
- CNPq
- FAPERJ
Zitieren
APA
Cruz, V. H., Carvalho, L. K., & Basilio, J. C. (2026). Generalized Sensor Mapping for Fault Diagnosis of Discrete Event Systems. IEEE Transactions on Automation Science and Engineering, 23, 8710-8721. https://doi.org/10.1109/tase.2026.3686409
Vancouver
Cruz VH, Carvalho LK, Basilio JC. Generalized Sensor Mapping for Fault Diagnosis of Discrete Event Systems. IEEE Transactions on Automation Science and Engineering. 2026;23:8710-8721. doi:10.1109/tase.2026.3686409
BibTeX
@article{Cruz2026,
author = {Cruz, Victor Hugo and Carvalho, Lilian Kawakami and Basilio, João Carlos},
title = {Generalized Sensor Mapping for Fault Diagnosis of Discrete Event Systems},
journal = {IEEE Transactions on Automation Science and Engineering},
year = {2026},
volume = {23},
pages = {8710--8721},
doi = {10.1109/tase.2026.3686409}
}
EUVIMED