Active Diagnosis for Switched Systems Using Mealy Machine Modeling - Archive ouverte HAL
Chapitre D'ouvrage Année : 2018

Active Diagnosis for Switched Systems Using Mealy Machine Modeling

Résumé

Generally, fault diagnosis schemes play an important role in ensuring the safety of physical or engineering systems. The study of diagnosis problem for switched systems is interesting and allows considering a more wide range of systems. This chapter deals with the active diagnosis for a class of switched systems which may not satisfy the classical diagnosability conditions usually considered in the Discrete-Event-Systems setting. In the first part, the modeling approach we propose is introduced. We propose to use an abstract representation of a switched system using a Mealy Machine where discrete faults may occur. An appropriate diagnoser is designed in order to reduce the uncertain state subset. In the second part, some diagnosability conditions are deduced. Based on the Mealy Machine, a new active diagnosis strategy is designed in order to ensure the fault detection and isolation for a class of switched systems. An algorithm combining the proposed di-agnoser and a testing procedure is introduced in order to solve the fault identification problem. A study on the cascade multicellular converter is carried out to detect and isolate faulty cells. Illustrative simulation results, on a two cells converter, show the details of the algorithm and experimental results, on a three cells converter, present the effectiveness, in real time, of the proposed scheme.
Fichier principal
Vignette du fichier
Chap_Diag_2018.pdf (3.08 Mo) Télécharger le fichier
Origine Fichiers produits par l'(les) auteur(s)
Loading...

Dates et versions

hal-02466641 , version 1 (26-03-2020)

Identifiants

Citer

Jérémy van Gorp, Alessandro Giua, Michael Defoort, Mohamed Djemai. Active Diagnosis for Switched Systems Using Mealy Machine Modeling. Diagnosability, Security and Safety of Hybrid Dynamic and Cyber-Physical Systems, Springer International Publishing, pp.147-173, 2018, ⟨10.1007/978-3-319-74962-4_6⟩. ⟨hal-02466641⟩
113 Consultations
83 Téléchargements

Altmetric

Partager

More