A data-driven approach for actuator servo loop failure detection - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2017

A data-driven approach for actuator servo loop failure detection

Résumé

This paper studies a data-driven approach to detect faults in flight control systems of civil aircraft. A particular class of failures, referred to as Oscillatory Failure Cases (OFC), impacting the actuator servo loop has motivated the authors to consider a data-driven approach based on distance and correlation measures (see reference [Goupil et al.(2016). A data-driven approach to detect faults in the Airbus flight control system. IFAC-PapersOnLine, 49(17), 52-57] of this paper) leading to promising results compared to the state-of-the-art methods based on analytical redundancy. The present paper extends the formulation and the results of the considered OFC detection approach investigating Support Vector Machine (SVM) techniques to define a more accurate detector based on distance and correlation measures.

Dates et versions

hal-03128340 , version 1 (02-02-2021)

Identifiants

Citer

Simone Urbano, Eric Chaumette, Jean-Yves Tourneret. A data-driven approach for actuator servo loop failure detection. 20th World Congress of the International Federation of Automatic Control (IFAC 2017), Jul 2017, Toulouse, France. ⟨10.1016/j.ifacol.2017.08.2353⟩. ⟨hal-03128340⟩
26 Consultations
0 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More