A diagnosis strategy for FDI in wireless networked control system

Cédric Berbra (1) , Zeashan Hameed Khan (1) , Sylviane Gentil (1) , Suzanne Lesecq (1) , Jean-Marc Thiriet (1)

1 GIPSA-SA-IGA - GIPSA - Signal et Automatique pour le Diagnostic et la Surveillance

This paper describes a wireless based networked control system (NCS) and its diagnosis. The purpose is to show the specific characteristics of a wireless network compared to a classical wired network and their consequences on the diagnosis. A comparison of residuals generated to detect and isolate sensor faults in a control loop is made when two network protocols are used: a CAN network and a ZigBee wireless network. An additional residual is proposed to differentiate sensor faults from packet losses. A reconfiguration procedure is used to improve control performance in case of packet losses. A D.C motor control and diagnosis are simulated with Matlab/Simulink while the networks are simulated with TrueTime 1.5. A distributed diagnosis strategy is also proposed from overall system perspective

Mots clés: Fault Detection and Isolation.; Fault Detection and Isolation; TrueTime; CAN; Wireless Networks; ZigBee; Networked Control Systems (NCS); Distributed Diagnosis

Domaines: Automatique / Robotique

Cédric Berbra, Zeashan Hameed Khan, Sylviane Gentil, Suzanne Lesecq, Jean-Marc Thiriet. A diagnosis strategy for FDI in wireless networked control system. FeT'2007 7th IFAC International Conference on Fieldbuses and Networks in Industrial and Embedded Systems, Nov 2007, Toulouse, France. pp.235-240. ⟨hal-00199492⟩