Actuator-fault Detection and Isolation based on Interval Observers and Invariant Sets

Feng Xu (1) , Vicenc Puig (1) , Carlos Ocampo-Martinez (1) , Florin Stoican (2) , Sorin Olaru (3, 4)

1 UPC - Universitat Politècnica de Catalunya [Barcelona]
2 UPB - University Politehnica of Bucharest [Romania]
3 DISCO - Dynamical Interconnected Systems in COmplex Environments
4 E3S - Supélec Sciences des Systèmes

This paper proposes an interval observer-based actuator fault detection and isolation (FDI) approach. An interval observer matching the healthy system mode is designed to monitor the system. When the system is in different modes, state or output interval vectors predicted by the interval observer manifest different dynamical behaviors. To guarantee reliable FDI, a collection of invariant set-based FDI conditions are established. Under these conditions, actuator faults can be accurately detected and isolated during the transition between different modes. At the end, the effectiveness of this proposed approach is presented by using a numerical example.

Domains: Engineering Sciences [physics]; Automatic

Feng Xu, Vicenc Puig, Carlos Ocampo-Martinez, Florin Stoican, Sorin Olaru. Actuator-fault Detection and Isolation based on Interval Observers and Invariant Sets. 52nd IEEE Conference on Decision and Control CDC 2013, Dec 2013, Firenze, Italy. pp.CD-Rom. ⟨hal-00918190⟩