• DocumentCode
    661035
  • Title

    Design of fault residual functions for systems stabilized by static output feedback

  • Author

    Krokavec, Dusan ; Filasova, Anna

  • Author_Institution
    Dept. of Cybern. & Artificial Intell., Tech. Univ. of Kosice, Kosice, Slovakia
  • fYear
    2013
  • fDate
    9-11 Oct. 2013
  • Firstpage
    596
  • Lastpage
    600
  • Abstract
    A new approach to fault residual functions design, giving a simple algorithmic support potentially applicable in embedded diagnosis structures, is proposed in the paper. Based on LMI approach to control design as well as to inverse discrete-time LQ problem, the method blends static control principle with checking and evaluation of unavoidable degradation in an equivalent performance index value sequence due to faults in actuators, sensors or system dynamic matrix parameters. The methodology exploits special fault residual functions build on weighting matrices of the equivalent LQ criterion. A simulation example subject to different type of failures is used to demonstrate the properties and effectiveness of the method.
  • Keywords
    control system synthesis; discrete time systems; fault tolerant control; feedback; linear matrix inequalities; linear quadratic control; performance index; stability; LMI approach; LQ criterion; actuator dynamic matrix parameter faults; control design; embedded diagnosis structures; fault residual function design; inverse discrete-time LQ problem; linear matrix inequalities; performance index value sequence; sensor dynamic matrix parameter faults; static control principle; static output feedback; system dynamic matrix parameter faults; system stability; weighting matrices; Actuators; Control design; Fault detection; Output feedback; Sensors; Symmetric matrices;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Control and Fault-Tolerant Systems (SysTol), 2013 Conference on
  • Conference_Location
    Nice
  • Type

    conf

  • DOI
    10.1109/SysTol.2013.6693851
  • Filename
    6693851