• DocumentCode
    2636203
  • Title

    Quasi Analytical State Observer in Real Time Lerneable System

  • Author

    Alexik, Mikulas

  • Author_Institution
    Dept. of Tech. Cybern., Univ. of Zilina, Zilina, Slovakia
  • fYear
    2009
  • fDate
    7-9 Sept. 2009
  • Firstpage
    170
  • Lastpage
    175
  • Abstract
    Application of state variables in description of linear and non-linear dynamic systems brings many advantages therefore that description is frequently used in modeling and simulation. For digital simulation using PC such system needs fundamental matrix for chosen sampling interval to be calculated. Such calculation, which uses finite number of input elements from infinite sequence, is therefore biased with computation error. Various ways of computation varies in precision and speed of calculation. If there is no need to realize real-time computation, precision can be improved using convenient type and number of sequences. The modern control of non-linear dynamic systems (positional servomechanisms) requires "real time" fundamental matrix calculation. In such a case the most suitable way is to apply a procedure developed by the author named quasi-analytical calculation of fundamental matrix, described in this paper.
  • Keywords
    learning systems; nonlinear dynamical systems; observers; optimal control; real-time systems; variable structure systems; digital simulation; nonlinear dynamic system; positional servomechanism; quasi analytical state observer; real time fundamental matrix calculation; real time learnable system; state variables application; Computational modeling; Control systems; Friction; Laboratories; Optimal control; Real time systems; Servomechanisms; Sliding mode control; Springs; State estimation; Time optimal control; real time; sliding mode; state extimator;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Computational Intelligence, Modelling and Simulation, 2009. CSSim '09. International Conference on
  • Conference_Location
    Brno
  • Print_ISBN
    978-1-4244-5200-2
  • Electronic_ISBN
    978-0-7695-3795-5
  • Type

    conf

  • DOI
    10.1109/CSSim.2009.47
  • Filename
    5350115