• DocumentCode
    2402929
  • Title

    A geometric perspective on adaptive system performance

  • Author

    Marcelle, Kenneth W.A. ; Lawrence, Dale A.

  • Author_Institution
    Dept. of Electr. & Comput. Eng., Cincinnati Univ., OH, USA
  • fYear
    1992
  • fDate
    1992
  • Firstpage
    3632
  • Abstract
    A geometric perspective is presented for the frequency domain performance prediction technique for adaptive systems. The geometry is simply that of an n-dimensional sphere, intersecting the origin, bisected by a hyperplane of lower dimension, which also intersects the origin. The amplitudes of all the significant frequency components of the adaptive prediction error are shown to lie on this hyperplane, and are thus bounded by the extrema of the intersection of the hyperplane and sphere. The hyperplane is determined by the original non-adaptive system, while the sphere can be determined largely by the system designer. This simple geometry allows straightforward, accurate computation of adaptive system performance bounds, and suggests simple design techniques for improving performance in desired frequency ranges. A simulation example is provided that illustrates the use of this geometry in designing adaptive systems to meet prespecified performance objectives
  • Keywords
    adaptive control; frequency-domain analysis; simulation; adaptive prediction error; adaptive system performance; frequency components; frequency domain performance prediction technique; geometric perspective; hyperplane; n-dimensional sphere; performance bounds; simulation example; system designer; Adaptive algorithm; Adaptive systems; Aerospace engineering; Computational geometry; Computational modeling; Frequency; Frequency domain analysis; Predictive models; Solid modeling; Stochastic processes; System performance;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Decision and Control, 1992., Proceedings of the 31st IEEE Conference on
  • Conference_Location
    Tucson, AZ
  • Print_ISBN
    0-7803-0872-7
  • Type

    conf

  • DOI
    10.1109/CDC.1992.370972
  • Filename
    370972