• DocumentCode
    2102013
  • Title

    A circuit theory of the Kalman filter

  • Author

    Carter, David W.

  • Author_Institution
    Dept. of Anal. & Software, Charles Stark Draper Lab. Inc., Cambridge, MA, USA
  • fYear
    1993
  • fDate
    15-17 Dec 1993
  • Firstpage
    1224
  • Abstract
    This note gives a short heuristic derivation of the discrete Kalman filter by considering simple voltage measurement circuits. The idea is to obtain conclusions about the general estimation problem by reasoning about an electric analog. The usual Kalman filter and the formulation known as the information filter are both quickly derived by this method. A basis for this approach is Nyquist´s result (1928) that any resistance R exhibits a thermal noise voltage-the Johnson noise in radio engineering-of constant power spectral density proportional to R. It follows that multivariate white noise may be realized as the thermal noise of a resistive n-port. We model the general noisy measurement problem as a measurement problem on a resistive circuit. Covariance matrices arise in a concrete fashion as resistance matrices
  • Keywords
    Kalman filters; estimation theory; heuristic programming; matrix algebra; multiport networks; thermal noise; voltage measurement; Johnson noise; circuit theory; constant power spectral density; covariance matrices; discrete Kalman filter; discrete sequential estimation; heuristic derivation; information filter; multivariate white noise; resistive n-port; thermal noise voltage; voltage measurement circuits; Circuit noise; Circuit theory; Covariance matrix; Electric resistance; Electrical resistance measurement; Information filters; Power engineering and energy; Thermal engineering; Thermal resistance; Voltage measurement;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Decision and Control, 1993., Proceedings of the 32nd IEEE Conference on
  • Conference_Location
    San Antonio, TX
  • Print_ISBN
    0-7803-1298-8
  • Type

    conf

  • DOI
    10.1109/CDC.1993.325378
  • Filename
    325378