Performance evaluation of derivative free Kalman filter for non-linear estimation problem

Kashyap, SK and Raol, JR (2005) Performance evaluation of derivative free Kalman filter for non-linear estimation problem. In: ICECON 05, 2005, Tiruchirappalli, Chennai, India.

Full text available as:
[img] PDF
Download (941Kb)

    Abstract

    For the estimation of the states of non-linear systems,13; Extended Kalman Filter (EKF) is frequently used. It13; has been observed by estimation community that EKF13; is only reliable for systems that are almost linear on13; the time scale of the updates (i.e. sampling interval).13; To overcome this problem, derivative free Kalman13; filter (DFKF) or more popularly known as Unscented13; Kalman filter, a method that propagates mean and13; covariance using non-lineal transformation. is13; frequently used. In this paper two schemes: 1)13; factorized version of EKF (UD Extended Kalman13; Filter or UDEKF) and ii) DFKF are studied and13; evaluated using various sets of simulated data of the13; non-linear systems. This method as compared to EKF13; is more accurate, easier to implement and has same13; order of calculations.13; 13;

    Item Type: Conference or Workshop Item (Paper)
    Uncontrolled Keywords: Target tracking;Extended kalman filter;UD factorization; Derivative free transformation;Derivative free kalman filter13;
    Subjects: ENGINEERING > Fluid Mechanics and Thermodynamics
    Division/Department: Flight Mechanics and Control Division, Flight Mechanics and Control Division
    Depositing User: Ms. Alphones Mary
    Date Deposited: 29 Feb 2008
    Last Modified: 17 Jun 2010 10:32
    URI: http://nal-ir.nal.res.in/id/eprint/4641

    Actions (login required)

    View Item