tutorialpoint.org

Engg. tutorials

  • Instrumentation and Control Lab
  • Control systems assignment
  • Shape memory alloy SMA actuator
  • Dielectric elastomer
  • EM theory lecture notes
  • GATE question papers
  • JAM question papers
  • Kalman filter tutorial
  • Nonlinear estimation
  • Linear Estimation

    Kalman Flowchart

    1. Kalman Filter Tutorial (Cont'd...)

    1.4 Properties of Kalman filter

    Unbiased estimator

    An estimator is said to be unbiased, if the statistical expectation of an error is equal to zero. i.e. \begin{equation}\label{4k} E[e_{k|k}]=E[e_{k|k-1}]=0. \end{equation} The expression of expectation of prior error could be written as: \begin{equation}\label{abc} E[e_{k|k-1}]=E[\hat{X}_{k|k-1}-X_{k}]=E[\hat{X}_{k|k-1}]-E[X_{k}]. \end{equation} We know that the predicted value depends on current state and previous measurement, i.e. \begin{equation*} E[\hat{X}_{k|k-1}]= E[E[X_{k}|Y_{k-1}]], \end{equation*} and also $X_{k}$ and $Y_{k-1}$ are independent to each other. So, we write \begin{equation*} E[E[X_{k}|Y_{k-1}]]=E[X_{k}]. \end{equation*} Now, substitute this value in the above equation, \begin{equation} E[e_{k|k-1}]=0. \end{equation}

    < Prev.Page   1   2   3   4  5   6   7   8   9   10   11   12   13   14   15   16   17   Next page>