No label defined (Q4551114)

From MaRDI portal

!

This is the item page for this Wikibase entity, intended for internal use and editing purposes. Please use the normal view instead:

scientific article; zbMATH DE number 1796078
Language Label Description Also known as
default for all languages
No label defined
    English
    No label defined
    scientific article; zbMATH DE number 1796078

      Statements

      0 references
      9 October 2002
      0 references
      Kalman filter
      0 references
      robustness
      0 references
      optimal state estimation
      0 references
      optimization
      0 references
      discrete system
      0 references
      training signal
      0 references
      learning
      0 references
      teaching signal
      0 references
      Let a linear discrete time-invariant (LTI) system be described by state equations of the form \[ x[n+1]=Ax[n]+Bu[n], y[n]=Cx[n], \] where the matrices \(A,B,C\) are known. Also, let the observations of a noisy output \(z[n]\) be described by \(z[n]=y[n]+v[n]\) and the training signal \(r[n]\) be given; here \(v[n]\) is a sequence of independent identically distributed variables with a small contaminated Gaussian distribution. The aim of the paper is the development of an approach for a robust recursive estimation of states \(y_{1}[n]=y_n, y_{2}[n]=y_{n-1},\dots,y_{m}[n]=y_{n-m+1}\) of the LTI system using the observations of the noisy output \(z[n]\) as well as a teaching signal \(r[n]\) in the learning phase and only the observations of \(z[n]\) in the working phase. A technique based on robust filtering by means of a bank of parallel Kalman filters and on the procedure of optimization of the state estimation itself is used. At each time moment, a current estimate is chosen that ensures a minimal absolute deviation from the current value of the teaching signal. An approach based on the relation between the mean squared deviation of state estimates from the true state and the innovation sequence variance as well as on the fact that both variables achieve their minimum for the same filter from the respective Kalman filter bank, is proposed for a working phase, where a training signal will be absent. The results of numerical simulation of the LTI system with the computer using a bank of Kalman filters are proposed.
      0 references

      Identifiers

      0 references
      0 references
      0 references
      0 references
      0 references
      0 references
      0 references