Asymptotic distribution theory for the kalman filter state estimator (Q3218980)
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 3885172
| Language | Label | Description | Also known as |
|---|---|---|---|
| default for all languages | No label defined |
||
| English | Asymptotic distribution theory for the kalman filter state estimator |
scientific article; zbMATH DE number 3885172 |
Statements
Asymptotic distribution theory for the kalman filter state estimator (English)
0 references
1984
0 references
state space model
0 references
non-Gaussian disturbances
0 references
model validation
0 references
Kalman filter state estimator
0 references
linear discrete-time systems
0 references
stability properties of the state transition matrix
0 references
borderline stable-unstable systems
0 references
asymptotic normality
0 references
measure of deviation from normality
0 references
0.8894807
0 references
0.88489944
0 references
0.8801103
0 references
0.87989414
0 references
0.8794361
0 references
0.87887335
0 references
0.8758132
0 references
This interesting and well-written paper presents an asymptotic distribution theory for the Kalman filter state estimator in the case of linear discrete-time systems. In the absence of the usual Gaussian assumption the distributional properties of the state estimate from the Kalman filter are shown to depend on the stability properties of the state transition matrix. Specifically, for borderline stable-unstable systems the state estimate is asymptotically normal under fairly weak conditions. For stable or unstable systems, however, it is shown that at least some of the random terms in the system equation must be normally distributed beyond some finite time, before the state estimate is asymptotically normal. The paper ends with a discussion of two interesting problems left for future research. The first one refers to the rate at which the distribution of the state estimate approaches the normal distribution, when asymptotic normality holds. The second problem relates to a quantitative measure of deviation from normality, when asymptotic normality does not hold.
0 references