DocumentCode :
263018
Title :
Measures of non-Gaussianity in unscented Kaiman filter framework
Author :
Straka, O. ; Dunik, J. ; Simandl, Miroslav
Author_Institution :
Dept. of Cybern., Univ. of West Bohemia, Pilsen, Czech Republic
fYear :
2014
fDate :
7-10 July 2014
Firstpage :
1
Lastpage :
8
Abstract :
The paper deals with state estimation of nonlinear stochastic dynamic systems. In particular, the stress is laid on the recursive computation of the higher order moments of the state estimate in a local filter. The higher order moments are then used as a cornerstone of non-Gaussianity measures having the ability to indicate a possible decrease of the estimate quality. A technique is proposed for a general algorithm of the local filters, detailed for the unscented Kalman filter, and illustrated in numerical examples.
Keywords :
Kalman filters; higher order statistics; nonlinear filters; stochastic processes; higher order moments; local filter; nonGaussianity measure; nonlinear stochastic dynamic systems; recursive computation; state estimation; unscented Kaiman filter framework; Approximation methods; Covariance matrices; Kalman filters; Noise; State estimation; Time measurement; Vectors; Kalman filtering; non-Gaussianity measures; nonlinear filters; nonlinearity measures; state estimation;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Information Fusion (FUSION), 2014 17th International Conference on
Conference_Location :
Salamanca
Type :
conf
Filename :
6916118
Link To Document :
بازگشت