Share Email Print
cover

Proceedings Paper

Influence of a non-Gaussian state model on the position estimation in the nonlinear filtration
Author(s): Stanisław Konatowski; Barbara Pudlak
Format Member Price Non-Member Price
PDF $14.40 $18.00

Paper Abstract

In navigation systems with nonlinear filtration algorithms extended Kalman filter is being used to estimate position. In this filter, the state model distribution and all relevant noise destinies are approximated by Gaussian random variable. What is more, this approach can lead to poor precision of estimation. Unscented Kalman filter UKF approximates probability distribution instead of approximating nonlinear process. The state distribution is represented by a Gaussian random variable specified using weighted sigma points, which completely capture true mean and covariance of the distribution. Another solution for the general filtering problem is to use sequential Monte Carlo methods. It is particle filtering PF based on sequential importance sampling where the samples (particles) and their weights are drawn from the posterior distribution.

Paper Details

Date Published: 28 December 2007
PDF: 8 pages
Proc. SPIE 6937, Photonics Applications in Astronomy, Communications, Industry, and High-Energy Physics Experiments 2007, 69373K (28 December 2007); doi: 10.1117/12.784892
Show Author Affiliations
Stanisław Konatowski, Military Univ. of Technology (Poland)
Barbara Pudlak, Consultant (Poland)


Published in SPIE Proceedings Vol. 6937:
Photonics Applications in Astronomy, Communications, Industry, and High-Energy Physics Experiments 2007

© SPIE. Terms of Use
Back to Top