Publication Date:
2019-07-13
Description:
Martingale theory and appropriate smoothing properties of Loeve (1953) have been used to develop a modified Gaussian second-order filter. The performance of the filter is evaluated through numerical simulation of a Jupiter flyby mission. The observations used in the simulation are on-board measurements of the angle between Jupiter and a fixed star taken at discrete time intervals. In the numerical study, the influence of each of the second-order terms is evaluated. Five filter algorithms are used in the simulations. Four of the filters are the modified Gaussian second-order filter and three approximations derived by neglecting one or more of the second-order terms in the equations. The fifth filter is the extended Kalman-Bucy filter which is obtained by neglecting all of the second-order terms.
Keywords:
SPACE SCIENCES
Type:
AD-765232
,
AFOSR-73-1436TR
,
Symposium on Nonlinear Estimation Theory and its Applications; Sep 11, 1972 - Sep 13, 1972; San Diego, CA
Format:
text
Permalink