Guaranteed estimation in the problems of navigation information processing

Abstract

A new approach for guaranteed state estimation of a linear system under parameter uncertainties in correlated disturbances and measurement errors is suggested. Conditions for adjustment of a Kalman-type filter with guaranteed estimation are derived for the case that the correlated disturbances and measurement errors are first-order Markov processes. The efficiency of the suggested method is illustrated by a navigation problem solved for aiding the inertial navigation system.