26.2 Remarks concerning Kalman filtering
Many processes can be modeled by a multivariate state-space model [see Catlin (1989) and Grewal and Andrews (2001)]. This linear dynamic system consists of a system equation
and an observation equation
The vector xt ∈ n describes the state of the system at time t, and the system equation models the dynamics of the system. The dynamics is influenced by the disturbance term t ∈ l. In this model the state of the system cannot be observed directly but only a linear function yt ∈ m can be observed. This linear dependence changes with time and is described by Equation (26.12) which is also influenced by a disturbance ηt ∈ m.
(t)t ≥ 0 and (