Steady state analysis for discrete tracking filters

Adaptive-gain-and-tau tracking filters for correlated target maneuvers

“…The dynamical and observation models associated with the target motion can be given a common state-space equation form (1) (2) where y i -observation, h = (1 0 0…) -1×N measurement matrix, g -control vector, w i and v i are mutually uncorrelated process and measurement noises, respectively, with variances Q= w 2 and R= v 2 . The transition matrix F describing a kinematic filter for the exponentially-correlated maneuver case becomes (3) where N is a flipped vector of the polynomial coefficients, …”

“…Tracking filters of the 2 nd , 3 rd and even 4 th order [4] are usually handled in the state-space form while the Kalman time-varying gain is replaced by a properly specified constant. Accordingly, most of past studies were focused on the steady-gain optimality [3,6]. In the literature, one can find closed-form and numeric solutions providing optimal gains for particular types of the process and observation noise, maneuver type and other specifications.…”

“…The target tracking (variously kinematic, polynomial, trend) filter is in the focus of interest over decades [1][2][3][4][5][6]. Tracking filters of the 2 nd , 3 rd and even 4 th order [4] are usually handled in the state-space form while the Kalman time-varying gain is replaced by a properly specified constant.…”

Aircraft Trajectory Tracking and Prediction

“…These trackers are similar to the acceleration model in other center of gravity methods [8,11,23,25,26,40]. A third order Gauss-Markov model is used to approximate the jerk in each coordinate direction.…”