Design of a tracking filter suitable for 2D motion dynamics of an off-road target
Kangwagye Samuel, Jae Weon Choi · 2017
Two linear Kalman filters arranged in series with each other which are used to estimate the system states, with the first filter for turn rate and the second one for target states for an off-road vehicle with unpredicted maneuvers is proposed. A Coordinated Turn model with unknown turn rate is used as a system dynamic equation. The root mean square error for all the estimates is small which indicates accurate and reliable results. This algorithm can more accurately track unpredicted target maneuvers with minimized errors. In this paper, the state of tracked object consists both of its longitudinal and lateral positions, and velocities. The turn rate which is also an unknown parameter is estimated.