UKF-STF tracking with correlated noises for the nonlinear system
Wenbin Li, Quanbo Ge · 2010
Nonlinear filtering is one of hot and difficult topics in the target tracking. The classical Extended Kalman Filter (EKF) has some shortages such as tracking accuracy and stability because of the using of linearization. So, the Unscented Kalman Filter (UKF) based on the nonlinear transformation and sampling has been widely applied. In this paper, we devote to research the nonlinear Unscented Kalman Filter on the systems with Correlated Noise (UKF-CN). Synchronously, the Strong Tracking Filter (STF) is introduced and a novel UKF-STF with Correlated Noise (UKF-STF-CN), which can deal with the systems with correlated noise and sudden change of state, is proposed .Compared with the existing method, the new algorithms can not only deal with the nonlinear target tracking problem with correlated noise, but also have better estimation accuracy and stability. Two simulation examples based on bearings-only tracking are illustrated to verify the efficiency of the proposed algorithms.