Partitioned time-varying smooth variable structure filter for airport target tracking

In practical applications of airport target tracking, in order to balance the performance and real time, the filters considered are usually comparatively simple, such as Kalman filter (KF). The smooth variable structure filter (SVSF) is a relatively new estimator and a new form of the SVSF utilizes a time-varying, robust and stable smoothing boundary layer (SVSF-VBL). In this paper, a partitioned SVSF-VBL (PSVSF-VBL) method is proposed for airport target tracking. For systems with fewer measurements than states, the derivations of a posterior covariance and a time-varying smoothing boundary layer width of the PSVSF-VBL are given. In the end, experimental results of real measured data from airport images are shown that the proposed algorithm compared with KF has less filtering errors, better tracking effects and its run-time meets the real-time requirement of airport systems.