클러터 환경에서 표적 상태 추정을 위한 다중스캔 베이시안 스무딩 알고리듬 연구

Abstract

This Ph. D. thesis presents a fixed interval smoothing multiscan algorithm - smoothing integrated track splitting (SITS)- for target tracking in clutter. Furthermore, the multi-target multi-scan smoothing algorithm, -smoothing joint integrated track splitting (SJITS)- is also proposed, which is the extended version of the SITS algorithm. The proposed algorithms provide both the probability of target existence and the target trajectory probability density function which are calculated using all available measurements. The proposed algorithms improve both the false track discrimination (FTD) and the target trajectory estimate. The fixed interval smoothing fuses the forward and the backward multi-scan predictions in order to obtain the smoothing predictions and smoothing innovations. Both trajectory estimates and the data association probabilities are calculated using the smoothing innovations. An overlapping batch procedure is described which limits the smoothing delay. Specially, the iterative joint algorithm for efficient measurement-to-track assignments is applied to the proposed SJITS algorithm and the forward joint data association probabilities are also used for the backward ITS update. Comparative assessments of the smoothing/non-smoothing and single-target/multi-target data association algorithm are carried out under various target tracking scenarios to verify the FTD, trajectory estimation and track retention performances.