Multisensor data fusion for manoeuvring target tracking

The aim of this paper is to describe an approach that performs data fusion on the output of the multiple sensors engaged in the manoeuvre target tracking. A common approach is to use the extended Kalman filter (EKF) for manoeuvre tracking problems, and the probabilistic data association (PDA) filter was adopted for the multisensor case. However, certain assumptions made in the derivation of the EKF algorithms render it suboptimal for track estimation. An efficient tracker that can use data from a host of sensing modalities and are capable of reliably tracking even a target may accelerate at non-uniform rates and may also complete sharp turns within a short time period. Further, the target may be missing from successive scans during the turns. A tracker incorporating radial basis function (RBF) network in a conventional EKF-PDA tracker is proposed, which has several advantages over existing nonlinear estimation algorithms in tracking applications. The main advantage is to gain the capability of adaptability and robustness from the RBF network in order to realize improved tracking performance while at the same time keeping the data fusion computational structure of the tracker as simple as possible.