An Unscented Kalm an Filter-Based MultisensorTrack Fusion Algorithm

Abstract

In this paper, an unscented Kalman filter (UKF)-based track fusion algorithm is developed for tracking targets in a nonlinear multisensor system. Employing the unscented Kalman filter and the measurements of the individual sensor in the multisensor system, the means and the variances of the states of a tracked target can be estimate. Based on these estimate results, an optimum state fusion scheme is obtained in terms of minimum mean square error (MMSE). The scheme can make the variance of the fused states smaller than that of the states estimated by UKF with any individual sensor in this multisensor system. Simulation results confirm the efficiency of the presented algorithm