UKF and Its Application to 3-D Underwater Target Tracking System
Fubin Zhang · Ship Engineering · 2005
This paper introduces the newly proposed unscented Kalman filter (UKF). In UKF, a minimal set of carefully chosen sample points is used to represent random variables distribution. And when propagated through the true nonlinear system, these sample pointes capture the mean and covariance accurately to the 3rd order for nonlinear transformation. In this paper, UKF is applied to a 3-D underwater target tracking system. The Monte Carlo simulation demonstrates that the UKF has higher filtering accuracy than conventional extended Kalman filter.