Compensation of navigation uncertainty for target tracking on a moving platform

Abstract

Established state-of-the art methods for target tracking assume perfect knowledge of the sensor position and orientation. This assumption is violated when the tracking sensor is mounted on a moving platform such as a ship. Two methods for solving this problem are compared. The Schmidt-Kalman filter maintains correlations between the ownship and the target, while a converted measurement approach merely translates the navigation uncertainty into the measurement model of the target. Simulation results indicate that the Schmidt-Kalman filter yields the best improvements with regard to consistency, while the converted measurement approach yields better improvements in root mean square error.