In this paper, performance of a non-linear filter named high-degree cubature kalman filter is studied for Bearings-Only Tracking (BOT) problem. Objective of BOT problem is to find the kinematics of a target using noise corrupted sensor bearing measurements. In many situations, ownship maneuver is required which makes the target states observable. Herein a maneuvering ownship estimates the target kinematics using a high-degree cubature kalman filter. In this filter the spherical-radial cubature rule has been expanded from third degree to fifth degree and the resulting, filter accuracy has improved. The bearings-only target tracking simulation shows that this filter provides better performance over the conventional algorithms.

Keywords