Maneuvering target tracking of underwater autonomous vehicles based on bearing-only measurements assisted by inequality constraints

Abstract

This paper addresses the problem of enabling a constant velocity observer to track a maneuvering target using inequality constraints. In this paper, an observer measures the bearing of sound emitted from the target at each time step. We consider an observer which can access the inequality constraints imposed on target's speed, range, target's course, or target location. As far as we know, this paper is unique in making a constant velocity observer, which measures the bearing angle of a maneuvering target, track the target assisted by range, course, speed, or target location constraints. The observer restarts the filter using inequality constraints to reduce the estimation error considerably, even after the filter diverges. As the observer restarts the filter, stored target information is reused by running the filter backward in time. We verify the effectiveness of our filter restart approach by comparing it with the interacting multiple model (IMM) Kalman filter using MATLAB simulations.