Bearings-Only Tracking with Biased Measurements

Abstract

This paper focuses on particle filtering techniques for tracking a single target using bearings-only measurements. The problem is formulated as fusing information collected from two or more sensors in the presence of additive noise and multiplicative/additive biases. Assuming the biases are nuisance parameters and marginalizing them out from the estimation problem, we propose an algorithm that combines a standard particle filter and one Kalman filter to efficiently resolve the fusion problem. The algorithms are tested and compared by computer simulations which offer insight into the advantages and disadvantages of the proposed method.