IMM Box Particle Filtering for Bearings-Only Maneuvering Target Tracking

Abstract

Aiming at bearings-only maneuvering target tracking, a new multisensor interacting multiple model box particle filtering algorithm is proposed, which uses the interactive multiple model algorithm to estimate the state of each box particle. In the update stage of the proposed filter, the quadratic contraction algorithm is put forward to increase the interval of innovation and the survival probability of the box particle, which achieves the purpose to improve the diversity of the box particle and the performance of target state estimation. Simulation results are presented to demonstrate the improved performance of the proposed filtering algorithm.