| In recent years,multi-target tracking based on the random finite set has received extensive attention in the field of information fusion,and has become a focus of the research in the field of multi-target tracking.Based on the random finite set theory framework,the labeled multi-Bernoulli filter achieves trajectory-level filtering by introducing label ideas into the random finite set theory,providing efficient and stable tracking performance.In radar target tracking research,radar observations are often interfered by glint noise.However,measurement noise is mostly modeled by Gaussian noise,which is not applicable.In order to accurately track multiple targets under glint noise,this paper mainly studies the multi-target tracking algorithm based on the labelel multi-Bernoulli filter under glint noise.The research work is as follows:1)We introduce the probability hypothesis density filter(PHD),the cardinality balanced multi-target multi-Bernoulli filter(CBMe MBer),the δ-generalized labeled multi-Bernoulli filter(δ-GLMB)and the labeled multi-Bernoulli filter(LMB),which are based on the random finite set theory.In view of the non-Gaussian and long tailed characteristics of glint noise,this paper studies the problem of glint noise and uses the Student’s t-distribution to model the glint noise.2)In order to track multiple targets under glint noise,an efficient labeled multi-Bernoulli filter for glint noise is proposed.Then we develop its implementation method in linear system and nonlinear system under glint noise.Simulation experiments verify the multi-tracking performance of the filter in the glint noise environment.3)To track multiple maneuvering targets under glint noise,a novel efficient labeled multi-Bernoulli filter for jump Markov system model under glint noise is proposed by introducing the multi-model method into the filter.In addition,the implementations of the proposed filter in linear system and nonlinear system under glint noise are proposed.The simulation results illustrate that the proposed filter performs better than other filters in tracking multiple maneuvering targets under glint noise. |
