A Research On Multi-target Tracking Algorithm For MIMO Radar Based On Optimized Particle Filter

As the fast development of contemporary science and technology,radar's detection and tracking of targets is facing more and more challenges.Multi-target tracking technology has widely research area.Distributed MIMO radar relies on spatial diversity technology which can efficiently solve the impact of RCS flicker.It has unique advantages for improving detection performance.Therefore,the development of multi-target tracking technology for distributed MIMO radar system has attracted much attention.In order to improve the multi-target tracking performance of distributed MIMO radar,the paper focuses on particle filter and related improved algorithms,as well as probability hypothesis density filtering for multi-targets processing.The main research work includes:Firstly,for the multi-targets tracking problem in MIMO radar scene,this paper introduces the motion models of the target in different scenarios,analyzes the working principle of MIMO radar and the principle of detection and tracking,establishes the signal model and observation model,and compare the advantages and disadvantages of some current popular tracking algorithms.In this paper,the research mainly focus on particle filter tracking algorithm and random finite set theory.Aiming at the problems of particle filter algorithm in multi-target tracking,this paper proposes an improved algorithm based on probability hypothesis density.Secondly,aiming at the tracking problem of time-varying or unknown scenes,the paper focuses on the probability hypothesis density filtering.On the basis of random set theory,the multi-objective model is established.An auxiliary particle improvement algorithm that introduces observations into the distribution function is mentioned in the paper.In addition,the auxiliary particle filter algorithm based on probability hypothesis density is further optimized by setting the threshold to the preprocessing method of observation data in this paper,By marking particles improves the probability of particle distribution in the high likelihood region.The simulation results show that the proposed optimization algorithm reduces the tracking error and improves both the target state and the number estimation accuracy,which improves the particle utilization efficiency to some extent.Finally,for the multi-targets cross trajectory tracking consistency and coherence problem,in this paper,the target feature vector is extracted,and the feature vector is updated by the likelihood function to track the particle set.By optimizing the particle set step,the cross track detection and tracking performance is improved to some extent,and the particle set stability is improved.Aiming at the problem of increasing the computational complexity and complexity caused by the increase of the number of particles in particle filter,this paper proposes an improved algorithm of adaptive particle number by analyzing information entropy and error ellipse theory,and analyzes the relationship between state estimation error and measurement error covariance.The error ellipse confidence interval is obtained from the estimation error,and the lower bound of the adaptive particle number is derived,and the relationship between the error elliptic parameter and the particle number is analyzed.The simulation results show that the improved algorithm can improve the tracking accuracy when multi-objective overlapping trajectories.The improved algorithm is about 1.2 times shorter than the traditional PF algorithm,the target estimation error is reduced,and the tracking real-time and tracking accuracy are improved by the way.