Research On Tracking Layer Technology In Tracking Algorithm Before Detection By Double-layer Particle Filter

Early warning detection system often uses multiple radars to detect and track multiple targets.By effectively using the measurement information of multiple radars,the detection and tracking effect of targets can be improved.The pre-detection tracking algorithm based on the double-layer particle filter structure is a common target detection and tracking algorithm.Through the accumulation of multi-frame information of the target,the detection and tracking of multiple weak targets are realized,which is commonly used in the joint detection and tracking of multiple radars and multiple targets.However,when the radar detects multiple long-distance targets,due to the high false alarm rate of the target,the initial tracking accuracy of some targets is low.Therefore,when multiple targets are close,it is easy to cause the target's track deviation,and even cause the target.Wrong combination of tracks.The multi-radar multi-target tracking algorithm based on double-layer particle filter realizes the detection of new targets and the tracking of discovered targets by the detection layer and the tracking layer,respectively.Filtering false targets is mainly implemented by the target tracking layer.Therefore,in view of the above problems,this paper studies the target tracking layer of the multi-radar multi-target pre-detection tracking algorithm based on the double-layer particle filter structure.Through the correction of the target echo signal and the tracking particle swarm in the target tracking layer,the tracking track target tracking layer.The main research work of this paper is as follows :1.A tracking correction Patricle Filter Track Before Detect(TC-PF-TBD)algorithm for multi-radar multi-target particle filter based on tracking correction is proposed to solve the problem of target track deviation when the target is relatively close.When tracking a certain target,the algorithm modifies the amplitude of radar echo according to the state information of other targets in the overlapping area of radar,then calculates the particle weights and fuses them,so as to reduce the interference of the closer target on the particle weights of the target.Secondly,in the process of particle resampling,the offset particles interfered by other targets are filtered,and the particle swarm distribution is corrected,thereby weakening the interference of closer targets.The simulation results show that this algorithm can improve the tracking accuracy at the initial stage of the target and reduce the false alarm rate.2.In order to solve the problem of low tracking accuracy and high false alarm rate in the early stage of long-range tracking,a multi-radar multi-target particle filter track pre-detection tracking algorithm based on target tracking is proposed.After calculating the target probability in the target tracking layer,the algorithm adds a target retracking method.If the target probability is greater than the threshold,the particles with larger weights in the original particle swarm are retained,and the state estimation of the target frame is taken as the center to provide new particles for the particle swarm,which improves the tracking accuracy of the target in the early stage.Then,in the process of target screening,by calculating the short-term motion distance of the target,the motion frame length of the false target is reduced,so as to better filter out the false target.The simulation results show that compared with the traditional multi-objective PF-TBD algorithm,this algorithm can improve the tracking accuracy in the initial stage of the target and reduce the false alarm rate.3.Aiming at the defects of the two proposed algorithms,a Hybrid Multi-radar Multi-target Patricle Filter Track Before Detec(HM-PF-TBD)algorithm is proposed,which fuses TC-PF-TBD and TR-PF-TBD,and gives full play to the advantages of the two algorithms,improves the target tracking accuracy and reduces the false alarm rate.In the tracking layer,the algorithm firstly uses the radar echo correction method to reduce the interference between the targets,and then calculates the particle weights through the modified radar echo signal and resamples the particle swarm,and then calculates the target probability.If the target does not disappear,it retains the high weight particles in the particle swarm before sampling and provides new particles for the particle swarm with the target state estimation as the center,which improves the tracking accuracy in the early stage of target detection.Finally,in the process of target screening,the false target track is filtered by calculating the short-term moving distance of the target,so as to reduce the false alarm rate of the target.The simulation results show that the proposed algorithm can improve the signal interference between targets,target tracking accuracy and target false alarm problem in the multi-target particle filter algorithm under the condition of long-distance observation.