Research On Cognitive Multi-target Tracking Method In Distributed Radar Network

In order to adapt to the increasingly complexity of target tracking,how to make full use of distributed radar network so as to obtain the desired target tracking accuracy has become a hot topic.The thesis discusses this topic and focuses on the distributed radar network cognitive multi-target tracking method:Above all,the basic problems of cognitive multi-target tracking in distributed radar network are studied,including state estimation filtering algorithm,distributed fusion algorithm with feedback,target tracking performance description parameter,which lay some foundations of the cognitive multi-target tracking algorithm in distributed radar network.Next,the cognitive multi-target tracking based on distributed phased array radar network is studied,and exhaustive matching relation with the golden section resource optimization method,the heuristic matching relation with the golden section resource optimization method,and the heuristic matching relation with the polynomial resource optimization method are given respectively.Among them,the method that enumerates all target-radar matching relations,and the transmitted energy resources required by each radar are calculated.Then the relation with least sum of tracking accuracy or the relation with least sum of resources consumed is selected.In order to reduce the computation,the heuristic allocation method is used where the radar with shorter distance is selected for the target with large deviation of tracking accuracy.On this basis,the golden section method and polynomial solution method are used to calculate the radar emission energy respectively.Then,the cognitive multi-target tracking based on collocated multiple input multiple output(CMIMO)radar network is studied,and a dual objective function optimization problem is established.Based on the heuristic allocation method,hypothesis allocation method and the golden section resource optimization method,a distributed CMIMO radar network cognitive multi-target tracking method is proposed.The method adaptively controls the subarray division,beam direction and emission energy.When the expected tracking accuracy is strict,the sum of tracking accuracy deviations can be minimized,which means that the tracking accuracy is close to the expected value with limited system resources.When the tracking accuracy is loose,the sum of resource consumption can be minimized while ensuring the target tracking accuracy.Finally,according to the different work modes of CMIMO radar,the cognitive multitarget tracking framework of a distributed heterogeneous radar network containing phased array radars and multi-beam CMIMO radars,a distributed heterogeneous radar network containing phased array and single-beam CMIMO are established respectively to form a dual objective function optimization problem,and the corresponding cognitive multi-target tracking method is proposed.The proposed method can adaptively adjust the subarray division,beam direction and emission energy to affect the allocation relation,so as to reduce the emission energy of the system while keeping the target tracking accuracy close to the expected value.