Research On Tracking Methods Of Maneuvering Targets

With the continuous progress of science and technology,the emergence of high-speed and high-mobile aircraft represented by aerospace aircraft,ballistic missiles,and fifth-generation aircraft has brought great challenges to radar target detection.Existing target tracking algorithms can achieve better tracking performance when the target moves stably.However,when the target is maneuvering,these algorithms usually show significant performance degradation.How to track the maneuvering target stably and effectively has become an important issue to be solved.Aiming at the problem of maneuvering target tracking,this paper focuses on the maneuvering target tracking algorithm combining AR model,smoothing filter and interactive multi-model structure,a method of target tracking and massto-drag ratio estimation for reentry cone based on maneuvering model,and the design of target tracking module in radar system is realized by programming language.The main contents of the paper are as follows:1.Research on the modified method of model probability in IMM algorithm.Several commonly used motion models and the basic principles of EKF filters are introduced.On this basis,an RTS smoother based on EKF filter algorithm is presented.The AR model is deduced,the method of solving the AR model coefficients is given,and the advantages of the AR model in maneuvering target tracking are analyzed.The basic principle of IMM is introduced.Based on the ERTSS smoother,a method of using smooth state covariance to modify the model probability in IMM algorithm is given.On this basis,combined with AR model,an interactive multi model algorithm of AR model based on ERTSS smoothing filter is proposed,and the specific implementation steps of the algorithm are given.The advantages of AR model in target tracking compared with traditional discrete-time model are verified by simulation,and the excellent performance of an interactive multi-model algorithm of AR model based on ERTSS smooth filter for maneuvering target tracking is verified.2.Research on tracking of reentry targets and estimation of mass-to-drag ratio.Firstly,the definition of the mass-to-drag ratio of the reentry target and the modeling method of atmospheric density are given.Then,the motion state equation of the reentry target and the radar measurement equation are established through the force analysis of the reentry target.The basic principle of the UKF filter is introduced,and an improved UKF filter is given to the defects of the UKF filter.The prediction covariance of UKF filter is modified by introducing the fading factor to adjust the filter gain,and the superiority of the algorithm is verified through simulation.The principle and characteristics of the LSTM network are introduced.On this basis,a method for estimating the mass-to-drag ratio of the reentry target based on LSTM is proposed.The training method of LSTM network is given and the excellent performance of LSTM network to estimate the mass-to-drag ratio of reentry target is verified by experimental simulation.3.Aiming at the engineering realization of target tracking technology,the design of target tracking module in radar system is realized by programming language.The methods of dynamic storage,data encapsulation,memory alignment,parameter transfer and other methods related to the design of the target tracking module are introduced.The implementation flow of the target tracking module and the interface and function of the important functions are given.Finally,the target tracking module is integrated with other modules in the radar system,and the correctness of the target tracking module is verified through joint debugging,which has high engineering application value.