The Research On Radar Target Tracking Algorithm

Radar is an indispensable electronic device in modern warfare,target tracking of radar is an important function of radar equipment,and the performance of target tracking directly affects the performance of radar.So the radar target tracking technology becomes the focus of researchers.This thesis discusses the key technology of radar target tracking;simulates and analyzes target motion model and target tracking algorithm(such as kalman filtering algorithm and particle filtering algorithm),and evaluates their performance.The main contents of this thesis are described as follows:Firstly,introduces the fundamentals of target tracking.In this part,we introduce Bayesian theory and radar coordinate system(including geodetic coordinate system,Cartesian coordinate system,polar coordinate system and the conversion between Cartesian coordinate system and polar coordinate system),and introduce the fundamentals and key steps of measurement data preprocessing,including wild value preprocessing and data association.Secondly,studies several common target motion models in radar target tracking,including CV(constant velocity)model,CA(constant acceleration)model,Singer model and "current" correlation statistical model,etc.For these models,we discuss the system equations,the scope of application,the advantages and problems in application of these models in depth,and test and verify their performance according to simulation.The results show that,CV model has better tracking performance for uniform linear motion,and CV model has better tracking performance for uniform acceleration linear motion and uniform circular motion.Finally,the thesis discusses filtering algorithms in radar target tracking,and mainly studies kalman filtering algorithm and particle algorithm.In the kalman filtering algorithm part,we discuss its system model and its filtering model and study the initialization problem of the kalman filter.On this basis,we simulate and study linear kalman filtering algorithm,test and verify that linear kalman filtering algorithm has good tracking performance in linear system.In the particle filtering algorithm part,we introduce its basic theory and core module,the sequential importance sampling.To solve the problem of particle degeneration in particle filter,resampling is used.At last,applies particle filtering algorithm into target tracking for simulation and analysis,and the simulation proves that particle filtering algorithm still has good tracking performance in nonlinear system.