Research And Implementation Of Target Tracking Algorithm Based On Passive Time Difference Location

Because of the increasing complexity of international relations and the development of electronic information technology,the research interest in the subject of target positioning and target tracking remains high.Interference of high maneuvering clutter environment introduces problems of low target positioning accuracy and difficulty in determining the number of real targets.It requires target positioning and tracking to be more accurate,real-time and efficient.This thesis optimizes the related passive time difference positioning and maneuvering target tracking algorithms from two aspects: improving positioning accuracy and improving target tracking performance.In passive time difference positioning,the Chan algorithm is widely used in engineering because of its small amount of calculation and high accuracy.However,there are fuzzy and unsolvable problems when using this method to solve the positioning solution.This paper analyzes the baseline length,included angle,layout of the base station and the influence of the time difference measurement error on the special solution area.It puts forward a solution to the problem of special solution.This thesis also studies the core content of time difference positioning: positioning accuracy.This thesis first introduces the angle relationship between the target and the base station to derive the calculation process of the plane positioning accuracy.It obtains the theoretical basis of the influence of the base station layout and target height on the accuracy.Based on the above,the research object is the air target.It introduces the distance difference differential method to calculate the spatial positioning accuracy.It also analyzes the relationship between positioning accuracy and base station layout,baseline length,time difference measurement error,station site error,target height and other factors through simulation.Then it gives suggestions to improve positioning accuracy.In the study of single target tracking problem.First of all,this thesis analyzes the target motion characteristics and the application scope of each motion model.In addition,the tracking performance of CV,CT and IMM models in the same target motion state is compared by simulation.From the results of filtering,position error and RMSE,IMM has better filtering performance,and the performance is related to the probability of transfer matrix.Then,aiming at the nonlinear filtering algorithms EKF,UKF and PF based on Bayesian recurrence formula.On the basis of principle description,the tracking performance of EKF,UKF and PF in Gaussian and non Gaussian noise environment is compared.The final results show that PF algorithm has the best filtering performance.In multitarget tracking,aiming at the problem of poor estimation accuracy of speckle-type outlier elimination,a method of outlier elimination based on the points supplemented by the Spline algorithm after kalman identification is proposed.Through the verification of the real data collected,it is found that the target eliminated by the correction method is more convenient for subsequent tracking and processing.For the problem of data association,the theoretical analysis of association gate,similarity measurement method,association judgment criterion,and association algorithm is given.It compares the filtering performance of PDAF and NNDA algorithms.Aiming at the problem of track initiation,firstly,we study the initiation principle of the logic method.In order to solve the defects of the method with high false initiation rate and easy to produce "V"-shaped track,this thesis proposes a modified logic method based on self-reporting points.According to the self-reporting points existing in different initial stages,the initial steps of the track are corrected.Through qualitative simulation experiments and quantitative calculations,it is found that the false initiation probability of the correction method is about 20% lower than that of the traditional method.Finally,the algorithm is applied to the project of multi-target tracking system.It can realize the functions of target location,track association,track initiation and track maintenance within the monitoring range.At present,the system has been developed.The system can cooperate with the base station for joint debugging and put into use in the outfield to meet the requirements of real-time monitoring and displaying the target track.