Research On Multi-target Localization And Tracking Based On Direction Of Arrival

The Passive location and tracking technology uses the intercepted signal from the target radiation source to determine the position and speed of the radiation source.It does not need to transmit electromagnetic signal,has good concealment,can improve the survivability of the defense system in the complex electronic warfare environment,and is one of the hotspots of detection and reconnaissance technology research in recent years.In various passive location and tracking systems,the passive location and tracking technology based on Angle of Arrival(AOA)uses the arrival angle of the radiation source to locate and track the radiation source,which is widely used in single station and multi-station network observation scenarios.However,in the over the horizon observation and asynchronous observation of multi-target tracking scenarios,the traditional AOA based localization and tracking technology is faced with such problems as large localization error and easy to produce false target.In view of the particularity of AOA based passive sensor localization and tracking,this paper mainly studies the bias compensation of localization,multi passive sensors over the horizon localization,multi passive sensors multi target AOA cross data association,multi passive sensors asynchronous AOA multi target tracking.In chapter 2,the methods of AOA fusion localization and bias compensation for known altitude target is studied.First,the target observation model is established and the analytical solution of the target position is derived.Based on this,a method of bias estimation and compensation is proposed by the expansion of the third-order Talyor series on the analytical solution.This method can keep the higher order Taylor series expansion terms,the error estimation accuracy is higher,and the expression of the error is simpler.Then,to solve the problem of multiple AOA measurements fusion localization,the performance of two fusion lozalization methods based on original AOA measurments and single localization result is analyzied.Finally,the effectiveness of the proposed method is verified by computer simulation.In chapter 3,the method of over the horizon cross localization based on multi passive sensors is studied.Aiming at the problem of over the horizon localization caused by atmospheric refraction,a passive sensor over the horizon observation model is established,which can adapt to the scene where the over the horizon and sight distance observation coexist.A maximum likelihood estimation algorithm based on constrained pseudo-linear least square(CPLS)initialization is proposed to solve the problem of locating target with known altitude.Aiming at the problem that the localization bias has great influence on the localization performance,a Gauss-Newton iterative localization algorithm based on bias compensation is proposed.The effectiveness of the proposed algorithms are verified by computer simulaitons.In chapter 4,the method of multi-passive sensor multi target AOA crossover data association is studied.First,the shortcoming of the traditional data association algorithms based on observation domain is analyzed,and the conclusion is drawn that the time complexity of the traditional data association algorithms are exponentially correlated with the number of sensors and targets.To solve this problem,a direct data association algorithm based on target state domain is proposed,and the framework,cost function model,candidate target initialization method,algorithm termination condition and time complexity are analyzed in detail.This algorithm transforms the data association problem from the observation domain to the target state domain,and the time complexity changes from the original exponential relationship with the number of sensors and targets to the present linear relationship,which greatly improves the efficiency of data association.Finally,the effectiveness of the proposed algorithm is verified by computer simulation.In chapter 5,the method of multi passive sensor asynchronous AOA multi target tracking is studied.First,the multi passive sensor multi target observation model is established,and the motion model of cruise target with constant velocity at constant altitude in the geodetic coordinate system is derived.Then,the difficulties and problems of multi target tracking under the condition of multi passive sensor asynchronous AOA measurements are analyzed.On this basis,an algorithm of multi passive sensor asynchronous AOA multi target tracking is proposed.The multi target tracking problem is transformed into measurement to angle track association,angle track to angle track association,measurement to track association and so on.The algorithm uses the direct data association algorithm based on the stated domain to carry out track initiation and track update,which has higer timeliness.Then,in view of the situation that synchronous observation may occur in asynchronous observation,an algorithm of track updating for variable dimension observation under general conditions is proposed.Finally,the effectiveness of the proposed algorithm is verified by computer simulation.