Research On Space Target And Targets Group Tracking Methods Of Wideband Radar

With the development of space technology, space surveillance and early warning abilityhave become the core foundation of the international military competition. Wideband radarcan get higher range resolution, more target information and has better clutter restrainedability. It has important functions in the aspects of tracking, recognition, and guidance, so ithas great effect to the performance of whole space surveillance system. The widebandradars in use are measurement radars and mainly utilized for target imaging in China. Andthe space target range from radar is long and has low signal-to-noise (SNR). Thus, thetraditional target tracking method with large SNR and narrowband radar are not suitable forspace target tracking with wideband radar. The study aims at space target and space targetsgroup tracking with wideband radar. For nonlinear tracking of the space target withwideband radar under low SNR, nonlinear filters have been studied and designed toimprove tracking precision. Moreover, the space targets group tracking methods are studiedgiven that the targets in a group have the same motion and share the similarity of orbit. Byutilizing multiple targets information, new methods for space targets group tracking andweak target tracking in group are presented. The specific content and conclusion of thesestudies are as follows.The nonlinear cause of space target tracking and filtering methods are studied. Firstly,the optimal Bayesian filtering and multiple sub-optimal nonlinear filtering methods forspace target tracking are derived based on the motion and measurement models of spacetarget. And the effect of nonlinear filtering methods is examined from theory andsimulation. In addition, aiming at the nonlinearity of space target motion and measurement,the curvature is utilized to measure the target state transition function. And the linearitydegree of a measurement function is proposed and quantified based on correlationcoefficient and Taylor series expansion of nonlinear function. From nonlinearity analysis,the nonlinearity of target motion curve is extremely small. The nonlinear filtering of spacetracking is mainly decided by radar nonlinear measurements. And the nonlinearity of radar measurement increases with the increase of radar bandwidth, beam width and decrease ofSNR.Aiming at the nonlinear space target tracking problem, a mixed nonlinear filteringmethod based on forward-backward filtering for space target state estimation is proposed. Aforward-backward-forward filtering frame with sliding fixed-interval is utilized in thepresented method. Within the frame, many filtering methods can be chosen and employedfor forward filtering and backward filtering. And then, the motion characteristic andmeasurements can be fully used, thus the affect of nonlinearity and SNR on state estimationaccuracy can be decreased. From the nonlinearity of space target tracking and differentcharacter of filtering methods, the forward and backward filtering methods are reasonablychosen and the lower bound PCRLB of estimation accuracy with mixed-nonlinear filteringmethod is derived. From the simulation, the estimation accuracy of the proposed is muchbetter than forward filtering. And the PCRLB with the new method is better than that offiltering.The motion of space targets group is studied, and aiming at dispersed-spaced spacetargets group, a layered filtering method for space targets group is proposed. From theanalysis of space targets group movements, space targets tend to travel in groups are aredispersed-spaced most of the time, except for a short time after being released. With theincrease of time, these targets in a group are much more dispersed and then travelseparately. For timely target discrimination and attack, space target should be tracked assoon as possible. And the group information can to be utilized for tracking target inadvance. At the top layer of the proposed layered filtering method, target measurements arefused to improve the group centre measurement accuracy, and then the centre state isestimated with high accuracy. At the bottom layer of the proposed method, the targettracking is assisted by the group centre with the introduced local motion which is defined asthe space target motion relative to the group centre. From simulation, it is proved that thegroup centre filtering accuracy is more accurate than that of space target. And theestimation accuracy of the target in the group is improved with the layered method.Moreover, the gain of target state estimation accuracy with the proposed method increaseswith the decrease of target SNR. There might be weak and strong space targets with different radar cross sections (RCS)in a group which are close to each other and move following the same motion model. Thestrong space targets can be prior detected and tracked. With these, a novel method of weakspace target tracking assisted by strong space target is proposed. By introducing the relativemotion tracking, the nonlinear tracking of weak space target is converted to linear relativemotion state estimation. In addition, the errors in the motion model for space target trackingare counteracted through the relative motion. Compared with weak target tracking withoutstrong target assistance, it is proved that the proposed tracking method can provide higherstate estimation accuracy, faster convergence rate and better track validity probability forweak target. In addition, the gain of weak target estimation improvement with assistanceover that without strong target assistance increases with lower SNR of weak target andhigher SNR of strong target.