Research On Maneuvering Target Tracking Filtering Algorithms

With the great change of maneuvering features and the increasing complicated environment,many new technologies have been applied in target tracking system for fitting complex targets' motility and circumstances.With growth of maneuvering target tracking technology,the need of the tracking performance quota has been satisfied,credible and precise target tracking is obtained,and after that a later command decisions can be achieved,which is founded on precise tracking information of the target.Hence,the maneuvering target tracking is more and more being the research crux all over the world,which can be summarized as follows:The motion models and tracking algorithms are expounded.This thesis suggests several familiar motion models,and sums up the features of each model.Some kinds of filtering algorithms are stated in theory,then the filtering performance of each nonlinear filtering algorithms are analyzed theoretically,which provides the theoretical basis for adaptive filtering algorithm study on the nonlinear maneuver targets tracking system later.The cubature Kalman filter(CKF)algorithm and the its improved algorithm are applied to the GPS/DR integrated navigation system and tested the nonlinear estimation performance of EKF algorithm,UKF algorithm and CKF algorithm.To solve the problem of declining evaluate precision of spherical simplex-radial cubature Kalman filter(SSRCKF)when target motion state suddenly change,so a novel adaptive spherical simplex-radial cubature Kalman filter(ASSRCKF)algorithm based on strong tracking filter(STF)algorithm is proposed.By bringing fading factor into SSRCKF to revise the forecast states' deviation covariance,then filter gain is revised,and the tracking deviation declines adaptively,so the robustness is reformed.The simulation results show that the proposed method has better tracking performance than AIMM-SRCKF to deal with unknown maneuvering target tracking problem.Secondly,a improved HMSSRCKF algorithm is proposed to overcome the falling filtering precision of SSRCKF when the system noise statistic is unknown.The measurement renewal of SSRCKF algorithm is carried out by the Huber M estimation.Due to not calculated measurement covariance directly,HMSSRCKF is robust to deal with this uncertain situation.The experimental results indicate that the new algorithm has better tracking performance than SSRCKF.To solve the problem of low accuracy and slow model switching speed of IMM-CKF algorithm caused by unknown system noise statistics in maneuvering target tracking,so a IMM-ASSRCKF algorithm based on the simplified Sage-Husa algorithm was proposed.The new method adopted SSRCKF algorithm based on the spherical simplex-radial(SSR)cubature rule to obtain better accuracy than conventional CKF,meanwhile a simplified noise estimator for nonlinear system based on Sage-Husa algorithm was added to SSRCKF algorithm,therefore the IMM-ASSRCKF target tracking algorithm was generated by combining IMM algorithm.The proposed IMM-ASSRCKF algorithm was applied to track the established anti-ship missile trajectory to test its performance.The simulation result shows that the improved tracking algorithm can achieve faster convergence speed and better robustness than conventional IMM-CKF.