Research On Maneuvering Target Tracking Based On Gauss Filter

In recent years,with the rapid development of science and technology,ballistic missiles and various types of aircraft in addition to conventional maneuver,but also in different sports models between the instantaneous switch,take the initiative to complete more complex maneuver,the movement becomes more uncertain and volatile The Correspondingly,the various types of reconnaissance,early warning and interception systems as command layers also put forward higher requirements for the tracking system,requiring modern target tracking technology to track the maneuvering target motion more quickly and accurately.Therefore,in the face of the advanced flight technology,this paper deeply studies the Kalman filter tracking algorithm and its instability phenomenon.Based on the Gaussian filtering algorithm,this paper studies the key technologies of the three aspects of the non-cooperative maneuvering target tracking problem : Maneuvering target tracking in linear model,maneuvering target tracking of nonlinear model,and maneuvering target tracking in clutter environment.The main research work and contribution are as follows:(1)Kalman Filtering Tracking Algorithm and Its Instability.Based on the existing literature research,the motion equations and the measurement equations of the two different situations,including the filtering model with process noise and without process noise,are systematically summarized and deduced.Depth analysis and analysis of the Kalman filter process noise and filter inconsistency of the phenomenon,whether the configuration process and configure the different process noise power on the Kalman filter algorithm performance of the degree of a comprehensive simulation analysis.Finally,the Kalman filter performance and process noise configuration are systematically expounded,and the contradiction between Cramer-rao low bound(CRLB)and filter consistency is estimated.(2)Maneuvering target tracking under linear model.The Gaussian-Aikten filter is studied,The Based on the test of squared sum of the filtered weighted residuals,the control model of the proposed model is matched and the real motion model of the maneuvering target is matched.The control algorithm of Gaussian filtering is proposed,and the detailed steps and algorithm flow chart of the algorithm are given.The algorithm of Gaussian-Aikant filter and three improved interactive multi model(IMM)Kalman filter are simulated and compared.The control algorithm is used to control the Gaussian-Kettle filter.(3)Maneuvering Target Tracking in Nonlinear Model.The nonlinear tracking model of Gauss-Newton filter algorithm is studied,and the nonlinear tracking problem of maneuvering target is transformed into the approximate linear filtering problem of disturbance vector by Newton local linearization method.The realization of the minimum variance method in the nonlinear system is deduced,and the Gaussian-Newton filter algorithm for maneuvering target tracking is deduced in detail.The Gauss-Newton filter algorithm and the IMM-PF algorithm are simulated and compared in the typical maneuvering target scene.The tracking performance of the Gauss-Newton filter algorithm in the nonlinear maneuvering target tracking problem is analyzed.(4)Maneuvering Target Tracking in Clutter Environment.On the basis of the existing literature,the validity of the nearest neighbor algorithm for the selection of the measurement is analyzed.And the advantages of Gaussian filtering on maneuvering target tracking are fully utilized.The PSNF algorithm based on Gaussian filtering is proposed based on the strongest neighbor algorithm based on power feature.The algorithm is compared with IMM-PSNF and IMM-PDAF in different intensity clutter environment,and the tracking stability and track loss rate of the three algorithms are studied.