Nonlinear Filtering, Target Tracking

The nonlinear filtering problem of target tracking of fire control system is concernedin this paper. By analyzed the nonlinear problems of target tracking of fire control system,some methods based on unscented Kalman filter (UKF) are presented. The maincontents of this paper are as follows:Firstly, taking thought the characteristic of real time demand of fire control system,the main nonlinear filtering algorithms are studied based on Bayesian framework. And theappropriate algorithms are selected for the fire control system;Secondly, nonlinear filtering problem with general measurement model is discussed.Under such situation, the extended Kalman filter (EKF) is often adopted. Big error will beimported, if the linearization error can not be ignored. Under the different accuracy ofmeasurement, this nordinear problem is studied based on EKF and UKF;Thirdly, under such accuracy of measurement, to solve the problem that tracking lowvelocity and strong maneuver targets is difficult. For this problem, the reason of thisquestion is analyzed and Doppler measurement of radar is introduced to the fire controlsystem, and derive filter algorithm based on UKF;Finally, low repetition frequency laser range finder, an efficient estimation method isgiven in this paper, which could obviously improve the state estimation precision of targetwithout increasing any hardware cost. That possesses of significant engineering meaningand economic value.Through Theoretical analysis and digital simulation demonstrate that UKFperformance better than EKF in fire control system. The derivative filtering algorithmsbased on UKF are effective for the above-mentioned nonlinear problems.