Research On On-Orbit Target Tracking Algorithm Based On Kalman Filter

With the deepening of human space exploration,space has gradually become the main battlefield of the world’s major powers wrestling.The control and utilization of space has become a symbol of national science and technology.And the accurate and stable tracking of a large number of orbiting objects in space is a prerequisite for observing or controlling them.Generally,elements of orbit are used to reasonably reason out the position vector and velocity vector of orbit targets,then according to the state,observation information of orbit targets are calculated,and then the observation equipment is guided for tracking on this basis.In engineering application,as elements of orbit will creep over time or indeterminacy of elements of orbit or other reasons,there are some errors in the states of the targets in orbit,such errors will cause the target not to be in the center of the field of view,that is,the miss distance generated,and then affect the tracking effect,even in extreme cases,the target will be lost.Aiming at this problem,this paper focuses on discussing the filter tracking problem of non cooperative targets under the condition of bearing-only angle.Firstly,the time scale of on orbit target motion and transformation of celestial coordinate system and earth coordinate system are studied,and the algorithm is implemented by using SOFA software library;Secondly,this paper proposes a filtering algorithm based on complex mechanical model,in this mechanical model,the main perturbation factors for the motion of an orbiting object in space are considered,and some perturbation models are briefly described,then on this basis,the perturbation equation is solved by numerical integration,which realizes the track extrapolation algorithm,and this algorithm is compared with computing results of high precision orbit prediction function in STK software;then,the algorithm takes position vector and state vector of orbit targets as system state quantity,and based on the height angle and azimuth angle of the target relative to the ground observation station,the filter based on extended Kalman filter algorithm and unscented Kalman filter algorithm is designed.Finally,the simulation test platform is built using MATLAB and C.The STK software and the measured data are used as the input of the algorithm,and the influence of the different frame rate data and filter parameters on the filtering results is studied.Simulation results show,in the orbit target filter tracking system studied by this paper,extended Kalman filter algorithmand unscented Kalman filter algorithm all can correct the number of orbital elements with multiple errors by means of the data of multiple sets of theodolite,the unscented Kalman filtering algorithm is slight better than extended Kalman filter algorithm in robustness,convergence speed and filtering accuracy,but to a certain extent,it sacrifices the algorithm efficiency.