Kinetic States Real-time Estimation For Space Targets

Kinetic states real-time estimation for space target is one of the most significant supporting technologies in aviation and aeronautic area, it's the bases of launch security control, vehicle separation, orbit and attitude control and maintenance, shuttle re-entry, missile defense and penetration. To establish a real-time system for high precision kinetic states estimation, many problems should be tackled. Under the present measurement conditions, this dissertation is mainly bout how to obtain high precision kinetic states real-time estimation for space target during different flight phases. Three key points were focused on kinetic states modeling, measurement acquirement and utilization, real-time estimation methods.Concerned problems about kinetic states real-time estimation were analyzed firstly, laying emphasis on states modeling, measurement and estimation methods. Then, considering missiles and satellites as typical space targets, different flight phases of them were studied in detail, taking all the three points above into accout.In boost phase of a trajectory, studies are mainly about parmameterized methods of modeling and estimantion. According to the difficulty of dynamic modeling and new problems in novel measurement system that position by range rate information, superiorities of parameterized methods are prominent, as analyzed in this thesis. Unfortunately, present parameterized methods often deficiency in real-time needs. In this thesis, for quasi real-time purpose, a modeling method based on sliding polynomial was adpoted, it's already been pilot used in data processing and distribution designing of observation stations in shooting range. An algorithm based on spline EKF was also ameliorated for real-time requires, the improvement made the algorithm more robust and more suitable for practical instance.The discussion of re-entry phase was focused on re-entry trajectory modeling. With a simple model that no dynamics exist, a real-time estimation algorithm using EKF was derived and validated by simulation firstly. Then, time series modeling and error compensation using combinations of nonlinear filters were discussed tentatively for high precision purpose.For the free-flight phase, researches centralized on formation flying satellites. Firstly, Base on dynamic modeling, state equation of absolute orbit was given. A method called baseline spreading for relative orbit modeling was proposed, it does not use assumptions, which are prerequisited in Hill equation. As to attitude estimation, a kinematic modeling for absolute attitude was obtained using attitude rate informantion of gyroscopes, and method for relative attitude modeling was suggested too. Then, taking GPS precise single satellite positioning as an example, measurement acquirement and joint utilization of single sensor were studied. After this, Multi-sensor measurement fusion was introduced in state estimation of satellite, and fusion estimantion about absolute orbit, relative orbit and absolute attitude of formation statellites were achieved. In addition, three mode of measurement for intersatellite sensing were devised. Finally, a fusion algorithm using both absolute and relative measurement was proposed, the algorithm avoids relative orbit modeling, and has a great improvement for absolute orbit estimation, which was testified by simulation.