Study On Inter-satellite Baseline Determination Method Of Formation-Flying Satellite

Satellite formation flying is currently one of the focuses being studied in the domain of domestic and overseas spaceflight. Satellite formation flying uses all kinds of sensors loaded on the satellite to realize many high-precision measurement missions which cannot be achieved by a single large satellite by the cooperation of inter-satellite and inter-sensor, so it will greatly improve the general performance and expand the application fields of the custom satellite system. One of the key technologies of satellite formation flying which need to be resolved is the determination of high-precision inter-satellite baseline. High-precision inter-satellite baseline is not only the key of the digital elevation measurement and ground moving target indication(GMTI) of the distributed synthetic aperture radar(SAR), but also very important to the control of formation satellite.Aiming at dual-satellite formation, this thesis studies the methods of inter-satellite baseline determination, and its main achievements will include three parts mostly:High-precision baseline determination methods based on autonomous measurement means are studied. By analyzing the characteristic of the common state measure methods combined with the formation requirement, this thesis puts forward a combined measurement method, which can not only improve the autonomy of formation, but also realize the high-precision of the determination of inter-satellite baseline. Because of the difference between the directly measured baseline and that actually used one by the formation, taking the distribute InSAR for example, this paper gives the transform flow between them and constructs the corresponding error propagation model, and analyses all sorts of factors influencing the altimeter baseline determination precision by simulation.The post-facto inter-satellite relative positioning method based on global positioning system(GPS) measurement is studied. Based on the continuity of relative position and the theory of function approximation, spline model of GPS relative positioning is constructed. Because the traditional random model is not identical with that of real GPS double-difference measurements, an iterative random modeling method based on spline function model is put forward, and the concrete algorithm flow and its implementation steps of single-frequency and dual-frequency GPS are given. In addition, in order to improve the precision of relative positioning and to increase the robustness of system, an relative positioning method based on augmented GPS is proposed, and its corresponding spline function model is constructed, and the concrete algorithm of state estimation is given.The real-time inter-satellite relative positioning based on augmented GPS is studied. Based on with and without the dynamic model of formation satellites, three methods are proposed, naming dynamic sequential point estimation method, adaptive unscented kalman filtering(UKF) relative positioning method based on triple-differenced measurement time propagation and adaptive UKF relative positioning method based on high-precision dynamic model of satellites . Because the measurements of current epoch are only used during the parameter estimation of the first method, the level of decimeter of relative positioning precision can be attained. The second method uses the least square method to achieve the time propagation relationship based on GPS tripe-differenced measurements, and uses adaptive UKF to estimate the inter-satellite relative positioning. Simulations show that this method can effectively improve the precision when the inter-satellite distance is short. Compared to the former two methods, the third method not only uses all measurements before the current epoch, but also includes the state information of the system (naming satellite orbit dynamic model) during the estimation of state, so its precision of relative positioning is further improved.