中文摘要：

空间飞行器需要实时的高精度轨道信息来完成对栽荷的指令操作和遥感数据的实时处理。除了星栽GPS技术，星载多普勒无线电定轨定位系统（DORIS，Doppler Orbitography and Radio—positioning Intergrated by Satellite）是仅有的有能力提供分米级精度的实时在轨轨道确定技术，它可通过测量星地相对多普勒频移，在星上完成实时定轨和预报，目前该技术已在国外多个卫星上实现，达到了较好的效果，而我国还没有建立这样实时自主定轨系统。为此，结合我国高分辨率空间对地观测系统的建设需求和我国航天器对实时自主定轨及其精度的要求，利用扩展卡尔曼滤波算法对多普勒测量进行了实时自主定轨仿真计算，分析了频率偏差估计与否、初轨误差、地面信标站地理分布以及观测精度等对实时自主定轨的滤波收敛时间和定轨精度的影响，为我国利用DORIS技术进行实时在轨轨道确定提供方案和软件原型。仿真计算表明，基于28个全球分布的地面站，对于高度为800km的卫星，在忽略其动力学模型误差的假设下，若初轨三维位置、速度误差分别为100m（或差至1km）、1m／S（1d），2h后滤波可以达到稳定收敛，收敛后的实时定轨误差可以达到0．1m（1d）。滤波估计参数除了6个卫星轨道状态参数，还估计了地面信标相对于卫星超稳定振荡器的频率偏差；

英文摘要：

Earth observation satellites require highly precise orbit information to realize control of onboard equipment and realtime process of remote sensing data. Based on beacons distributed homogeneously on the earth surface and accurate measurements of the Doppler shifts, CNES has developed DORIS （Doppler Orbitography and Radio-posi tioning Integrated by Satellite） and it is the only technique after onboard GPS that can achieve realtime on-board or- bit determination and forecast with decimeter precision. Considering the demand of domestic high-revolution earth observation system, a project is carried out using extended Kalman filter to determine orbit in realtime through pro cessing of Doppler measurement. To investigate the stability and accuracy of the approach, a set of simulation ex periments have been performed to analyze various error sources, e.g. priori error, the distribution of the ground- based beacons and observation errors that would affect the convergence of the filter and the accuracy. On the basis of 28 beacons distributed homogeneously on the earth surface and neglecting the dynamic model error, the state of posi- tion and velocity as well as a frequency bias due to frequency offsets of the beacons~ USO are estimated. Simulation result indicates that the orbit determination in reahime for satellite at a height of 800kin can converge in 2 hours and reach 0. lm （la） in accuracy of position with priori error of 100m and lm/s in three directions of position and velocity.