测绘学报 ›› 2017, Vol. 46 ›› Issue (10): 1308-1315.doi: 10.11947/j.AGCS.2017.20170380

• 大地测量学与导航 • 上一篇    下一篇

动力学法的卫星重力反演算法特点与改进设想

沈云中   

  1. 同济大学测绘与地理信息学院, 上海 200092
  • 收稿日期:2017-07-04 修回日期:2017-08-11 出版日期:2017-10-20 发布日期:2017-10-26
  • 作者简介:沈云中(1962-),男,教授,研究方向为大地测量数据处理及其卫星重力和卫星定位应用。E-mail:yzshen@tongji.edu.cn
  • 基金资助:
    国家自然科学基金(41474017);中国科学院战略性先导科技专项(B类)(XDB23030100);高分遥感测绘应用示范系统资助项目

Algorithm Characteristics of Dynamic Approach-based Satellite Gravimetry and Its Improvement Proposals

SHEN Yunzhong   

  1. College of Surveying and Geo-informatics, Tongji University, Shanghai 200092, China
  • Received:2017-07-04 Revised:2017-08-11 Online:2017-10-20 Published:2017-10-26
  • Supported by:
    The National Natural Science Foundation of China(No. 41474017);The Strategic Priority Research Program of the Chinese Academy of Sciences(No. XD1323030100);The Program of Application and Demonstration System of High Resolution Remote Sensing in Surveying and Mapping.

摘要: 根据卫星轨道计算的积分公式,导出了以参考轨道为初值的线性化解算地球重力场的观测方程,给出了其系数矩阵的积分计算公式,阐明了动力学法本质上是观测值相对于参考轨道的线性摄动方法,因此其变分方程力模型参数的偏导数初值必定为0。在此公式的基础上,分析了动力学法观测方程的主要特点,即线性化误差随轨道弧段增长而快速增大,其观测方程的性质也随弧段增长而变差,且积分计算误差将是下一代重力卫星数据处理的重要瓶颈问题。提出了进一步提高动力学法重力反演精度的方法,主要归结为:改进以几何轨道为初值的线性化方法以减小线性化误差,改变参数化方式以改善观测方程的性质,综合应用解析公式与数值积分公式以提高轨道计算精度。

关键词: 卫星重力反演, 动力学法, 轨道积分, 解析轨道, 线性化方法

Abstract: By using the integration equation for computing satellite orbit, this paper derives the observational equation for computing gravity field model linearized with respect to the reference orbit, provides the integration equation of calculating the design matrices of the observational equation, and clarifies that the dynamic approach is in principle the perturbation method relative to the reference orbit, therefore the initial partial derivatives with respect to the force model parameters must be zero. Based on the derived formulae this paper analyzes the main characteristics of dynamic approach-based observational equation, i.e. the linearization error will rapidly increase and the property of observational equation becomes worse as the integration arc extends longer, and the numerical integration error will be the bottle-neck problem for the data processing of next generation of satellite gravity exploration. Then this paper proposes the methods for improving the accuracy of gravity recovery, which can be summarized as that refining the linearization method relative kinematic orbit to reduce the linearization error, modifying parameterization method to improve the property of observational equation, and combined using analytic formula and numerical integration formula to increase the accuracy of orbit computation.

Key words: satellite gravimetry, dynamic approach, orbit integration, analytic orbit, linearization method

中图分类号: