Methods for Precise Determination of Geopotential from Satellite Missions

该研究项目的主要新成果可简要概括如下：（i）作为利用GPS卫星确定重力场的第一步，我们必须利用载波相位观测值精确确定低卫星的位置。为了准确、快速地解算GPS整周模糊度未知量，提出了一种新的GPS整周解相关方法，并通过随机模拟的方法对不同的相关方法进行了科学的比较。结果表明，新的方法是目前最好的，并表明高斯整数解相关是不是最好的;（ii）我们已经研究了非线性滤波器的精密定轨以及。新的结果包括提出一个无偏二阶预测器和一个截断无偏二阶滤波器。在非线性连续系统的情况下，导出了滤波器及其协方差矩阵的耦合微分方程。新 ...更多信息 结果表明，连续非线性系统40年的非线性滤波可能存在根本性的缺陷：（3）通过模拟，我们测试了在相同的姿态下，以不同的速率对多颗卫星进行采样，以提高重力场的空间和时间分辨率。我们还研究了ECMWF地面气压数据的采样率和混叠对时变重力场的影响。结果验证了仿真的有效性;（iv）由于从空间观测确定重力场是非线性的，并且记住模拟退火和遗传算法-两种最广泛使用的全局优化方法实际上通常不能正确地产生全局最优解，本文重点研究了新的非随机算法，提出了一种新的非线性非凸地质反演的混合全局优化方法。通过实例验证，理论结果得到了验证，并证实了其实用性.少

The major new results of this research project can be briefly summarized as follows: (i) As a first step to use GPS satellites in the determination of the gravitational field, we have to precisely determine the positions of a low satellite by using carrier phase observables. In order to correctly and quickly resolve the GPS integer ambiguity unknowns, we have proposed a new GPS integer decorrelation method and a new random simulation approach to scientifically compare different methods of correlation. The results have indicated that the new method is currently the best and showed that Gaussian integer decorrelation is not the best; (ii) We have investigated the nonlinear filter for precise orbit determination as well. The new results include proposing an unbiased predictor of second order and a truncated and unbiased filter of second order. We have also derived the coupled differential equations of the filter and its covariance matrix in the case of nonlinear continuous system. The new … More results imply that forty years of nonlinear filtering of continuous nonlinear systems may have a fundamental defect in the foundation; (iii) By means of simulation, we have tested the improvement of spatial and temporal resolution in gravity field by sampling more than one satellite at the same attitude but at different rates. We have also investigated the effect of the sampling rates of ECMWF surface pressure data and aliasing on the time varying gravity field. The results have confirmed the effectiveness of simulation; and (iv) Since the determination of the gravitational field from space observations is nonlinear, and keeping in mind that simulated annealing and genetic algorithms --- two most widely used global optimization methods are actually not capable of correctly producing the global optimal solution in general, we have focused on new non-random algorithms and proposed a new hybrid global optimization method for nonlinear nonconvex geo-inversion. The theoretical results have been established and practical applicability confirmed through the demonstration of examples. Less