Homogeneous and Heterogeneous Observational Data Weighting Method of Combined Orbit Determination of Near-earth Satellites of Bi-satellite Positioning System and its Applications

In the light of the problem of amalgamation and processing of multisource observational data in the combined orbit determination of near-earth satellites of the bi-satellite positioning system, the optimal weighting method of the improved variance component estimation of the two-step systematic error correction of homogeneous observational data is proposed. Analyses show that the multi-source amalgamation measurement model of the heterogeneous observational data essentially is a multi-structure, multi-parameter non-linear regression model, and the optimal weighting method of the combination of model structure characteristic analysis and variance component estimation of the heterogeneous observational data is established. The realization algorithms of the optimal weighting and the combined orbit determination parameter estimation of the two sorts of observational data are designed, and the simulation experiments of the combined orbit determination are carried out by taking the distances among the two satellites and the backup satellite and the homogeneous observational data and the distance between the two satellites and the heterogeneous observational data of satellite sensor angle measurements as the examples. The results of theoretical analysis and simulation calculation show that for the combined orbit determination of homogeneous observational data, the accuracy of orbit determination obtained by adopting the variance component estimation method of the two-step systematic error correction can be more superior than that obtained by means of the traditional empirical weighting method. For the combined orbit determination of heterogeneous observational data, through the introduction of the weighting factor by which the model structure is characterized the accuracies of the combined orbit determination of the near-earth satellite and geostationary satellite are both improved to a certain extent in comparison with the mean weighting mode.     

The Bootstrap Estimation of the Accuracy of Preliminary Orbit Determination

From the point of view of the non-parametric statistics, a general estimation method of the accuracy and con?dence interval of preliminary orbit determination is proposed for the occasion without any other information but observational data. Based on the bootstrap method, the estimation relies only on the observational data and does not require the precise orbit determination as a reference, or the assumption of normal distribution of observational errors. Numerical experiments show that this method is very simple in implementa- tion, and may serve as an easy accuracy evaluation for the preliminary orbit determination and for the follow-up employments.     

A Robust Method of Preliminary Orbit Determination

The preliminary orbit determination with optical angular measure- ments plays an important role in the survey of space objects. The classical method of orbit computation based on the least square error estimation is not robust while outliers occur in the observation. A robust method is proposed by employing the least absolute deviation estimation. The method reduces the problem of orbit determination to a linear programming problem, and gives the variance of the estimation with the bootstrap method. Numerical check shows that the method is effective and robust, and has a high breakdown point.     

The Data Depth-weight-kernel Estimation of the Model Error of Satellite Orbit Determination

It is an objective fact that there exists error in the satellite dynamic model and it will be transferred to satellite orbit determination algorithm, forming a part of the connotative model error. Mixed with the systematic error and random error of the measurements, they form the unitive model error and badly restrict the precision of the orbit determination. We deduce in detail the equations of orbit improvement for a system with dynamic model error, construct the parametric model for the explicit part of the model and nonparametric model for the error that can not be explicitly described. We also construct the partially linear orbit determination model, estimate and fit the model error using a two-stage estimation and a kernel function estimation, and finally make the corresponding compensation in the orbit determination. Beginning from the data depth theory, a data depth weight kernel estimator for model error is proposed for the sake of promoting the steadiness of model error estimation. Simulation experiments of SBSS are performed. The results show clearly that the model error is one of the most important effects that will influence the precision of the orbit determination. The kernel function method can effectively estimate the model error, with the window width as a major restrict parameter. A data depth-weight-kernel estimation, however, can improve largely the robustness of the kernel function and therefore improve the precision of orbit determination.     

Simulation of precise orbit determination of lunar orbiters

Based on the ongoing Chinese lunar exploration mission, i.e. the “Chang'e 1” project, precise orbit determination of lunar orbiters is analyzed for the actual geographical distribution and observational accuracy of the Chinese united S-band (USB) observation and control network as well as the very long baseline interferometry (VLBI) tracking network. The observed data are first simulated, then solutions are found after including the effects of various error sources and finally compared. We use the space data analysis software package, GEODYN, developed at Goddard Space Flight Center, NASA, USA. The primary error source of the flight orbiting the moon is the lunar gravity field. Therefore, the (formal) error of JGL165P1, i.e. the model of the lunar gravity field with the highest accuracy at present, is first discussed. After simulating the data of ranging and velocity measurement as well as the VLBI data of the time delay and time delay rate, precise orbit determination is carried out when the error of the lunar gravity field is added in. When the orbit is determined, the method of reduced dynamics is adopted with the selection of appropriate empirical acceleration parameters to absorb the effect of errors in the lunar gravity field on the orbit determination. The results show that for lunar missions like the “Chang'e 1” project, that do not take the lunar gravity field as their main scientific objective, the method of reduced dynamics is a simple and effective means of improving the accuracy of the orbit determination of the lunar orbiters.     

自适应Lp估计及其在照相天体测量中的应用

Ｐ范分布是一个广义的测量误差分布族．众所周知的正态分布只是其ｐ＝２时的一个特例．因此，与正态分布和最小二乘估计相比，Ｐ范分布和相应的Ｌｐ估计（最小ｐ范估计）可描述和处理类型更广泛的测量误差．针对具体的测量数据，推断其实际误差之Ｐ范分布的Ｐ值，然后依此进行最或然估计，此即自适应Ｌｐ估计．本文介绍了Ｌｐ估计的性质．较为详细地讨了Ｌｐ估计的效率和自适应Ｐ值的具体求解方法；并且还针对鬼星团照相天体测量数据的归算，给出了一个实际应用的例子．结果表明，所处理的观测误差并不服从正态分布；此时自适应Ｌｐ估计的引入将带来显著优于最小二乘估计的结果．     

LP估计在星载GPS运动学定轨中的应用及精度分析

为得到高精度的星载GPS运动学定轨,必须利用观测精度高的相位观测值,但是相位观测值预处理后,仍然存在残余小周跳．在残差服从正态分布情况下LS法是最佳参数解算方法,但该方法不能解决资料的系统误差消除问题,LP估计是处理资料残差分布含有系统误差的有效方法之一．基于LS、LP方法的有效条件和GPS数据预处理的特性,将LP估计方法引入星载GPS运动学定轨数据处理中,以CHAMP卫星资料为例,研究了LP估计在星载GPS运动学定轨中的应用及其精度分析．实践表明:在处理含有残余小周跳的相位观测值时,LP估计比LS更有效,提高了星载GPS运动学定轨精度,但随着残余周跳的进一步修复,LP估计相对于LS估计的优越性越来越弱,在资料完全没有系统误差,残差服从正态分布的情况下,LP估计不能很好地体现其优越性,精度反而低于LS估计．     

抗差估计在星载GPS卫星非差运动学定轨中的应用

针对星载GPS卫星非差运动学定轨的特点,提出将抗差估计应用于星载GPS卫星运动学定轨,一方面尽量保留来自每颗GPS卫星的观测值,保持较强的卫星几何强度,避免轨道求解出现奇异;另一方面,采用等价权思想,有效地降低质量较差的观测值对定轨结果的影响,保证定轨精度．还采用CHAMP卫星的实测GPS数据验证了新方案的可行性和有效性．     

A Research on the Orbit Determination Method by Means of Sparse Data of Electronic Fences

Based on the Lambert equation and knowledge of space geometry a method of orbit determination is given using the sparse observational data provided by the space monitoring electronic fence device. Our simulated experiment of a large number of targets shows that the initial orbit determined by this method can be improved and can converge to a final accuracy better than 100 m, so proving that the method can be applied to the orbit determination of an overwhelming majority of space targets with the observed data of the electronic fence. Finally, the effect of the latitude of the observing station on the application of the method is discussed.     

Determination of the parabolic orbit for a body moving in the plane of the ecliptic,by the Laplace method

A modification of the Laplace method for the determination of a preliminary orbit of a body moving in the plane of the ecliptic is presented. It is shown that on the basis of three close positions on the celestial sphere, the parabolic orbit can be determined in cases, when the classical Laplace method is not suitable. The numerical estimation of the number of possible solutions is carried out, and their spatial distribution depending on initial conditions is shown. An example of the determination of an orbit for the comet C/2007 N3 (Lulin) is given.     

初轨计算中角资料观测误差传播的一些讨论

用非线性规划方法处理初轨计算问题,将观测误差看成参量的微小扰动量。在无摄情况下,建立了非线性规划中的灵敏度分析模型（ｆｈｐ）。由此推导,在无摄情况下关于角资料观测误差对角资料观测真值的影响的误差传播分析公式（一阶近似）。     

A new method for the determination of an intermediate orbit using three positions of a small body on the celestial sphere

We propose a new method for the determination of the preliminary orbit of a small celestial body using three pairs of its angular coordinates in three moments of time. The method is based on the use of the intermediate orbit we constructed earlier using three position vectors and the corresponding time moments. This intermediate orbit accounts for the main part of the perturbations of the motion of the body under study. We compare the results obtained by the classical Lagrange-Gauss method, Herrick-Gibbs method, generalized Herrick-Gibbs method, and the new method by the examples of the determination of the orbit of the small planet 1566 Icarus. The comparison showed that the new method is a highly efficient tool for the study of perturbed motion. It is especially efficient when applied to high-precision observational data covering short arcs of the orbit.     

The Effect of GPS Ephemeris on the Accuracy of Precise Orbit Determination of LEO Satellite-borne GPS

The satellite-borne GPS receivers dedicated to precise orbit determination are now being carried by more and more low earth orbit (LEO) satellites and the satellite-borne GPS has become one of the main means for the precise orbit determination of low earth orbit satellites. The accuracy of satellite-borne GPS precise orbit determination depends on the accuracies of the GPS ephemeris and the clock error. Based on the orbit determination function of SHORDEIII zero-difference dynamics and using the observational data obtained by the GRACE satellites for the week from 2005 August 1 to 7 as an example, three versions of GPS ephemerides (igs, igr and igu) are used to carry out orbit determination under the same conditions and to estimate the effect of the GPS ephemeris accuracy on the accuracy of orbit determination of low earth orbit satellites. Our calculated results show that the two ephemerides, igs and igr, are equivalent to each other in orbit determination accuracy (about 9.5 cm), while igu is slightly less accurate, at about 10.5 cm. The effect produced by the data of the high frequency GPS satellite clock error on the accuracy of orbit determination is 1–6 cm.     

环月飞行器精密定轨的模拟仿真

以中国正在实施的探月计划“嫦娥1号”工程为背景，分析了在中国联合S波段(USB)测控网和甚长基线射电干涉(VLBI)跟踪网的现有空间分布、观测精度水平下的环月飞行器精密定轨．采用的方法是模拟仿真计算，即首先模拟观测数据，然后在计入各误差源的影响后进行求解，并对解算结果进行比较．模拟仿真的工具是美国宇航局哥达德飞行中心的空间数据分析软件系统GEODYN．环月飞行的主要误差源是月球重力场，为此首先讨论了目前精度最高的月球重力场模型JGL165P1的(形式)误差．在模拟了测距、测速以及VLBI的时延、时延率数据后，计入月球重力场的误差进行精密轨道确定．定轨时采用了减缩动力学(reduced dynamic)方法，即选用合适的经验加速度参数吸收重力场误差对定轨的影响．结果表明对于一个不将月球重力场作为主要科学目标的探月计划(如“嫦娥1号”)，减缩动力学方法是一个简单、有效地提高环月飞行器定轨精度的方法．     

测角观测误差在初轨确定中的传布

本文用非线性规划方法探讨了初轨确定问题。测角观测误差看作为参数的微摄动,而非线性规划的灵敏性分析模型（ｆｈｐ） 建立在无摄动的情况下。由此,得出了在一阶近似下轨道根数计算值和测角观测误差值间关系的分析表达式。     

Evaluation of LEO based GPS slant TEC using a simulation technique

With the increased number of low Earth orbit (LEO) satellites equipped with Global Positioning System (GPS) receiver, the LEO based GPS slant total electron content (STEC) data play a more important role in ionospheric research due to better global coverage. The accuracy of LEO TEC is hardly evaluated by comparison with the independent TEC measurement simultaneously. We propose an approach based on the simulated data to verify the accuracy of TEC determination. The simulated data (i.e., the pseudorange and carrier phase observations) was generated based on the consideration of the effect of the ionosphere, the so-called differential code bias (DCB) and observational noise. The errors of carrier phase to code leveling process and DCB estimation are analyzed quantitatively. Also, the effect of observational noise, solar activity and LEO orbit altitude on the accuracy of TEC determination will be discussed in detail. The accuracy of TEC determination is relative to solar activity and LEO orbit altitude, the higher LEO orbit and lower F10.7 index, the higher accuracy of TEC determination. It is found by the first time that, with the amplification of the pseudorange noise, the accuracy of leveling process and TEC determination declines almost linearly. With the LEO missions in the near future, it is hoped that the GPS satellite DCBs estimated based on LEO observations would be better than those based on ground-based observations.     

已知e的先验值的初轨寻优算法

本文讨论由短弧测角资料,已知偏心率e的先验值时,初轨的寻优算法。把偏心率的先验值看作是对e的约束条件,再固定观测弧段两端点的观测量与作为约束条件,本文导出了一维(对M_0)非线性寻优算法;针对观测资料中可能存在异常值的情形,本文提出了一种初轨的稳健寻优算法。仿真结果表明,当e的先验值有足够高的精度时,能明显提高周期P的估计精度;稳健算法还可用作剔除观测资料的异常值。     

望远镜静态指向模型的基本参数

对地平式望远镜静态指出模型的基本参数进行了讨论，给出了一种参数较少，模型稳定的望远镜修正模型，与常用的球谐函修正模型相比较，基本参数模型具有对观测资料分布不敏感，模型的系随资料分布变化较小和修正精度较高优点，并且其残差序列服从正态分布，基本上能够满足高精度定轨的需要。     

A robust estimation method in orbit improvement

Orbit improvement is studied via processing simulated optical data of 20 satellites. Using our criterion the fraction of outliers falls to less than 10% in most cases. Our method, which ensures the accuracy of orbit improvement, consists in using the Huber estimator twice, followed by iterating the Hampel estimator until convergence. The parameters C₀ and C₁ in the Hampel estimation may be given the fixed values of 2.2 and 3.6, while the value of C₂ varies in a definite way with the size of the outlier considered.