摘要
在卫星精密定轨中 ,对资料的加权一般是采用经验的方法 .本文采用赫尔默特(Helmert)方差分量估计方法 ,对Lageos2卫星 6个月的SLR资料进行试算 ,统计了各台站的验后方差 .结果表明 ,用验后方差估计的方法可以得到各台站大致合理的权重 。
The determination of the precise orbits of artificial satellites by taking advantage of several types of satellite tracking techniques has become a tendency today. Especially for multidisciplinary satellites, in order to guarantee their success, several tracking equipments are usually collocated. In the precision orbit determination (POD) by using either single or multiple kinds of tracking techniques, there always exists a problem of how to weight the observations. For using single kind of tracking techniques, observation accuracy would be different not only among each station, but also for the same station due to temporal variations. For using multi technique to determine satellite orbits, there might exist technical differences. In POD for an artificial satellite with dynamic statistic method, the data weights are defined by W=σ 2 0R -1 with variance covariance matrix R and the unit variance σ 2 0, which can be an arbitrary positive number. Generally, R is unknown and so W cannot be determined properly. In practical computations, there are usually two approaches to weight the observations. One utilizes directly the standard deviation σ usually supplied in the data file and another comes from experience. Because standard deviation in data file denotes only the internal consistency but not the real observation precision, many institutes use only the latter or both of them. Neither the data file supplied σ nor the empirical method are good in determining the proper data weights due to the variational situations. This paper uses the Helmert method of variance component estimation (VCE) in order to obtain a set of proper data weights. The VCE method can be used to get the a posteriori variance components of the observations for data weighting. According to this method, observations can be divided into several groups in terms of instruments, error sources, accuracies, and so on. Through VCE, the variance factors for each group can be estimated after POD and the a priori data weights can be adjusted for the next POD process. After several POD VCE iterations untill VCE converges, a set of optimal data weights could be obtained. In order to inspect the effect of the VCE method on the satellite POD, an example of Lageos2 satellite is analyzed. Six arcs each with 31 days length are computed according to the POD VCE iteration process. Two or three times of iteration are needed before the VCE converges. The observations are classified in terms of SLR stations. After the VCE convergence, the adjusted standard deviations σ c for each station are obtained. The results indicate that the σ c′s basically represent the real accuracies of each SLR station. Moreover, by using the VCE method, the root mean square (RMS) of observation residuals diminishes in each trajectory arc. Another valuable result is that the VCE method can be used to automatically calibrate the improper data weighting and used as an auxiliary approach to find unreasonable selection of solved for parameters.
出处
《天文学报》
CSCD
北大核心
2000年第2期129-138,共10页
Acta Astronomica Sinica
基金
国家攀登项目!( 970 2 3 1 0 0 3 )
国家自然科学基金!重点项目 ( 1 983 3 0 3 0 )
中科院"九五"基础性研究重大项目!(KJ951 -1- 1
