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．

Investigation and Precision Analysis on Space-borne GPS KPOD with LP Estimation

To get precise space-borne GPS kinematic orbit, carrier phase observations must be used rather than pseudo-range ones, it is bothersome that we can not detect cycle-slip completely, and there are still small nuisance cycle-slips left after GPS data preprocessing. It is well known to us that Least Square (LS) is an optimal method to estimate parameters if the residual distribution is Gauss normally distributed, but it is not the optimal one to process data with systematic error. LP estimation is an effective method to process data with systematic error. Based on the validation condition of LS and LP, and the property of space-borne GPS data, LP method is used to process space-board GPS KPOD. Taking CHAMP data from 16th, July, 2001 for example, the precision using LP estimation is investigated in this paper. It is showed that LP estimation is superior to that of LS in processing GPS carrier phase data with remnant cycle-slips, and the precision of space-borne GPS KPOD can be improved significantly. But, contrast to LS estimation, the superiority of LP estimation is weaken with the cleaning of remnant cycle-slips, furthermore, the orbit precision estimated by LS estimation will be the optimal estimation when the data residual distribution obeys Gauss normally distributed and without systematic error, that is to say, LP is not the optimal estimation when residual distribution obeys normal distribution, the precision will be poor than that of LS.