摘要
在实测数据较少的情况下,采用何种模型能对GPS Block IIR(M)卫星钟差进行最佳预报?经研究发现,采用GM(1,1)-AR(p)复合模型进行1天短期预报的精度在1 ns之内,进行10天长期预报的精度在10 ns之内,这不仅优于二次多项式和GM(1,1)等传统钟差预报模型,而且好于IGS(the International GPS Service for Geodynamics)提供的预报钟差7 ns的精度。
Which models can be used to predict optimum GPS Block ⅡR (M) satellite clock errors were researched in the case of few surveying data. The research shows that the accuracy for predicting in one day is a 1 ns and it in ten days is 10 ns with the GM( 1,1 )-AR(p) compound model. The prediction results with better than the results with quadratic polynomial model and GM ( 1,1 ) model and the accuracy is better than that from IGS ( the International GPS Service for Geodynamics) of the accuracy of 7 ns. Because the atomic clocks of COMPASS satellite navigational system are similar as these satellite atomic clocks, the accuracy can be assured using the GM ( 1,1 ) - AR(p) compound model in COMPASS satellite clocks error prediction.
出处
《大地测量与地球动力学》
CSCD
北大核心
2009年第4期111-115,共5页
Journal of Geodesy and Geodynamics