GPS高程拟合的方法比较  

Comparison of methods in GPS elevation fitting

作　　者：何美琳[1] 文鸿雁[1] 潘元进[1] 李超[1] 

机构地区：[1]桂林理工大学测绘地理信息学院/广西空间信息与测绘重点实验室,广西桂林541004

出　　处：《测绘科学》2013年第3期63-65,共3页Science of Surveying and Mapping

基　　金：国家自然科学基金资助项目(41071294);广西科学基金资助项目(桂科基0991023);广西区研究生科研创新项目(2011105960816M13)

摘　　要：GPS高程拟合的方法有很多,各有其优点和适用范围。本文在最小二乘估计算法的基础上,讨论主成分估计法和半参数模型估计法及各自的优点,并利用六参数模型的最小二乘估计、主成分估计和半参数模型三种拟合方法对GPS高程进行拟合,通过实例计算结果比较分析,得出半参数估计方法优于其他方法,在一定条件下精度高的优点。最后得出半参数估计方法精度达毫米级,论证了其在实际应用中的优越性。There are many fitting methods in GPS elevation fitting with their own advantages and application scope. This paper based on the algorithm of least-squares estimation discussed principal component estimation method, semi-parametric model estimation method and their respective advantages. Then GPS elevation fitting was made with three methods： least-square estimation method in six parameter model, principal component estimation and semi-parametric model estimation method. Comparing with and analyzing the cal- culation results, it was shown that the semi-parametric model estimation method would have a higher precision than others under certain conditions, and the semi-parametric model estimation method could reach the precision of millimeter level and demonstrate its superiori- ty in practical application.

关 键 词：最小二乘 主成分估计 半参数估计 GPS高程拟合 

分 类 号：P228.3[天文地球—大地测量学与测量工程]