摘要
本文介绍一种求平差参数的有偏估计方法——主成分估计。当法方程系数矩阵呈病态时,主成分估计的均方误差要小于最小二乘估计的均方误差,因此,采用主成分估计是有益的。
This paper discusses a biased estimation of adjustment parameters——princial components estimates. When (the matrix of the normal equation coefficient) is ill conditioned, the mean square error of principal components estimates is smaller than the larst square estimnte. Therefore, we suggest that principal estimates should be used.
出处
《武测科技》
北大核心
1989年第1期5-10,共6页
