摘要
本文根据具有先验信息的大地监测网的特点,在估值均方差分析的基础上,提出最小二乘最小加权偏差范数估值和最小二乘最小均方差迹估值,并证明了它们与贝叶斯估值在参数先验方差因子σ_x^2→+∞的极限状态下的一致性。作者认为:这两种方法都利用先验信息补充了平差基准,同时克服了先验统计模型中所含缺陷的不良影响。
Combining apriori information of geodetic monitoring networks and based on the analysis of the Mean Square Error of estimations, the Least Square Minimum Weighted Deviation Norm Estimation and the Least Square Minimum Mean Square Error Trace Estimation are advanced, and proof of their consistence with Bayes estimation in the extreme state when parameters' a priori variance factor σ_x^2→+∞ is given.The author thinks that both estimations utilize the a priori information to supplement the datum of the network adjustment, and, at the same time, eliminate the undesired effects of defects in the a priori statistic model.
出处
《地壳形变与地震》
CSCD
1989年第1期28-37,共10页
Crustal Deformation and Earthquake
