基于伪逆-抗差联合估计模型的概率积分法预计参数求取  

Probability integral method of predicted parameters acquisition based on generalized inverse-robust joint estimation

作　　者：胡海斌 郭新慧[2] 

机构地区：[1]山西建筑职业技术学院,山西太原030006 [2]山西省遥感中心,山西太原030001

出　　处：《矿山测量》2017年第1期51-53,84,共4页Mine Surveying

摘　　要：为了克服病态法矩阵和异值点干扰而使结果出现发散等问题,文中提出了一种伪逆-抗差联合估计算法。结合工程实例将其与常用的抗差岭估计结果进行对比分析。结果表明:伪逆-抗差联合估计不仅有效抑制异值点和病态法矩阵的干扰,而且易于在matlab中编程实现,操作简便。To solve the divergent problems caused by the morbidity of normal matrix and the outliers in measurement data processing, this paper introduces a generalized inverse - robust joint estimation. The paper, based on practical projects, makes a comparative analysis on the result between this method and the often used generalized inverse - robust joint estimation method. Results show that the often used generalized inverse - robust joint estimation can deal with the heterogeneous value interference and the ill - conditioned matrix problems effectively and can be achieved easily by programming in matlab.

关 键 词：伪逆-抗差联合估计 概率积分法 预计参数 

分 类 号：TD325.2[矿业工程—矿井建设]