总体最小二乘联合平差

发布时间：2018-03-13 17:38

  本文选题：联合平差　切入点：相对权比　出处：《武汉大学学报(信息科学版)》2016年12期 　论文类型：期刊论文

【摘要】：提出了将总体最小二乘方法应用于联合平差的模型,推导了附有相对权比的总体最小二乘联合平差方法。采用了多种方案来确定相对权比的大小。以参数估值与真值的差值范数作为评价指标,分析比较了单一数据总体最小二乘平差和两类数据总体最小二乘联合平差的模拟算例;通过给各类数据加入不同大小的随机噪声,分析了判别函数最小化法中随机噪声大小对确定相对权比的影响。模拟算例表明,平差结果的质量与相对权比的选取有关;当先验信息准确时,验前单位权方差法的结果最好,而当先验信息不准确时,判别函数为∑n_1i=1|V1_i|+∑n_2j=1|V2_j|及∑n_1i=1|V1_i|+∑n_2j=1|V2_j|/(1+X~TX)法均能取得有效的平差的判别函数最小化结果。
[Abstract]:In this paper, a model of combined adjustment using the total least square method is proposed. The total least squares combined adjustment method with relative weight ratio is derived. Several schemes are used to determine the magnitude of relative weight ratio. The difference norm between parameter estimation and true value is used as the evaluation index. This paper analyzes and compares the simulation examples of single data total least square adjustment and two kinds of data total least squares combined adjustment, by adding different sizes of random noise to all kinds of data, The influence of random noise size on determining relative weight ratio in minimization method of discriminant function is analyzed. The simulation example shows that the quality of adjustment results is related to the selection of relative weight ratio, and when the prior information is accurate, the result of prior unit weight variance method is the best. When the priori information is not accurate, the discriminant function is 鈭，

本文编号：1607426