稳健估计下的坐标系统转换总体最小二乘算法  

Total least squares algorithm for coordinate system conversion

作　　者：杨娟[1,2] 陶叶青[1,2] 

机构地区：[1]宿州学院地球科学与工程学院,安徽宿州234000 [2]中国矿业大学环境与测绘学院,江苏徐州221008

出　　处：《测绘科学》2015年第4期15-18,共4页Science of Surveying and Mapping

基　　金：安徽省优秀人才基金项目(2013SQRL085ZD);安徽省煤矿勘探工程技术研究中心平台项目(2014YKF04;2014YKF08)

摘　　要：针对传统的应用最小二乘法建立高斯-马尔科夫(G-M)模型实现坐标系统转换的方法导致转换模型参数精度低下的问题,该文提出一种基于总体最小二乘算法的坐标系统转换方法。考虑到粗差会导致控制点坐标精度差异较大,因此根据稳健估计理论进行迭代定权,在总体最小二乘算法下建立G-M模型,以便求解转换模型参数,并通过算例比较不同算法的转换精度。实验结果表明:基于稳健估计的总体最小二乘抗差算法实现的空间坐标转换精度高于传统方法的转换精度。The traditional method of applying least squares（LS）algorithm to establish G-M model for coordinate system conversion has an influence on the precision of transformation model parameters.In order to solve this problem,a coordinate system conversion method using total least squares（TLS）algorithm was proposed.Considering control points accuracy is different because of gross error,through iteration and fixed weight method based on robust estimation,conversion model parameters were calculated by G-M model using TLS algorithm.Experimental results showed that the conversion precision of establishing G-M model based on TLS was higher than that of applying LS algorithm.

关 键 词：稳健估计 总体最小二乘 最小二乘 高斯-马尔科夫模型 坐标系统转换 精度 

分 类 号：P221[天文地球—大地测量学与测量工程]