Three-Dimensional Coordinate Transformation Model and Its Robust Estimation Method Under Gauss-Helmert Model
-
摘要: 对三维坐标转换的高斯-赫尔默特(Gauss-Helmert,GH)模型,采用牛顿-高斯(Newton-Gauss)迭代算法构建了该模型的拉格朗日目标函数,推导了其解算方法,并给出了具体的计算步骤。在此基础上,考虑到可能出现的粗差对观测空间与结构空间的综合影响,基于标准化残差构造权因子函数,推导了该模型的抗差解法。仿真实验结果表明,GH模型用于三维坐标转换时不受旋转角度大小和其他附加条件限制,解算结果与现有算法一致,且估计参数的维数大大降低,计算效率有一定程度的提高;所提出的抗差解法效果良好,与现有基于整体最小二乘的三维坐标转换的抗差解法相比,表现出了更好的稳健性。Abstract: For three-dimensional coordinate transformation, it's impossible using the Gauss-Markov model to obtain optimal parameter estimation from the functional model with error in its coefficient matrix. On the other hand, errors-in-variables model has difficulty expressing the functional model, and partial errors-in-variables model is complex as well as too much parameters to be estimated for the quasi-observation method. Therefore, Gauss-Helmert model is employed for three-dimensional coor-dinate transformation. The target function of the proposed model is established based on Newton-Gauss iterative algorithm, and the estimated method and its derivation procedure also are presented in this paper. Beyond the above process, we proposed a new robust estimation method for the proposed model, which is based on the normalized residual error and takes the influence of gross error on both observation and structure spaces into consideration. Meanwhile, derivational process of statistical tests and iterative algorithm are presented. The simulation experiment results show that the proposed estimation method has the same accuracy as the traditional method, which has robust with angular dimension and other additional conditions, but less estimation parameters. In addition, the new robust estimation method has effective robustness when comparing with the other existing robust total least square methods for the coordinate transformation.
-
-
表 1 存在1个粗差时各方案的均方根误差对比
Table 1 Root Mean Square Errors of Estimated Parameters for Different Schemes with One Outlier
方案 X0 /m Y0 /m Z0 /m μ/(10-5 rad) α1 /(10-5 rad) α2 /(10-5 rad) α3 /(10-5 rad) 方案1 19.302 1 10.579 6 15.580 1 2.600 1.510 1.160 1.750 方案2 3.922 1 2.666 3 3.613 2 0.632 0.405 0.351 0.366 方案3 3.92 21 2.666 3 3.613 2 0.632 0.405 0.351 0.366 方案4 6.470 3 3.981 9 5.938 8 1.010 0.644 0.557 0.570 方案5 4.502 9 3.163 6 4.381 6 0.696 0.453 0.410 0.452 方案6 4.194 9 2.907 2 4.106 6 0.666 0.435 0.382 0.407 表 2 存在3个粗差时各方案的均方根误差对比
Table 2 Root Mean Square Errors of Estimated Parameters for Different Schemes with Three Outliers
方案 X0 /m Y0 /m Z0 /m μ/(10-5 rad) α1 /(10-5 rad) α2 /(10-5 rad) α3 /(10-5 rad) 方案1 10.307 9 12.340 7 18.474 0 1.460 1.390 1.730 1.130 方案2 3.752 9 2.956 4 2.844 9 0.507 0.289 0.311 0.417 方案3 3.752 9 2.956 4 2.844 9 0.507 0.289 0.311 0.417 方案4 8.797 8 6.880 5 6.709 9 1.230 0.664 0.770 0.952 方案5 7.068 0 5.744 2 5.810 6 0.978 0.592 0.628 0.798 方案6 4.695 8 3.896 1 3.931 2 0.671 0.404 0.413 0.518 表 3 存在5个粗差时各方案的均方根误差对比
Table 3 Root Mean Square Errors of Estimated Parameters for Different Schemes with Five Outliers
方案 X0 /m Y0 /m Z0 /m μ/(10-5 rad) α1 /(10-5 rad) α2 /(10-5 rad) α3 /(10-5 rad) 方案1 12.748 6 11.219 6 15.764 9 2.250 1.790 1.150 1.400 方案2 3.414 0 3.380 1 4.576 3 0.610 0.469 0.380 0.424 方案3 3.414 0 3.380 1 4.576 3 0.610 0.469 0.380 0.424 方案4 10.173 6 10.600 6 14.184 5 1.960 1.480 1.120 1.320 方案5 8.900 8 8.467 6 11.919 8 1.580 1.290 0.956 1.110 方案6 6.064 6 6.039 3 7.999 3 1.090 0.818 0.639 0.757 表 4 方案5和方案6计算检查点的均方根误差(1个粗差)/m
Table 4 Root Mean Square Errors of Check Points for Schemes 5 and 6 with One Outlier/m
方向 方案 点位1 点位2 点位3 点位4 点位5 点位6 点位7 X 方案5 0.079 2 0.081 3 0.072 9 0.077 6 0.081 5 0.080 3 0.080 3 方案6 0.078 4 0.080 3 0.072 4 0.077 1 0.080 8 0.079 0 0.078 7 Y 方案5 0.096 2 0.040 7 0.076 9 0.039 8 0.065 6 0.068 2 0.041 0 方案6 0.095 9 0.039 1 0.077 0 0.038 6 0.065 3 0.068 2 0.040 1 Z 方案5 0.084 8 0.052 2 0.067 5 0.059 8 0.077 6 0.082 8 0.057 3 方案6 0.084 7 0.052 2 0.066 8 0.058 3 0.076 4 0.081 9 0.056 1 表 5 方案5和方案6计算检查点的均方根误差(3个粗差)/m
Table 5 Root Mean Square Errors of Check Points for Schemes 5 and 6 with Three Outliers/m
方向 方案 点位1 点位2 点位3 点位4 点位5 点位6 点位7 X 方案5 0.078 5 0.081 1 0.073 9 0.082 2 0.082 7 0.080 3 0.082 9 方案6 0.077 8 0.078 5 0.071 7 0.078 2 0.080 4 0.078 5 0.078 0 Y 方案5 0.098 7 0.048 1 0.082 7 0.044 6 0.065 1 0.073 0 0.044 9 方案6 0.094 6 0.041 9 0.078 4 0.038 7 0.063 9 0.070 7 0.039 2 Z 方案5 0.089 3 0.056 9 0.068 3 0.061 8 0.076 8 0.083 4 0.059 3 方案6 0.087 1 0.052 9 0.066 2 0.057 6 0.076 2 0.082 1 0.057 2 表 6 方案5和方案6计算检查点的均方根误差(5个粗差)/m
Table 6 Root Mean Square Errors of Check Points for Schemes 5 and 6 with Five Outliers/m
方向 方案 点位1 点位2 点位3 点位4 点位5 点位6 点位7 X 方案5 0.084 4 0.09 22 0.080 6 0.089 3 0.087 3 0.090 7 0.097 6 方案6 0.079 7 0.083 5 0.075 0 0.082 1 0.081 3 0.081 6 0.085 5 Y 方案5 0.104 0 0.057 7 0.087 7 0.052 2 0.072 8 0.079 4 0.057 6 方案6 0.097 9 0.047 5 0.083 1 0.041 9 0.069 5 0.072 1 0.046 8 Z 方案5 0.099 5 0.066 1 0.078 1 0.073 4 0.085 4 0.087 9 0.068 1 方案6 0.091 9 0.060 1 0.070 8 0.062 1 0.078 9 0.083 5 0.060 0 -
[1] Grafarend E W. Nonlinear Analysis of the Three-Dimensional Datum Transformation[J].Journal of Geodesy, 2003, 77(1):66-76 doi: 10.1007-s00190-002-0299-9/
[2] 姚宜斌, 黄承猛, 李程春, 等.一种适用于大角度三维坐标转换参数求解算法[J].武汉大学学报·信息科学版, 2012, 37(3):253-256 http://ch.whu.edu.cn/CN/Y2012/V37/I3/253 Yao Yibin, Huang Chengmeng, Li Chengchun, et al.A New Algorithm for Solution of Transformation Parameters of Big Rotation Angle's 3D Coordinate[J].Geomatics and Information Science of Wuhan University, 2012, 37(3):253-256 http://ch.whu.edu.cn/CN/Y2012/V37/I3/253
[3] 杨元喜, 徐天河.不同坐标系综合变换法[J].武汉大学学报·信息科学版, 2001, 26(6):509-513 http://ch.whu.edu.cn/CN/abstract/abstract5223.shtml Yang Yuanxi, Xu Tianhe.The Combined Method of Datum Transformation Between Different Coordinate Systems[J].Geomatics and Information Science of Wuhan University, 2001, 26(6):509-513 http://ch.whu.edu.cn/CN/abstract/abstract5223.shtml
[4] 刘大杰, 施一民, 过静珺.全球定位系统(GPS)的原理与数据处理[M].上海:同济大学出版社, 1996 Liu Dajie, Shi Yimin, Guo Jingjun.The Theory and Data Processing of Global Positioning System (GPS)[M].Shanghai:Tongji University Press, 1996
[5] 曾文宪, 陶本藻.三维坐标转换的非线性模型[J].武汉大学学报·信息科学版, 2003, 28(5):566-568 http://ch.whu.edu.cn/CN/abstract/abstract4861.shtml Zeng Wenxian, Tao Benzao.Non-linear Adjustment Model of Three-Dimensional Coordinate Transformation[J].Geomatics and Information Science of Wuhan University, 2003, 28(5):566-568 http://ch.whu.edu.cn/CN/abstract/abstract4861.shtml
[6] 陆珏, 陈义, 郑波.总体最小二乘方法在三维坐标转换中的应用[J].大地测量学与地球动力学, 2008, 28(5):77-81 http://d.old.wanfangdata.com.cn/Periodical/dkxbydz200805016 Lu Jue, Chen Yi, Zheng Bo. Total Least Squares to Three-Dimensional Datum Transformation[J].Journal of Geodesy and Geodynamics, 2008, 28(5):77-81 http://d.old.wanfangdata.com.cn/Periodical/dkxbydz200805016
[7] Golub H G, Vanloan F C.An Analysis of the Total Least Squares Problem[J].SIAM Journal on Numerical Analysis, 1980, 17(6):883-893 doi: 10.1137/0717073
[8] Schaffrin B, Felus Y A.On the Multivariate Total Least-Squares Approach to Empirical Coordinate Transformations[J].Journal of Geodesy, 2008, 82(6):373-383 doi: 10.1007/s00190-007-0186-5
[9] Felus F, Burtch R.On Symmetrical Three-Dimensional Datum Conversion[J]. GPS Solutions, 2009, 13(1):65-74 doi: 10.1007/s10291-008-0100-5
[10] Shen Yunzhong, Li Bofeng, Chen Yi. An Iterative Solution of Weighted Total Least-Squares Adjustment[J]. Journal of Geodesy, 2011, 85(4):229-238 doi: 10.1007/s00190-010-0431-1
[11] Xu Peiliang, Liu Jingnan, Shi Chuang.Total Least Squares Adjustment in Partial Errors-in-Variables Models:Algorithm and Statistical Analysis[J].Journal of Geodesy, 2012, 86(8):661-675 doi: 10.1007/s00190-012-0552-9
[12] Xu Peiliang, Liu Jingnan.Variance Components in Errors-in-Variables Models:Estimability, Stability and Bias Analysis[J].Journal of Geodesy, 2014, 88(8):719-734 doi: 10.1007/s00190-014-0717-9
[13] 魏二虎, 殷志祥, 李广文, 等.虚拟观测值法在三维坐标转换中的应用研究[J].武汉大学学报·信息科学版, 2014, 39(2):152-156 http://ch.whu.edu.cn/CN/abstract/abstract2918.shtml Wei Erhu, Yin Zhixiang, Li Guangwen, et al.On 3D Coordinate Transformation with Virtual Observation Method[J].Geomatics and Information Science of Wuhan University, 2014, 39(2):152-156 http://ch.whu.edu.cn/CN/abstract/abstract2918.shtml
[14] 方兴, 曾文宪, 刘经南, 等.三维坐标转换的通用整体最小二乘算法[J].测绘学报, 2014, 43(11):1139-1143 http://www.cqvip.com/QK/90069X/201411/663183594.html Fang Xing, Zeng Wenxian, Liu Jingnan, et al.A General Total Least Squares Algorithm for Three-Dimensional Coordinate Transformations[J].Acta Geodaetica et Cartographica Sinica, 2014, 43(11):1139-1143 http://www.cqvip.com/QK/90069X/201411/663183594.html
[15] Neitzel F.Generalization of Total Least-Squares on Example of Unweighted and Weighted 2D Similarity Transformation[J].Journal of Geodesy, 2010, 84(12):751-762 doi: 10.1007/s00190-010-0408-0
[16] 方兴, 曾文宪, 刘经南, 等.基于非线性高斯-赫尔默特模型的混合整体最小二乘估计[J].测绘学报, 2016, 45(3):291-296 http://d.old.wanfangdata.com.cn/Periodical/chxb201603007 Fang Xing, Zeng Wenxian, Liu Jingnan, et al.Mixed LS-TLS Estimation Based on Nonlinear Gauss-Helmert Model[J].Acta Geodaetica et Cartographi-ca Sinica, 2016, 45(3):291-296 http://d.old.wanfangdata.com.cn/Periodical/chxb201603007
[17] Chang Guobin.On Least-Squares Solution to 3D Similarity Transformation Problem Under Gauss-Helmert Model[J].Journal of Geodesy, 2015, 89(6):573-576 doi: 10.1007/s00190-015-0799-z
[18] 欧吉坤.一种检测粗差的新方法——拟准测定法[J].科学通报, 1999, 44(16):1777-1781 doi: 10.3321/j.issn:0023-074X.1999.16.021 Ou Jikun.A New Method to Detect Gross Errors:Quasi-Accurate Detection Method[J].Chinese Science Bulletin, 1999, 44(16):1777-1781 doi: 10.3321/j.issn:0023-074X.1999.16.021
[19] 欧吉坤.粗差的拟准检定法(QUAD法)[J].测绘学报, 1999, 28(1):15-20 doi: 10.3321/j.issn:1001-1595.1999.01.004 Ou Jikun. Quasi-Accurate Detection of Gross Errors(QUAD)[J].Acta Geodaetica et Cartographica Sinica, 1999, 28(1):15-20 doi: 10.3321/j.issn:1001-1595.1999.01.004
[20] Yang Yuanxi.Robust Estimation of Geodetic Datum Transformation[J].Journal of Geodesy, 1999, 73(5):268-274 doi: 10.1007/s001900050243
[21] 周江文, 黄幼才, 杨元喜, 等.抗差估计最小二乘法[M].武汉:华中理工大学出版社, 1997 Zhou Jiangwen, Huang Youcai, Yang Yuanxi, et al.Robust Least Squares Method[M].Wuhan:Huazhong University of Science and Technology Press, 1997
[22] 陈义, 陆珏.以三维坐标转换为例解算稳健总体最小二乘方法[J].测绘学报, 2012, 41(5):715-722 http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=QK201205237128 Chen Yi, Lu Jue. Peforming 3D Similarity Transformation by Robust Total Least Squares[J].Acta Geodaetica et Cartographica Sinica, 2012, 41(5):715-722 http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=QK201205237128
[23] Lu Jue, Chen Yi, Li Bofeng, et al.Robust Total Least Squares with Reweighting Iteration for Three-Dimensional Similarity Transformation[J].Survey Review, 2014, 46(334):28-36 doi: 10.1179/1752270613Y.0000000050
[24] Mahboub V, Amirisimkooei A R, Sharifi M A.Ite-ratively Reweighted Total Least Squares:A Robust Estimation in Errors-in-Variables Models[J].Survey Review, 2013, 45(329):92-99 doi: 10.1179/1752270612Y.0000000017
[25] 龚循强, 李志林.稳健加权总体最小二乘法[J].测绘学报, 2014, 43(9):888-894 http://d.old.wanfangdata.com.cn/Periodical/kckxjs201301005 Gong Xunqiang, Li Zhilin.A Robust Weighted Total Least Squares Method[J].Acta Geodaetica et Cartographica Sinica, 2014, 43(9):888-894 http://d.old.wanfangdata.com.cn/Periodical/kckxjs201301005
[26] 王彬, 李建成, 高井祥, 等.抗差加权整体最小二乘模型的牛顿-高斯算法[J].测绘学报, 2015, 44(6):602-608 http://d.old.wanfangdata.com.cn/Periodical/chxb201506002 Wang Bin, Li Jiancheng, Gao Jingxiang, et al.Newton-Gauss Algorithm of Robust Weighted Total Least Squares Model[J].Acta Geodaetica et Cartographica Sinica, 2015, 44(6):602-608 http://d.old.wanfangdata.com.cn/Periodical/chxb201506002
[27] Yang Yuanxi. Robust Estimation for Correlated Observations Based on Bifactor Equivalent Weights[J].Journal of Geodesy, 2002, 76(6):353-358 doi: 10.1007/s00190-002-0256-7
-
期刊类型引用(4)
1. 刘晓云,郭春喜,靳鑫洋,蒋涛. 2020年珠峰高程测量与确定流程解析. 大地测量与地球动力学. 2024(02): 111-115+127 . 百度学术
2. 张建华,张庆涛,张涛,郑文科,陈小英. 顾及水准起算重力异常差异的珠峰地区垂直形变分析. 测绘科学. 2023(06): 1-8 . 百度学术
3. 郭春喜,靳鑫洋,蒋涛,王斌,刘晓云. 2020与2005珠峰测量与高程确定异同. 测绘科学. 2023(07): 10-15 . 百度学术
4. 党亚民,蒋涛,杨元喜,孙和平,姜卫平,朱建军,薛树强,张小红,蔚保国,罗志才,李星星,肖云,章传银,张宝成,李子申,冯伟,任夏,王虎. 中国大地测量研究进展(2019—2023). 测绘学报. 2023(09): 1419-1436 . 百度学术
其他类型引用(0)