Weighted Total Least Squares with Constraints: Alternative Derivation without Using Lagrange Multipliers

被引:10
作者
Amiri-Simkooei, A. R. [1 ]
机构
[1] Univ Isfahan, Fac Civil Engn & Transportat, Dept Geomat Engn, Esfahan 8174673441, Iran
关键词
Constrained weighted total least squares (CWTLS); Errors-in-variables (EIV) model; Linear equality constraints; IN-VARIABLES MODELS; INEQUALITY CONSTRAINTS; TRANSFORMATION;
D O I
10.1061/(ASCE)SU.1943-5428.0000253
中图分类号
TU [建筑科学];
学科分类号
0813 ;
摘要
This technical note presents alternative derivations for the weighted total least-squares (WTLS) problem subject to weighted and hard constraints. The derivations do not take into account the Lagrange multipliers, and the final results are shown to be identical to those presented by a recently published article in the same journal. The final WTLS estimates are formulated by the standard least-squares theory. The method, which is formulated using the Gauss-Newton method, provides an alternative to the Newton method. The formulation is generally presented for the weighted linear(ized) constraints. As a by-product, hard constraints turn out to be a special case of the general formulation of the weighted constraints. An alternative derivation is presented for the WTLS problem subjected only to hard constraints. For this purpose, the parametric representation of the hard constraints is used, the general solution of which is obtained as the sum of a particular solution and the homogeneous solution. As a special case, the WTLS formulation (without the constraints) can also directly be followed. (C) 2018 American Society of Civil Engineers.
引用
收藏
页数:6
相关论文
共 39 条
[1]   Total least squares solution of coordinate transformation [J].
Akyilmaz, O. .
SURVEY REVIEW, 2007, 39 (303) :68-80
[2]   Weighted total least squares formulated by standard least squares theory [J].
Amiri-Simkooei, A. ;
Jazaeri, S. .
JOURNAL OF GEODETIC SCIENCE, 2012, 2 (02) :113-124
[3]   Weighted Total Least Squares with Singular Covariance Matrices Subject to Weighted and Hard Constraints [J].
Amiri-Simkooei, A. R. .
JOURNAL OF SURVEYING ENGINEERING, 2017, 143 (04)
[4]   On the Covariance Matrix of Weighted Total Least-Squares Estimates [J].
Amiri-Simkooei, A. R. ;
Zangeneh-Nejad, F. ;
Asgari, J. .
JOURNAL OF SURVEYING ENGINEERING, 2016, 142 (03)
[5]   Weighted total least squares applied to mixed observation model [J].
Amiri-Simkooei, A. R. ;
Mortazavi, S. ;
Asgari, J. .
SURVEY REVIEW, 2016, 48 (349) :278-286
[6]   Estimation of straight line parameters with fully correlated coordinates [J].
Amiri-Simkooei, A. R. ;
Zangeneh-Nejad, F. ;
Asgari, J. ;
Jazaeri, S. .
MEASUREMENT, 2014, 48 :378-386
[7]   Application of least squares variance component estimation to errors-in-variables models [J].
Amiri-Simkooei, A. R. .
JOURNAL OF GEODESY, 2013, 87 (10-12) :935-944
[8]   Data-snooping procedure applied to errors-in-variables models [J].
Amiri-Simkooei, Ali Reza R. ;
Jazaeri, Shahram .
STUDIA GEOPHYSICA ET GEODAETICA, 2013, 57 (03) :426-441
[9]  
Amiri-Simkooei AR., 2007, Least-Squares Variance Component Estimation. Theory and GPS Applications, DOI [10.54419/fz6c1c, DOI 10.54419/FZ6C1C]
[10]  
[Anonymous], 2000, Series on Mathematical Geodesy and Positioning