Abstract:Total least squares estimation based on Gauss–Newton method in nonlinear errors-in-variables (NEIV) model will encounter the problems of convergence, correctness and accuracy of solution related to the selected initial parameter values. In this contribution, a new total least squares estimator is introduced to solve NEIV model. This method is an extension of the homotopy nonlinear weighted least square (HNWLS) method, which is used in the nonlinear Gauss–Markov model where only the dependent variables contain random errors. The new estimator is called homotopy nonlinear weighted total least squares (HNWTLS), because it adopts weighted total least squares adjustment criterion and homotopy method to estimate nonlinear model parameters. The homotopy function of HNWTLS is constructed by using the normal equation of weighted total least squares adjustment criterion. By taking the error vector of independent variables as a parameter vector, the NEIV model is transformed into a classical nonlinear adjustment model. Then, according to the conclusion of HNWLS, the calculation formula of HNWTLS is derived, and the corresponding calculation algorithm is developed accordingly, where the standard Euler prediction and Newton correction method are introduced into it to tracks the homotopy curves. Finally, three examples to demonstrate the advantage and efficiency of HNWTLS estimator are given and some conclusions are drawn.

A Gauss--Newton iteration for Total Least Squares problems

A Gauss–Newton method for mixed least squares-total least squares problems

Approximate Gauss–Newton methods for solving underdetermined nonlinear least squares problems

Least Squares Solution with the Minimum-Norm to General Matrix Equations Via Iteration.

Convergence and Uniqueness Properties of Gauss-Newton's Method

A Proximal Iteratively Regularized Gauss–Newton Method for Nonlinear Inverse Problems

A Riemannian under-determined BFGS method for least squares inverse eigenvalue problems

Incremental Gauss--Newton Methods with Superlinear Convergence Rates

Gauss-Southwell type descent methods for low-rank matrix optimization

Gauss-Newton Methods for A Class of Nonsmooth Optimization Problems

Inexact Gauss-Newton methods with matrix approximation by sampling for nonlinear least-squares and systems

ON CONVERGENCE OF A TRUNCATED GAUSS-NEWTON METHOD FOR SOLVING UNDERDETERMINED NONLINEAR LEAST SQUARES PROBLEMS

A convergence analysis of the iteratively regularized Gauss-Newton method under Lipschitz condition

An iterative algorithm for the least Frobenius norm least squares solution of a class of generalized coupled Sylvester-transpose linear matrix equations

On an Iterative Method for Solving the Least Squares Problem of Rank-Deficient Systems

A Rapidly Convergent Iterative Method for the Solution of the Generalised Nonlinear Least Squares Problem

Linearly convergent adjoint free solution of least squares problems by random descent

An alternating iterative minimisation algorithm for the double-regularised total least square functional

Homotopy nonlinear weighted total least squares adjustment

Weighted least squares collocation methods

Extended Gauss-Newton and ADMM-Gauss-Newton Algorithms for Low-Rank Matrix Optimization