ALGORITHM BASED ON LANDWEBER ITERATION FOR SOLVING RANK DEFICIENCY NONLINEAR LEAST SQUARES PROBLEM

Tang Limin,Zhu Jianjun

DOI: https://doi.org/10.3969/j.issn.1671-5942.2010.01.021

2010-01-01

Abstract:More methods such as the gauss-newton method or modified gauss-newton method will failure when the iteration matrix is rank-deficient or very ill-conditioned in solving ill-posed nonlinear least squares problem.Nonlinear Landweber iteration formula xδk+1=xδk-f'(xδk)(f(xδk)-yδ)is analyzed and a new method is derived.On the basis of the conversion relation of inverse matrix and adjoint matrix,by using 1/ω instead of |(B'(xk)B(xk))|,a new Landweber iteration formula xδk+1=xδk-ω(B'(xδk)B(xδk))B'(xδk)(f(xδk)-yδ)is constructed for solving rank deficiency nonlinear least squares problem,with which the phenomenon that leads to ill-posed problem because the iteration matrix is rank-deficient and very ill-conditioned in numerical iterative process is avoided.The numerical experiment showes that the new Landweber iteration formula is accurate and of applicability for nonlinear adjustment of free networks with rank deficiency and rank deficiency nonlinear least squares problems.