ScholarLens Back to search
2002•Journal of Guangxi Normal UniversityRequires access

AN INTERIOR TRUST REGION ALGORITHM WITH NONMONOTONIC BACK TRACKING TECHNIQUE FOR BOX CONSTRAINED OPTIMIZATION

Zhu De

Open publisher page 0 citations

Abstract

This paper modifies the trust region interior point algorithm proposed by Coleman Li for solving the optimization problem of a smooth nonlinear function subject to bounds on the variables.A mixed strategy using both trust region and nonmonotonic line search techniques is adopted which switches to back tracking steps produced by the trust region subproblem.The global convergence and fast local convergence rate of the improved algorithm are established under some reasonable conditions.A nonmonotonic criterion is used to speed up the convergence progress in some illconditioned cases.

About this research paper

What this paper is about

This paper modifies the trust region interior point algorithm proposed by Coleman Li for solving the optimization problem of a smooth nonlinear function subject to bounds on the variables.A mixed strategy using both trust region and nonmonotonic line search techniques is adopted which switches to back tracking steps produced by the trust region subproblem.The global convergence and fast local convergence rate of the improved algorithm are established under some reasonable conditions.A nonmonotonic criterion is used to speed up the convergence progress in some illconditioned cases.

Why it matters

A significance statement is not available in the OpenAlex record.

Key contribution

A contribution statement is not available in the OpenAlex record.

Method / approach

Method details are not available in the OpenAlex metadata.

Main findings

Findings are not separately available in the OpenAlex metadata.

Limitations

Limitations are not available in the OpenAlex metadata.

Applications

Application details are not available in the OpenAlex metadata.

Available abstract

This paper modifies the trust region interior point algorithm proposed by Coleman Li for solving the optimization problem of a smooth nonlinear function subject to bounds on the variables.A mixed strategy using both trust region and nonmonotonic line search techniques is adopted which switches to back tracking steps produced by the trust region subproblem.The global convergence and fast local convergence rate of the improved algorithm are established under some reasonable conditions.A nonmonotonic criterion is used to speed up the convergence progress in some illconditioned cases.

Key concepts: Trust region, Line search, Convergence (economics), Mathematical optimization, Algorithm, Function (biology), Computer science, Nonlinear system

Related papers

Related paper 1Related paper 2Related paper 3
Back to paper searchBrowse research topicsOriginal source
AN INTERIOR TRUST REGION ALGORITHM WITH NONMONOTONIC BACK TRACKING TECHNIQUE FOR BOX CONSTRAINED OPTIMIZATION — Research Paper | ScholarLens