ScholarLens Back to search
2007•Mathematica ApplicataRequires access

A Modified SQP-filter Method

Ke Su

Open publisher page 0 citations

Abstract

The current sequential quadratic programming (SQP) type algorithm may fail if the QP subproblem is infeasible. Moreover the choice of penalty function in line search is difficult. In [1], the author modified the SQP method for the former problem. In this paper, on the base of paper [1], a line search filter search technique is presented to conquer the second problem. Under some conditions, the global convergence of the algorithm in this paper has shown at the end.

About this research paper

What this paper is about

The current sequential quadratic programming (SQP) type algorithm may fail if the QP subproblem is infeasible. Moreover the choice of penalty function in line search is difficult. In [1], the author modified the SQP method for the former problem. In this paper, on the base of paper [1], a line search filter search technique is presented to conquer the second problem. Under some conditions, the global convergence of the algorithm in this paper has shown at the end.

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

The current sequential quadratic programming (SQP) type algorithm may fail if the QP subproblem is infeasible. Moreover the choice of penalty function in line search is difficult. In [1], the author modified the SQP method for the former problem. In this paper, on the base of paper [1], a line search filter search technique is presented to conquer the second problem. Under some conditions, the global convergence of the algorithm in this paper has shown at the end.

Key concepts: Sequential quadratic programming, Mathematical optimization, Line search, Convergence (economics), Filter (signal processing), Penalty method, Mathematics, Divide and conquer algorithms

Related papers

Related paper 1Related paper 2Related paper 3
Back to paper searchBrowse research topicsOriginal source
A Modified SQP-filter Method — Research Paper | ScholarLens