ScholarLens Back to search
2010•Journal of Shantou UniversityRequires access

A SQP Method for Inequality Constrained Optimization and Its Convergence

Ning Zhu

Open publisher page 0 citations

Abstract

In this paper,a SQP method,in which the merit function is nondifferentiable exact penalty function,is presented to solve inequality constraint.The penalty is adjusted automatically.The arc search and the second order correction,which is obtained by solving an auxiliary linear equation system,are used to obtain a feasible descent algorithm.Under some suitable assumptions,it is proved that the convergence of the algorithm is global as well as superlinear.

About this research paper

What this paper is about

In this paper,a SQP method,in which the merit function is nondifferentiable exact penalty function,is presented to solve inequality constraint.The penalty is adjusted automatically.The arc search and the second order correction,which is obtained by solving an auxiliary linear equation system,are used to obtain a feasible descent algorithm.Under some suitable assumptions,it is proved that the convergence of the algorithm is global as well as superlinear.

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

In this paper,a SQP method,in which the merit function is nondifferentiable exact penalty function,is presented to solve inequality constraint.The penalty is adjusted automatically.The arc search and the second order correction,which is obtained by solving an auxiliary linear equation system,are used to obtain a feasible descent algorithm.Under some suitable assumptions,it is proved that the convergence of the algorithm is global as well as superlinear.

Key concepts: Penalty method, Sequential quadratic programming, Convergence (economics), Mathematics, Mathematical optimization, Constraint (computer-aided design), Descent (aeronautics), Constrained optimization

Related papers

Related paper 1Related paper 2Related paper 3
Back to paper searchBrowse research topicsOriginal source
A SQP Method for Inequality Constrained Optimization and Its Convergence — Research Paper | ScholarLens