Scholarlink

<p>This paper introduces a new function that builds upon a smoothed and symmetrized version of the Fischer-Burmeister function. Based on this proposed function, we present an approach for solving the second-order cone complementarity problems (denoted as SOCCPs) that combines smoothing and regularization techniques. To determine the step size, this proposed method adopts a new nonmonotone line search. We also show the global and local quadratic convergence of the method under suitable assumptions. In addition, for analysing its local quadratic convergence, we employ the local error bound condition - a weaker requirement than standard nonsingularity assumptions. Moreover, we consider <InlineEquation ID="IEq1"> <EquationSource Format="TEX">\(\nu \)</EquationSource> <EquationSource Format="MATHML"><math> <mi>ν</mi> </math></EquationSource> </InlineEquation> as the smoothing parameter as well as the regularization parameter and treat <InlineEquation ID="IEq2"> <EquationSource Format="TEX">\(\nu \)</EquationSource> <EquationSource Format="MATHML"><math> <mi>ν</mi> </math></EquationSource> </InlineEquation> as independent variables in the proposed method. Numerical experiments are conducted to validate the theoretical properties of our proposed method.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

A Unified Approach to Smoothing and Regularization for SOCCPs under Local Error Bound

  • 期刊论文

  • 发表日期：

    2026年5月22日

查看全文
  • Rashmi Kumari,
  • C. Nahak

摘要

This paper introduces a new function that builds upon a smoothed and symmetrized version of the Fischer-Burmeister function. Based on this proposed function, we present an approach for solving the second-order cone complementarity problems (denoted as SOCCPs) that combines smoothing and regularization techniques. To determine the step size, this proposed method adopts a new nonmonotone line search. We also show the global and local quadratic convergence of the method under suitable assumptions. In addition, for analysing its local quadratic convergence, we employ the local error bound condition - a weaker requirement than standard nonsingularity assumptions. Moreover, we consider \(\nu \) ν as the smoothing parameter as well as the regularization parameter and treat \(\nu \) ν as independent variables in the proposed method. Numerical experiments are conducted to validate the theoretical properties of our proposed method.