大理大学学报 ›› 2022, Vol. 7 ›› Issue (6): 1-8.

• 数学与计算机科学 •    下一篇

信赖域子问题求解方法及其数值试验研究


  

  1. 滁州城市职业学院教育学院,安徽滁州 239000
  • 收稿日期:2022-02-14 修回日期:2022-03-05 出版日期:2022-06-15 发布日期:2022-07-04
  • 作者简介:袁远,讲师,主要从事数学教育与教学研究。
  • 基金资助:

    滁州城市职业学院校级高职扩招专项项目(2021kzzx01);滁州城市职业学院校级质量工程项目(2021zdjyxm10

Study on the Solution Methods and Numerical Tests of Trust Region Subproblem

#br#   

  1. Department of Education, Chuzhou City Vocation College, Chuzhou, Anhui 239000,China
  • Received:2022-02-14 Revised:2022-03-05 Online:2022-06-15 Published:2022-07-04

摘要: 信赖域算法是目前求解无约束优化问题的一种重要的数值计算方法,而信赖域子问题的求解则是实现信赖域算法的关键。阐述求解信赖域子问题的3种方法(不定折线法、Moré-Sorensen法以及截断共轭梯度法),利用国际上广泛采用的无约束优化测试函数包对以上3种方法进行大量的数值试验。结果表明截断共轭梯度法的数值计算效率在一定程度上优于其他两种方法,非单调的信赖域算法在一定程度上优于传统的单调算法。

关键词:

 , 信赖域算法;信赖域子问题;数值计算效率;Matlab

Abstract: Trust region algorithm is an important numerical method to solve unconstrained optimization problems at present and the solution of trust region subproblem is the key to realizing the trust region algorithm. In this paper three methods for solving trust region subproblemsindeterminate broken line method Moré-Sorensen method and truncated conjugate gradient method are described. A large number of numerical tests are carried out on the above three methods by using the unconstrained optimization test function package widely used in the world. The results show that to some extent the efficiency of the truncated conjugate gradient method is better than the other two methods and the non-monotone trust region algorithm is better than the traditional monotone algorithm to some extent.

Key words: trust region algorithm, trust region subproblem, numerical efficiency, Matlab

中图分类号: