数学物理学报 ›› 1999, Vol. 19 ›› Issue (5): 531-536.

• 论文 • 上一篇    下一篇

参变极值问题的信息凝聚分布与Boltzmann极大熵函数

  

  1. (大连理工大学应用数学系 大连 116024)

  • 出版日期:1999-12-05 发布日期:1999-12-05

An Information Condensing Distribution and a Boltzmann Entropy Function for Continuous Maximin Problems

  1. (Applied Mathematics Department, Dalian University of Technology, Dalian 116024)

  • Online:1999-12-05 Published:1999-12-05

摘要:

该文利用Boltzmann熵概念给出了参变极值问题最优解的一种积分极限表达式和极值函数的极大熵函数,讨论了它们一致收敛性的要求并给出了极大熵函数一致收敛的一个充分条件,将之应用到全局最优解问题得到了全局最优解和最优值的一种显表示,最后还探讨了极大熵函数在一类双层规划问题求解中的应用.

关键词: 极大极小问题, Boltzmann熵, 极大熵函数, 全局最优解, 双层规划.

Abstract:

An information condensing distribution via a prior experence distribution about the solution and a maximum entropy function for continuous maximin problem are presented in the means of Boltzmann entropy concept. An explicit expression for the global optimization solution is obtained and an application of the maximum entropy function approach in sloving a class of bilevel programming problem is also discussed.

Key words: Maximin problem, Boltzmann entropy, Maximum entropy function, Global optimal solution, Bilevel programming.

中图分类号: 

  • 90C