| 摘要: |
| 本文研究了基于神经网络的二层规划问题.利用互补松弛条件的扰动,获得了二层规划问题局部最优解的充分条件,克服了互补松弛条件不满足约束规格的局限性,并给出了相应的神经网络求解方法,从而求解原二层规划问题,数值实验表明算法有效. |
| 关键词: 二层规划 扰动 约束规格 神经网络 |
| DOI： |
| 分类号:O221.2 |
| 基金项目: |
|
| NEURAL NETWORKS SOLUTION OF BILEVEL PROGRAMMING PROBLEM |
|
PENG Ai-min
|
|
School of Mathematics and Quantitative Economics, Hubei University of Education, Wuhan 430205, China
|
| Abstract: |
| Here is the abstract of bilevel programming based on neural networks. By using complementarity constraints, a sufficient condition is considered. It improves the drawback that constraint qualification does not hold at any feasible point when a bilevel programming is changed into a single one. We arrive at a conclusion of neural networks solution for bilevel programming. The validity of the networks is demonstrated by several numerical examples. |
| Key words: bilevel programming perturted constraint qualification neural networks |
