摘要
当双层规划(BLP)的下层问题存在不确定性时,运用鲁棒优化方法可转化成双层二阶锥规划问题(SOCBLP).由于SOCBLP通常是非凸不可微问题,难以直接处理.本文将二维线性SOCBLP转化为线性BLP,并给出一些理论性质.基于这些性质,给出求解二维线性SOCBLP的一种Kth-best算法.算例表明该算法的有效性.
When there is uncertainty in the lower level problem of a bilevel programming problem（BLP）,it can be formulated as a second-order cone bilevel programming problem（SOCBLP） by a robust optimization method.Since SOCBLP is generally a problem of nonconvexity and nondifferentiability,it is difficult to deal with directly.In this paper,we reformulate the 2-dimensional linear SOCBLP as a linear BLP and give some theoretical properties.Based on these properties,a Kth-best approach is proposed for solving the 2-dimensional linear SOCBLP.The preliminary example shows the effectiveness of our approach.
出处
《黄冈师范学院学报》
2012年第6期1-4,共4页
Journal of Huanggang Normal University
基金
supported by the Natural Science Foundation of Hubei Province(2011CDC028)
the Key Project of Hubei Provincial Department of Education(D20122701)
the Excellent Youth Project of Hubei Provincial Department of Education(Q20122709)
the Doctorial Foundation of Huanggang Normal University(2011CD292,08CD158)
关键词
二阶锥双层规划
非凸
不可微
Kth-best算法
second-order cone bilevel programming
nonconvex
nondifferentiability
Kth-best approach
