基于松弛策略解半无限规划模型的显式修正算法

讨论了一类线性半无限最优规划模型的求解算法.采用松弛方法解其系列子问题LP(T_k)及DLP(T_k),基于松弛策略和在适当的假设条件下,提出了一个我们称之为显式算法的新型算法.新算法的主要改进之处是算法在每一步迭代计算时,允许丢弃一些不必要的约束.在这种方式下,算法避免了求解系列太大规模的子问题.最后,基于提出的显式修正算法,并与传统割平面方法和已有文献中的松弛修正算法、对同一问题作了初步的数值比较实验.

A Modified Explicit Algorithm Based-Relaxation-Strategy for Solving Semi-infinite Programming

In this paper,we investigate a class of linear semi-infinite programming.We adopt the relaxation-approach such that a sequence of linear programsLP(T_k)and DLP(T_k) is solved.Based on the relaxed strategies,and under some conditions,a new explicit algorithm is developed.The major improvement of the proposed new algorithm is that it allows the unnecessary constraints to be dropped at each iteration.In this way,the proposed algorithm avoid solving too large problems.Finally,three different algorithms(namely,the proposed explicit algorithm,the traditional cutting plane method,and the modification algorithm published literature)have been implemented for same problems.