期刊文献+

整数规划中割平面法的研究 被引量:4

Study on Cutting Plane Method in Integer Programming
原文传递
导出
摘要 割平面法是求解整数规划问题常用方法之一.用割平面法求解整数规划的基本思路是:先用单纯形表格方法去求解不考虑整数约束条件的松弛问题的最优解,如果获得的最优解的值都是整数,即为所求,运算停止.如果所得最优解不完全是整数,即松弛问题最优解中存在某个基变量为非整数值时,就从最优表中提取出关于这个基变量的约束等式,再从这个约束式出发构造一个割平面方程加入最优表中,再求出新的最优解,这样不断重复的构造割平面方程,直到找到整数解为止.主要研究以下四个关键点:一是研究从最优表中提取出的、关于基变量的约束等式出发,通过将式中的系数进行整数和非负真分数的分解,从而得到一个小于等于0的另外一个不等式的推导过程;二是总结出从小于等于0的那个约束不等式出发构造割平面方程的四种方法;三是分析构造割平面方程的这四种方法相互之间的区别和联系;四是探讨割平面法的几何意义.通过对这四个方面的分析和研究,对割平面法进行透彻的剖析,使读者能够全面把握割平面法. Cutting Plane Method (CPM for short) is one of the approaches adopted frequently for solving Integer Programming problems. Its basic idea is: the optimal solution is found firstly for the relaxed linear program problem in which integer conditions are cancelled from the original integer program problem. Moreover, if the solution's values are all integer, this solution is also optimal for the original problem. If the solution's values are not all integer, that is, there exists non-integer value for certain basis variable, an equality constraint is obtained from the corresponding row of the optimal simplex table, then according to this constraint, an cutting plane equation is constructed which is subsequently joined in the optimal simplex table.Further, by using dual simplex method in the table, a new optimal solution is obtained. Continuously constructing cutting plane equations, until an optimal solution in which all values are integer is derived. This paper mainly investigate the following four critical aspects: the first is to consider the process that: according to the equality constraint obtained from the optimal simplex table, by decomposing each coefficient into an integer and a nonnegative proper fraction, an inequality which is less than or equal to zero is obtained. The second is that: four methods are summarized, under which cutting plane equations are constructed. The third is to analyze the differentiation and relation among the four methods. The fourth is to study the geometry significance of CPM. From the analysis and study of these four aspects, CPM is thoroughly explored so that people can comprehensively understanding it.
出处 《数学的实践与认识》 CSCD 北大核心 2011年第11期82-90,共9页 Mathematics in Practice and Theory
基金 北京市优秀人才培养项目(2009D005003000001) 北京市属高校科技创新平台项目(201098)
关键词 整数规划 割平面法 割平面方程的构造 四个关键点 integer programming cutting plane method construction of cutting plane equation four critical aspects
  • 相关文献

参考文献8

二级参考文献9

  • 1运筹学编写组.高等学校使用教材运筹学(2)[M].北京:清华大学出版社,1990.124.
  • 2Sivaramakrishnan. k. , Mitchell. J. E. Properties of a Cutting Plane Method for Semidefinite Programming [R]. 2007. http://www. rpi. edu/-mitchj.
  • 3Fang, S. C. An Analytic Center Cutting Plane Method for Solving Semi-Infinite Variational Inequality Problems [J]. Journal of Global Optimization. 2004, 28: 141-152.
  • 4Hanif D. Sherali, Barbara M. P. Fraticelli Enhancing RLT relaxations via a new class of semidefinite cuts [J]. Journal of Global Optimization 22:233 - 261, 2002.
  • 5Elhedhli, S. , Goffin, J.-L. The integration of an interior-point cutting plane melhod within a branch-and-price algorithm. [J]. Math. Program. Ser. A 100: 267-294, 2004.
  • 6闵嗣鹤 严士健.初等数论[M].北京:高等教育出版社,1982..
  • 7张建中,许绍吉.线性规划[M].科学出版社,1997.
  • 8Shuzhong Zhang. Quadratic maximization and semidefinite relaxation[J] 2000,Mathematical Programming(3):453~465
  • 9王新辉,刘三阳,刘红卫.半定规划的割平面算法及其应用[J].西安电子科技大学学报,2004,31(1):140-142. 被引量:2

共引文献6

同被引文献25

  • 1刘振航,王全文,吴振奎.割平面法的改进[J].天津轻工业学院学报,2003,18(B12):67-70. 被引量:4
  • 2苟格.整数规划中的割平面法与分枝定界法比较[J].达县师范高等专科学校学报,2005,15(2):18-21. 被引量:3
  • 3邢清华,刘付显.区域防空部署优化系统建模[J].系统工程与电子技术,2006,28(5):712-715. 被引量:33
  • 4谭洁群.整数规划割平面法解题新探[J].广西师院学报(自然科学版),1996,13(3):40-47. 被引量:4
  • 5任继业,陈明,张小水.基于概率的兵力部署模型[J].空军工程大学学报(自然科学版),2007,8(5):45-47. 被引量:10
  • 6Gomory R E. Outline of an algorithm for integer solutions to linear problem [ J ]. Bulletin of the American Mathematical Society, 1958, 64 (5) :275 - 278.
  • 7SHERALIH D, DRISCOLL P J. Evolution and state-of-the- art in integer programming [ J ]. Journal of Computational and Applied Mathematics, 2000, 124 (1/2) :319-340.
  • 8GOMORY R E. Outline of an algorithm for integer solu- tions to linear problem [ J ]. Bulletin of the American Mathematical Society, 1958, 64(5) :275-278.
  • 9HUA Z S, HUANG F H. A variable-grouping based ge- netic algorithm for large-scale integer progranmming [ J ]. Information Sciences, 2006, 176 ( 19 ) : 2869-2885.
  • 10SCHLUTER M, EGEA J A, BANGA J. Extended ant colo- ny optimization for non-convex mixed integer nonlinear programming [ J ]. Computers and Operations Research, 2009, 36(7) : 2217-2229.

引证文献4

二级引证文献7

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部