首页 | 本学科首页   官方微博 | 高级检索  
     

线性视划鞍点算法原理与实际计算
引用本文:尚毅,成孟金,张国光,李鹏辉,刘冬. 线性视划鞍点算法原理与实际计算[J]. 沈阳化工学院学报, 2004, 18(2): 138-143
作者姓名:尚毅  成孟金  张国光  李鹏辉  刘冬
作者单位:沈阳化工学院,辽宁,沈阳,110142
摘    要:阐述了线性规划鞍点算法原理与实际计算.讨论了在基不变条件下,算法的收敛性和迭代步长的选择问题.指出算法性质为q-线性收敛.鞍点算法软件与美国优化技术中心编制的内点算法软件PCx进行了比较,用两种算法计算网上NETLEB 的线性规划问题,公布了比较结果.线性规划鞍点算法已应用到石化企业的优化生产工作.

关 键 词:线性规划   梯度   鞍点   并行算法

Principle and Computational Experience of a Saddle Point Algorithm for Linear Programming
Abstract. Principle and Computational Experience of a Saddle Point Algorithm for Linear Programming[J]. Journal of Shenyang Institute of Chemical Technolgy, 2004, 18(2): 138-143
Authors:Abstract
Abstract:This paper proposes a principle of new algorithm for linear programming based on saddle point. The convergence and choice of the step size under the condition of the changeless base set is discussed. The computational complexity is q-linear. The NETLIB problems were computed by using the proposed algorithm and the interior-point working code PCx respectively, and results were compared. This code has been successfully used to solve problems in petrochemical industry.
Keywords:linear programming  gradient  saddle point  parallel algorithm
本文献已被 CNKI 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号