首页 | 本学科首页   官方微博 | 高级检索  
     检索      


A Finite Branch-and-Bound Algorithm for Linear Multiplicative Programming
Authors:Takahito Kuno
Abstract:On the basis of Soland's rectangular branch-and-bound, we develop an algorithm for minimizing a product of p (ge2) affine functions over a polytope. To tighten the lower bound on the value of each subproblem, we install a second-stage bounding procedure, which requires O(p) additional time in each iteration but remarkably reduces the number of branching operations. Computational results indicate that the algorithm is practical if p is less than 15, both in finding an exact optimal solution and an approximate solution.
Keywords:global optimization  nonconvex optimization  linear multiplicative programming  branch-and-bound algorithm  continuous knapsack problem
本文献已被 SpringerLink 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