首页 | 本学科首页   官方微博 | 高级检索  
     


Improving a Lagrangian decomposition for the unconstrained binary quadratic programming problem
Authors:Geraldo Regis Mauri  Luiz Antonio Nogueira Lorena
Affiliation:a Federal University of Espírito Santo - UFES, Alto Universitário s/n, Guararema, Alegre-ES 29500-000, Brazil
b National Institute for Space Research - INPE, Av. dos Astronautas 1758, Jd. da Granja, São José dos Campos-SP 12210-970, Brazil
Abstract:This paper presents a new alternative of Lagrangian decomposition based on column generation technique to solve the unconstrained binary quadratic programming problem. We use a mixed binary linear version of the original quadratic problem with constraints represented by a graph. This graph is partitioned into clusters of vertices forming subproblems whose solutions use the dual variables obtained by a coordinator problem. Computational experiments consider a set of difficult instances and the results are compared against other methods reported recently in the literature.
Keywords:Column generation   Lagrangian decomposition   Quadratic programming
本文献已被 ScienceDirect 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