Upravlenie Bol'shimi Sistemami
 RUS  ENG JOURNALS   PEOPLE   ORGANISATIONS   CONFERENCES   SEMINARS   VIDEO LIBRARY   PACKAGE AMSBIB
 General information Latest issue Archive Impact factor Search papers Search references RSS Latest issue Current issues Archive issues What is RSS

 UBS: Year: Volume: Issue: Page: Find

 UBS, 2016, Issue 62, Pages 60–74 (Mi ubs880)

Analysis and Synthesis of Control Systems

Parallel algorithms with feasible set partitioning for large-scale optimization problems

A. S. Velichko

Institute for Automation and Control Processes of FEB RAS

Abstract: We consider parallel algorithms for the class of constrained optimization problems. The algorithms are based on the two ideas: gradient projection method and decomposition of the set of constraints into disjoint subsets with different number of constraints. The proposed approach is used for a class of large-scale linear programming problems. A number of simulations was made to analyze computational complexity, convergence speed and the parallelization efficiency. The algorithms show a good performance for the special special computationally difficult benchmark problem. The most efficient decomposition strategy is to partition the feasible set into a relatively small number of subsets. The decomposition by separate constraints demonstrates less efficiency of parallelization.

Keywords: parallel algorithm, gradient projection method, decomposition, large-scale problem

Full text: PDF file (248 kB)
References: PDF file   HTML file
UDC: 519.85
BBK: 22.18
Received: February 1, 2016
Published: July 31, 2016

Citation: A. S. Velichko, “Parallel algorithms with feasible set partitioning for large-scale optimization problems”, UBS, 62 (2016), 60–74

Citation in format AMSBIB
\Bibitem{Vel16} \by A.~S.~Velichko \paper Parallel algorithms with feasible set partitioning for large-scale optimization problems \jour UBS \yr 2016 \vol 62 \pages 60--74 \mathnet{http://mi.mathnet.ru/ubs880} \elib{https://elibrary.ru/item.asp?id=26608672}