RUS  ENG
Full version
JOURNALS // Zhurnal Vychislitel'noi Matematiki i Matematicheskoi Fiziki // Archive

Zh. Vychisl. Mat. Mat. Fiz., 2007 Volume 47, Number 9, Pages 1524–1537 (Mi zvmmf247)

This article is cited in 2 papers

Application of parallel heuristic algorithms for speeding up parallel implementations of the branch-and-bound method

M. A. Posypkina, I. Kh. Sigalb

a Institute of Systems Analysis, Russian Academy of Sciences, pr. Shestidesyatiletiya Oktyabrya 9, Moscow, 117312, Russia
b Computing Center, Russian Academy of Sciences, ul. Vavilova 40, Moscow, 119991, Russia

Abstract: A scheme for the parallel implementation of the combined branch-and-bound method and heuristic algorithms is proposed. Results of computations for the one-dimensional Boolean knapsack problem are presented that demonstrate the efficiency of the proposed approach. The main factors that affect the speedup of the solution when local optimization is used are discussed.

Key words: knapsack problem, branch-and-bound method, algorithms for parallel computations, discrete optimization, local optimization.

UDC: 519.6:519.852.6

Received: 08.02.2007


 English version:
Computational Mathematics and Mathematical Physics, 2007, 47:9, 1464–1476

Bibliographic databases:


© Steklov Math. Inst. of RAS, 2024