Exact Solution of the Single Machine Total Tardiness Problem : the Power of Memorization - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2017

Exact Solution of the Single Machine Total Tardiness Problem : the Power of Memorization

Résumé

This paper provides new experimental results on the exact solution of the single machine total tardiness problem (SMTT). The currently fastest algorithm reported in the literature is due to Szwarc et al \cite{szwarc2001}, which solves instances with up to 500 jobs in size. The key elements of their algorithm are the use of \techmemo{} and the exploitation of an ``Algorithmic Paradox''. Recently, we revisited the problem and proposed a \techbm{} algorithm that runs in $\ostar{(2+\epsilon)^n}$ in time, with $\epsilon$ arbitrarily small, and in polynomial space. During the study of the computational performance of this algorithm, the authors occasionally found that the \techmemo{} technique used in \cite{szwarc2001} can be further boosted and their ``Paradox'' can be enhanced. The resulting algorithm is able to solve all instances with up to 1200 jobs in size. We also discuss the worst-case running time of the algorithm.
Fichier non déposé

Dates et versions

hal-01550022 , version 1 (29-06-2017)

Identifiants

  • HAL Id : hal-01550022 , version 1

Citer

Lei Shang, Vincent t'Kindt, Federico Della Croce. Exact Solution of the Single Machine Total Tardiness Problem : the Power of Memorization. International Conference on Industral Engineering and System Management (IESM’17), Oct 2017, Saarbrucken, Germany. ⟨hal-01550022⟩
143 Consultations
0 Téléchargements

Partager

Gmail Facebook X LinkedIn More