An iterated greedy algorithm for the parallel blocking flow shop scheduling problem and sequence-dependent setup times

This paper deals with the problem of scheduling jobs in a parallel flow shop configuration under the blocking constraint, in which the setup time of machines depends not only on the job to be processed but also on the previously processed one, i.e., there are sequence-dependent setup times. The perf...

Full description

Bibliographic Details
Authors: Ribas Vila, Immaculada|||0000-0002-3701-118X, Companys Pascual, Ramón, Tort-Martorell Llabrés, Xavier|||0000-0003-1167-6703
Format: article
Publication Date:2021
Country:España
Institution:Universitat Politècnica de Catalunya (UPC)
Repository:UPCommons. Portal del coneixement obert de la UPC
Language:English
OAI Identifier:oai:upcommons.upc.edu:2117/351547
Online Access:https://hdl.handle.net/2117/351547
https://dx.doi.org/10.1016/j.eswa.2021.115535
Access Level:Open access
Keyword:Parallel processing (Electronic computers)
Mathematical optimization
Industrial management
Parallel Flow Shop
Sequence-dependent Setup times
Makespan
Distributed Flow Shop
Blocking
Processament en paral·lel (Ordinadors)
Optimització matemàtica
Empreses -- Direcció i administració
Àrees temàtiques de la UPC::Economia i organització d'empreses
Description
Summary:This paper deals with the problem of scheduling jobs in a parallel flow shop configuration under the blocking constraint, in which the setup time of machines depends not only on the job to be processed but also on the previously processed one, i.e., there are sequence-dependent setup times. The performance analysis of several iterated greedy algorithms with different initial solution procedures and local searches lets us define an efficient algorithm to minimize the maximum job completion time. Moreover, the computational evaluation showed the efficiency of searching in different neighborhood structures and noted the significant influence of the initial solution. However, contrary to other scheduling problems, starting with a high quality solution does not guarantee better performance of the algorithm.