An effective multi-start iterated greedy algorithm to minimize makespan for the distributed permutation flowshop scheduling problem with preventive maintenance

•We study a distributed flowshop problem with preventive maintenance.•We present a novel multi-start iterated greedy algorithm.•An improved NEH with dropout is proposed to generate searching starting-point.•A destruction with the tournament selection is well designed.•The effective of the iterated g...

Full description

Saved in:
Bibliographic Details
Published in:Expert systems with applications Vol. 169; p. 114495
Main Authors: Mao, Jia-yang, Pan, Quan-ke, Miao, Zhong-hua, Gao, Liang
Format: Journal Article
Language:English
Published: New York Elsevier Ltd 01.05.2021
Elsevier BV
Subjects:
Distributed permutation flowshop scheduling problem
Greedy algorithms
Heuristic methods
Iterated greedy algorithm
Job shops
Makespan
Permutations
Preventive maintenance
Scheduling
Iterated greedy algorithm
Makespan
Distributed permutation flowshop scheduling problem
Heuristic methods
Preventive maintenance
ISSN:0957-4174, 1873-6793
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:•We study a distributed flowshop problem with preventive maintenance.•We present a novel multi-start iterated greedy algorithm.•An improved NEH with dropout is proposed to generate searching starting-point.•A destruction with the tournament selection is well designed.•The effective of the iterated greedy is proved by extensive experiments. In recent years, distributed scheduling problems have been well studied for their close connection with multi-factory production networks. However, the maintenance operations that are commonly carried out on a system to restore it to a specific state are seldom taken into consideration. In this paper, we study a distributed permutation flowshop scheduling problem with preventive maintenance operation (PM/DPFSP). A multi-start iterated greedy (MSIG) algorithm is proposed to minimize the makespan. An improved heuristic is presented for the initialization and re-initialization by adding a dropout operation to NEH2 to generate solutions with a high level of quality and disperstiveness. A destruction phase with the tournament selection and a construction phase with an enhanced strategy are introduced to avoid local optima. A local search based on three effective operators is integrated into the MSIG to reinforce local neighborhood solution exploitation. In addition, a restart strategy is adpoted if a solution has not been improved in a certain number of consecutive iterations. We conducted extensive experiments to test the performance of the presented MSIG. The computational results indicate that the presented MSIG has many promising advantages in solving the PM/DPFSP under consideration.
Bibliography:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:0957-4174
1873-6793
DOI:10.1016/j.eswa.2020.114495