Optimizing fuzzy makespan and tardiness for unrelated parallel machine scheduling with archived metaheuristics

This research presents two simulated annealing (SA) and a greedy randomized adaptive search procedure (GRASP) to solve unrelated parallel machine scheduling problems (UPMSPs) with two fuzzy optimization objectives—makespan and average tardiness. Few studies have employed fuzzy approach to solve mult...

Full description

Saved in:
Bibliographic Details
Published in:International journal of advanced manufacturing technology 2011-11, Vol.57 (5-8), p.763-776
Main Authors: Chyu, Chiuh-Cheng, Chang, Wei-Shung
Format: Article
Language:English
Subjects:
Adaptive search techniques
Algorithms
CAE) and Design
Computer simulation
Computer-Aided Engineering (CAD
Decoding
Engineering
Fitness
Industrial and Production Engineering
Job shops
Matching
Mechanical Engineering
Media Management
Multiple objective analysis
Optimization
Original Article
Production scheduling
Scheduling
Simulated annealing
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:This research presents two simulated annealing (SA) and a greedy randomized adaptive search procedure (GRASP) to solve unrelated parallel machine scheduling problems (UPMSPs) with two fuzzy optimization objectives—makespan and average tardiness. Few studies have employed fuzzy approach to solve multi-objective UPMSPs. In the research, several schemes are incorporated into the algorithm, including (1) matching-based decoding; (2) acceptance rule based on Pareto dominance, objective fitness, or Pareto reference point distance; (3) random or fixed weighted direction search. The matching-based decoding scheme has two phases: first max–min matching and then Hungarian method. Experiments were conducted to evaluate the algorithms’ performance for moderate to large problem size instances. The results indicate that matching-based decoding scheme significantly improves solution quality, but will require more computation time. GRASP with path relinking performs slightly worse than objective fitness-based multi-objective simulated annealing algorithms (MOSA), but better than Pareto dominance-based MOSA in terms of several Pareto performance measures.
ISSN:0268-3768
1433-3015
DOI:10.1007/s00170-011-3317-3