Switching strategy-based hybrid evolutionary algorithms for job shop scheduling problems

Shahed Mahmud (), Ripon K. Chakrabortty (), Alireza Abbasi () and Michael J. Ryan ()
Additional contact information
Shahed Mahmud: University of New South Wales
Ripon K. Chakrabortty: University of New South Wales
Alireza Abbasi: University of New South Wales
Michael J. Ryan: Capability Associates

Journal of Intelligent Manufacturing, 2022, vol. 33, issue 7, No 4, 1939-1966

Abstract: Abstract Since production efficiency and costs are directly affected by the ways in which jobs are scheduled, scholars have advanced a number of meta-heuristic algorithms to solve the job shop scheduling problem (JSSP). Although this JSSP is widely accepted as a computationally intractable NP-hard problem in combinatorial optimization, its solution is essential in manufacturing. This study proposes performance-driven meta-heuristic switching approaches that utilize the capabilities of multi-operator differential evolution (MODE) and particle swarm optimization (PSO) in a single algorithmic framework. The performance-driven switching mechanism is introduced to switch the population from an under-performing algorithm to other possibilities. A mixed selection strategy is employed to ensure the diversity and quality of the initial population, whereas a diversity check mechanism maintains population diversity over the generations. Moreover, a Tabu search (TS) inspired local search technique is implemented to enhance the proposed algorithm’s exploitation capability, avoiding being trapped in the local optima. Finally, this study presents two mixed population structure-based hybrid evolutionary algorithms (HEAs), such as a predictive sequence HEA (sHEA) and a random sequence HEA (rHEA), and one bi-population inspired HEA, called bHEA. The comparative impacts of these varied population structure-based approaches are assessed by solving 5 categories of the standard JSSP instances (i.e., FT, LA, ORB, ABZ and TA). The performance of these hybridized approaches (i.e., sHEA, rHEA and bHEA) is compared and contrasted with its constituent algorithms (MODE, PSO and TS) to validate the hybridization’s effectiveness. The statistical analysis shows that sHEA ranked first with mean value 1.84 compared to rHEA (1.96) and bHEA (2.21). Moreover, the proposed sHEA is compared with 26 existing algorithms and ranked first with a mean value 5.09 compared to the near-best algorithms. Thus, the simulation results and statistical analysis prove the supremacy of the sHEA.

Keywords: Job shop scheduling problem; Multi-operator differential evolution; Particle swarm optimization; Evolutionary algorithm; Local search (search for similar items in EconPapers)
Date: 2022
References: View references in EconPapers View complete reference list from CitEc
Citations:

Downloads: (external link)
http://link.springer.com/10.1007/s10845-022-01940-1 Abstract (text/html)
Access to the full text of the articles in this series is restricted.

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:spr:joinma:v:33:y:2022:i:7:d:10.1007_s10845-022-01940-1

Ordering information: This journal article can be ordered from
http://www.springer.com/journal/10845

DOI: 10.1007/s10845-022-01940-1

Access Statistics for this article

Journal of Intelligent Manufacturing is currently edited by Andrew Kusiak

More articles in Journal of Intelligent Manufacturing from Springer
Bibliographic data for series maintained by Sonal Shukla () and Springer Nature Abstracting and Indexing ().