Diversity oriented multi-objective island based genetic algorithm for flexible job shop scheduling considering setup operator regulation

• Published in: Journal of Human Environmental Studies Vol. 13; no. 1; pp. 1 - 12
• Main Authors: Nagao, Masahiro, Sugimoto, Takuya, Morinaga, Yasuhiko, Sano, Mitsuru
• Format: Journal Article
• Language: English
• Published: Society for Human Environmental Studies 2015
• Subjects: flexible job shop scheduling problem; genetic algorithm; multiogjective; pareto optimal; setup operator
• ISSN: 1348-5253; 1883-7611
• DOI: 10.4189/shes.13.1

The flexible job-shop scheduling problem has received considerable attention in the production scheduling field. Setup operators are inseparable from the system of a small-volume and widely diverse producing. The work load balance among setup operators is important to increase the satisfaction at work. In this paper, a diversity-oriented multi-objective island based genetic algorithm is proposed to solve the FJSP considering the setup operators as a constraint. The total tardiness of each product and the difference of the setup number between setup operators are regarded as objectives and optimized using the Pareto-based method. When the proposed GA is applied to a test instance with moderate size of the problem complexity, it obtained the two exact solutions within a few seconds. Results show that the proposed algorithm, which was applied to the realistic production planning of the model enterprise, improves diversity of sub-solutions and the local search performance by the introduction of the diversity-oriented evaluation of each island, and emigration strategy among islands.