Unrelated Parallel Machine Scheduling Problem with Sequence Dependent Setup Times and Release Dates and Machines Eligibility with Minimization of Total Tardiness and Earliness

Please enable javascript in your browser.

Feizi, Yasser | 2013

693 Viewed

Type of Document: M.Sc. Thesis
Language: Farsi
Document No: 45687 (01)
University: Sharif University of Technology
Department: Industrial Engineering
Advisor(s): Salmasi, Naser
Abstract:
This research deals with the Unrelated Parallel Machine Scheduling (UPMS) problem with sequence dependent setup times and release dates and machines eligibility. The studied objective is the minimization of total tardiness and earliness.The due dates and release times of the jobs are distinct which complicates the problem. Two mixed linear integer programming models proposed to solve the research problem optimally in small size instances.The problem is shown to be NP-hard in the strong sense.As it is not possible to solve medium and large size instances in approporaite time, a metaheuristic algorithm based on tabu search developed. Several versions of algorithm are obtained based on different neighborhood searchs and tabu strategies. We develop a benchmark of small, medium and large instances. After solving all benchmark instances we used the design of experiment (DOE) approach to statistical analysis. With ANOVA analysis and tukey test, we were able to identify the best scenario between several versions of proposed tabu search metaheuristicwhich it is insert-swap semi-random reduced neighborhood search
Keywords:
Preparation Time ; Sequence Dependent Setup ; Machine Eligibility Constraint ; Unrelated Parallel Machines ; Total Tardiness and Earlines

No TOC