Scheduling of Virtual Cellular Manufacturing Systems: A Biogeography-Based Optimization Algorithm. Issue 7 (7th June 2019)

• Record Type: Journal Article
• Authors: Zandieh, M.
• Is Part Of: Applied artificial intelligence. Volume 33:Issue 7(2019)
• Page Start: 594
• Page End: 620
• Publication Date: 2019-06-07
• Subjects: Artificial intelligence -- Periodicals; 006.3
• DOI: 10.1080/08839514.2019.1577021
• Languages: English
• ISSNs: 0883-9514

Abstract

ABSTRACT: Virtual cellular manufacturing system (VCMS) is one of the modern strategies in the production facilities layout, which has attracted considerable attention in recent years. In this system, machines are located in different positions on the shop floor and virtual cells are a logical grouping of machines, jobs, and workers from the viewpoint of the production control system. These features not only enhance the system's agility but also allow a dynamic reassignment of cells as demand changes. This paper addresses the VCMS scheduling problems where the jobs have different orders on machines and the objective is to simultaneously minimize the weighted sum of the makespan and total traveling distance in order to create a balance between criteria. The research methodology firstly consists of a mathematical programming model with regard to the production constraints in order to describe the characteristics of the VCMS. Secondly, a basic genetic algorithm (GA), a biogeography-based optimization (BBO) algorithm, an algorithm based on hybridization of BBO and GA, and the BBO algorithm accompanied by restart phase are developed to solve the VCMS scheduling problems. The developed algorithms have been compared to each other and their performance are evaluated in terms of their best solution and computational time as effectiveness and efficiency criteria, respectively. Consequently, the performance of the best algorithm has been evaluated by the state-of-the-art algorithm, GA, in the literature. The results show that the best algorithm based on BBO could find solutions at least as good as the last famous algorithm, GA, in the literature.