Tire mixing process scheduling using particle swarm optimization. (August 2017)

• Record Type: Journal Article
• Authors: Kim, Hwang Ho; Kim, Do Gyun; Choi, Jin Young; Park, Sang Chul
• Is Part Of: Computers & industrial engineering. Volume 110(2017)
• Page Start: 333
• Page End: 343
• Publication Date: 2017-08
• Subjects: Tire mixing process -- Compound -- Makespan -- Sequence-dependent setup time -- Flexible job shop scheduling problem -- Particle swarm optimization, particle encoding scheme -- Error rate; Engineering -- Data processing -- Periodicals; Industrial engineering -- Periodicals; 620.00285
• DOI: 10.1016/j.cie.2017.06.012
• Languages: English
• ISSNs: 0360-8352

Abstract

Highlights: We propose an efficient scheduling method for a tire mixing process using particle swarm optimization (PSO). We suggest a special particle structure for tire mixing process. We design an efficient particle encoding scheme of PSO for tire mixing process scheduling problem. We develop a method to derive a schedule from a particle in PSO. We show the superiority of the suggested algorithm through numerical experiments. Abstract: The tire manufacturing process consists of four major steps, including mixing, component preparation, building, and curing. Among these, the mixing process is the most important, since it produces a type of synthesized rubber called a compound, which determines various features of the tire and consumes 70% of the overall manufacturing budget. However, despite the importance of mixing, there is little research on this process. Thus, in this study, we consider a scheduling problem for a tire mixing process and suggest an efficient particle swarm optimization (PSO) algorithm for minimizing makespan. Specifically, we design a system in which particle coordination and velocity are used to generate a processing sequence for compounds and machine allocation information, while being updated in successive steps. The superiority of the proposed PSO algorithm is shown with numerical experiments that compare the solutions found by the proposed algorithm with those obtained from a mixed integer linear programming model developed in previous research.