An ant colony optimization algorithm for setup coordination in a two-stage production system

This paper is concerned with the coordination of setup times in a two-stage production system. The problem is derived from a furniture plant, where there are two consecutive departments including cutting and painting departments. Items with the same levels of both attributes are grouped into a single batch in advance. A sequence-dependent setup time is required in a stage when a new batch has a different level of attribute from the previous one. The objective is to minimize the total setup time. In this paper, we first propose a simple dispatching rule called the Least Flexibility with Setups (LFS) rule. The LFS rule can yield a solution better than an existing genetic algorithm while using much less computation time. Using the LFS rule as both the initial solution method and the heuristic desirability, an Ant Colony Optimization (ACO) algorithm is developed to further improve the solution. Computational experiments show that the proposed ACO algorithm is quite effective in finding the near-optimal solution.