The reliable design of one-piece flow production system using fuzzy ant colony optimization

In this work, one-piece flow production system is designed with the purpose of ensuring just-in-time production. Three approaches are applied to achieve the goal: adopting straightforward schedule policies, relaxing the Takt time and decreasing the risk of machine failures and operator mistakes. Consequently, a multi-objective design model is proposed, whose aim is to minimize cycle time, changeover count, cell load variation and the number of cells and maximize the extent to which items are completed in a cell. The fuzzy ant colony optimization (FACO) is also presented to solve the formulated problem. In FACO, the fuzzy logic controller (FLC) is used to adapt the evaporated and deposited value of pheromone trail based on the ant's fitness and pheromone trail age. Furthermore, domain knowledge of facility layout, generated based on the travel chart method, is also adaptively injected to improve the performance of FACO. The proposed method is evaluated with the real-world data and experimental results demonstrate that our method outperforms many other methods in efficiency, solution quality and facilitation measures.