Applying Metaheuristic Algorithms for Output Rate Analysis in Two-Machine Robotic Manufacturing Cells

Abstract

This article analyses the output rate in two-machine flexible robotic manufacturing cells. The flexible CNC machines in this manufacturing cell can process different operations. The manufactured parts in the cell are identical and it is assumed that different operations are required to manufacture each part. Moreover, loading/unloading time of a part by the robot (ε), robot movement time between the machines and input and output areas (σ), and processing time of the jth part on the machines (tj) are considered to be fixed. The main objective of this article is to minimize cycle time in order to increase the output rate of the manufacturing cell. To achieve this goal, it is important to optimally assign operations required for manufacturing a part to each machine and to determine the optimal robot moves sequence. Accordingly, existing feasible movement policies in the cell and their cycle times have been reviewed, and then these policies have been considered in a new machine layout and their cycle times have been calculated based on the new robot moves sequence. Afterwards, a mathematical model has been presented to select optimal cycle time in the manufacturing cell and this model has been solved by a branch and bound exact algorithm; since the mathematical model is non-linear and the optimal solution cannot be obtained, two metaheuristic algorithms—genetic and simulated annealing algorithms—have also been proposed to solve the model and their results have been compared.