Scheduling algorithms for a two-machine flexible manufacturing system

Scheduling algorithms for a two-machine flexible manufacturing system

0.00 Avg rating0 Votes
Article ID: iaor19952044
Country: United States
Volume: 7
Issue: 3
Start Page Number: 207
End Page Number: 227
Publication Date: Jul 1995
Journal: International Journal of Flexible Manufacturing Systems
Authors: , ,
Keywords: scheduling
Abstract:

The flexible manufacturing system (FMS) considered in this paper is composed of two CNC machines working in series-a punching machine and a bending machine connected through rollers acting as a buffer system of finite capacity. The main difference between the present problem and the standard two-machine flow shop problem with finite intermediate capacity is precisely the buffer system, which in the present problem consists of two stacks of parts supported by rollers: the first stack contains the output of the punching machine, while the second stack contains the input for the bending machine. When the second stack is empty, the first stack may be moved over. Furthermore, the capacity of each stack depends on the particular part type being processed. The FMS can manufacture a wide range of parts of different types. Processing times on the two machines are usually different so that an unbalance results in their total workload. Furthermore, whenever there is a change of the part type in production, the machines must be properly reset-that is, some tools need to be changed or repositioned. A second important difference between the present problem and the usual two-machine flow shop problem is the objective. Given a list of p part types to be produced in known quantities, the problem considered here is how to sequence or alternate the production of the required part types so as to achieve various hierarchical targets: minimize the makespan (the total time needed to complete production) and, for instance, compress the idle periods of the machine with less workload into a few long enough intervals that could be utilized for maintenance or other reasons. Although Johnson’s rule is optimal in some particular cases, the problem addressed in the paper is NP-hard in general: heuristic procedures are therefore provided.

Reviews

Required fields are marked *. Your email address will not be published.