Distributed-intelligence approaches for weighted just-in-time production

This paper considers the dynamic flexible-flow shop scheduling problem in which the manufacturing environment consists of multiple stages with multiple machines at each stage. All jobs must go through all stages in sequence in order to become a product, and the processing time for each product on each machine is different. There exists a delivery time between machines of neighboring stages, and the release time of each job is unknown, which means that a new job may be released into the system at any time. The objective of this work is to minimize the total earliness and tardiness penalties of all jobs, or to achieve just-in-time production. Previous researches did much on the static scheduling of such problem with different objectives, whereas little effort has been made on dynamic scheduling for such problem, which is more difficult than the static problem but becomes more and more important under the increasingly competitive manufacturing industry. Therefore, to address such need, efforts are been made on dynamic scheduling and several distributed-intelligence approaches are proposed in this paper. Featured by concurrent computing, the distributed-intelligence approaches have been tested on two different manufacturing environments, whose result tells that the proposed approaches deliver competitive performance for the targeted problem.