A Scheduling Algorithm for Distributed Manufacturing with Fixed Output Constraints

Abstract

The customized products such as electromechanical prototype products are a type of product with trial manufacturing characteristics, and the constraints between tasks are more complex than traditional job shop. In this case, the production of one product involves the question of how multiple geographically located machine cells work together to produce it, and not only the coordinated transport time of multiple machines, but also the fact that the machine environment has a designated output machine, should be considered. In practice, this is a kind of distribution manufacturing problem and has become an important role in the integration of industrial clusters. In response to the question, the RSS (Reverse Scheduling Strategy) is proposed to resolve the fixed output issue in such a kind of distributed manufacturing problem. The RSS allocates the tasks to the machines in reverse order, namely, from the ending task to the beginning task. In this manner, the output machine can process the ending task to guarantee the Fixed Output Constraint. In the scheduling process, we utilize the MPS (Measure of Production Structure) and MDM (Machine Distribution Metric) to determine the allocation of the tasks. The MPS reflects the impact of the successive tasks on the schedulable tasks, and the MDM represents the influence of the machine on the distributed machine network. Therefore, the proposed RSS utilizing the MPS and MDM to determine the allocation considers the optimization on the 2 aspects of production and distribution. The parameters of the MPS and MDM are demonstrated by experimental analysis. Finally, we performed experiments to verify the effectiveness and rationality of the proposed RSS.