Autonomous Decentralized Scheduling System for Multi-stage Parallel-Unit Production Processes

Abstract

This paper describes an autonomous decentralized scheduling system for the processes consisting of a sequence of production stages, each of which has parallel units. In the proposed system, an individual scheduling system belonging to each production stage gradually generates a plausible schedule by repeating the schedule generation and the data exchange among the production stages. In the process of schedule generation at each stage, the jobs processed at each unit are also decided by exchanging the scheduling data among the units in the stage. The results of example problems show that the schedule generated by the proposed system has almost the same level of performance as the one obtained by the conventional method, despite the fact that in the proposed system only the local information was used to derive the schedule at each unit