Abstract
Recent advances in computer technology and information and communication technology are realizing highly-distributed manufacturing systems (HDMSs) where each machine is computerized and can communicate with other machines. For this type of manufacturing system, a distributed method of discrete event simulation and that of production scheduling were proposed, respectively. Because these methods have the same architecture based on common distributed sequencing algorithms for HDMSs, they can be integrated and the integration makes it possible to generate a better operation plan efficiently. Considering this point, an integrated method of the distributed scheduling and simulation for HDMSs was developed. This paper discusses enhancement of this method. In the integrated method, several dispatching rules for job selection and machine selection are utilized and the weights of the rules are optimized using simulated annealing by an external monitoring system in a centralized manner. This optimization process is improved so that it is performed in a decentralized manner by each machine with synchronization by the monitoring system. The optimization process is enhanced further so that not only non-delay schedules but active schedules can be searched. The effectiveness of this enhanced method was shown by numerical experiments.
