Optimization-based simulation algorithm for predictive-reactive job-shop scheduling of reconfigurable manufacturing systems
Manufacturing industry is now moving forward rapidly towards reconfigurability and reliability to meet the hard-to-predict global business market, especially job-shop production. However, even there is a proper planned schedule for production, and there is also technique for scheduling in Reconfigur...
Saved in:
| 总结: | Manufacturing industry is now moving forward rapidly towards reconfigurability and reliability to meet the hard-to-predict global business market, especially job-shop production. However, even there is a proper planned schedule for production, and there is also technique for scheduling in Reconfigurable Manufacturing System (RMS) but jobshop production will always come out with errors and disruption due to complex and uncertainty happening during the production process, hence fail to fulfill the due-date requirements. This study proposes a generic control strategy for piloting the implementation of a complex scheduling challenge in a RMS. This study is aimed to formulate an optimization-based algorithm with simulation tool to reduce the throughput time of complex RMS, which can comply with complex product allocations and flexible routings of the system. Predictive-reactive strategy was investigated, in which Genetic Algorithm (GA) and dispatching rules were used for predictive scheduling and reactivity controls. This research also provided some results in combining the rule-based simulation with optimization: first, a feasible schedule was computed and then fine-tuned with the rule-based simulation system, then tested with RMS which is the reactive part. Simulation experiments were run using different parameters to analyze the performance of the proposed algorithm with the system. The results showed that the proposed optimizationbased algorithm had successfully reduce the throughput time of the system. In this case, the effectiveness and reliability of RMS is increase by combining the simulation with the optimization algorithm. |
|---|