| Summary: | Manufacturing systems are dynamic, non-linear and often chaotic environments, subjected to the occurrence of unexpected disturbances that provokes deviations from the initial plans and degrades the system’s performance. Traditionally, disturbance management is performance in a centralized manner, using a fail-fix approach and considering only one type of disturbance, the machine failure. However, the new generation of intelligent manufacturing systems should be capable to treat emergency as earlier as possible and in an effective way, to minimize its negative impact in the production performance. This paper addresses this challenge by introducing a holonic disturbance management architecture based on the ADACOR foundations. The disturbance handling functions are performed in a distributed manner, introducing a predictive dimension to the re-scheduling problem and considering the major types of shop floor disturbances. An experimental implementation was performed and the obtained results showed the applicability of the proposed approach.
|
|---|