A Reactive Approach for Reducing the Myopic and Nervous Behaviour of Manufacturing Systems

摘要

Scheduling is a crucial activity for the successful control and piloting of manufacturing activities. Manufacturing systems operate in dynamic environments vulnerable to real-time events, which frequently force a reactive revision of pre-established schedules. In an uncertain environment, it has become preferable to adapt the scheduling from pre-established schedules as the latter may become vastly degraded at unexpected events. For this, a reactive module is often included to update the scheduling in order to assure manufacturing execution. However, due to the need of rapid response, this updating might be executed experiencing myopia and/or nervousness issues. This paper aims to develop a proof of concept of a decision support system for practitioners, attempting to minimize the degradation and lack of information during these scenarios. This approach starts with a metaheuristic technique to generate a predictive schedule for establishing an initial scheduling, named pre-established schedule. Then, during disruptions, it executes a scheduling updating through a reactive module based on a heuristic technique. The proposed approach was tested in a simulated scenario of a real flexible manufacturing system located in Valenciennes (France), called AIP-PRIMECA Valenciennes.