Predictive disturbance management in manufacturing control systems

The manufacturing systems are dynamic, non-linear and often chaotic environments, subject to the occurrence of unexpected disturbances that leads to deviations from the initial plans and usually degrades the performance of the system. The treatment of exceptions and disturbances is one major requirement to the next generation of intelligent manufacturing control systems, that should be able to treat emergency as a normal situation. In this paper, a predictive disturbance management approach that transforms the traditional “fail and recover” practices into “predict and prevent” practices, improving the control system performance, will be presented. The predictive mechanism is based in the frequency analysis of each type of disturbance to find repetitive patterns in their occurrence