An Autonomous Control Strategy Alternative for Critical Industrial Automated Systems

In automatized industrial systems, inactivity due to unplanned resource shortage or processes failures have a great influence on the system's performance because it generates discontinuities and instabilities. Distributed and autonomous control systems may help to cope with these kinds of problems because of the higher performance, but safety issues and real time constraints must be tightly addressed in these systems because of the risks involved (human, financial and environmental). Thus, an intelligent control system instantiated at the local level to allow each controller to take critical decisions in an autonomous way is proposed here. Above that, a supervisory system manages more complex situations beyond the capabilities of local control. The proposed control aims to enable auto-adjustment in the system to improve performance and to prevent and treat unexpected faults. It therefore has to learn through events which have occurred in the system and environment using Artificial Intelligence tools.