Intelligent Supervisory Control of Continuous Processes exposed to Faults:: a practical approach

In order to carry out Supervisory Control in continuous industrial processes exposed to faults it is required to provide intelligence to the super-vision mechanism, to cope with the identification of a diversity of faults, the origin of them and to be able to anticipate the consequences derived by their occurrence. In this article we propose to extend the idea of using of multiresolutional models [1] to construct a framework for supervisory control based upon a fuzzy and agent event detection approach. The suggested approach is validated using discrete-event simulation techniques applied to a hydro-pneumatic system.