Wastewater treatment process supervision by means of a fuzzy automaton model

Describes a methodology for the design of a supervisory system applied to a wastewater treatment process. A behavioural model is built by the joint participation of the process expert along with clustering techniques applied to measured signals. A fuzzy automaton becomes an heuristic model of the process under supervision. In the application example real data for an activated sludge wastewater plant was used. The automaton states are identified by the expert as significant operation situations and the transitions are generated from input measured variables. Thus, this model reflects the dynamics of the continuous system underneath