DECISION METHOD FOR STATES VALIDATION IN DRINKING WATER PLANT MONITORING

This paper proposes a transitions validation method between system functional states in drinking water plant monitoring. The method is based in fuzzy entropy measure. The water plant is monitored by means of fuzzy classification method. The classification algorithm provides the membership degrees of the instant system data (individual) to all possible identified functional states (classes) of the water plant. Usually, the higher membership degree determines the recognized class. However, this criterion leads to uncertainty levels when decision is based on a bad conditioned fuzzy set due to similar membership degrees. The proposed method validates the recognized functional state in presence of uncertainty.