A Water Treatment Application for Model-based Decision Support System

Abstract

It lays out some basic ideas for how model-based systems technology can be used in a decision support system for a drinking water treatment plant. The purpose of this system is to discover abnormalities in plant performance, pinpoint potential reasons, and suggest suitable corrective actions. A library of model fragments representing desired processes, disturbances, and potential interventions forms its foundation. As a way to solve the diagnostic problem, these fragments and the observations are used to automatically make possible models of how the damaged plant might behave. A potential therapy idea would be an extension of such a model by models of interventions such that the outcome is consistent with remedial goals. We go into greater detail on how the fundamental reasoning processes can be used to leverage the model and consistency-based problem solving.