KNOWLEDGE ENGINEERING FOR A PROCESS DESIGN DOMAIN

This paper describes knowledge engineering using a conceptual modellin g tool called the Inferential Modelling Technique, which supports the elicitation and analysis phases of the knowledge acquisition process. The technique is derived from a domain independent categorization of k nowledge types and inferences, called the Inferential Model. To illust rate use of the technique, we apply it in constructing a knowledge bas ed system in the chemical engineering design problem of solvent select ion for carbon dioxide removal processes. We show that the technique f acilitates construction of the conceptual model for the domain. Since the Inferential Model is object-based, translation of the conceptual m odel formulated with the Inferential Modelling Technique into a workin g system is straightforward provided an object-oriented tool is used. Finally, we discuss implementation of the knowledge-based system calle d the Solvent Selection Advisory System using the real-time expert sys tem shell G2.