Representation of Strategic Choices in Structural Modeling

Abstract : Structural modeling requires the construction of an appropriate mathematical description to describe the behavior of a physical object. Because conflicting goals and uncertainty permeate the process of structural modeling, structural model derivation is a complicated process. Therefore, to effectively and efficiently model any structure, one must have a method for planning actions, for proceeding in the face of uncertain information, and for dealing with uncertainty. This paper presents a method for representing structural modeling as a strategic process involving decision-making in an environment where the results of any decision may not be known with complete certainty. Specifically, an existing task-level architecture developed to manage uncertainty in treating cardiac disease is adapted to embody the strategic knowledge of a structural engineer in formulating and solving structural problems. The MUMS system focuses on plates as the structure of interest in its pilot implementation. Since the system is concerned with structural modeling at the strategic level (as opposed to a detailed design, for instance), the ideas presented are applicable to modeling any structure. Keywords: Knowledge based systems.