A Novel Approach to Extract Knowledge from Simulation Results

The extraction of knowledge from simulation results is becoming increasingly important, as numerical simulation is being widely used in the engineering design process. Knowledge extraction systems face challenging problems as the databases of simulation results tend to be dynamic, incomplete, redundant, sparse, and very large. This paper describes a novel approach for handling them. A consistent object-oriented data model for finite-element analysis results has been created using EXPRESS-G, which has facilitated the construction of a database for the knowledge mining procedure. After briefly introducing Rough Sets Theory (RST) and principal component analysis (PCA), this paper investigates the capabilities and implementation of both methods for extracting knowledge from simulation results. The methodology developed has been applied to a real application in sheet metal forming simulation and the results are presented.