An Emergent Simulation Modeling Approach for Discovery of Knowledge on Phenomena in Chemical Systems

Abstract

A system-based approach is effective in conducting simulation modeling of complex chemical systems. However, traditional centralized or monolithic modeling approaches are restricted to the knowledge domain of the modeler, and therefore these approaches cannot lead to the discovery of potentially important knowledge from other areas. Furthermore, knowledge of important phenomena, which are unfamiliar to the researcher who is designing the chemical system but may be well-known in other domains, are often overlooked in the conventional modeling process, leading to incorrect predictions of the modeled chemical system characteristics. An emergent simulation modeling approach for knowledge discovery (emergent simulation approach) is proposed to evaluate phenomenological behavior in chemical systems. In this emergent simulation approach, knowledge experts who have developed computational models that simulate the phenomena occurring under different design and operation conditions make those models available to a software system. Other engineers and scientists can access the software system to search for models simulating phenomena that match certain conditions and specifications describing a particular chemical system. In order to support the search and match process, an ontological markup language for chemical systems is proposed. Researchers can use this markup language to create a specification of the chemical system that they are investigating. That specification is used to discover which computational models can simulate phenomena that could occur in the given specification. This paper presents 1) the concept of the emergent simulation approach, 2) an ontological markup language developed for the emergent simulation of phenomena in chemical systems, and 3) a design for a software system that supports the proposed emergent simulation approach.