A prototype knowledge-based simulation support system

Publisher Website
Google Scholar
Abstract
As a preliminary step toward the goal of an intelligent automated system for simulation modeling support, we explore the feasibility of the overall concept by generating and testing a prototypical framework. A prototype knowledge-based computer system was developed to support a senior level course in industrial engineer ing so that the overall feasibility of an expert simulation support system could be studied in a controlled and observable setting. The system behavior mimics the diagnostic (intelligent) process performed by the course instructor and teaching assistants, find ing logical errors in INSIGHT simulation models and recom mending appropriate corrective measures. The system was pro grammed in a non-procedural language (PROLOG) and designed to run interactively with students working on course homework and projects. The knowledge-based structure supports intelligent behavior, providing its users with access to an evolving accumula tion of expert diagnostic knowledge. The non-procedural ap proach facilitates the maintenance of the system and helps merge the roles of expert and knowledge engineer by allowing new knowledge to be easily incorporated without regard to the ex isting flow of control. The background, features and design of the system are described and preliminary results are reported. Initial success is judged to demonstrate the utility of the reported approach and support the ultimate goal of an intelligent model ing system which can support simulation modelers outside the classroom environment. Finally, future extensions are suggested.

This publication has 1 reference indexed in Scilit: