KNOWLEDGE-BASED SYSTEM-DEVELOPMENT TOOL SELECTION FOR PROCESS-CONTROL

This paper presents an overview of an established artificial intellige nce (AI) technology, as are traditional knowledge-based systems (KBSs) , reviews their features, and attempts to put them in perspective as t o their demonstrated and potential usefulness, apparent limitations, a nd main areas of application, with an emphasis on real-time process co ntrol. The focus is an eminently applied one, thought to be of use to the practicing engineer interested in formulating a set of intelligent criteria for evaluating competing alternatives among KBS development tools; as such, relevant theoretical issues, while briefly discussed, are not addressed in depth, nor are black-box alternatives to tradiona l expert systems, such as artificial neural networks (ANNs) or genetic algorithms (GAs).