Domain Knowledge and Decision Time: A Framework for Soft Computing Applications

Abstract

We analyze the issue of decision-making using soft computing (SC) models. We define a natural framework in the cross product of the decision's time horizon and the type of domain knowledge used by the SC models. Within this framework, we analyze the progression from simple lexicon to annotated lexicon, morphology, syntax, semantics, and pragmatics. We compare this progression with the injection of domain knowledge in SC to perform tasks in the context of prognostics & health management (PHM), such as anomaly detection and identification (unsupervised clustering), failure mode analysis (supervised learning), prognostics of remaining useful life (prediction), on-board fault accommodation (realtime control), and off board logistics actions (decision support). Finally, we analyze evolutionary fuzzy systems (EFS) and determine their position and role in this framework