A HYBRID AI-BASED APPROACH FOR MECHANICAL SYSTEM-DESIGN PROBLEMS

A hybrid, Al-based system that integrates the knowledge-learning process with the analysis/synthesis process in mechanical system design problems is developed and presented. Unlike the traditional expert systems, where the knowledge engineers are usually responsible for sorting out and setting up rules, this approach includes the experts and the end users in the learning and rule-creating process. Another major contribution is that this novel algorithm couples qualitative expert knowledge with quantitative simulation representations, which is a critical step toward real-world engineering applications.