Study on Knowledge Expression Methods and Reasoning Strategies in Intelligent CAD System

摘要

As the core of intelligent computer aided design (ICAD), the expert system with the purpose of product design automation and intelligence has been developed rapidly. Knowledge expression, management, utilization, and access method has always been the core of artificial intelligence and expert system technology , Substantial literature both at home and abroad involved in the content and achieved certain results in the area of design. In the aspect of knowledge expression and management, including: the knowledge expression based on gene model[3], the knowledge expression based on design pattern[4], the knowledge expression based on Petri net, mixed knowledge expression, and so on. In the aspect of system reasoning, including: fuzzy reasoning[5,6], cease-based reasoning[7], integration reasoning, and so on. This paper mainly introduces some technologies and methods about electrical products at ICAD system with the above-mentioned two aspects. As a sub-system of ICAD system, the expert system for electrical apparatus designing was studied in the artificial intelligent technologies, including the way to build a large knowledge database managed by multi level frames and expressed by various knowledge expression methods, and the reasoning strategy suitable for the structure of knowledge database and integrated rule-based reasoning (RBR) with case-based reasoning (CBR) together. Based on these technologies, two-level expert system controlled by meta reasoning machine was built so that the reasoning efficiency was boosted very strongly.