冷挤压工艺设计系统模糊知识表示和推理机制的研究

研究了冷挤压工艺设计领域的知识表示以及冷挤压工艺设计系统的模糊推理机制。利用面向对象表示法、模糊产生式表示法以及模糊框架表示法等知识表示方法,实现了基于模糊的面向对象的异构知识混合表示,该方法利用面向对象的思想构建领域知识的基类,利用模糊产生式规则表示领域知识库中的模糊知识规则。在此基础上构建了冷挤压工艺设计领域知识库。提出了广义可信度的概念和一种基于广义可信度的模糊推理方法,该方法通过对知识库中的知识设置可信度和构建隶属函数,用隶属度来定量表示冷挤压工艺设计领域知识的不确定性,实现领域知识的不确定性匹配,用广义可信度来表示不确定性匹配的程度,并利用不确定性传递算法计算最终结论的可信度值。通过典型实例进行了验证。

Research on Fuzzy Knowledge Representation and Reasoning Mechanism of Cold Extrusion Process Planning System

The knowledge representation methods and fuzzy reasoning mechanism of a cold extrusion process planning system were researched. A kind of fuzzy-based object-oriented knowledge representation method was realized based on object-oriented method, fuzzy-rules and fuzzy-frame knowledge representation methods. The base classes of the dominant knowledge were constructed by object-oriented method, the rules in the dominant knowledge base were represented through fuzzy production rules, and the dominant knowledge base of the cold extrusion process design system was established. The concept of generalized reliability and the generalized reliability-based fuzzy reasoning method were put forward. The fuzziness and uncertainty of the cold extrusion process planning knowleges were represented quantitatively through setting reliability to knowleges and constructing subjection functions, and the uncertainty matching of dominant knowleges was realized. The uncertainty matching degree was represented by using generalized reliability, and the reliability values of the conclusions were calculated through uncertainty transfer algorithm.The research work was validated through a typical cold extruded part.