论文部分内容阅读
工艺设计主要是应用生产实践中逐步积累起来的工艺知识来进行创造性的实践活动 ,因此 ,相当一部分知识是经验型的知识 ,用基于规则的知识表达方法以及基于框架的知识表达方法描述工艺设计领域中大量的数值型和字符型经验知识时 ,存在知识量大、表达效率低和专家推理效率低等缺陷 .为此 ,论述了工艺知识两种表示形式的特点 ,提出了采用人工神经网络的隐式表示与规则的显式表示相结合的工艺知识表示法 ,通过神经网络的BP算法的学习功能 ,把工艺知识的外部规则表示转化为内部网络权值表示形式 ,并利用该训练好的网络生成供加工方法推理选择的模糊规则矩阵 .
Therefore, a considerable part of knowledge is experience-based knowledge, which uses rule-based knowledge representation and frame-based knowledge representation to describe the field of process design A large number of numerical and character experience knowledge, there is a large amount of knowledge, the expression of low efficiency and expert reasoning efficiency and other shortcomings.Therefore, the presentation of the two forms of knowledge of process characteristics, the proposed artificial neural network hidden The expression of process knowledge combined with the explicit expression of the rules transforms the external rule representation of process knowledge into the representation form of the weight of the internal network through the learning function of the BP algorithm of the neural network and uses the trained network to generate Fuzzy Rules Matrix for Reasoning and Selection of Processing Methods.