Abstract: It is time-consuming to determine appropriate machining parameters in milling process.The paper aims to propose an approach with hybrid technology to establish knowledge base for automation of the determination.Firstly，the relationship among milling parameters is discussed.And then the knowledge contained in these relations is classified into three kinds.As rule based reasoning technology cannot represent the quantitative kind knowledge in an efficient and full-scale way，Aartificial Neural Network（ANN） is brought forward to solve the problem in selection of these parameters.A modified Vogl learning strategy is presented to support the developed ANN model.To demonstrate the performance of the proposed approach，the most complicated patterns from handbook literature are used to train ANN.In the end，some key techniques with the method used to develop a knowledge base system for milling operation are introduced.

Key words: machining parameters, knowledge base, knowledge acquisition, Artificial Neural Network, Rule

摘要： 分析铣削加工参数匹配关系及其知识表示，针对产生式规则难以全面、高效表示加工参数定量匹配知识的问题，提出应用规则推理与人工神经网络（ANN）混合技术构建知识库的方法，给出了参数定量匹配知识表示的神经网络模型和改进的Vogl知识获取方法，运用手册上提供的最复杂样本集数据进行实验验证，结果表明提出的方法具有较好的知识表示性能。最后就如何应用该技术开发面向铣削加工的参数匹配知识库系统展开论述。

关键词: 加工参数, 知识库, 知识获取, ANN, 规则