Knowledge Based System Fusing Experimental Knowledge & Theoretical Model Methods for Cutting Operation

在构建智能制造系统时，建立一种知识表示体系的形式是非常重要的。在模糊推理与方程、仿真等理论模型方法相结合的基础上，建立了一种以经验补偿模型误差的知识表示系统。应用该方法对车削作业方案中的颤振进行了补偿和预测。首先，提出了用经验补偿模型误差的知识表示和用系统预测颤振稳定边界的概念。在切割作业计划方面，还需要有知识表示系统，该系统有效地采用了多种相互关联的模型。描述了具有层次结构的高级系统的构造，其中高层模型的参数由低层模型决定，即多阶段推理。应用该方法预测了二维和三维切削过程中的振动常数和颤振极限。

It is important to establish a form of the knowledge representation system when we construct an intelligent manufacturing system. A knowledge representation system which compensates for errors of models by experiences has been developed, based on a combination of fuzzy inference and theoretical model methods like equations, simulation and so on. It has been applied in a case study to compensate and to predict for chatter vibration in an operation plan for turning.First, it is described that a conception of the knowledge representation for compensating errors of models by experience and the system to predict for stability boundary of chatter vibration.On cutting operation plan, it is also necessary to have the know ledge representation system which adopts effectively many models which are related mutually. It is described that the construction of the advanced system which has a hierarchical structure in which parameters of higer-level models are decided by lower-level models, so called the multistage inference. It has applied to predict vibration constants, and the limit of chatter vibration on both two and three dimensional cutting.