运用混合推理的数控加工刀具智能选择的研究

Research of NC machining tools intelligent selection based on hybrid reasoning

实例推理(CBR)与规则推理(RBR)是知识工程中两类重要的基于知识的问题求解和学习方法,结合这两种方法并将其用于刀具的自动选择,有利于机械加工的智能化发展。提出一种将实例推理与规则推理相集成的混合推理方法,首先运用规则推理得出刀具类型及刀具材质,并将其作为初始条件检索出候选实例集,再以实例推理得出与目标实例最相似加工实例。此外,针对传统的基于距离函数的相似度算法的复杂性及不确定性等缺点,基于模糊数学理论应用隶属函数求解实例间属性的相似度。同时考虑了属性权重问题,以层次分析(AHP)法来确定各个属性相对应的权重。最后,以电梯业中磁铁安装板的加工为例,验证了该方法的可行性。

Case-Based Reasoning （CBR） and Rule-Based Reasoning （RBR） are two important knowledge-based problems＇ sol- ving and learning methods in knowledge engineering, and the combination of the two methods for automatic selection of cut tools is benefit to the intelligence development for mechanical processing, proposed a hybrid reasoning that integrated by rule-based rea- soning and case-Based reasoning, first, use rule-based reasoning select tool types and the cutting tool material, and use it as the initial conditions to retrieve and get the instance candidate, and then obtain the instance that most similar to the target problem. Besides, in view of the complexity and uncertainty of the shortcomings of traditional similarity algorithm based on distance func- tion, applied subordinate function that based on the theory of fuzzy mathematics to solving the similarity between instances attrib- ute. And considered the problem of attribute weights, by the method of AHP to determine the corresponding weights for each at- tribute. Weights. Finally, applied this method in the elevator industry＇ s magnet mounting plate processing, verified the feasibility of this method.