基于扩展粗糙集的工艺偏好知识发现  

KNOWLEDGE DISCOVERY OF PREFERENTIAL PROCESS BASED ON EXTENTED ROUGH SET MODEL

作　　者：王忠浩[1] 邵新宇[1] 张国军[1] 李培根[1] 

机构地区：[1]华中科技大学教育部智能制造技术重点实验室,武汉430074

出　　处：《机械工程学报》2005年第7期84-89,共6页Journal of Mechanical Engineering

基　　金：国家自然科学基金(50275056)国家863高科技(2003AA411042)资助项目

摘　　要：工艺偏好知识是工艺决策者和工艺设计者关注的重要知识。针对工艺属性复杂多样的特点,采用扩展粗糙集模型,将经典粗糙集模型进行扩充,综合运用等价、相似和偏好关系,即复合关系进行工艺偏好知识的挖掘。接着给出工艺属性的预处理方法、工艺属性的约简算法、工艺偏好规则的表达方法和工艺偏好知识的归纳学习算法。最后以孔加工工艺信息表为例进行实例验证,结果证明所获得的工艺偏好知识是能够直接指导工艺决策者的关键知识,并且可节省大量的基于支持度、可信度和确定度的工艺规则评价环节。The preferential process knowledge is the most important knowledge which the decision-makers and the designers are concerned with. As to the complex characteristics of the process attributes, it utilizes the extended rough set model, extends the classical rough set model used for knowledge discovery and discovers the preferential process knowledge with indiscernibility relation, similarity relation and dominance relation synthetically, i.e. multiple relations. Secondly, it presents some key methods and algorithms, such as the pretrea-tment methods and reduction algorithm for the process attributes, the presentation of the preferential rules and induction algorithm of the preferential knowledge. Finally, an instance is verified based on the process information table of the bore machining. The result proves that the preferential process knowledge discovered based on the extended rough set model is the key one that can guard the decision-makers to carry out decisions directly and efficiently, for it has left out the evaluation procedure of the process rules based on support, confidence and certainty.

关 键 词：粗糙集 工艺设计 知识发现 偏好知识 

分 类 号：TP391.7[自动化与计算机技术—计算机应用技术]