基于可拓分类挖掘的产品配置设计  

Product Configuration Design Based on Extension Classification Mining

作　　者：赵燕伟[1] 王欢[1] 郭明[1] 洪欢欢[1] 周建强[1] 

机构地区：[1]浙江工业大学特种设备制造与先进加工技术教育部浙江省重点实验室,杭州310014

出　　处：《机械设计与制造》2012年第6期26-28,共3页Machinery Design & Manufacture

摘　　要：基于实例推理的产品配置设计往往是在整个产品模型上,而非各个性能属性上,匹配出最相似产品并调整相应的方案解;同时,由于客户对产品内部结构、实现原理及使用环境等的不充分认识,性能需求仍存在不确定性和模糊性。鉴于此,基于可拓分类挖掘的产品配置设计方法被提出,其针对每一性能属性进行产品实例划分,当没有产品公共实例存在于各个性能的正域时,则需在结构配置与性能改进的共同约束下,选择适当实例并作相应的调整,并将静、动态分类知识共同存储于关系型数据库中。Generally speaking,product configuration design based on case based reasoning（CBR）is to find out the most similar product and adjust its scheme solution in terms of the entire product model other than each performance.Moreover,owing to an insufficient understanding of internal structure,scheme solution and working environment of products,performance requirements are both uncertain and fuzzy to customers and designers.Taking these into account,product configuration design based on extension classification mining is put forward.That’s to say,classify existing products according to each performance,select and adjust a proper product under constraints of structure configuration and performance improvement when there are no existing products in positive domains for every performance.Finally,static and dynamic classification knowledge are together stored in the relational databases.

关 键 词：实例推理 性能 配置设计 分类挖掘 

分 类 号：TH16[机械工程—机械制造及自动化]