混合案例推理机制在变型设计中的应用  

Application of Hybrid Case Reasoning in Variant Design

作　　者：李海庆[1] 殷国富[1] 方辉[1] 

机构地区：[1]四川大学制造科学与工程学院,四川成都610065

出　　处：《四川大学学报（工程科学版）》2007年第2期164-169,共6页Journal of Sichuan University (Engineering Science Edition)

基　　金：国家自然科学基金资助项目(50575153);高等学校博士学科点专项科研基金(20060610033)

摘　　要：为了将智能技术有效地应用在CAD设计领域,提出了一种支持基于特征的产品变型设计和改写知识自学习的混合案例推理机制。该混合机制以CBR为基础,综合了模糊集理论和SCBR(secondary-CBR)技术。其中模糊集理论被用来进行相似性匹配从而降低噪音在案例检索时的干扰,SCBR技术用来实施对案例的修改和重用。应用实例表明,在变型设计中使用这种混合机制可以为设计人员提供有效的设计决策并降低其对相关领域知识的依赖程度。To apply the intelligence technology in CAD design domain effectively, a hybrid case reasoning mechanism, which supports product variant design based feature and the self-learning of adaptation knowledge, was proposed. The mechanism was based on CBR and combined Fuzzy set theory and SCBR（secondary-CBR） together. Fuzzy set theory was employed to perform approximate matching and thus to reduce potential noise in case retrieval, and the SCBR technology was used to implement the adaptation and reuse for case. The application result showed that the hybrid mechanism could provide the designers with effective design decision and reduce the requirement of correlative domain knowledge for them.

关 键 词：混合案例推理 变型设计 CBR 模糊集 SCBR 

分 类 号：TP391.72[自动化与计算机技术—计算机应用技术]