华南理工大学学报(自然科学版) ›› 2005, Vol. 33 ›› Issue (12): 61-65.

• • 上一篇    下一篇

基于覆盖性和适应度的CBR实例库维护

汤廷孝1 刘勇2 黄翔1 廖文和2   

  1. 1.南京航空航天大学 机电学院,江苏 南京 210016;2.成都飞机工业(集团)有限责任公司,四川 成都 610092
  • 收稿日期:2005-02-25 出版日期:2005-12-25 发布日期:2005-12-25
  • 通信作者: 汤廷孝(1976-),男,博士生,主要从事数字化设计、模具CAD等研究 E-mail:tomhn2001@yahoo.com.cn
  • 作者简介:汤廷孝(1976-),男,博士生,主要从事数字化设计、模具CAD等研究
  • 基金资助:

    国家“863”高技术研究发展计划项目(2002A-A411030)

Case Base M aintenance of CBR System Based on Case Cover Capability and Case Fitness Degree

Tang Ting-xiao1  Liu Yong2  Huang Xiang1  Liao Wen-he2   

  1. 1.College of Mechanical and Electrical Engineering,Nanjing Univ.of Aeronautics and Astronautics,Nanjing 210016,Jiangsu,China;2.Chengdu Airplane Industrial(Group)Corporation,Chengdu 610092,Sichuan,China
  • Received:2005-02-25 Online:2005-12-25 Published:2005-12-25
  • Contact: 汤廷孝(1976-),男,博士生,主要从事数字化设计、模具CAD等研究 E-mail:tomhn2001@yahoo.com.cn
  • About author:汤廷孝(1976-),男,博士生,主要从事数字化设计、模具CAD等研究
  • Supported by:

    国家“863”高技术研究发展计划项目(2002A-A411030)

摘要: 针对实例推理(CBR)系统中实例无限增加会导致实例库冗余和实例检索效率降低,从而使系统总体性能下降的问题,对现有的实例库维护方法进行了分析,给出了基于实例覆盖性的实例保存策略,以此来限制实例库的无限扩大,进而提出以实例适应度函数对实例在实例库中的存在性进行描述,精简实例库中适应性差的实例,从而优化了实例库的存储结构,提高了系统性能.最后以冲裁模具实例库的维护为例,验证了上述方法的合理性和有效性.

关键词: 实例推理, 实例库维护, 实例覆盖性, 实例适应度, 冲裁模具

Abstract:

In a Case-Based Reasoning(CBR)system,the case learning makes the case base expand quickly,which results in the redundancy of the case base and the decrease of the searching eficiency.Thus,the system function is weakened.To SO]Ve this problem.this paper analyzes the traditional methods of case base maintenance and proposes a case-saving strategy based on case cover capability to limit the infinite expanding of the case base.Moreover,a case fitness function is defined to describe the subsistence of a case in the base,by which some cases with small fitness degree aye deleted.Therefore,the storage structure of the case base is optimized and the system performance is improved.A CAD system of punch die design is finally presented to verify the correctness and the rationality of the proposed method.

Key words: case-based reasoning, case base maintenance, case cover capability, case fitness degree, punch die