• DocumentCode
    572877
  • Title

    Research on case retention strategy for industrial case-based reasoning (CBR) system: A practical case study

  • Author

    Zhou, Ping ; Yin, Wei-Da ; Zhao, Jin-Hui

  • Author_Institution
    State Key Lab. of Synthetical Autom. for Process Ind., Northeastern Univ., Shenyang, China
  • fYear
    2012
  • fDate
    24-26 Aug. 2012
  • Firstpage
    376
  • Lastpage
    380
  • Abstract
    Case retention is an important unit in developing a case-based reasoning (CBR) system. It not only directly relates to the difficulty of the case base maintenance, but also affects the reasoning efficiency and correctness of the CBR system. However, the case retention is still not attracted enough attention, and the researches on it is very little. In this paper, case retention strategy for the industrial CBR system is studied by taking an industrial soft-sensor CBR system as the research case. According to the different ways of case creation, two different case retention strategies are proposed, respectively. The improved CBR system with the proposed case retention method has been applied in the grinding product particle size soft-sensor CBR system. The successful practical application results have shown the effectiveness and prospect of the proposed method.
  • Keywords
    case-based reasoning; grinding; industrial engineering; particle size; production engineering computing; case base maintenance; case retention strategy; grinding product particle size soft-sensor CBR system; industrial case-based reasoning system; industrial soft-sensor CBR system; Cognition; Databases; Educational institutions; Feeds; case learning; case retention; case-based reasoning (CBR); grinding production particle size; soft-sensor;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Computer Science and Information Processing (CSIP), 2012 International Conference on
  • Conference_Location
    Xi´an, Shaanxi
  • Print_ISBN
    978-1-4673-1410-7
  • Type

    conf

  • DOI
    10.1109/CSIP.2012.6308872
  • Filename
    6308872