ESTABLISHMENT OF A CEMENT QUALITY PREDICTIVE SYSTEM─ANALYSES OF THE RELATIONSHIPS BETWEEN CHARACTERISTICS AND PROPERTIES OF CEMENT─

セメント品質予測システムの構築─セメントのキャラクターとプロパティーの関係解析─

A cement quality predictive system that utilizes neural network was developed. This system predicts properties from characteristics by using historical data of cement. The system verification was made by using a simulated function, which demonstrated the validity and the operational directionality of this system. Thus, when cement quality data including X-ray diffraction (XRD)/Rietveld analysis results were applied to this system, the results generally coincided with the previous studies and the accuracy was improved compared to that of a conventional method. Continuous operations and adding more data to this system will contribute to the stabilization of cement quality, the increase in the usage of recycle materials and the optimization of cement manufacturing process.ニューラルネットワークを利用したセメント品質予測システムを開発した。本システムは既往のデータを用いて、セメントのキャラクターからプロパティーを予測する。模擬的な関数を用いた検証により、本システムの有効性と実運用の方向性が示された。XRD/リートベルト法分析結果等のセメント品質データを本システムに適用することで、既往の知見と一致する傾向の予測結果、及び、従来手法同等以上の予測精度が得られた。本システムの継続的運用及び更なるデータ追加により、これまで以上の品質の安定化、リサイクル原単位向上、及び製造プロセスの最適化が可能となる。

• Record URL:
  • https://doi.org/10.14250/cement.66.87
• Record URL:
  • https://www.jstage.jst.go.jp/article/cement/66/1/66_87/_pdf
• 
• Summary URL:
  • http://scholar.google.com/scholar_lookup?title=ESTABLISHMENT+OF+A+CEMENT+QUALITY+PREDICTIVE+SYSTEM%e2%1d%acANALYSES+OF+THE+RELATIONSHIPS+BETWEEN+CHARACTERISTICS+AND+PROPERTIES+OF+CEMENT%e2%1d%ac&author=M.+OHNO&author=D.+KUROKAWA&author=H.+HIRAO&publication_year=2012
• Availability:
  • Find a library where document is available. Order URL: http://worldcat.org/issn/09163182
• Authors:
  • OHNO, Maiko
  • KUROKAWA, Daisuke
  • HIRAO, Hiroshi
• Publication Date: 2012

Language

• English
• Japanese

Media Info

• Media Type: Digital/other
• Features: Figures; References; Tables;
• Pagination: pp 87-94
• Serial:
  • Cement Science and Concrete Technology
  • Volume: 66
  • Issue Number: 1
  • Publisher: Japan Cement Association
  • ISSN: 0916-3182
  • EISSN: 2187-3313
  • Serial URL: https://www.jstage.jst.go.jp/browse/cement/-char/en
• Publication flags:

  Open Access (libre)

Subject/Index Terms

• TRT Terms: Cement; Clinkers; Data collection; Neural networks; Optimization; Quality control; Recycled materials
• Uncontrolled Terms: X-ray diffraction
• Subject Areas: Data and Information Technology; Highways; Materials; I32: Concrete;

Filing Info

• Accession Number: 01522873
• Record Type: Publication
• Source Agency: Japan Science and Technology Agency (JST)
• Files: TRIS, JSTAGE
• Created Date: Apr 24 2014 11:44AM