Earticle

초록

영어

Demolition waste generation has exceedingly increased in the world, including korea. It is estimated the approximately 115 thousand tons of waste concrete are occurred in 2009 in this country. And it is estimated that the waste concrete will be more generated more each year than those of today. Therefore, it is necessary to use waste concrete for aggregate of concrete. Many studies have been conducted to reuse of recycled aggregate for construction materials. To make use of recycled aggregate, the elastic modulus of recycled aggregate concrete must bo assured. The purpose of this study is to predict elastic modulus of recycled aggregate concrete using neural network which was applied to predict properties of engineering problems. New learning algorithms for the prediction on the elastic modulus of recycled aggregate concrete is proposed focusing on input layer components and a normalization method for input data and their validity is examined through elastic modulus data of recycled aggregate concrete. In artificial neural network algorithm, the main input data to be trained are the water content, fine and coarse aggregate ratios, curing temperature of recycled aggregate concrete, water/cement ratios and etc. Through this study, it was confirmed that artificial neural network successfully predict the elastic modulus of recycled aggregate concrete.

목차

Abstract
 1. 서론
  1.1 연구배경 및 목적
 2.3 뉴랄-네트워크의 구조
 3. 인공신경망에 의한 기존 순환골재 콘크리트의 탄성계수 학습
 4. 순환골재콘크리트의 탄성계수 예측 및 검증
  4.1 배합 및 사용재료
  4.2 탄성계수 측정방법
  4.3 학습데이타의 검증
 5. 결론
 참고문헌

키워드

저자

• 박선규 [ Park, Sun-Gyu | 목원대학교 건축학부 조교수, 공학박사 ] 교신저자

참고문헌

발행기관

간행물

표지보기 관심저널 등록

• 간행물명

  대한건축학회연합논문집 [Journal of the Regional Association of Architectural Institute of Korea]

• 간기

  격월간

• pISSN

  1229-5752

• 수록기간

  1999~2026

• 등재여부

  KCI 등재

• 십진분류

  KDC 540 DDC 690