摘要
采用定量体视学图像分析法,测定掺有9种引气剂的混凝土28d孔结构参数,运用人工神经网络理论,建立由总孔隙率和6个不同孔径范围的分级孔隙率预测混凝土28d抗压强度的BP神经网络模型,该模型预测的平均误差只有6.7%。结果表明:混凝土28d抗压强度的预测值与试验值符合良好,较传统回归分析,预测的准确度有所提高,体现其优越性。
air void structure parameters of 28 d concrete were measured by image analysis method. A model for predicting the compressive strength of concrete is established through total porosity and 6 graded porosity based on BP neural networks. The average error of the model is only 6. 7%. The results show that the predicted value of compressive strength is in good agreement with the experimental value. The prediction model is superior due to its small predicting error compared with the traditional regression analysis.
出处
《施工技术》
CAS
北大核心
2017年第S1期292-295,共4页
Construction Technology
关键词
混凝土
孔结构
强度
BP神经网络
concrete
air void structure
strength
BP neural networks
