摘要
为准确预测混凝土的碳化深度,开展了不同水灰比、粉煤灰掺量、矿渣掺量混凝土的制备与碳化深度测试,进行了数据采集。根据数据及BP算法,建立了3-7-1型三层BP网络,包含三因子网络输入量(水灰比、粉煤灰掺量、矿渣掺量)及单因子网络输出(碳化深度),提出了基于人工神经网络的混凝土碳化深度预测模型。采用最小二乘法建立了线性及伪线性两种预测模型与人工神经网络预测模型进行对比。结果显示:基于BP神经网络建立的混凝土碳化深度预测模型,相比较于常用的最小二乘法线性、伪线性模型更适用于多因素影响条件下的混凝土碳化深度预测,误差仅为线性模型的63.6%,伪线性模型的61.9%,采用BP神经网络能达到理想的预测结果。
In order to predict the carbonation depth of concrete accurately,concrete preparation and carbonation depth test with different water-cement ratio,content of fly ash and content of slag were carried out.Based on the data of carbonation depth,a 3-7-1 three-layer BP network was established,including three factor network input(water-cement ratio,fly ash content,slag content)and single factor network output(carbonation depth).Finally,a prediction model of concrete carbonation depth based on artificial neural network was proposed.In addition,linear and pseudo-linear prediction models are established by least square method and compared with the artificial neural network prediction model.The results show that the prediction accuracy of the pseudo-linear model is better than that of the linear model,and the prediction accuracy of the BP neural network model is better than that of the least square model,which indicates that the established neural network model is more accurate to predict the carbonation depth of concrete.
作者
徐飞
张凯
陈正
陈犇
XU Fei;ZHANG Kai;CHEN Zheng;CHEN Ben(Nanning Rail Transit Group Co.,Ltd.,Nanning 530029,China;Guangxi Guiwu Electromechanical Group Co.,Ltd,Nanning 530000,China;Key Laboratory of Engineering Disaster Prevention and Structural Safety of Ministry of Education College,Civil Engineering and Architecture,Guangxi University,Nanning 530004,China;Guangxi Key Laboratory of Disaster Prevention and Engineering Safety,Guangxi University,Nanning 530004,China)
出处
《混凝土》
CAS
北大核心
2022年第5期57-60,共4页
Concrete
基金
国家自然科学基金联合重点项目(NSFC U2006224)。
关键词
混凝土
碳化深度
BP人工神经网络
最小二乘法
预测模型
concrete
carbonation depth
BP artificial neural network
least square method
prediction model