陈 爽1,元艳妃2,易 金1,2
摘 要:通过快速碳化试验研究了水灰比、水泥用量、粉煤灰掺量、碳化龄期四个因素对珊瑚混凝土碳化深度的影响,试验龄期分别为3 d、7 d、14 d、28 d。结果表明:珊瑚混凝土的碳化深度与水灰比、粉煤灰掺量呈正比,与水泥用量呈反比;碳化深度均随碳化龄期增长而增大。基于MATLAB中双隐含层BP神经网络,建立了珊瑚混凝土碳化深度预测模型,编写了循环训练算法程序,经过40次循环近百万次网络训练筛选出了双隐含层最佳神经元节点数,分别为11、5,最小均方根误差为0.67。经验证,该预测模型预测平均误差为5.57%,预测精度良好。
Abstract: The effects of water-cement ratio, cement amount, fly ash content, and carbonization age on the coral concrete by using fast carbonation experiment were studied. The experimental ages were 3 d, 7 d, 14 d and 28 d, respectively. The results show that the carbonation depths of coral concrete are positive correlation with water-cement ratio, fly ash content and carbonation age, while negative correlation with the cement content. The carbonation depth increases with the increase of carbonization age. A forecast model of coral concrete carbonation depth was established based on double hidden layer BP neural network in MATLAB. The cyclic-training algorithm program was written, and the optimal neurons number of hidden layer was found to be 11 and 5 after 40 cycles nearly million times network training. It is verified the average prediction error of the prediction model is 5.57%, and the prediction accuracy is good.
陈爽,元艳妃,易金.珊瑚混凝土碳化深度影响因素及其预测模型研究[J].混凝土与水泥制品,2020(11):15-19.
CHEN S,YUAN Y F, YI J.Study on Influencing Factors and Forecasting Model of Coral Concrete Carbonation[J].CHINA CONCRETE AND CEMENT PRODUCTS,2020(11):15-19.
珊瑚混凝土碳化深度影响因素及其预测模型研究
陈 爽1,元艳妃2,易 金1,2
摘 要:通过快速碳化试验研究了水灰比、水泥用量、粉煤灰掺量、碳化龄期四个因素对珊瑚混凝土碳化深度的影响,试验龄期分别为3 d、7 d、14 d、28 d。结果表明:珊瑚混凝土的碳化深度与水灰比、粉煤灰掺量呈正比,与水泥用量呈反比;碳化深度均随碳化龄期增长而增大。基于MATLAB中双隐含层BP神经网络,建立了珊瑚混凝土碳化深度预测模型,编写了循环训练算法程序,经过40次循环近百万次网络训练筛选出了双隐含层最佳神经元节点数,分别为11、5,最小均方根误差为0.67。经验证,该预测模型预测平均误差为5.57%,预测精度良好。
Abstract: The effects of water-cement ratio, cement amount, fly ash content, and carbonization age on the coral concrete by using fast carbonation experiment were studied. The experimental ages were 3 d, 7 d, 14 d and 28 d, respectively. The results show that the carbonation depths of coral concrete are positive correlation with water-cement ratio, fly ash content and carbonation age, while negative correlation with the cement content. The carbonation depth increases with the increase of carbonization age. A forecast model of coral concrete carbonation depth was established based on double hidden layer BP neural network in MATLAB. The cyclic-training algorithm program was written, and the optimal neurons number of hidden layer was found to be 11 and 5 after 40 cycles nearly million times network training. It is verified the average prediction error of the prediction model is 5.57%, and the prediction accuracy is good.
陈爽,元艳妃,易金.珊瑚混凝土碳化深度影响因素及其预测模型研究[J].混凝土与水泥制品,2020(11):15-19.
CHEN S,YUAN Y F, YI J.Study on Influencing Factors and Forecasting Model of Coral Concrete Carbonation[J].CHINA CONCRETE AND CEMENT PRODUCTS,2020(11):15-19.
摘 要:通过快速碳化试验研究了水灰比、水泥用量、粉煤灰掺量、碳化龄期四个因素对珊瑚混凝土碳化深度的影响,试验龄期分别为3 d、7 d、14 d、28 d。结果表明:珊瑚混凝土的碳化深度与水灰比、粉煤灰掺量呈正比,与水泥用量呈反比;碳化深度均随碳化龄期增长而增大。基于MATLAB中双隐含层BP神经网络,建立了珊瑚混凝土碳化深度预测模型,编写了循环训练算法程序,经过40次循环近百万次网络训练筛选出了双隐含层最佳神经元节点数,分别为11、5,最小均方根误差为0.67。经验证,该预测模型预测平均误差为5.57%,预测精度良好。
Abstract: The effects of water-cement ratio, cement amount, fly ash content, and carbonization age on the coral concrete by using fast carbonation experiment were studied. The experimental ages were 3 d, 7 d, 14 d and 28 d, respectively. The results show that the carbonation depths of coral concrete are positive correlation with water-cement ratio, fly ash content and carbonation age, while negative correlation with the cement content. The carbonation depth increases with the increase of carbonization age. A forecast model of coral concrete carbonation depth was established based on double hidden layer BP neural network in MATLAB. The cyclic-training algorithm program was written, and the optimal neurons number of hidden layer was found to be 11 and 5 after 40 cycles nearly million times network training. It is verified the average prediction error of the prediction model is 5.57%, and the prediction accuracy is good.
版权所有:中国混凝土与水泥制品网 苏ICP备10086386号 网站建设:中企动力 苏州