摘要
混杂纤维混凝土受冻融作用后的损伤值受多种因素的影响,现阶段难以建立各影响因素与损伤值之间的数学模型。通过神经网络的自适应、自学习和非线性映射,可以建立以影响因素为输入变量、以损伤值为输出变量之间的非线性关系。采用相关试验数据,基于MATLAB软件建立AdaBoost-BP和BP神经网络预测模型,利用这两种预测模型对混杂纤维混凝土受冻融作用后的损伤值进行了预测,并将各自的预测值和实测值进行了对比分析。结果表明:AdaBoost-BP神经网络的预测精度较BP神经网络的预测精度更高,该模型为工程上研究混杂纤维混凝土受冻融循环损伤后的损伤程度提供了新方法。
The damage value of HFRC after freeze-thaw is affected by many factors,so it is difficult to establish the mathematical model between each factor and damage value at present.Through the self-adaptive,self-learning and nonlinear mapping of neural network,the nonlinear relationship between the influencing factors as input variables and the damage value as output variables can be established.Based on the relevant test data and MATLAB software,the prediction models of AdaBoost-BP and BP neural network are established.The damage value of hybrid fiber concrete after freeze-thaw is predicted by using these two prediction models,and the predicted value and the measured value are compared and analyzed.The results show that the prediction accuracy of AdaBoost-BP neural network is higher than that of BP neural network,which provides a new method for engineering research on the damage degree of hybrid fiber concrete after freeze-thaw cycle damage.
作者
郭少龙
赵丽红
柳晓科
刘有志
Guo Shaolong;Zhao Lihong;Liu Xiaoke;Liu Youzhi(School of Science,China Tianjin Chengjian University,Tianjin 300384,China;College of Water Conservancy and Hydropower Engineering,Hohai University,Nanjing 210098,China;Tianjin Hengde Labor Service Co.,Ltd.,Tianjin 300350,China;North China Municipal Engineering Design and Research Institute Co.,Ltd.,Tianjin 300381,China;Tianjin Yuanxu Engineering Consultation and Management Co.,Ltd.,Tianjin 300191,China)
出处
《山西建筑》
2021年第12期84-86,共3页
Shanxi Architecture
基金
天津市技术创新引导专项(基金)(20YDTPJC00180)。