基于深度学习的氯盐环境下高性能混凝土氯离子扩散系数的预测  

Prediction of chloride diffusion coefficient of high performance concrete based on depth learning in chloride environment

作　　者：周双喜[1] 盛伟 何顺爱[2] ZHOU Shuangxi;SHENG Wei;HE Shun′ai(East China Jiaotong University,Nanchang 330013,China;Sichuan Academy of Building Sciences,Chengdu 610081,China)

机构地区：[1]华东交通大学,江西南昌330013 [2]四川省建筑科学研究院,四川成都610081

出　　处：《混凝土》2019年第7期27-31,共5页Concrete

基　　金：国家自然科学基金（51662008）;江西省重点研发计划（20171BBG70078）;江西省自然科学基金项目（20181BAB206007）

摘　　要：氯盐的扩散系数是一个控制混凝土内部氯离子迁移状况的关键参数,为了评估氯盐在混凝土中的扩散情况,经常需要对其进行有效的预测.由于时间和成本的限制,实际工程中该系数很难从试验中获得.为了评估氯盐在混凝土中扩散的情况,基于深度学习卷积神经网络方法研究了氯盐在高性能混凝土中的扩散系数.收集了掺粉煤灰(FA)和粒化高炉矿渣(GGBFS)300种不同的数据,针对两种类型的高性能混凝土建立了卷积神经网络模型(CNN).CNN模型包括4个输入参数,分别是W/B比、水泥含量、粉煤灰或矿渣粉掺量和养护龄期,输出参数为氯盐的扩散系数.与传统的神经网络相比,基于深度学习的卷积神经网络预测模型具有高精度、高效率和高泛化能力的优点,可以作为氯盐环境下混凝土氯离子扩散预测的新方法。Chloride diffusion coefficient is a key parameter to control chloride migration in concrete.In order to evaluate the diffusion of chloride in concrete,it is often necessary to predict the chloride transport.Due to the limitation of time and cost,it is difficult to obtain the coefficient from the test in practical engineering.In this study,the diffusion coefficient of chloride salt in high performance concrete is evaluated based on the deep learning convolutional neural network.A total of 300 different data of fly ash(FA)and granulated blast furnace slag(GGBFS)were collected from the literature.A convolutional neural network model(CNN)is established for two types of high performance concrete.The CNN model consists of 4 input parameters,which include W/B ratio,cement content,fly ash or slag powder content and curing age,and the output parameter is the diffusion coefficient of chloride.The results show that compared with the traditional neural network,the prediction model based on depth learning has the advantages of high precision,high efficiency and high generalization ability,and can be used as a new method for predicting chloride diffusion in concrete under chloride environment.

关 键 词：氯离子扩散系数 卷积神经网络 粉煤灰 矿渣粉 预测 

分 类 号：TU528.01[建筑科学—建筑技术科学]