基于随机森林的支持向量机混凝土抗渗性预测模型研究  

Prediction of Impermeability of Concrete Structures by Support Vector Machines Based on Random Forest

作　　者：李铁军[1] 胡毅 刘茜[3] 吴贤国[3] 王雷[3] LI Tie-jun;HU Yi;LIU Xi;WU Xian-guo;WANG Lei(China Communications Construction Co.,Ltd,Beijing 100088,China;China Communications Construction Group Second Aviation Engineering CO.,Ltd.,Qingdao 266071,China;School of Civil and Hydraulic Engineering,Huazhong University of Science and Technology,Wuhan 430074,China)

机构地区：[1]中国交通建设股份有限公司,北京100088 [2]中国交通建设集团第二航务工程有限公司,山东青岛266071 [3]华中科技大学土木与水利工程学院,湖北武汉430074

出　　处：《水电能源科学》2021年第2期119-122,118,共5页Water Resources and Power

基　　金：国家自然科学基金项目(51378235,71571078,51308240);国家重点研发计划(2016YFC0800208);教育部博士后基金(2015M570645);湖北省自然科学基金重点项目(zrz2014000104);武汉市建委科技项目(201414,201619,201334)。

摘　　要：为了快速有效预测混凝土的耐久性,引入随机森林并结合支持向量机(RF-SVM)算法,从混凝土材料配比层面选择了影响混凝土抗渗性的16种因素,以抗氯离子扩散系数作为混凝土抗渗性评价指标,构建了基于RF-SVM的混凝土抗渗性预测模型,预测了松通桥梁项目工程混凝土抗渗性,同时与其他预测模型进行对比,分别得出均方根误差和拟合优度。结果表明,该模型回归拟合效果更佳,预测结果精度更高。In order to predict the durability of concrete quickly and effectively,random forest combined with support vector machine(RF-SVM)algorithm was introduced.From the level of concrete material ratio,16 factors that affect the impermeability of concrete were selected,and the anti-chloride ion diffusion coefficient was used as the evaluation in-dex of concrete impermeability,and a concrete impermeability prediction model based on RF-SVM was constructed.The concrete impermeability of the Songtong Bridge Project was predicted.Compared with other prediction models,the root mean square error and goodness of fit were obtained respectively.The results show that the regression fitting effect of this model is better,and the prediction accuracy is higher.

关 键 词：混凝土 抗渗性 预测 随机森林 支持向量机 

分 类 号：TU528.2[建筑科学—建筑技术科学]