基于SVM的混凝土抗硫酸盐侵蚀系数预测模型
Prediction model of sulfate resistance coefficient of concrete based on SVM
刘亮 1赵越 2任文杰 2李增浩2
作者信息
- 1. 广州地铁设计研究院股份有限公司,广东 广州 510000
- 2. 河北工业大学 土木与交通学院,天津 300401
- 折叠
摘要
硫酸盐侵蚀会对混凝土结构造成严重危害,目前工程上以硫酸盐侵蚀下达到规定干湿循环次数时的混凝土抗硫酸盐侵蚀系数作为评价结构抗硫酸盐侵蚀性能的指标.以试验数据和文献数据为样本数据,以水胶比、粉煤灰取代率、矿粉取代率、砂率、减水剂掺量、硫酸盐浓度和干湿循环次数为输入向量,以混凝土抗硫酸盐侵蚀系数为输出向量,利用支持向量机(SVM)建立了混凝土抗硫酸盐侵蚀系数的预测模型.设计了2种容量的样本集,分别计算了SVM模型的预测误差,结果表明:SVM模型可利用较少数量的训练样本很好地预测混凝土抗硫酸盐侵蚀系数,方便实际工程应用.
Abstract
Sulfate attack will cause serious damage to concrete structure.At present,the sulfate resistance coefficient of concrete subjected to the specified number of dry-wet cycles is used to evaluate the sulfate resistance capacity of struc-tures.The experimental data and the literature data are taken as sample data.Considering water-binder ratio,fly ash re-placement ratio,slag powder replacement ratio,sand ratio,superplasticizer dosage,sulfate concentration and dry-wet cy-cles as input and sulfate resistance coefficient of concrete as output,the prediction model for sulfate resistance coeffi-cient of concrete is established through the support vector machine(SVM)method.Two kinds of sample sets are de-signed,and the prediction errors of SVM model are calculated respectively.Results show that SVM model can attain higher prediction accuracy with a small number of training samples,which is convenient for practical engineering appli-cation.
关键词
混凝土/硫酸盐侵蚀/支持向量机/预测模型Key words
concrete/sulfate attack/support vector machine/prediction model引用本文复制引用
出版年
2024
NETL
NSTL
万方数据