基于灰色关联支持向量机的半柔性路面灌注性水泥砂浆强度预测  

Predicting strength of poured cement mortar in semi-flexible pavement based on grey relational-support vector machine

作　　者：朱文邦 温勇[1] 马蕾 楚文涛 李承诺 生秋平 ZHU Wenbang;WEN Yong;MA Lei;CHU Wentao;LI Chengnuo;SHENG Qiuping(School of Architecture and Civil Engineering,Xinjiang University,Urumqi 830047,China)

机构地区：[1]新疆大学建筑工程学院,新疆乌鲁木齐830047

出　　处：《混凝土》2021年第11期126-129,共4页Concrete

基　　金：国家自然科学基金(51768068)。

摘　　要：针对半柔性路面灌注性水泥砂浆的强度受多种影响因素共同耦合作用,传统预测模型存在固有的缺陷,不能对小样本、非线性问题进行很好预测,提出灰色关联支持向量机模型,利用灰色关联分析对半柔性路面灌注性水泥砂浆的各影响因素进行筛选,将筛选后保留的影响因素作为支撑向量机的输入变量,构建灰色关联支持向量机的半柔性路面灌注性水泥砂浆强度预测模型。通过实例验证,并同BP神经网络和GM(1,1)2种预测模型的结果进行对比分析。结果表明:基于灰色关联支持向量机建立的预测模具有较好的操作性和预测精度,为提高半柔性路面灌注性水泥砂浆强度预测精度开拓新的途径。As the strength of semi-flexible pavement cement mortar is affected by a variety of factors,the traditional prediction model has inherent shortcomings and cannot predict small samples and nonlinear problems well,it proposes a grey relational-support vector machine model.The gray correlation analysis was used to screen the influencing factors of the semi-flexible pourable cement mortar.The influencing factors retained after the screening were used as the input variables of support vector machines,a grey correlation-support vector machine was used to construct a semi-flexible pavement cement mortar strength prediction model.It is verified by an example and compared with the results of BP neural network and GM(1,1)prediction models.The results show that the prediction model based on the grey relational-support vector machine has better operability and prediction accuracy,and opens a new way to improve the prediction accuracy of the cement paste strength of semi-flexible pavement.

关 键 词：半柔性路面 水泥砂浆强度 灰色关联分析 支持向量机 

分 类 号：TU528.01[建筑科学—建筑技术科学]