基于MK-SVM算法的大坝位移快速预测模型研究  

Research on Rapid Prediction Model of Dam Displacement Based on MK-SVM Algorithm

作　　者：陈锦 王琛 Chen Jin;Wang Chen(Hunan No.1 Engineering Co.,Ltd.,Changsha 410001,Hunan;Qingzhou Water Engineering Construction Limited Company,Qingzhou 262500,Shandong)

机构地区：[1]湖南省第一工程有限公司,湖南长沙410001 [2]青州水建工程建设有限公司,山东青州262500

出　　处：《陕西水利》2024年第8期21-24,共4页Shaanxi Water Resources

摘　　要：大坝的位移趋势对其安全运行十分重要,而在非常规条件下快速得到大坝的预测模型的研究则相对较少。为使管理人员可以高效地掌握大坝位移的变化性态,基于多核支持向量机(Multiple-kernels support vector machine,MK-SVM)算法构建了一种快速实现大坝位移预测的模型,并讨论该模型参数的范围,利用相关评价指标计算度量模型的精度。由工程案例可知,该模型预测大坝变形的变化趋势与实测变形基本一致,模型预测精度可以满足工程运行需要。结论可为大坝管理人员提供有益借鉴。The displacement trend of the dam is very important for its safe operation,but there are relatively few studies on the rapid prediction model of the dam under unconventional conditions.In order to enable managers to efficiently grasp the changing behavior of dam displacement,a model for rapid prediction of dam displacement is constructed based on the multiple-kernels support vector machine(MK-SVM)algorithm.The range of model parameters is discussed,and the accuracy of the model is calculated by using relevant evaluation indicators.It can be seen from the engineering case that the trend of dam deformation predicted by the model is basically consistent with the measured deformation,and the prediction accuracy of the model can meet the needs of engineering operation.The results can provide useful reference for dam managers.

关 键 词：大坝 位移 快速预测 MK-SVM算法 

分 类 号：TV689.11[水利工程—水利水电工程]