灰色动态聚类法在大坝监测中的应用

Application of Grey Dynamic Clustering Method to Dam Monitoring

灰色动态聚类法采用"段"数据作为基本研究对象有别于常规聚类方法中的"点"数据,利用该方法对面板堆石坝面板挠度变形监测数据开展研究,可确定关键的面板变形段,需要重点监测和分析,同时结合matlab软件建立灰色模型GM(1,1)和多元回归模型对关键段监测数据进行预报拟合,结果表明,对于短期或者残缺的资料,利用灰色动态聚类法处理大坝监测数据具有便捷、高效的优点,模型预报结果拟合度较高.

Grey dynamic clustering method using ＂section＂ data is different from conventional clustering methods used in the ＂point＂ data as a basic research obiect; the method is used to carry out the research on the deformation monitoring data of the deformation of the face slab of the face slab rockfill dam. The determination of key section of the panel deformation sections, it needs to be focused on monitoring and analysis; at the same time, the GM （1,1） model and multiple regression model are established by using MATLAB soft- ware, which is used to forecast the key segment monitoring data. The results show that the grey dynamic clustering method is used to deal with the short term or incomplete dam monitoring data, which has the advantages of convenient, high efficiency and high precision.