土石围堰表面变形监测数据自动聚类分析

Automatic Cluster Analysis on Deformation Monitoring of Earth Rockfill Cofferdam Surface

针对传统变形监测方法在土石围堰表面仅从单点位置进行监测,而不能反映土石围堰整体变形情况的不足,本文提出采用基于基因表达式(GEP)的自动聚类方法(GEP-Cluster)。运用GEP在解决聚类问题时强大的全局搜索能力,及在用户不需数据先验知识的状态下仍可在未知簇个数等先验知识的前提下进行自动聚类分析并获得精确的聚类结果。本文通过对某土石围堰进行变形监测获得监测数据,利用GEP-Cluster方法对所获数据进行研究分析,从而预测监测区土石围堰的变形情况,为新建水利工程提供技术参考。

As the traditional method of earth rockfill cofferdam surface deformation monitoring is simply monitored from a single point and does not reflect the lack of overall deformation of the earth rockfill cofferdam,this paper proposes based on gene expression(GEP)of automatic clustering method(GEP-Cluster).Use of GEP in solving the problem of clustering powerful global search capability and GEP-Cluster in the user state data without prior knowledge of the cluster is still unknown number of prior knowledge of the premise automatic clustering analysis and obtain accurate clustering results.This paper based on the deformation monitoring of a cofferdam to obtain monitoring data,GEPCluster method using data obtained from research and analysis,in order to predict the deformation monitoring area cofferdam to provide technical reference for the new water project.