基于有序样品聚类最优二分割算法的滑坡演化阶段划分  

Evolution Stages Division of Landslide Based on the Optimal Segmentation Algorithm of Orderly Sample Cluster

作　　者：黄丽[1,2] 樊孝菊[1] 罗文强[2] 

机构地区：[1]湖北文理学院数学与计算机科学学院,湖北襄阳441053 [2]中国地质大学(武汉)数学与物理学院,湖北武汉430074

出　　处：《湖北文理学院学报》2015年第2期13-16,共4页Journal of Hubei University of Arts and Science

基　　金：国家重点基础发展计划资助973计划(2011CB710605);国家自然科学基金重点项目(41230637)

摘　　要：利用有序样品聚类的最优二分割算法,对新滩滑坡检测累计位移数据进行有序聚类划分,再结合加速度变化特征对滑坡演化阶段进行划分.结果显示新滩滑坡初始变形阶段和中等变形阶段划分点为1979年8月,中等变形阶段和加速变形阶段划分点为1982年9月,而加速变形阶段中初等加速与中等加速的分界点为1983年7月,中等加速与临滑阶段的分界点为1984年10月,与实际监测基本吻合.表明根据有序样品聚类最优二分割算法和滑坡变形加速度变化特征可以对滑坡演化阶段进行准确划分,并为滑坡预警预报提供依据.The deformation of landslide from the beginning to the end is divided into three stages, the initial deformation, constant speed deformation and accelerated deformation. The displacement and acceleration of landslide at different stages have different characteristics. The optimal segmentation algorithm of orderly sample cluster is used to divide the stages of Xintan landslide in this paper, August, 1979 is the split point between initial deformation and constant speed deformation, September, 1982 is the split point between constant speed deformation and accelerated deformation, July, 1983 is the split point between initial accelerated deformation and constant accelerated deformation, October 1984 is the split point between constant accelerated deformation and critical sliding stage. The results show that this method can correctly divide the stages of landslide accurately and provide powerful criterion for landslide warning and forecast.

关 键 词：有序样品聚类 最优二分割算法 滑坡演化阶段划分 新滩滑坡 

分 类 号：P642.22[天文地球—工程地质学] O212.4[天文地球—地质矿产勘探]