摘要
滑坡的演化一般要经历初始变形、等速变形和加速变形3个阶段,准确地划分这3个阶段是进行滑坡预测预报的基础。依据大量滑坡从初始变形到失稳破坏的监测数据,采用数据挖掘中的关联规则,用经典的Apriori算法挖掘滑坡变形过程中累积位移、累积加速度等属性与演化阶段的关联关系,发现滑坡演化规律与累积加速度变化特征关联性强,在滑坡从初始变形进入等速变形,以及从等速变形到加速变形阶段时,都存在相应的阈值,即累积加速度在滑坡进入等速和加速变形阶段前后呈现出不同的特点。针对此特点,以湖北秭归县鸡鸣寺滑坡为例,将累积加速度作为划分鸡鸣寺滑坡演化阶段的指标,得出划分阈值,并利用关联规则分析其相关性。最后将此指标用于其他滑坡,均得到良好效果。因此根据该指标,可望实现滑坡演化阶段的自动判识。
Landslide generally experiences three phases including initial deformation, uniform deformation and accelerated deformation. Precise classification of these three phases is the base for landslide forecast. Based on a large amount of monitoring data of landslide evolution from initial deformation to instability failure phase, using association rules of data mining, correlation of cumulative displacement, cumulative acceleration and evolution phases of landslide deformation are revealed by classical Apriori algorithm. And it is found that there is a high correlation between the rule of landslide evolution and the variation charac- teristics of accumulative acceleration. Once the landslide changes from initial deformation into uniform de- formation phase, or from uniform deformation into accelerated deformation phase, there is a corresponding threshold. That is to say, the accumulative acceleration shows completely different characteristics before and after the landslide getting into other phases. Taking Jimingsi landslide as an example, accumulative ac- celeration is taken as a stage division index for landslide evolution and division threshold is obtained. Then the correlation between them is analyzed by association rule. This index is applied well as well to other ca- ses of landslides. Therefore according to this indicator, automatic stage division of landslide evolution is expected to achieve for disaster prevention.
出处
《地质科技情报》
CAS
CSCD
北大核心
2014年第2期160-164,共5页
Geological Science and Technology Information
基金
国家重点基础研究发展计划"973"计划(2011CB710605)
国家自然科学基金项目(40372120)
关键词
滑坡演化
关联规则
累积加速度
阶段判识
landslide evolutions association rules accumulative acceleration l stage division
