模糊聚类及其在滑坡监测中的应用研究  

Research on fuzzy clustering and its application

作　　者：邓勇[1,2] 张正禄[1,3] 谢年生[2] 吕耕超[2,3] 

机构地区：[1]武汉大学测绘科学与技术学院,武汉430079 [2]中国水电顾问集团中南勘测设计研究院,长沙410014 [3]武汉大学测绘遥感信息工程国家重点实验室,武汉430079

出　　处：《测绘科学》2010年第4期163-165,共3页Science of Surveying and Mapping

基　　金：国家重点基础研究发展计划(973计划)项目(2003CB716705)

摘　　要：模糊聚类的关键在于确定评价模糊聚类效果的阈值λ。本文针对滑坡形变,研究了监测点的模糊聚类算法,对于模糊聚类效果,提出了基于先验信息和后验信息的评价方法,并进行了实例验证。本文的研究有助于对滑坡变形势态、原因和趋势的了解,有助于对滑坡变形监测方案的修改和完善。The sticking points of fuzzy clustering are constructing fuzzy semblable matrix and confirming the threshold value λ to evaluate the clustering effect. Aiming at landslide deformation, this paper researched on the arithmetic of fuzzy clustering for monitoring points, and put forward two methods to evaluate the effect of fuzzy clustering respectively based on prior information and posterior information, and testified them with instances. The research could help realize the states, reason, and the trend of landslide deformation, and also help amend and complete the landslide deformation monitoring scheme.

关 键 词：滑坡监测 模糊聚类 模糊相似矩阵 阈值 聚类效果评价 

分 类 号：P258[天文地球—测绘科学与技术]