集对分析聚类预测法在大坝变形分析中的应用

Set Pair Analysis-cluster Predication Method in Deformation Analysis of Dams

集对分析聚类预测法融合了集对分析中的同异反模式识别的"择近原则"和聚类分析的基本思想,是解决模糊、随机以及其他不确定性问题的一种新方法。简要介绍了集对分析聚类预测法的基本原理及其在大坝变形监测数据处理中的应用。通过对丹江口水电站坝体监测的连续观测数据的计算分析,论证了该方法在大坝变形监测中的可行性和有效性。

The set pair analysis-cluster prediction method is proposed to classify and predict objects. It combines the thought of clustering analysis and the principle of closeness optimization of identical-discrepant-contrary pattern recognition about the set pair analysis. It is a new method to solve the ambiguous, uncertain and other random problems. The proposed method is analyzed and applied to the analysis and prediction of the dam deformation. A case study of dam deformation is given, and the periodical settlement data at its Monitoring point are analyzed. The results show that the method is feasible and effective during the monitoring process of dam deformation.