中文摘要：

滑坡变形是一个高度复杂的非线性过程,常用的单点灰色预测模型没有充分利用各变形监测点间的相关信息,不足以反映滑坡整体的变形规律。多点灰色模型可以发挥相互影响的多个监测点间的整体积极作用,从而更加完整和全面的反映坡体的变形状况。以三峡库区某滑坡为研究对象,对布设在坡表的6个监测点进行聚类分析,将彼此高度相关的3点纳入多点模型中。针对传统多点灰色模型在背景值取值方面存在的缺陷,本文将基于背景值优化的多点灰色模型引入到滑坡变形预测中。计算结果表明,与传统灰色模型相比,背景值优化的多点灰色模型是解决贫信息条件下滑坡整体变形预测较理想的方法。

英文摘要：

Landslide deformation is a highly complex and nonlinear process. Single point forecasting models which is used frequently don＇t make full use of relevant information of each deformation monitoring points and can＇t reflect the integrated deformation regularity of landslide. Multi-gray model can play an integral positive role between a number of interacting monitoring points and reflect the slope of the deformation conditions more completely and comprehensively. Taking a landslide in the Three Gorges reservoir area for the study,six monitoring points on the surface of slope are analyzed by clustering,putting three points of highly correlated with each other into the multi-point model. In terms of the defect of background value in the traditional multipoint grey model,a multi-point model based on background value optimization be introduced into this paper.The results shows that,compared with the traditional gray model,the background value optimized gray model is ideal for solving the landslide overall deformation prediction under conditions of poor information.