基于滑坡地表位移监测数据,运用多重分形理论中的广义关联维计算方法,分析滑坡变形演化过程中所表现出的多重分形特征。首先,从具有蠕变特性的滑坡的理想演化过程出发,构造简化函数,探明不同变形阶段滑坡位移的多重分形特征及其变化规律,其次,结合滑坡实例,验证运用滑坡累积位移曲线的多重分形特征来判别滑坡的变形阶段和稳定性状态的可行性。研究结果表明:对于蠕变型滑坡,状态越稳定,其相应累积位移曲线的多重分维值均值越大,反之,多重分维值均值越小,且滑坡进入加速变形前期,此多重分维值分布会伴随出现突变点;蠕变型滑坡累积位移多重分维值能够作为表征滑坡变形趋势及稳定性状态的特征量,为该类型的滑坡局部稳定性评价及预测提供依据。
According to the landslides surface displacement monitoring data, the multi-fractal characteristics of landslide deformation evolution were analyzed using correlation dimension method based on the multi-fractal theory. Firstly, the ideal deformation evolution process of creep moving landslide was studied, and different functions were simplified to represent the displacement curve of different landslide deformation stages. Secondly, based on the analysis of real landslide samples, the acceptance of using the multi fractal characteristics of landslide displacement as a proxy to estimate landslide deform stage and stability was proved. The results show that for creep moving landslide, the more stable the landslide is, the higher the multi fractal dimension of cumulative displacement curve will be obtained, and usually before an accelerate deformation stage of this kind of landslide, there will be a turning point of its multi fractal dimension distribution. In addition, for creep landslide, multi fractal dimension can be adopted as a proxy to represent the local deforms situation and stability of this kind of landslide, which is meaningful for the assessment and prediction of landslide local stability.