准确地判别滑坡的变形演化阶段,特别是对进入加速变形阶段突变点的识别,对及时捕捉滑坡前兆,尽早开展滑坡预警报和进行工程处理都具有十分重要的意义。将小波的多分辨分析与突变点识别原理相结合,依据滑坡的累积位移监测资料,结合3类典型滑坡(光滑型、阶跃型和振荡型)实例对滑坡进入加速变形阶段的突变点进行了识别研究。研究结果发现,滑坡在进入加速变形阶段时,其累积位移监测信号在4个连续尺度上分解重构后的细节信号的幅值均存在着明显的突变特征,而且4个尺度的突变特征均具有较好的一致性。据此特征,即可较为准确地识别滑坡进入加速变形阶段的突变点。

Accurate identification of the deformative evolution stage of landslides,especially the accelerative deformation stage,is very important for seizing landslide precursors and carrying out landslide early warning,forecast and engineering control as quickly as possible. Using the multi-resolution analysis combined with the identification principle of abrupt change points of wavelet analysis method,this paper identified the abrupt change points of the accelerative deformation stages of the landslides based on the accumulative displacement data of the 3 types of typical landslides ( smooth type,step type and oscillation type) . Results show that,there exists significant abrupt amplitude change in the detailed signals of the decomposed and reconstructed accumulative displacements in 4 continuous scales,and the abrupt change features of the 4 scales are in good consistency when the landslides enter the accelerative deformation stages. Therefore,by this characteristic,the wavelet analysis method can identify the abrupt change points at which when landslides enter the accelerative deformation stage correctly.