一种抑制边界效应的EMD方法及其在滑坡监测中的应用

滑坡位移序列的正确分解对滑坡的预测、预警有重要影响。经验模态分解是一种常用的时间序列分解方法,但该方法在分解过程中存在端点效应,建立的预测模型端点发散,导致预测误差出现较大偏差。为此,本文基于函数延拓的原理对经典经验模态分解方法进行改进,提出一种抑制端点效应的经验模态分解方法,利用金沙江白格滑坡数据对改进方法进行分析验证。试验结果表明,改进方法的端点效应抑制效果良好,预测精度较经典方法有一定的提高。

An EMD method to suppress boundary effect and its application in landslide monitoring

The correct decomposition of landslide displacement sequence has important influence on landslide prediction and early warning.Empirical mode decomposition (EMD) is a commonly used time series decomposition method, but it has endpoint effect in the decomposition process.Therefore, this paper improves the classical EMD method and proposes an EMD method to suppress endpoint effect.The improved method is analyzed and verified by using the data of Baige landslide in Jinsha River. The experiment results show that the improved method has good endpoint effect suppression effect and the prediction accuracy is also improved compared with the classical method.