基于EMD-ARIMA模型的边坡位移分析与预测

An EMD-ARIMA Model-Based Analysis and Prediction of the Slope Displacement

针对波动性大、规律性差且含噪声的高速公路边坡位移监测时间序列,提出了一种基于经验模态分解技术(EMD)和自回归移动平均(ARIMA)相结合的预测算法,基本思想为"数据分解-数据预测-数据合成"。通过对重庆奉云高速公路247 d边坡位移监测数据的预测和分析,表明EMD-ARIMA具有较高的精度,好于单一ARIMA模型,可为工程建设及防灾减灾提供技术指导。

Aiming at solving the problem of the time series of monitoring expressway slope displacement with large fluctuation,poor regularity and noise,proposed in the present paper is a prediction algorithm based on the combination of the empirical mode decomposition(EMD)and the auto-regressive moving average(ARIMA),the basic idea of which is"data decomposition-data prediction-data synthesis".Through the prediction and analysis of the monitored data of 247-day slope displacement of the Fengyun Expressway in Chongqing,it is found that the emd-arima is higher in precision,better than a single ARIMA model,and thus,can provide technical guidance for engineering construction and disaster prevention and mitigation.