ARIMA模型和GM(1,1)与ARMA(p,q)组合模型在滑坡短期预报中的对比分析  

Comparative Analysis of ARIMA Model and GM(1,1)-ARMA(p,q) Combination Model in the Short-term Landslide Prediction

作　　者：赵鑫[1] 郭嘉[1] 吴建川[1] 

机构地区：[1]中国地质大学工程学院,武汉430074

出　　处：《安全与环境工程》2012年第3期24-28,共5页Safety and Environmental Engineering

摘　　要：滑坡的短期预报研究是当前国内外滑坡领域的重要研究方向之一。在滑坡的中短期时间预报中,基于统计学的预测模型是主要的分析预测工具。采用两种理论上比较成熟的ARIMA模型和GM(1,1)与ARMA(p,q)组合模型来模拟滑坡的累积位移量,并对这两种模型的优缺点以及各自适用条件进行了对比分析。结果表明:两种模型都能较好地拟合滑坡累积位移量时间序列并做出一定精度的预测,但是两种模型的适用条件不同。本研究可为滑坡短期预报提供借鉴和参考。The research on the short-term prediction of landslide is one of the important directions in the landslide realm both at home and abroad.In the medium-short-term prediction of landslide,the prediction model based on statistics is the major tool for analysis and forecasting.This paper uses ARIMA model and GM（1,1）-ARMA（p,q） combination model to simulate the accumulated displacement of landslide,and makes a comparative analysis on the advantages and disadvantages of the two models and their respective adaptive conditions.The results indicate that both models can well fit the accumulated displacement temporal sequence of landslide and make the perdition with certain degree of accuracy,though they have different adaptive conditions.This study can provide reference for the short-term prediction of landslide.

关 键 词：滑坡 短期预报 累积位移量 ARIMA模型 组合模型 

分 类 号：X43[环境科学与工程—灾害防治] P642.22[天文地球—工程地质学]