基于小波变换与ARMA模型的建筑物沉降预报

对建筑物进行沉降监测并预报其变化趋势,能有效保障建筑物的安全性。本文提出一种基于小波变换的ARMA模型用于建筑物沉降预报。利用小波多尺度分析将沉降监测数据分解为高频信号和低频信号,并分别采用ARMA模型进行预测,然后将各序列预测结果进行合成,得到最终预测结果。并以青岛市某高层建筑物监测数据为例,分别采用传统ARMA模型以及基于小波变换的ARMA模型进行预报对比分析,结果表明基于小波变换的ARMA模型取得了较高的预报精度。

Prediction of Building Settlement Based on Wavelet Transform and ARMA Model

To monitor and forecast deformation of building can effectively guarantee its safety. In this paper, an ARMA model based on wavelet transform is proposed for building settlement prediction. The monitoring data are decomposed into high frequency and low fre-quency signals by wavelet multi - scale analysis. The ARMA model is used to predict both of them. Then, the forecasting results of each sequence are synthesized and the final forecasting results are obtained. Taking the monitoring data of a high - rise building in Qingdao as an example, the traditional ARMA model, the ARMA model based on wavelet transform are used to compare and analyze the results. The results show that the ARMA model based on wavelet transform has higher forecasting accuracy.