GM-ARIMA模型在大坝安全监测中的应用

Application of GM-ARIMA Model to Dam Safety Monitoring

针对大坝安全监测的小样本数据既有一定趋势性又有一定波动性的特点,把灰色模型和时间序列模型结合起来运用在大坝安全监测中.首先利用灰色模型进行拟合和预测,然后对灰色残差序列建立ARIMA模型,对残差进行预测,最后将两者结合起来即可得到预测值.本文以小湾拱坝坝顶某测点的径向位移为例,建立GM-ARIMA进行拟合和预测,并与实测值比较.计算结果表明,与GM模型相比,GM-ARIMA模型的精度高,预测值更接近于实测值.

In view of that the dam safety monitoring data sets have a certain characteristic of tendency and vol- atility, the grey model and the time series model are combined to use in dam safety monitoring. Firstly, the gray model is used to fit and forecast; and then ARIMA model is built for grey residuals to predict the residu- als; the final combination of two models can be used to predict values. Taking the radial displacements of Xi- aowan arch dam for example, the GM-ARIMA model is established to fit and predict, and then compare with the measured values. The results show that, compared with the traditional GM model, GM-ARIMA model is more accuracy and predictive values are closer to the measured values.