灰色组合模型在坝体水平位移预测中的应用

Application of the combination prediction model on horizontal displacement of dam

在灰色预测的基础上,引入BP神经网络模型,建立了GM(1,1)和BP神经网络组合模型,分别运用GM(1,1)模型、BP神经网络模型和它们的组合模型三种模型对坝体的非线性水平位移进行了分析,对比发现此组合模型兼有灰色预测和BP神经网络预测的优点,模型既克服了原始数据少的缺点,也增强了预测的自适应性,证明了组合模型的预测精度较高,可以应用于坝体水平位移的预测。

Combine the GM（1,1） and BP Neural Network in order to realize the combination prediction model.Application of the GM（1,1）,BP Neural Network and combination prediction model respectively on non-linear horizontal displacement of dam.After comparing the three models,the combination prediction model have advantage of GM（1,1） and BP,which get over the lack of original data and reinforce the self-adaptability of the prediction.The results prove that the combination prediction model is very accurate.So this model can apply on the prediction of horizontal displacement of dam.