基于CNN-LSTM的大坝变形组合预测模型研究

Research on Deformation Prediction Method of Concrete Face Rockfill Dam Based on CNN-LSTM

为了提高大坝变形预测模型精度和泛化能力,建立了一种基于卷积神经网络(Convolutional neural networks,CNN)与深度学习长短期记忆(Long short-term memory,LSTM)神经网络的组合预测模型CNN-LSTM。该模型先利用CNN提取大坝变形监测时间序列的特征,再利用LSTM生成特征描述,该模型精度高、泛化能力强。以柏叶口水库混凝土面板堆石坝为例,经过CNN-LSTM模型计算,将模型变形预测值与原型监测资料进行对比,再与LSTM模型及CNN模型的预测结果进行对比。结果表明,CNN-LSTM模型预测值最接近监测资料实测结果。

In order to improve the accuracy and generalization ability of dam deformation prediction models,a neural network combination prediction model CNN-LSTM is established based on the convolutional neural networks(CNN)and the deep learning long short-term memory.In this model,the CNN is firstly used to extract the features of dam deformation monitoring time series,and then the LSTM is used to generate the feature description.This model has high accuracy and strong generalization ability.Taking Baiyekou Face Rockfill Dam as an example,the CNN-LSTM model is used to calculate the dam deformation,and the calculation results are compared with the actual monitoring data.The calculation results of CNN-LSTM model are also compared with that of the LSTM model with the CNN model respectively.The comparisons show that the predicted value of CNN-LSTM model is more close to the actual monitoring data.