Application of Deep Learning and Distributed Fiber Optic Sensing in Monitoring and Predicting of Foundation Pit Deformation
FAN Cheng
PENG Yanli
ZHAO Jie
CHENG Shukai
FAN Yijiang
Abstract：With the increasing degree of urbanization in China,it is increasingly necessary to monitor and predict the deformation of deep foundation pit.A convolutional neural network-long short term memory-self attention mechanism(CNN-LSTM-SAM)combined neural network model with time series monitoring data as input was proposed in this paper.For a deep foundation pit project with internal support in Donggang business district,Dalian city,the CNN-LSTM-SAM model was used to predict the horizontal displacement of the pile top in combination with the distributed fiber optic sensing(DFOS)technology,the deformation prediction values were compared with the prediction results of the back propagation(BP)neural network,long short term memory(LSTM)neural network and convolutional neural network-long short term memory(CNN-LSTM)neural network models.The results show that the CNN-LSTM-SAM model has higher accuracy than the other three models.The monitoring data of other monitoring points are then selected for training and prediction to further verify the prediction effect of the model,proving the applicability and robustness of the CNN-LSTM-SAM model.The research results of this paper can provide a reference for the design and construction of similar deep foundation pit projects.
Keywords：convolutional neural networklong short term memory neural networkself attention mechanismdeformation predictiondistributed optical fiber
Publication Date：2025-06-25
Online Publishing Date：2025-08-15(First online date of this platform, not the publication date of the document)
Pages：12( 79-90 )