DEFORMATION PREDICTION METHOD OF DEEP FOUNDATION PIT OF CONSTRUCTION ENGINEERING BASED ON PARTICLE SWARM OPTIMIZATION CONVOLUTIONAL NEURAL NETWORK
Taking the Phase 5 general contract project of China Resources Fuyang Center Project as the research object,the convolution neural network method based on particle swarm optimization is used to predict the horizontal displacement and surface settlement of deep foundation pit envelope.With the increase of monitoring time,the predicted values of horizontal displacement and surface settlement of deep foundation pit envelope have consistent changes with the measured values.Compared with the measured value,the root-mean-square error of the predicted horizontal displacement of the envelope structure is 3.89%,the average percentage error is 5.92%,and the root-mean-square error of the predicted surface settlement is 4.53%and the average percentage error is 3.96%,both of which are less than the error limit of 8%.The results show that the convolutional neural network based on particle swarm optimization has high prediction accuracy for deep foundation pit deformation.
construction engineeringdeep foundation pitdeformation predictionconvolutional neural networkparticle swarm optimization
NETL
NSTL
万方数据
