Deformation Analysis of Adjacent Structures Caused by Diaphragm Wall Construction Based on Machine Learning

Diaphragm wall is widely used in foundation pit engineering as a kind of retaining structure. However, in the case of deep excavation and large width of the diaphragm wall, it is inevitable that the surrounding soil will be disturbed during the construction process, leading to the deformation or even destruction of adjacent structures. Therefore, it is of great significance to explore the surrounding soil disturbance and the foundation deformation of structures caused by the construction of the diaphragm wall. In this paper, based on the field monitoring data of an excavation in Nanjing, the improved genetic algorithm (IGA) optimized long-term memory neural network (LSTM) is used to predict the deformation of structures caused by the excavation of the diaphragm wall. The neural network model was trained, tested, and verified. The results show that the prediction results of the model are very close to the measured data.