Parameters' sensitivity analysis of deep foundation pit excavation based on RBF neural network

In order to study the deformation of retaining structure and surrounding environment influenced by deep foundation pit excavation,a three-dimensional numerical simulation for deep foundation pit excavation and a mapping relation between the displacement of foundation pit and the parameters are established with three-dimensional numerical analysis software FLAC3D and RBF neural network,respectively.The influence of different supports during foundation pit excavation on the deformation of the pit and surrounding environment is analyzed,as well as the sensitivity of the parameters of the pit.The analysis results show that the 3 D numerical simulation of the excavation reflects soil conditions and construction conditions;that different support ways will lead to different deformation of foundation pit and different impact on the surrounding environment;that the deformation of foundation pit under stable condition is mainly related to the supporting structure;that the sensitivity coefficients of parameters of different soils are different,and the coefficients decrease with the excavation.