论文部分内容阅读
In view of the characteristics of soft soil deep foundation pit for the construction and geotechnical characteristics of the special medium, it is difficult to calculate theoreti-cally accurately structural deformation of the foundation pit, so in the course of excavation on the construction of the information is particularly important. The analysis and compari-son of several popular non-linear forecasting methods, combined with the actual projects, set up a grey theoretical prediction model, time series forecasting model, improved neural network model to predict deformation of the foundation pit. The results show that the use of neural network to predict with high accuracy solution, it is the foundation deformation prediction effective way in underground works with good prospects.