| With the development of urbanization in China,the development and utilization of urban underground space has become a hot spot for research and application.In China's coastal areas,the groundwater level is generally high,and many of them are soft soil.The shallow water level of the shallow layer is more varied and the foundation is unstable.In such an environment,underground buildings tend to be resistant to water buoyancy brought certain difficulty to the construction,in order to overcome this problem,a lot of time to adopt pulling measures,pulling measures is most common in use of uplift pile foundation.The construction technology,load transfer mechanism,bearing capacity influencing factors and bearing capacity analysis of the uplift pile are very important links.At present,the calculation system of performance parameters such as bearing capacity of uplift piles is still to be improved.Artificial neural network,as an important part of artificial intelligence,has great advantages in nonlinear optimization.It analyzes the nonlinear relationship between the ultimate bearing capacity of pile and the ultimate bearing capacity of pile,and summarizes the empirical rule.has the important theoretical and application value.Therefore,using the relevant theory and experimental data of computer technology and bearing capacity of pile foundation,combined with the fault tolerance,intelligence and self-learning of neural network,the empirical data of pile foundation are obtained to obtain accurate neural network model,well applied to pile engineering.Based on the experimental data of the test pile of the static load test in the field of static load test,this paper analyzes and summarizes the five pile foundation projects of the Shanghai Construction Industry Hospital ward building reconstruction project,the Shanghai Pujiang Town 125-2 block underground garage,the mechanism of load transfer,the failure mode and the bearing capacity of the uplift pile in the bottom of the slurry forming mechanism;Based on the quantitative theory I,the mathematic model of the influencing factors of the bearing capacity of the uplift piles at the bottom of the grouting is established,and the contribution of the influencing factors is determined,and the main influencing factors are extracted.Based on the principle of BP neural network and the large data processing software MATLABr2016,a prediction model of bearing capacity and its main influencing factors of grouting pile is established.The main contents of this paper are as follows:(1)In view of the shortcomings of the existing construction technology of the expanded pile,the construction technology of the uplift pile at the bottom of grouting is introduced,and its structure,characteristics and application are analyzed.(2)Based on the load transfer mechanism and the failure mode of the equal section pile,the mechanism and failure mode of the uplift pile are studied.According to the failure mode of the uplift pile,the suitable calculation method of the bearing capacity of the uplift pile is selected.(3)Based on the principle of quantitative theory I,the test pile samples of five engineering projects in the underground garage of 125-2 plots in Pujiang Town of Shanghai area and the ward building of Shanghai Construction Engineering Hospital were analyzed and combined with large-scale data processing software MATLABr2016.The bearing capacity of uplift piles is analyzed,and the influencing factors of bearing capacity of uplift piles are analyzed.(4)Based on the principle of BP neural network and the large data processing software MATLABr2016,the prediction model of pile bearing capacity is established for six main influencing factors,the hidden layer is 15 and the output layer is 1.The relative error between the final predicted value and the measured value is calculated by using the model to test the test specimen,and the accuracy of the model and the practicability of the model are verified. |
PDF Full Text Request