The Optimization Of Feedback Analysis Of Foundation Engineering Based On GA-ANN

The 21st century is the century of underground works.With the people's increasing demand for space,and underground space as a valuable natural resource,its development and use is a reflection of progress of contemporary society.The large number of foundation works are excavated is one of the important aspects.With the development of computer technology,a large number of intelligent algorithms have being applied to foundation engineering,which have played a very good guide to design and construction of the foundation,Artificial Neural Network with Genetic Algorithm (GA-ANN)is one of them.At present,the use of GA-ANN on the feedback analysis of mechanical parameters of foundation,the optimization of supporting structure parameters and the optimization of anti-analysis of the permeability coefficient are also rare,especially the relationship between these three works in the pit are complementary to each other,the former value has a great impact on the value of the latter,so the use of GA-ANN to optimize this series of parameters has a great practical significance.In this paper,Artificial Neural Network with Genetic Algorithm(GA-ANN) is proposed which is a new type of optimization algorithm and combines both merits of Genetic Algorithm and Artificial Neural Network,it obtains good results in the practical engineering examples.In the use of GA-ANN to optimize and feedback-analyze mechanical parameters of the foundation,displacement back analysis method is chosen in which displacement is chosen as its indicator.Combining with a foundation engineering practice,the actual monitoring of displacements of foundation pit is used to anti-analyze the mechanical parameters of soil based on GA-ANN,then input these parameters into the numerical model,parameters will be calculated and the calculated displacements,which are compared with the actual displacements,can be read,the compared results show that the application has been very good.In the engineering example of Uniform Payment Center of the State Treasury in Xicheng District,Beijing,GA-ANN is carried out in detail design of the supporting structure,in which the optimized objective function is integrated with the displacement and plastic zone size as well as the supporting structure costs,then the last optimized feedback analysis parameters of the supporting structure are input the numerical model, the results of the calculation have very good safety,stability and economic benefits, which are better than results of design.GA-ANN is used to feedback-analyze the uneven soil permeability coefficient in the foundation of Suzhou Runjie Square which is based on the head values,the results of optimized feedback analysis of the permeability coefficient then are put into the numerical model to calculate the fluid-solid coupling effect of the foundation,the calculation results are compared with the results which are not considered the impact of ground water,it shows that fully considering the impact of groundwater on the foundation has great significance in security and stability of foundation.