| CFG pile composite foundation’s treatment scheme design is a very complicated process,that contains many influencing factors and tedious calculations,in addition to this the related theoretical research is also very lagging far behind.In real engineering,CFG pile composite foundation design schemes usually use a combination of trial and error method.First,based on past engineering experience,several design schemes are proposed and then it is checked whether the scheme meets the design specifications and engineering requirements.The most economical solution is selected as the final solution to be implemented.This design process has several drawbacks including lot of time and manpower,also the availability of most economical solution is not assured.With the development of computer information technology during the recent years,new ideas have emerged for the designing of CFG pile composite foundation schemes,promoting the intelligent development of CFG pile composite foundation treatment scheme.Based on the in-depth study of the knowledge system of CFG pile composite foundation design field,this paper uses genetic algorithms,non-parametric regression,fuzzy analytic hierarchy process,artificial neural network theories,utilizing programming language to develop a system that can assist engineering designers to quickly realize the intelligent optimization design of CFG pile composite foundation scheme.The main work and conclusions are as follows:1.In-depth study of CFG pile composite foundation design domain knowledge,analysis of knowledge constituent elements,extraction of key knowledge elements,thereby drawing entity connection diagrams and constructing a domain knowledge expansion framework;Proposed the means of production method and first-order predicate logic representation to represent CFG pile composite foundation design domain knowledge;2.Established mathematical models for CFG pile composite foundation optimization design,a computer program based on an optimization algorithm was compiled to achieve the optimal design,and an example is trialed to verify that the optimized design model can reasonably and quickly infer a more economical foundation treatment solution,achieve the purpose of improving design efficiency and economic benefits.3.Two methods,including smooth spline nonparametric regression method and fuzzy analytic hierarchy process are proposed to map the optimal design of CFG pile composite foundation of heterogeneous soil layer into the design of homogeneous soil to solve the design under heterogeneous subsoil situation,and an example is trialed to verify that this method can reduce the design complexity under heterogeneous foundation and improve the design efficiency.4.Based on CFG pile composite foundation optimization design model,a large number of optimized design solutions were screened out,compile the neural network reasoning program with two hidden layers to realize faster and more intelligent reasoning of optimal design scheme based on engineering data,The analysis of the network training results shows that the network model has a good training effect,small network errors and fast convergence speed,and can be effectively used to predict the design parameters.5.Created the GUI interface of the CFG pile composite foundation scheme intelligent optimization design system to realize the control of each section and data flow in the intelligent optimization system,to let it become an intelligent optimization design system that is more user friendly,compact and noncomplex.This article aims to enrich the theoretical research of CFG pile composite foundation design,improve design efficiency and economic benefits,the research results of this article are conducive to improving the digitalization and intelligence of CFG pile composite foundation design,and provide some help for engineering designers to carry out design work. |
