Research On The History Of Tuxi Railwa

"Innovation is the first driving force to lead development",with the country vigorously advocating innovation,the combination of construction civil engineering and sustainable,green,ecological,intelligent and other key words gradually increase in frequency.As a part of "intelligence",intelligent algorithms have been combined with construction civil engineering,resulting in many new research topics.One of them is the analysis and prediction of settlement and bearing capacity of composite foundation by intelligent algorithm model.As a common kind of composite foundation in coastal soft ground engineering,vibratory gravel pile composite foundation has the effect of crowding,replacement and drainage on soft foundation,and has strong antiliquefaction and shear resistance.However,the working environment of this foundation is complex,and the calculation of its settlement and bearing capacity is mostly considering the influence of the main factors such as the upper structure,geological situation and the nature of the pile,and it is difficult to quantify the influence of other factors such as construction technology and climate change,which leads to the deviation of the calculation result of settlement and bearing capacity of the foundation from the actual measurement result of the project.Based on this,the researchers proposed to use intelligent algorithm model to analyze and predict the settlement and bearing capacity of foundation based on the engineering actual measurement data,but the traditional single algorithm model has more restrictive conditions,poor adaptability and insufficient prediction accuracy,and there is room for optimization.Therefore,in order to improve the prediction accuracy of settlement and bearing capacity of composite foundation with vibratory pile,this study carries out the improvement,combination and optimization of relevant algorithmic models.In this paper,based on the study of various related algorithm models,the traditional algorithm models are improved,combined and optimized in two parts: single influencing factor and multiple influencing factors,and three kinds of composite foundation settlement and bearing capacity prediction methods are proposed,namely,using gray Markov model to predict the foundation settlement under static load test,and then specifying the corresponding base bearing capacity under settlement;using unbiased gray Markov model to predict foundation settlement and analyze the settlement law of foundation with time;using gray correlation analysis-stepwise regression-support vector machine regression(GRA-SR-SVR)prediction model to analyze the main influencing factors of foundation bearing capacity,based on which,the foundation bearing capacity under the influence of multiple factors is predicted.The main research results are as follows.(1)Using traditional gray model and gray Markov model to predict the settlement of composite foundation under static load test,the average relative errors between the predicted and measured values of traditional gray model in two cases are 0.0512 and 0.0231 respectively,while the gray Markov model is 0.0203 and 0.0050.Markov theory can correct the relative errors of traditional gray model and strengthen the traditional gray model’s ability to deal with the "volatility" of foundation settlement.(2)The time-deposition data of foundations were predicted using the traditional gray model,unbiased gray model,gray Markov model,and unbiased gray Markov model.The average relative errors between the predicted and measured values of the traditional gray model,unbiased gray model,gray Markov model,and unbiased gray Markov model were 0.2439,0.229,0.1186,0.0992,and 0.1602,0.1449,0.0525,and 0.0246 in two cases,respectively.improved on the basis of the traditional gray model,which improves the prediction accuracy of the traditional gray model for foundation time-settlement data,and combined with Markov optimization,further increases the prediction accuracy of the model.(3)The main influencing factors of the bearing capacity of the composite foundation of vibratory pile were clarified by gray correlation and stepwise regression analysis.Among them,the gray correlation of effective pile length,dense current,filling coefficient,natural density,bedding thickness,replacement ratio and diameter are all greater than 0.7,and the stepwise regression of pore ratio is significant,all of which are the main influencing factors of bearing capacity of vibratory pile composite foundation.(4)The GRA-SR-SVR model is used to predict the bearing capacity of foundation under multiple influencing factors.The average relative errors of model prediction and measured value in two cases are 0.0177 and 0.0122 respectively.the prediction accuracy of both cases is higher than 98%,GRA-SR-SVR model can complete multi-factor analysis,sample screening,nonlinear fitting of composite foundation bearing capacity under large data volume,and accurately predict the bearing capacity of composite foundation.