Construction And Application Analysis Of Surrounding Settlement Prediction Model For Deep Foundation Pit Excavation

With economic development driving the acceleration of urbanization,underground garages and underground shopping malls have been built in cities in order to save ground space and satisfy people's daily lives.However,compared with the above-ground construction,underground construction is more difficult and has a higher risk factor.The excavation of the foundation pit will cause changes in the original soil mass,which may cause damage to the surrounding ground surface,surrounding buildings,and other degrees of damage.Therefore,it is an important issue to study the influence of excavation on the surrounding environment.This paper uses theoretical analysis,measured data,intelligent algorithms and computer technology to study the deep foundation pit,surrounding surface and adjacent buildings as a whole,trying to find a suitable method to predict the environmental changes caused by excavation of foundation pits.First of all,from the background of engineering accidents caused by excavation of foundation pits,it is important to study the law of deformation of foundation pits.Secondly,we read the literature to understand the research status of the changes in the surrounding environment caused by the construction of pits at home and abroad.Analyze the shortcomings of various methods and find out the key research content of this article.Read the relevant materials,combine the knowledge learned in the classroom,sort out the relevant theoretical knowledge about the changes in the surrounding environment caused by the excavation of foundation pits,further deepen the understanding,and solidify the theoretical basis for subsequent research.Thirdly,the basic theory related to the establishment of predictive model is expounded,and the theory is more perfect and suitable for small sample,multidimensional,nonlinear least squares support vector machine(LSSVM)as the prediction method.A LFOA-LSSVM prediction model was established by optimizing the parameters of LSSVM using the Fruity Fly Optimization Algorithm(FOA).Finally,combining the measured data with the empirical study of the model,the FOA-LSSVM prediction model and the LSSVM prediction model are established to compare the results.The root mean square error and the relative error are used as evaluation indicators.The results show that the LFOA-LSSVM prediction model has the highest accuracy.The improved forecasting model has reference value for similar projects and has an early warning effect on quality and safety accidents in excavation of deep foundation pits.