Reliability Analysis Of Foundation Pit Retaining Pile Based On Bayesian Principle

In the process of urban-rural integration,a large number of shallow foundation pit projects appear in soft soil areas.Considering factors such as excavation depth,support awareness,and project costs,they are often not supported,leading to a large number of accidents.To solve this problem,based on the national green development concept and the pursuit of ecological and environmentally friendly buildings,bamboo is used as the support material for shallow foundation pits.In this paper,the probabilistic limit state method is introduced.Based on the relevant codes,the lateral displacement of pile top is taken as the index to judge the supporting function of pile foundation.In addition,considering the uncertain factors such as soil parameters and pile foundation parameters,a calculation program is developed by using MATLAB to analyze the reliability of foundation pit and supporting pile.The main work and achievements are as follows:(1)Based on Rankine earth pressure theory,the safety factors of pit bottom anti-heave and wall anti-overturning stability were calculated.Considering the variability of soil parameters,the JC method was used to calculate the anti-overturning reliability index.The safety of foundation pit is judged by small probability events.When the soil parameter variability is large,but the safety factor is slightly larger than the standard requirements,the conventional safety factor method is not accurate,and the reliability index should be combined to determine the safety.Taking the variability of soil parameters as a variable,the results show that the reliable index is more sensitive to the friction angle in the soil and the ground load.(2)Based on different service functions of supporting piles,the limit state was divided into normal service limit state and bearing capacity limit state,and the corresponding function was established.The central point method was used to derive the calculation formula of reliability index respectively.Try to introduce dimensionless normalized processing meter ratio on pile top displacement,using the K-S distribution type of inspection,inspection data,combined with the principle of the bayesian optimization,estimate the mathematical characteristics of central point method and JC method and mapping transformation method to calculate reliability index,the final inspection results to the MC method.It is proved that the use of Bayesian principle to optimize the estimation of the mathematical characteristics of the model factors can effectively solve the influence of data errors on the reliability index.(3)The calculation formula is deduced,and the uncertainty of the parameters in the formula and the influence of the load change on the reliability index are analyzed.The results show that the influence of the allowable displacement value on the reliability index is greater than the calculated displacement value,and the difference continuously increases in the range of 0.11～0.5.When the soil condition is good and the displacement variation coefficient is less than 0.3,the safety factor can be 2.0～2.5(4)Taking 2.5 as the target reliability,analyze the values of the safety factors and displacement coefficients of the supporting piles at different positions.The results show that the supporting side piles can be taken from 1.5 to 2.0;the supporting inner piles can be taken from 2.0 to 2.5;the supporting piles should be taken from 2.5 to 3.5.When the data obeys the normal distribution and the lognormal distribution,the difference of the original data calculation sub-factor coefficient can reach up to 0.6,while the data optimized by the Bayesian principle has a change range of less than 0.1.