基于PSO-RF的扩底灌注桩极限抗拔承载力预测  

The ultimate uplift bearing capacity of expanded bottom bored pile prediction based on PSO-RF

作　　者：张博航 薛新华[1] Zhang Bohang;Xue Xinhua(College of Water Resource and Hydropower,Sichuan University,Chengdu Sichuan 610065,China)

机构地区：[1]四川大学水利水电学院,四川成都610065

出　　处：《山西建筑》2023年第24期78-81,85,共5页Shanxi Architecture

摘　　要：地下建筑物需要采取一定的抗浮措施保证安全,根据工程需求设计出合适的抗拔桩基础具有十分重要的意义。从有关文献中收集了40组扩底灌注桩抗拔试验数据,分析不同影响因子之间的相关性,将主要影响因素作为输入变量、将扩底灌注桩极限抗拔承载力作为输出变量。利用粒子群优化(PSO)算法对随机森林(RF)方法中决策树个数及其最大深度的确定方式进行了优化,并在此基础上建立了扩底灌注桩极限抗拔承载力的预测模型。研究结果表明,PSO-RF模型的决定系数R^(2)为0.976,预测性能优于RF模型(R^(2)=0.954)与GMDH模型(R^(2)=0.88)。PSO-RF模型能够准确预测扩底灌注桩的极限抗拔能力,在实际的桩基工程中具有较好的适用性。The underground buildings need to take some certain anti-floating measures to ensure the safety,according to the engineering requirements to design the appropriate anti-pulling pile foundation is of great significance.In this paper,40 groups of uplift pile test data were collected from relevant literatures,and the correlation between different influencing factors was analyzed.The main influencing factors were taken as input variables and the ultimate uplift bearing capacity of the expanded bottom pile was taken as output variables.Particle swarm optimization(PSO)algorithm is used to optimize the determination of the number and the maximum depth of decision trees in random forest(RF)method,and on this basis,a prediction model of ultimate uplift bearing capacity of expanded bottom pile is established.The results show that the coefficient of determination(R^(2))of PSO-RF model is 0.976,and the prediction performance of PSO-RF model is better than that of RF model(R^(2)=0.954)and GMDH model(R^(2)=0.88).PSO-RF model can accurately predict the ultimate uplift bearing capacity of expanded bottom pile,and has good applicability in actual pile foundation engineering.

关 键 词：扩底灌注桩 极限抗拔承载力 粒子群优化 随机森林 PSO-RF 

分 类 号：TU473.1[建筑科学—结构工程]