基于BP神经网络的桥梁桩基水平位移预测  

Prediction of Horizontal Displacement of Bridge Pile Foundation Based on BP Neural Network

作　　者：王地 蒋敏敏[1] WANG Di;JIANG Min-min(School of Civil Engineering,Henan University of Technology,Zhengzhou 450001,China)

机构地区：[1]河南工业大学土木工程学院,郑州450001

出　　处：《南通职业大学学报》2023年第1期98-104,共7页Journal of Nantong Vocational University

摘　　要：基坑开挖会引起周围土体位移,导致桩身层产生附加弯矩、侧向位移等,从而影响邻近桩基正常工作。为更好地预测桩基位移并加以控制,以某桥梁基坑工程为例,建立BP人工神经网络预测模型,结合三维数值模拟及实际监测数据,对桩基水平位移进行预测对比。结果表明:基坑开挖对邻近桩基的影响随两者距离增大而缓慢减弱;开挖过程中桥梁产生的竖向位移始终大于其他方向位移;针对基坑工程中普遍存在的非线性复杂问题,通过输入一定组数影响变形的相关因素,进行反复迭代训练,能提高模型预测精度,减小预测误差,可为桩基位移的预测提供一定参考。The excavation of foundation pit causes the displacement of surrounding soil,resulting in additional bending moment and lateral displacement of pile layer,interfering with the normal work of the adjacent pile foundation.To better predict and control the displacement of pile foundation,a BP artificial neural network prediction model is established based on the foundation pit engineering of a bridge,and the horizontal displacement of pile foundation is predicted and compared by examining the 3D numerical simulation and actual monitoring data.The results show that the influence of excavation of foundation pit on adjacent pile foundation gradually weakens with the increase of the distance between them,and the vertical displacement generated by the bridge is always greater than the displacement in other directions during excavation.To deal with the nonlinear complex problems in foundation pit engineering,inputting a certain groups of relevant factors leading to deformation,and conducting repeated iterative training can improve the prediction accuracy and reduce the prediction errors of the model,which provides a reference for the prediction of pile displacement.

关 键 词：基坑工程 桥梁桩基 BP神经网络 有限元模拟 桩基位移 

分 类 号：U443.15[建筑科学—桥梁与隧道工程] TU753.3[交通运输工程—道路与铁道工程]