基于数据融合技术的桩基承载力预测方法研究  

Method for Prediction of the Influencing Factors of Single Pile Bearing Capacity

作　　者：郗锋[1] 翁光远[1] 

机构地区：[1]陕西交通职业技术学院公路工程系,陕西西安710018

出　　处：《华中科技大学学报（城市科学版）》2010年第4期26-30,共5页Journal of Huazhong University of Science and Technology

基　　金：陕西省教育科学"十一五"规划2009年度课题(SGH0903032)

摘　　要：为了能够快速并较准确的预测桩基的承载能力,达到在施工过程中减少或不做试桩的效果,以单桩为例,分析了影响竖向承载能力的量化因素及非量化因素,利用小波概率神经网络(WPNN)与数据融合技术的联想和预测功能,得出承载力和这些因素的关系。通过对钻孔灌注桩及钢筋混凝土预制桩的静载试验数据分析,选择了WPNN与数据融合技术的方法对分别对两组试验数据进行分析,建立了合理的模型进行承载力预测,60根单桩的承载力的预测值与实测值吻合较好,证明了该方法在预测桩基竖向承载能力时可以满足工程实际的需要。In order to quickly and accurately predict the bearing capacity of the pile foundation, single piles are taken as an example, the quantitative and qualitative factors which influence the vertical bearing capacity are analyzed. With the lenovo and prediction function of wavelet probability neural network and data fusion technology, the relationship between bearing capacity and the influence factors was obtained, therefore, test pile cannot or less be done during construction. Based on the analysis of dead-load experimental data of cast- in-situ pile and reinforced concrete prefabricated piles used by high rise building, the prediction is very close to the measured value. The prediction of bearing capacity can be used to adopt the method of neural networks technique. It is proved to be satisfied with engineering needs.

关 键 词：桩基承载力 小波概率神经网络(WPNN) 数据融合 承载力预测 

分 类 号：TU473.1[建筑科学—结构工程]