基于BP神经网络的振冲碎石桩复合地基承载力预测

Bearing Capacity Prediction of Vibro-replacement Crushed Stone Piles Composite Foundation Based on BP Neural Network

由于影响振冲碎石桩复合地基承载力的因素众多,各种因素又相互作用,很难准确地计算其地基承载力。目前较准确的复合地基承载力载荷试验方法需投入大量的人力物力且耗时长,很难满足现场施工质量实时检测及工程进度的需求。本文在总结分析振冲碎石桩复合地基承载力影响因素的基础上,利用BP神经网络强大的非线性映射能力,建立了基于BP神经网络的复合地基承载力预测模型。模型预测结果表明,利用BP神经网络来预测振冲碎石桩复合地基承载力的方法是可行的,为振冲碎石桩复合地基承载力的快速设计检算提供了一种人工智能解决方法。

It is difficult to accurately calculate the bearing capacity of vibro-replacement crushed stone piles composite foundation because there are many influencing factors which may interact with each other.Some more accurate load test methods for calculating bearing capacity of composite foundation cost a lot of manpower and material resources and are time-consuming,which could barely meet the needs of real-time detection of on-site construction quality and engineering progress.On the basis of summarizing and analyzing the influence factors for bearing capacity of vibro-replacement crushed stone piles composite foundation,the bearing capacity prediction model of composite foundation based on BPneural network was established by using the strong nonlinear mapping ability of BPneural network in this paper.The model prediction results showthat it is feasible to predict the bearing capacity of vibro-replacement crushed stone piles composite foundation by using BPneural network,which could provide an artificial intelligence method for the rapid design and calculation of the bearing capacity of vibro-replacement crushed stone piles composite foundation.