摘要
利用BP神经网络较强的非线性映射能力和学习功能,建立了单桩竖向极限承载力预测的人工神经网络模型.该模型依据实测静力触探资料建模,通过静载荷试验成果的学习与预测检验,证明其预测精度良好、适用性强,验证了神经网络方法的可行性,因而具有较大的工程实用价值.
Based on the strong nonlinearity and learning ability of artificial neural networks, a method for predicting vertical ultimate bearing capacity of single pile is presented. In this model data of static tests are trained and predicted by survey of static contacts. The prediction results agree with in site tests, showing that the method is feasible and effective.
出处
《上海大学学报（自然科学版）》
CAS
CSCD
2004年第6期639-642,共4页
Journal of Shanghai University:Natural Science Edition
关键词
BP神经网络
单桩承载力
静力触探
预测
BP neural network
bearing capacity of single pile
static contact survey
prediction
