摘要
运用BP神经网络建立了静力触探成果与单桩承载力的相关关系 ,并通过工程实例将BP神经网络预报值与线性回归数学模型、规范公式方法计算值进行对比 ,说明BP神经网络在利用静力触探成果确定单桩承载力中是一种行之有效的方法 ,具有精度高 ,信息处理智能化等特点 ,计算结果明显优于传统的线性回归数学模型和规范公式方法 ,具有较好的应用价值。
The relationship between the result of static contact survey and the bearing capacity of single pile is set up applying BP nerve net. By project examples, the comparison is made between the forecast result from BP nerve net and the calculation results from the mathematic model of linear regression and the standard formula. The conclusion of comparison shows that the method of applying BP nerve net to determine the bearing capacity of single pile using the result of static contact survey is effective, characteristic of high precision and intelligent information treatment, and is obviously superior to the traditional method of the mathematic model of linear regression and the standard formula, and so is good in application value.
出处
《山东农业大学学报（自然科学版）》
CSCD
北大核心
2002年第3期338-342,共5页
Journal of Shandong Agricultural University：Natural Science Edition
关键词
BP神经网络
静力触探
单桩承载力
应用
线性回归
BP neural net
static contact survey
linear regression
bearing capacity of single pile
