Displacement Level Forecast on Deep Foundation BP Algorithm Application based on Neural Network

被引：0

作者：

Yang Shibin ^{[1
]}

Mao Zhengli ^{[1
]}

机构：

[1] Henan Univ Urban Construct, Pingdingshan City 467044, Henan Province, Peoples R China

来源：

CIVIL ENGINEERING IN CHINA - CURRENT PRACTICE AND RESEARCH REPORT | 2010年

关键词：

The deep foundation; The level of displacement; Prediction; BP Algorithm;

D O I：

暂无

中图分类号：

S2 [农业工程];

学科分类号：

0828 ;

摘要：

Basing on the nonlinear mapping function of neural network, a new prediction method of the deep foundation horizontal displacement was put forward without the establishment of complex mathematical models. The variation and law of the level of displacement of the deep foundation was predicted by using the adaptive learning rate of BP algorithm, with a maximum error of +/- 5% between the predicted and actual values. The predicted results show that this method can predict the level of displacement of the deep foundation with a higher accuracy when mathematical model is not needed, it also can avoid the shortcomings of the traditional methods using fuzzy mathematics which is difficult to establish and hard to meet the requirements.

引用

页码：567 / 571

页数：5