摘要
概述了 BP神经网络的基本原理。通过对影响桩基承载力因素的分析并结合 BP神经网络特点 ,提出位置系数新概念 ,并建立了桩基工程中几种常用桩的单桩极限承载力预测模型。在工程实例的基础上 ,用 MATL AB的神经网络工具箱说明 BP神经网络预测单桩极限承载力可行性。
This paper deals with the BP neural network design for pile foundation.The main principle of BP neural network is decribed briefly. Through analyzing the affecting elements of the bearing loads of piles and the characteristics of BP neural network,a new conception-position parameter is advanced, and forecasting models of ultimate bearing loads of some kinds of piles are founded. Based on a practical example, the feasibility of such forecasting model is proved by means of neural network toolbox of MATLAB.
出处
《合肥工业大学学报（自然科学版）》
CAS
CSCD
2001年第2期198-202,共5页
Journal of Hefei University of Technology：Natural Science
