摘要
为了对基坑变形进行更准确的监测和预报 ,根据基坑变形的特点 ,提出了应用动态递归神经网络进行实时建模预报 ,并采用一种改进的在线学习算法 ,较好地描述了基坑变形的动态特性。通过对某工程基坑的监测 。
In accordance with the characteristic of deep excavation deformation,a new real time modeling method is presented for predicting the deformation,based on an improved Elman neural network.An improved on line learning algorithm is introduced to describe the dynamic behavior of deep excavation deformation.The new method has been verified to be more accurate and convenient than traditional one.
出处
《振动．测试与诊断》
EI
CSCD
2002年第3期217-220,共4页
Journal of Vibration,Measurement & Diagnosis
关键词
基坑变形
动态神经网络
实时建模
预报方法
excavation deformation analysis neural network real time modeling prediction
