摘要
基于反问题中的模型参数辨识理论,通过人工神经网络,建立不同应力路径下的砂土神经网络本构关系模型。这种本构关系的建模方法有别于传统的建模方法,它不仅可以充分利用试验数据所包含的信息,还可以反映出不同的应力路径对本构关系的影响,这对实际工程的数值模拟与仿真有着重要的意义。
Based on the parameter identification by inverse problem theory,neural network model is set up to model the constitutive relation for sandsoil,under different stress paths. The approach is different from the traditional ones. It can not only make full use of all information obtained from test data,but also reflect the effect of stress path,which is very important in numerial simulation for practical engineering.
出处
《岩石力学与工程学报》
EI
CAS
CSCD
北大核心
2002年第10期1487-1489,共3页
Chinese Journal of Rock Mechanics and Engineering
关键词
应力路径
砂土
本构关系
人工神经网络
有限元
constitutive relationship,artificial neutral net,finite element,stress path,inverse problem
