摘要
利用反向传播前馈神经网络 (BP网络 )的模拟能力来代替传统的方法 ,建立了在三轴单调比例加载情况下混凝土的全量型和增量型神经网络本构模型。通过对比模拟结果 ,对这些模型的差别进行了分析。从模型预测值和试验值逼近程度可以看出 ,训练后的模型具有很高的学习精度和良好的泛化能力 。
The application of the artificial neural networks to the constitutive model of concrete is carried out in the paper. The BP model of artificial neural network is used for the study. It is important to use the BP model to model the behavior of concrete. The triaxial proportional loading is also used for the study. A stress strain curve of concrete under the proportional loading is held through the study. The predicted results are in good agreement with that of the measured one.
出处
《土木工程学报》
EI
CSCD
北大核心
2003年第4期38-42,48,共6页
China Civil Engineering Journal
基金
国家自然科学基金重点基金项目 (50 1 390 1 0 )
