摘要
采用人工神经网络（ＢＰ网络）进行了沙的本构关系模拟，建立了沙体三点式神经网络本构模型，并通过网络自适应学习，使本构模型不断改进。试验表明，网络本构模型与实验结果吻合良好。
Artificial neural network methods are applied totwo distinct aspects of material modelling in civil engineering.The first aspect is the constitutive behavior modelling and constituive model development for sand by means of BP algorithm.The second aspect is the improvement of constitutive model for sand through the adaptive method .The experiment shows that the neural network material model has truly learned the behavior of triaxial specimens of sand.
出处
《贵州工业大学学报（自然科学版）》
CAS
1997年第3期89-93,共5页
Journal of Guizhou University of Technology(Natural Science Edition)
关键词
本构模型
沙体
工程地质
神经网络模拟
neural network
triaxial
constituive model
adaptive method
