论文部分内容阅读
采用BP(back propagation)神经网络算法对金属橡胶的本构关系系数进行学习训练,对BP神经网络做了比较详细的介绍,讨论了如何运用神经网络去预估金属橡胶的非线性本构关系系数,选择了BP神经网络参量,通过对金属橡胶的静态压缩实验数据进行参数识别,获得仅材料密度变化、材料密度和形状因子两种因素同时变化两种情况的BP神经网络预估模型,从而实现了对金属橡胶材料非线性本构关系的预估,通过实验进行验证,发现理论与实验结果吻合较好,说明采用BP神经网络预估金属橡胶材料的非线性本构关系是可行的.
BP (back propagation) neural network algorithm is used to study the constitutive relation coefficient of metal rubber. The BP neural network is introduced in detail. The neural network is used to estimate the nonlinear constitutive relation coefficient of metal rubber , BP neural network parameters are selected, BP neural network prediction model is obtained through static compression experimental data of metal rubber parameters identification, and only two changes of material density, material density and shape factor simultaneously, so as to achieve The prediction of non-linear constitutive relation of metallic rubber materials is verified by experiments. It is found that the theoretical and experimental results are in good agreement, which shows that it is feasible to predict the nonlinear constitutive relationship of metallic rubber materials by BP neural network.