岩土本构模型智能识别的若干研究

SOME STUDIES ONINTELLIGENT IDENTIFICATION OF GEO-MATERIAL CONSTITUTIVE MODEL

岩土本构模型识别是反分析研究的一个重要方面,智能方法的引入极大地推动了其研究。主要从以下6个方面介绍了目前本构模型智能识别的一些研究进展及可能方向:本构模型的神经网络逼近;本构模型的神经网络描述;模型类内进化识别:模型类间进化识别;模型形式的进化识别;模型形式的进化重构。

The identification of the geo-material constitutive model is very important in back analysis, and the introduction of the intelligent methods has greatly stimulated the research. In this paper, the present research advancement and possible directions of this study are introduced in the following aspects: neural network approach of the constitutive model, neural network description of the constitutive model, identification of within-class evolution of the model, identification of between-class evolution of the model, evolution identification of the model form, and evolutional reconstruction of the model form.