摘要
土的本构关系模型的研究是土力学研究中的一个关键问题。但由于土的性质受多方面因素的影响 ,如何找到一个既能准确描述土的本构关系又具有工程实用价值的模型一直是学者们探讨的课题。基于此 ,本文利用人工神经网络的方法建立了一个土的非线性本构关系模型 ,该模型考虑了剪应力对体应变和静水压力对剪应变的相互耦合作用 ,并成功的避免了传统本构模型的经验假设和简化 ,同时充分利用了实验的全部数据 ,提高了模型的精确度。模型亦具有很大的容错性。
The research of Soil Constitutive Relationship Model is an important problem in soil mechanics.Since soil behaviors are influenced by many factors,how to find a model,which can describe the soil constitutive relationship accurately,as well as be convenient for engineering application,is worthy our research energies all along.In this paper,making use of artificial neural network,we established a soil nonlinear constitutive model.This model has consitdered the coupling influence between q and ε\-v,and between p and ,meantime it successfully avoided the empirical assumptions and simplification in traditional modeling method.Using all the test data,definitely it has high precision and strong fault tolerance.
出处
《土工基础》
2000年第3期40-43,共4页
Soil Engineering and Foundation
