基于最小二乘支持向量机的岩土本构关系曲线拟合

摘要

最小二乘支持向量机（Least Square Support Vector Machine,简称LSSVM）是对支持向量机（SVM）的一种改进,是继人工神经网络（ANN）之后又一新的机器学习方法。利用LSSVM拟合了利用Duncan-Chang本构模型的E-V模式数值模拟岩土的三轴实验得到的应力-应变函数曲线。通过对比分析,验证了LSSVM用于岩土本构关系曲线拟合的可行性和精确性。%The least square support vector machine （LSSVM） is an improved Support Vector Machine （SVM）, which is another new machine learning method in Artificial Neural Network （ANN）. In this paper, using LSSVM to fit stress-strain function curve based on the EV mode numerical simulation of geotechnical tri- axial test of Duncan-Chang constitutive model. Through comparative analysis, the author verified the feasibility and accuracy of LSSVM used for geotechnical constitutive relation curve fitting.