本论文采用人工神经网络的一个模型—CP网络来进行岩质边坡稳定性判别和砂土液化预测。本论文有CP网络成功地建立土体的各影响因素与判别之间的高度非线性映射。应用结果表明 ,CP网络为木工结构模式分类提供了良好的应用前景 ,使模式分类过程更为直观、简化和精确 In this thesis, a model of artificial neural network-CP network is used to judge the stability of rock slope and predict the sand liquefaction. In this thesis, CP network successfully establishes a highly nonlinear mapping between all factors and discriminant of soil. The application results show that the CP network provides a good application prospect for the classification of carpentry structure patterns and makes the pattern classification process more intuitive, simplified and accurate