论文部分内容阅读
由于神经网络的非线性映射、自适应及自学习的能力,已越来越多地被应用于边坡的稳定性分析和变形预测中。文章建立了一个三层RBF神经网络模型,模拟了公路土质边坡各种地质影响因素与边坡稳定性之间的非线性关系;通过广泛收集原位测试数据、室内岩土实验数据作为训练及测试样本,并做归一化处理后作为神经网络的输入;利用训练好后的网络对公路边坡的稳定性进行了分析。
Due to the nonlinear mapping, adaptive and self-learning capabilities of neural networks, it has been increasingly used in slope stability analysis and deformation prediction. In this paper, a three-layer RBF neural network model is established to simulate the nonlinear relationship between various geological influencing factors and slope stability of highway earth slope. Through extensive collection of in situ test data and indoor geotechnical experimental data as training and Test samples, and make normalized input as neural network; using the trained network to analyze the stability of highway slopes.