基于BP神经网络的边坡稳定性分析

The Stability Analysis on the BP Neural Network of Rock Slope

阐述了BP神经网络的基本思想、学习算法的步骤,以构建的学习样本为基础,建立边坡稳定性分析的BP神经网络模型,对学习样本进行归一化和训练,建立输入向量与输出向量的非线性关系,把训练好的网络运用于某露天矿边坡,结果表明:BP神经网训练结果与现场实际情况相符,说明该方法对工程实际有指导意义。

Abstract： This paper expounds the basic thought of BP neural networks and the steps of learning algorithm, and constructs the learning samples and establishes a BP neural network model of learning samples based on the analysis of slope stability. Then we can establish the nonlinear relation between input vector and output vector on normalization and training, and the trained network is used in one slope. The results show that BP neural network training results coqlorm to the actual condition, and illustrate the method is signifi- cant to the practical engineering.