基于PCA-BP神经网络的水电站库区边坡稳定性分析  

Slope Stability Analysis of Hydropower Station Reservoir Area Based on PCA-BP Neural Network

作　　者：甘海龙[1] GAN Hailong(Guangxi Technological College of Machinery and Electricity,Nanning 530007,China)

机构地区：[1]广西机电职业技术学院,广西南宁530007

出　　处：《红水河》2023年第1期122-126,共5页Hongshui River

基　　金：2017年度广西高校中青年教师科研基础能力提升项目(2017KY075)。

摘　　要：为检验神经网络在水电站库区边坡稳定性预测的可行性,通过结合原始数据PCA及神经网络的模式识别特性,构建神经网络模型,对38组实测数据进行PCA处理,选取32组数据作为神经网络输入端数据、6组数据验证神经网络的工作性能。结果表明,神经网络对水电站库区边坡稳定性模式识别率达到83.3%。训练良好的神经网络可以用于工程实践中的水电站库区边坡稳定性预测。In order to test the feasibility of neural network in the prediction of slope stability in the reservoir area of hydropower station, a neural network model is constructed by combining the original data PCA and the pattern recognition characteristics of neural network. PCA processing is performed on 38 sets of measured data,and 32 sets of data are selected as the input data of neural network and 6 sets of data to verify the working performance of neural network. The results show that the recognition rate of neural network for slope stability of hydropower station reservoir area reaches 83.3%. The well-trained neural network can be used to predict the slope stability of hydropower station reservoir area in engineering practice.

关 键 词：PCA-BP 神经网络 边坡稳定性 水电站 库区 

分 类 号：TP642.2[自动化与计算机技术—控制理论与控制工程]