BP神经网络模型应用于大坝原型观测研究  

Application of BP Neural Network Model in Dam Deformation Monitoring

作　　者：李少华[1] 董增川[1] 

机构地区：[1]河海大学水资源环境学院,南京210098

出　　处：《水利与建筑工程学报》2004年第2期31-33,共3页Journal of Water Resources and Architectural Engineering

摘　　要：针对传统的多元线性回归方法的不足,提出将BP神经网络模型应用于大坝原型观测数据处理。并介绍了BP神经网络模型的结构和算法,以及训练样本的归一化处理方法;将BP模型应用于黑河金盆水库大坝原型观测,结果表明,其模拟和预测效果均优于多元线性回归模型。Aiming at the limitation of traditional method of statistical model, BP neural network model is advanced for the analysis of dam deformation monitoring data in this paper. Structure and algorithm of BP neural network model are introduced together with the normalization of sample data. The application of this model in Heihe Jinpen Reservoir shows that its precision is prior to that of statistical model.

关 键 词：大坝原型观测 BP神经网络 归一化 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术] TV697.2[自动化与计算机技术—计算机科学与技术]