摘要
在简要介绍遗传神经网络的基本概念及学习步骤的基础上,分别对大坝坝顶径向水平位移、切向水平位移和大坝坝顶沉降量监测数据进行了训练和预测。结果表明,利用遗传算法特有的全局优化能力,可以较好地完成网络的学习,而且还减少了网络训练次数,缩短了网络训练时间。
On the basis of brief introduction of the principle and algorithm of genetic neural network, the forecast model is applied to the monitoring data of tangential horizontal displacement, radial horizontal displacement and settlement of a dam. The result indicates that the forecast model could complete the train and forecast of the network with less training frequency and time by means of overall optimizing ability of genetic algorithm.
出处
《岩土力学》
EI
CAS
CSCD
北大核心
2003年第S1期130-133,共1页
Rock and Soil Mechanics
关键词
人工神经网络
遗传算法
大坝变形监测
预测
artificial neural network
genetic algorithm
dam deformation monitoring
forecast
