基于人工神经网络的崩岸预测模型研究

针对崩岸影响因素众多、难以定量评价等问题,采用一种基于人工神经网络的崩岸预测方法,介绍了其基本原理.在研究崩岸诸多影响因素的基础上,选取11种主要影响因素建立崩岸预测模型.以随机构造的20个岸坡模型为训练样本,以5个随机岸坡模型为检验样本,实现了岸坡稳定性的人工神经网络预测.初步结果表明,神经网络方法可以应用于崩岸的预测.

Artificial neural network-based bank collapse prediction model

In consideration of many influencing factors and some difficulties in quantitative evaluation,the artificial neural network(ANN)-based bank collapse prediction method was adopted,and its principle was introduced.By selection of 11 main influencing factors,a bank collapse prediction model was established. Then 20 bank-slope models,which were randomly generated,were taken as training samples,and 5 random bank-slope models as testing samples,the bank-slope stability was predicted by use of the present model.The result shows that the ANN method can be used for bank collapse prediction.