Elman神经网络在平原区降水入渗补给预测中的应用

Application of Elman Neural Networks in Rainfall Infiltration Recharge Predication in Plain Area

在概括性地介绍了Elman神经网络的基本原理的基础上,以吉林省中部某平原区降水入渗补给的多年动态变化为例,建立了5-7-1结构的Elman神经网络动态预测模型。模型的检验结果表明,该模型的预测精度较高且能够反映该地区降水入渗补给的周期性变化特征。借此说明Elman神经网络在降雨入渗补给的多年动态变化预测中具有一定的实用价值,为Elman网络在其他领域的动态模拟应用提供参考。

Based on the introduction of the basic principles of the Elman neural network,taking the dynamic change of the rainfall infiltration recharge in the middle plain area of Jilin province as an example,the Elman neural network dynamic prediction model with 5-7-1structure for rainfall infiltration recharge is established in this paper.The model test results show that the prediction accuracy of the model is high and the model can reflect the cyclical change characteristic of rainfall infiltration recharge.So Elman neural network has some practical value in the rainfall recharge predication.The study result provides a reference for dynamic simulation applications with Elman neural network in other fields.