摘要
基于BP神经网络,以Cl-、矿化度、电导率和地下水位为黄河口区域海水入侵评价指标,建立了具有8个隐含层节点、3层网络的海水入侵程度评价模型。应用海水入侵程度评价指标的各级评价标准作为模型的训练样本和检验样本,对BP神经网络进行训练和检验,并对黄河口区域的海水入侵程度进行评价。结果表明,BP神经网络对检验样本的模拟输出和期望输出一致,黄河口区域海水入侵程度比较严重。
We established a BP neural network based seawater intrusion degree evaluation model,which included 8 cryptic-layer nodes and 3 layers.We trained and tested the neural network with the rank assessment criteria of seawater intrusion degree assessment index as the training and test samples.We further evaluated seawater intrusion degree for Yellow River Mouth.Results show that the simulation result of BP neural network is consistent with the expected result.Seawater intrusion degree at Yellow River Mouth is more serious.
出处
《山东科学》
CAS
2014年第2期1-7,共7页
Shandong Science
基金
中央高校基本科研业务费专项资金(13CX06012A)
海洋公益性行业科研专项经费(200805063)
关键词
BP神经网络
海水入侵
黄河口
评价
BP neural network
seawater intrusion
Yellow River Mouth
evaluation
