BP模型在闽江岸线变形预测中的应用

河道岸线变形问题是一个复杂的非线性动力系统问题,文中利用神经网络处理非线性问题的优势,以闽江竹歧至侯官河段为研究对象,建立预测河道岸线变形的BP神经网络模型。通过对河道岸线变形影响因子的分析,确立了2种输入、输出因子模式。结果表明,模型输入、输出因子的选择对模型预测结果影响很大,选用合适的输入、输出因子,会得到比较好的预测效果。

Application of BP model to predicting deformation of bank-line of Minjiang river

Deformation of river bank-line is an complex problem of non-linear dynamical system.By using the advantage of neural network to handle non-linear problem,BP neural network model is established to predict deformation of river bank-line from Zhuqi to Houguan section of Minjiang River.By analyzing the factors' influencing the deformation of river bank-line,two input and output factors models are established.The study shows the selection of input and output models have significant influence on the results of model prediction.To select suitable input and output factors can get better prediction results.