Elman网络在长江口南支悬沙浓度预测中的应用

Simulation of Suspended Sediment Concentration in the South Branch of Yangtze Estuary Using Elman Network

河口悬沙浓度的测定和模拟具有重要的意义。利用Matlab软件所提供的神经网络工具箱建立长江口南支的悬沙浓度预测模型,并选取一站位(31.5°N,121.4°E)进行仿真实验。结果表明,此模型可基本满足沉积动力学研究需要。此外,还发现转流前后的拟合效果欠佳。

Measurement and simulation of mouths of rivers are of great importance. Trials have been made to set up suspended sediment concentration （SSC） simulation model using the neural network toolbox provided by the software named Matlab. And one station （31.5°N, 121.4°E） has been chosen to conduct simulation experiments. Results suggest that the model can basically meet the needs of studies of deposition dynamics. Moreover, simulations during turns of tidal current are less satisfactory.