岬间海湾岸线平衡形态神经网络模型

The Neural Network Model of Equilibrium Shape of Headland-Bay

讨论了目前海湾岸线平衡形态经验模型存在的不足之处。从沿岸输沙公式入手,阐述了平衡岸线的机理模式,重新定义了平衡海湾的“下岬角”,给出了模型参数与主波向的具体计算方法,并以华南典型海湾为学习样本,建立了岸线平衡形态的神经网络模型。通过模拟海湾与实际稳定海湾——乌场湾间的对比分析,表明所建神经网络模型是较抛物模型更为理想的平衡岸线模型。

Some disadvantages of the current empirical equilibrium form models are discussed in the paper. Based on the sediment longshore transport formulation, the paper expounds the equilibrium shoreline mechanical model, redefines the ＂down-headland＂, and provides the calculating method of the model parameters and resultant wave direction. With the South China typical headland-bays as samples, a neural network model of the equilibrium shape of headland-bay has been established. Through comparing the simulative bay and the factual one, the result indicates that the established model in the paper is better than the parabolic model.