以长江下荆江河段较为典型的石首弯道为研究对象 ,研究建立了一种基于BP神经网络的河道岸线变形预测模型。计算表明 ,该模型能较为准确地模拟预测石首弯道进出口段典型断面岸线变化。模拟精度较高 ,便于应用 Taking the typical Shishou bend in Jingjiang reach of the Yangtze River as the research object, a forecasting model of riverbank deformation based on BP neural network was established. The calculation shows that the model can simulate and predict the change of typical cross-section shoreline at the inlet-outlet section of Shishou curve more accurately. High simulation accuracy, easy to use