基于改进BP神经网络的坡面雨滴溅蚀量研究

Study of rain plash-erosion on the slope by artificial neural network

介绍了雨滴溅蚀是一种极为普遍的坡地侵蚀,也是构成土壤侵蚀的重要组成部分。建立了以坡度、雨强、水深、单宽流量为输入,溅蚀量为输出的45 1的BP神经网络模型,坡面雨滴溅蚀量预测值与试验值线性回归值大于0 .9,提出了采用改进型BP神经网络坡面降雨模型预测坡面雨滴溅蚀量。

Rain splash-erosion is a complex process in which many factors are interactional. This paper analyzes the factors which affect the rain splash-erosion quantity, brings forward the BP neural network model which forecast the rain splash-erosion quantity, and validates it by an example. The calculation proves and effective. The BP neural network model after ameliorated has ampletude application to sloping field rain splash-erosion.