论文部分内容阅读
坡面径流输沙能力是建立土壤侵蚀过程模型的重要水力学参数,研究定量计算坡面径流输沙能力的实用模型具有重要的理论和实践意义。通过室内模拟径流冲刷试验,计算不同坡度和流量条件下的裸地坡面径流输沙能力,利用平均影响值(MIV)方法对影响坡面径流输沙能力的因子进行分析,建立以干密度、能坡、进口流量、出口流量、水力半径、流速为输入,以坡面径流输沙能力为输出的广义回归神经网络(GRNN)模型,并应用Adaboost算法对模型进行优化。验证结果表明,所建模型能够用于对坡面径流输沙能力的模拟预测。与BP神经网络模型进行对比分析的结果表明:在试验训练样本条件下,广义回归神经网络(GRNN)模型的模拟预测结果优于BP神经网络模型;Adaboost算法能够有效减小广义回归神经网络(GRNN)模型的模拟预测误差。
The runoff and sediment transport capacity of slope runoff is an important hydraulic parameter for establishing the soil erosion process model. It is of great theoretical and practical significance to study the practical model for quantitative calculation of sediment transport capacity of runoff. Runoff scouring experiment was conducted in laboratory to calculate the runoff and sediment transport capacity of runoff on bare slope with different slope and flow rates. Factors influencing the runoff and sediment transport capacity of runoff were analyzed by means of average influence value (MIV) The model is generalized regression neural network (GRNN) with output slope, inlet flow rate, outlet flow rate, hydraulic radius and flow rate as input and output runoff and sediment capacity of slope runoff. Adaboost algorithm is used to optimize the model. The verification results show that the model can be used to simulate runoff sediment transport capacity. Compared with the BP neural network model, the results show that the generalized regression neural network (GRNN) model is superior to the BP neural network model under the experimental training sample conditions. The Adaboost algorithm can effectively reduce the generalized regression neural network (GRNN) ) Model of the simulation prediction error.