当前位置: X-MOL 学术Ain Shams Eng. J. › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
Best non scour channel section design using artificial neural networks
Ain Shams Engineering Journal ( IF 6 ) Pub Date : 2020-11-11 , DOI: 10.1016/j.asej.2020.08.020
Ahmed M. Tawfik

Water flowing in channels has the ability to scour sediments from bed, berm, and banks and sweep these sediments downstream causing a lot of problems. Using Shield’s diagram and Manning’s equation, the design for trapezoidal cross sections is determined. This design ensures no motion of sediments. Matlab code is written to solve the design equations using trial and error. Design charts that determine the best design are presented. The best hydraulic section design is the worst design from the scour point of view. The effect of each design variable is studied. Increasing particle diameter or specific gravity of sediments leads to decrease in the bed width and increase in the corresponding water depth. Artificial Neural Networks are then used to find the best design. About 30,000 synthetic realizations were used to train, validate, and test the network. The proposed design aids help the design engineer to find the best design easily.



中文翻译:

使用人工神经网络的最佳非冲刷通道截面设计

在河道中流动的水能够冲刷河床、护堤和河岸上的沉积物,并将这些沉积物扫到下游,造成很多问题。使用 Shield 图表和曼宁方程,确定梯形横截面的设计。这种设计确保没有沉积物的运动。编写 Matlab 代码以使用试错法求解设计方程。提供了确定最佳设计的设计图表。从冲刷的角度来看,最好的液压部分设计是最差的设计。研究了每个设计变量的影响。增加沉积物的粒径或比重导致床宽减小,相应的水深增加。然后使用人工神经网络来寻找最佳设计。大约 30,000 个合成实现被用于训练、验证、并测试网络。建议的设计辅助工具可帮助设计工程师轻松找到最佳设计。

更新日期:2020-11-11
down
wechat
bug