|
چکیده
|
In a compound channel, there is a noticeable diference in streamwise velocity considering the flow depth diference between the main channel and its floodplain. This velocity diference leads to momentum exchange and accordingly, creation of cross flow in the compound channel. Floodplain width infuences the dimensions and expansion of low-velocity zone in a compound channel; therefore, changes of width can afect the flow pattern in compound channels. The creation of cross flow changes the efective parameters on flow pattern in these channels. One of the most important parameters is cross flow power, which can indicate the strength of cross flow in a compound channel. To obtain these parameters, numerical and experimental models or numerical calculation-based models can be used. Artifcial neural network (ANN) considering various widths of floodplains and flow rates is used in this article to determine the cross flow power in compound channels. Then, the results of artifcial neural networks are compared with those of experimental ones, and the best simulation conditions are selected to calculate the cross flow power for various widths of floodplain which are inaccessible
|