Estimating of Scour in Downstream of the Water Level Regulation Structures

Scour in the downstream of hydraulic structures is a phenomenon which usually occurs due to exceed the velocity or shear stress from a critical level. In this research by using the laboratory information which was gathered by Borman- Jouline and De-Agostino, it was tried to get more accurate equations in order to calculate the maximum depth of scour in the downstream of the water level regulation structures. The comparing of these relations with the equations of the other researchers showed that these relations are much more accurate. Artificial neurotic networks (ANNs) with learning algorithm of error back propagation (BP) were used to estimate maximum water scour depth, and the model which has seven neurons in its latent layer was produced as the most appropriate model.Finally by using statistical parameters, the neurotic model was compared with optimized equations. The results of this study showed the more appropriate and higher accuracy of artificial neurotic networks.