نوع مقاله : مقاله پژوهشی
1 دانشگاه تبریز
2 Tabriz University, Agricultural faculty, Irrigation department
عنوان مقاله [English]
This study investigates the impact of filter which is located in downstream of the hydraulic structures for reduction of uplift pressure and hydraulic gradient. Effective parameters for design of filter are: length of filter (L), distance from downstream of structure (X) and upstream water head (H). The outcomes of this study showed that design of filter with L/H equal to 0.057, results 60% reduction of downstream uplift pressure and 10% reduction of upstream uplift pressure. Thus the effect of filter in uplift pressure in downstream of floor is impressive. By increasing the filter length, exit hydraulic gradient always decreases and with increasing the distance of downstream (X), the effect of filter in reduction exit gradient increases. Design of filter with L/H equal to 0.057 have a good impact on exit hydraulic gradient reduction (65%), witch by increasing the length of filter, hydraulic gradient reduction will be reduced. Finally, regression and artificial intelligence models (RBF, MLP and SVM) were used for prediction of uplift pressure and exit hydraulic gradient in structure with filter. Comparison of these models base on two error measurements (R2, RMSE and MAE) demonstrated that regression model is a suitable model and SVM as a poor model in prediction of uplift pressure and hydraulic gradient.