Prediction models for estimation of exit gradient and uplift under the influence of downstream Filter

In present study, the impact of filter which located in downstream of hydraulic structures in reduction uplift pressure and hydraulic gradient was investigated. Impact filter effectively parameters such as length and distance from downstream structures and also upstream head was evaluated. The results of this study showed by design of filter with equal 0.057, 60% of downstream uplift pressure and 10% of upstream uplift pressure will be reduced and the effect of filter in uplift reduction in downstream of floor are impressive. By increasing the filter length, exit gradient always decreasing and with increasing distance of downstream (x), the effect of filter in reduction exit gradient increases. Design of Filter with equal 0.057 have good impact on exit gradient reduction (65%), witch by increasing length, gradient reduction will be reduced. Finally was used from regression and artificial intelligence models such as RBF, MLP and SVM for prediction of uplift and exit gradient in structure with filter. With comparing of this models in base of RMSE, MAE and, was introduced regression model as suitable model and SVM as poor model in prediction of uplift and gradient.