نوع مقاله : برگرفته از پایان نامه
نویسنده
دانشجوی دکتری دانشگاه آزاد اسلامی واحد اراک
چکیده
کلیدواژهها
موضوعات
عنوان مقاله [English]
نویسنده [English]
Abstract: Scouring depends on several factors including the obstacles in water flow, sections, piers and foundations, disturbance of bed materials, soil permeability, non-parallelism between the piers and water flow, the type of river activity (static or dynamic), existence of a waterfall or an obstacle that results in the formation of a waterfall in the natural bed materials which causes the underlying bed materials to be washed away. Local scour around the bridge piers is caused by the collision of the flow with the piers and the separation of flow. The dimensions of the scour hole created around the bridge piers depend on the hydraulic characteristics of the flow, fluid characteristics, the geometry of pier and characteristics of the bed materials. Feed forward optimal network with the error back-propagation training algorithm with sigmoid transfer functions has been used for four models mentioned in the paper. To determine the number of neurons in the hidden layer, one and ten neurons were selected in the hidden layer according to the verification indicators. Considering the network structure of one neuron in the hidden layer, a comparison was made between dimensional and dimensionless parameter models effective in the dimensions of the scour hole. The comparison of artificial neural network and nonlinear regression models shows a higher accuracy and capability of ar-tificial neural networks in the simulation of phenomenon. Also the R and RMSE values were obtain from 0.93 to 0.98 and 0.18 to .013 respectively
کلیدواژهها [English]
)