OPTIMIZATION MODELING OF GRADE CONTROL STRUCTURES DESIGN TO STABILIZE RIVERS

Due to human activities or natural changes rivers are subject to erosion and sedimentation. One of the best ways to control erosion and sedimentation is through grade control structures. Grade control structures are, in fact, a kind of structures constructed along the river reach to decrease longitudinal bed slope and control erosion. They stabilize the riverbed by increasing resistance to ow shear stress. This research deals with developing an optimization model to design a special type of these structures with the above function, called bed sills. Bed sills are constructed transversely across a cross section with the same level as bed elevation to protect river bed from erosion. In this respect, optimization model is developed as composition of hydraulic and structure modules under objective functions of minimum construction cost and maximum structure stability. The result of model not only consists of optimized dimension of bed sill, which consists of width, height, distance and number, but also the costs of construction implementation and depth of scouring. The model is then calibrated for Dough river. Comparing the obtained results with Bed sills of Dough river shows that the optimization modeling of grade control structures is able to obtain variations of designs for the objectives of minimum construction cost and maximum structure stability. In the next step, sensitivity analysis is carried out on input variables of longitudinal slope and design discharge to understand the eect of variations of these on cost and dimensions of the structure. Further on, sensitivity analysis results are applied for validation of the model. Model's sensitivity in the longitudinal slope can be applied in the Mountain rivers. These rivers can choose equations such as Meyer-peter and Muller equation, because they reduce costs and increase stability of Bed sills. The results of this study can be used for the prediction of optimized dimensions of Bed sills.