عنوان مقاله [English]
The, so called, barrage diverts water from rivers and steers it to the main channel. This requires careful design and analysis of hydraulic, hydrological and structural engineering. Barrage construction cost is not constant and, for a variety of different soil and hydrological conditions, the cost varies. The depth of sheet-piles/cutoffs and the length and thickness of the floor are the parameters influencing the cost of barrages. When a sheet-pile depth is changed, floor length and uplift force change over the floor, which has an impact on floor thickness and makes the cost of the barrage change to a non-linear form. In the present study, an optimization model is presented to fix the basic barrage parameters, which are depth of sheet piles/cutoffs, and length, and the thickness of the floor is based on Khoslas theory for subsurface flow. Basic barrage parameters need to be optimized, so that barrage costs for filling, xcavation, dewatering, concreting, sheet-piling and driving will be minimal, provided that the output of the hydraulic gradient does not exceed the permissible hydraulic gradient. Therefore, note that the objective function is nonlinear and the multivariate and nonlinear constraint of the problem is optimized using the genetic algorithm. The lgorithm, udiciously searches for optimal design within the problem space. Matlab software is a complete toolbox in the field of genetic algorithms, therefore, all procedures relating to the genetic algorithm is performed in Matlab software. Results of parametric analysis using the genetic algorithm show that a barrage founded on sand would cost less than a barrage on silt or silty clay for the same head, and the depth of the upstream sheet pile is sensitive neither to the value of permissible exit gradient nor to the seepage head.