Optimization of horizontal drain dimensions in heterogeneous earth dams using Artificial Neural Network (ANN): A case study on Marvak dam

It is important to design and optimize the dimensions of the dam drainage system to keep the dam's downstream shell dry and to prevent the increase of pore water pressure in the earth dam body. It will also be possible to find the minimum factor of safety (FOS) to reduce construction costs by optimizing the drainage dimensions. In this study, Marvak earth dam was modeled by GeoStudio software with real material parameters, and by changing the dimensions of drainage, the material of the material, and slope of the dam, the minimum factor of safety of the dam was obtained. To predict the minimum factor of safety, the software results were used in different cases in the two-layer neural network. By training the neural network from the data obtained from the modeling of the Marvak dam, the minimum factor of safety for horizontal drainage was obtained. To optimize, a command appropriate to the neural network function is taught, by which the optimal values of the dam parameters are calculated. The results of the study show that the two factors of the internal friction angle of the drainage material and the slope of the dam have the greatest impact on determining the minimum factor of safety of the dam.