Multi-Objective Hydraulic Optimization of Diversion Dam’s Cut-Off

The seepage flow beneath a hydraulic structure is formed by the hydraulic head difference between the upstream and downstream sides. Cut-off walls are often applied, as an expedience, to reduce the seepage flow through the foundation of diversion dams and to enhance the efficiency of these dams. In this research, perhaps for the first time, a novel methodology is propounded to assess the optimum characteristics of cut-off walls in diversion dams in order to ameliorate hydraulic interactions between the diversion dam foundation and the cut-off walls behavior, also their construction cost is minimized. The results are used to train and validate the Multi-Layer Perceptron (MLP) simulation model. Then MLP, as a meta-model for simulation of the hydraulic behavior of cut-off walls, is coupled with a robust multi-objective optimization algorithm, Non-dominated Sorting Genetic Algorithm-ΙΙ (NSGA-ΙΙ), to create a trade-off between the intended goals. Finally, Preference Ranking Organization METHod for Enrichment Evaluation (PROMETHEE) decision making model and Nash-Harsanyi bargaining model are utilized to find the compromise design optimal solution on the trade-off curve. Results demonstrate that the best agreed-upon design optimal solution using PROMETHEE and Nash-Harsanyi bargaining models can be considered as (10, 3.84, 32) meters and (2.47, 10, 29.22) meters for optimum depth of the upstream and downstream cut-off walls and the optimum distance between them, respectively.