Multi-Objective Optimization of Detention Rockfill Dam Characteristics Considering Hydraulic Conditions and Application of Borda Count Social Choice and Nash- Harsanyi Bargaining Models

Floods inflict a large amount of loss on communities and installations in Iran due to the geological characteristics and environmental destructions. Detention rockfill dams are important structural methods to mitigate flood damages. So, it is necessary to assess the optimum characteristics of detention rockfill dams considering their hydraulic performance. In this paper, a novel methodology based on data-driven simulation model, NSGA-II multi-objective optimization model and game theory is suggested in order to find the optimum features of detention rockfill dams. The proposed model is on the basis of the multi-objective genetic algorithm model (NSGA- ΙΙ) and the multilayer perceptron neural network meta-model (MLP). The MLP meta model simulates the relationship between design variables of detention rockfill dams and the hydraulic parameters of flood hydrograph passing through the porous media of the dam. The MLP based simulation model is trained and validated on the basis of experimental results of a physical model of detention rockfill dam. The validated MLP model is then coupled with a robust multi objective optimization algorithm, NSGA-ΙΙ, to determine the trade-off curve between the conflict objectives of the optimization. Nash- Harsanyi bargaining model and Borda count social choice method are used to find the best agreed-upon design point on the trade-off curve of the optimization objectives. Results indicate that the optimum design variables can be considered as [27.1, 27.58] cm and [2.02, 4.52] cm for the optimum thickness of detention rockfill dam and the optimum diameter of aggregates in the porous media, respectively, using both the mentioned methods. The similar optimum characteristics of detention rockfill dam using Nash bargaining model and Borda count social choice method depict the stability of the applied methods.