Multi-Objective Optimization of Seepage Control Structures of RCC Dams Based on the Surrogate Model

The single-objective and single-parameter optimization method is commonly used in the optimization of seepage control structures of roller-compacted concrete(RCC)dams, for which a seepage control structure design scheme that simultaneously satisfies multiple target requirements is difficult to obtain. Moreover, the method cannot fully consider the correlation between different design parameters, and its computational efficiency is low. In this study, a multi-objective optimization method based on the surrogate model was proposed for the seepage control structures of RCC dams. In the proposed method, ①a mathematical model, which utilizes the design parameters of the seepage control structure as the design variables and the maximum hydraulic gradient of the impervious layers, overflow elevation, and seepage control structure cost as the optimization targets, was established for the multi-objective optimization of the seepage control structures of RCC dams and ② a model solution method based on the multi-objective particle swarm optimization(MOPSO)algorithm and the surrogate model, in which the MOPSO algorithm was applied to rapidly solve the Pareto solution set, was proposed. Moreover, the surrogate model was used to establish an approximate relationship between the design parameters and the optimization objectives to replace the time-consuming seepage numerical model in the optimization process. Thus, the solution efficiency was improved and the correlation between different design parameters was fully considered. The proposed method was applied to optimize the seepage control structure of a practical RCC dam, and its effectiveness and advantages were validated by comparing with traditional method. © 2019, Editorial Board of Journal of Tianjin University(Science and Technology). All right reserved.