Performance evaluation of a genetic algorithm-based linked simulation-optimization model for optimal hydraulic seepage-related design of concrete gravity dams

Concrete gravity dams (CGD) are classified as a strategic and essential structures in water resources management. Precise seepage analysis, construction cost and safety are the most important factors in the design and construction of CGD. The analytical solution and empirical seepage analysis methods under hydraulic structure are not sufficient or precise enough to provide an ideal solution for complex projects. This study concentrated on developing accurate surrogate models utilizing the Artificial Neural Network (ANN) technique, which are trained based on numerical simulated data sets generated by the seepage modelling software (SEEPW/Geo-Studio). The developed surrogate models are linked with the Genetic Algorithm (GA) optimization solver to optimize the hydraulic design considering the design safety factors and minimum construction cost of CGD. The performance of the linked simulation-optimization (S-O) model is evaluated for different design scenarios. The evaluation results demonstrate the potential applicability of the methodology for efficient, safe, and economical hydraulic design CGD on permeable soils.