Optimum Shape Design of Arch Dams by a Combination of Simultaneous Perturbation Stochastic Approximation and Genetic Algorithm Methods

Abstract

An efficient method is proposed to find optimal shape of arch dams subjected to response spectrum loading. The optimization is performed by a combination of simultaneous perturbation stochastic approximation (SPSA) and genetic algorithm (GA) methods. This new method is called simultaneous perturbation genetic algorithm (SPGA). Operation of SPGA includes three phases. In the first phase, a preliminary optimization is accomplished using SPSA. In the second phase, an optimal initial population is produced using the first phase results. In the last phase, GA is employed to find optimum design using the optimal initial population. The numerical results reveal the robustness and high performance of the proposed method for optimum shape design of arch dams. The optimum design obtained by SPGA is compared with those of SPSA and GA. It is demonstrated that the SPGA converges to better solution compared to SPSA and GA by spending lower computational cost.