Arch Dam Optimization Considering Fluid-structure Interaction with Frequency Constraints Using Artificial Intelligence Methods

ABSTRACT: An efficient method is proposed to find optimal design of arch dams on the basis of constrained natural frequencies utilizing continuous evolutionary algorithm. To extract natural frequencies of arch dam considering fluid-structure interaction, it is necessary to solve the unsymmetrical damped eigenproblem. This means that the process of natural frequencies extraction may impose much computational effort. This deficiency can be resonated when a grate number of structural analyses are needed during the optimization process. In order to reduce the computational cost of the optimization problem, the natural frequencies of arch dam are predicted by properly trained back propagation (BP) and wavelet back propagation (WBP) neural networks. The presented WBP network appears better performance generality than BP network. The numerical results reveal the computational advantages of the proposed methods for optimal design of arch dams.