عنوان مقاله [English]
In this paper, genetic algorithm has been used for liquid sloshing modeling in the rectangular storage tanks. For this purpose, a numerical model was developed to model the sloshing phenomenon in the rectangular storage tank with different widths and water depths exposed to the horizontal periodic sway motion with different amplitudes and angular frequencies. In the numerical model, coupled finite element and boundary element methods were used to solve the governing equations. The governing equations are the Laplace equation and the dynamic free surface boundary condition. In order to validate the model, a rectangular tank with 0.9 m width and 0.6 m water depth was exposed to a horizontal periodic sway motion with 0.002m amplitude and 5.5 rad/sec angular
frequency. Finally, the results of the numerical model were compared with those of Nakayama and Washizu. Then, the numerical model was used to model the sloshing phenomenon in the rectangular storage tanks with different widths and water depths exposed to horizontal periodic sway motions with different amplitudes and angular frequencies. In the next step, a genetic algorithm method was used to model the sloshing phenomenon using the results of numerical model, and some relationships are presented to estimate the maximum free surface and the horizontal force is exerted on the tank perimeter. The correlation coefficient is used to evaluate the accuracy of the presented relations. The results show that genetic algorithm has good accuracy in predicting these two parameters. However, the accuracy of the model in the prediction of the maximum horizontal force exerted on the tank perimeter is more than that of prediction of the maximum free surface displacement. Because the correlation coefficient in the training and testing steps are 82 and 93 percent and 99 and 5 percent for these parameters, respectively. Furthermore, the percent relative errors of prediction of maximum free surface and horizontal force exerted on the tank perimeter are 11 and 5 percent, respectively. Therefore, the present model can be a power tool to model the sloshing phenomenon in the storage tanks.