Design of a polyurethane underwater sound absorber based on the BP neural network and genetic algorithm

Abstract

A coordinate transformation theory for the acoustic cloak design based on pentamode material was proposed by Norris in 2008. This theory avoids the problem of infinite density at the inner boundary of the inertial cloak, but the structural design of the pentamode material was difficult to be achieved. In order to solve this manufacturing problem, a design scheme of a low scattering cloak with only a few layers of polyurethane materials was proposed in this study. First, the material parameter distribution of the acoustic cloak was obtained based on the coordinate transformation theory. On this basis, the material parameter distribution in the two-dimensional annular acoustic cloak was derived by using linear mapping. Through simulation comparison between the uncovered acoustic cloak and covered acoustic cloak, the equivalent scattering characteristics of obstacles covered with the acoustic cloak were significantly reduced, which shows that anisotropic concealed materials have obvious advantages in acoustic control. When the thickness was given and the frequency was fixed, the sound pressure field of the cloak was simulated under the condition of uniform stratification. With the increase in the number of layers, the stealth effect was better, and the complexity of the preparation of metamaterials was increased. In order to obtain the cloak structure with a simple structure and good stealth performance, the BP neural network genetic algorithm and sound field finite element modeling were used. The minimum scattering area was taken as the optimization goal, and we tried to find the design scheme of a low scattering cloak with only a few layers of polyurethane materials. The optimization results show that when the target frequency is 2 KHz and the number of cloak layers is five, the optimized scattering area is 73% less than that of the uniformly layered structure. This shows that anisotropic stealth materials have obvious advantages in sound control. This study can guide the cell configuration design of polyurethane materials in different positions and greatly simplify the complexity of the manufacturing process of stealth materials.