Optimized ultrathin hybrid sound absorption metasurfaces with preserved hydrostatic pressure-resistant

• The multi-parameter co-optimization is achieved in underwater hybrid metasurfaces by artificial neural network. • Differentiated mechanical energy flows from hybrid coupling effect promote sound absorption. • The introduction of honeycomb structure improves broadband impedance matching. • The synergistic resistance effect enhances hydrostatic pressure-resistant performance. High-efficiency waterborne sound absorption with a high hydrostatic pressure resistance is a crucial ability for underwater noise-control engineering. Herein, driven by artificial neural network (ANN), a desirable design method is proposed to construct ultrathin underwater acoustic hybrid metasurface with the characteristic of hydrostatic pressure resistance. As a demonstration, two hybrid metasurfaces (containing different proportions of cavities and scatterers) are designed, manufactured and experimentally measured, with all the functionalities displaying high-efficiency sound absorption (over 0.80 and 0.88 respectively) in 0.8–10 kHz and ultrathin thickness of 32 mm. The hybrid coupling effect reveals that the differentiation of mechanical energy flow (MEF) among the sub-surfaces can promote the sound absorption. Additionally, the introduced honeycomb structure plays an important role in good impedance matching of the hybrid metasurfaces. More importantly, due to the addition of matching cover layer, the synergistic resistance effect enhances the average sound absorption performances of the hybrid metasurfaces within 3 MPa hydrostatic pressure. This work provides more possibilities for the engineering applications of underwater metasurfaces.