Spherical segment-shaped concave surface underwater acoustic transducer matching layer based on asymmetric metasurface

Abstract

The invention provides a spherical segment-shaped concave surface underwater acoustic transducer matching layer based on an asymmetric metasurface, which comprises a cylinder structure with an inward concave spherical segment-shaped concave surface at the top, and is characterized in that the cylinder structure is formed by alternately and periodically arranging rigid material layers and flexible bonding layers along the thickness direction; the rigid material layer and the flexible bonding layer have the same concave surface radian; a plurality of hemispherical concave hole units are distributed on the spherical segment-shaped concave surface of the rigid material layer on the topmost layer, and the plurality of concave hole units form an asymmetric metasurface. Through periodic impedance transition and a geometric focusing effect, the sound wave transmission efficiency and the working bandwidth are remarkably improved, and an angle-frequency dependent unique transmission response is shown in a frequency band of 69-79 kHz. The problems that a traditional matching layer is low in transmission efficiency, limited in bandwidth and lack of directional recognition capacity are solved, and the method is suitable for underwater acoustic communication, detection and positioning systems.

Description

Technical Field

[0001] This invention relates to the technical fields of underwater acoustic engineering and materials science, specifically to the application of metamaterials in underwater acoustics, the optimized design of acoustic transducers, and in particular to a matching layer for a spherical cap concave underwater acoustic transducer based on an asymmetric metasurface. Background Technology

[0002] In underwater acoustic systems, the underwater acoustic transducer is the core component for converting acoustic energy into electrical energy, and it is widely used in underwater acoustic communication, underwater detection, marine monitoring, and military target identification. However, the transducer's performance in water is significantly affected by acoustic impedance mismatch. The acoustic impedance of water (≈1.5 × 10⁻⁶) 6 ) and commonly used materials for transducers (such as piezoelectric ceramics ≈ 3 × 10 7 The significant differences between the transducer and the transducer cause strong reflections of sound waves on the transducer surface, reducing the sound energy transmission efficiency and thus limiting the transducer's detection sensitivity and signal transmission quality.

[0003] Traditional methods typically employ matching layer technology, introducing a material with intermediate acoustic impedance between the transducer and water to reduce reflection and increase sound wave transmittance. To further improve the transmittance coefficient, a structure with gradient impedance matching is usually designed, allowing the sound wave to gradually transition between interfaces and reducing reflection. Common matching layer materials include epoxy resin and silicone rubber. While they improve transducer performance to some extent, they still have limitations in areas such as broadband matching, high-frequency response, and low loss.

[0004] In recent years, metamaterials have demonstrated excellent controllability in the field of acoustics. Acoustic metasurfaces are artificial structures with subwavelength scales, capable of precisely controlling the propagation direction, phase, and amplitude of sound waves by designing their geometry and material distribution. Asymmetric metasurfaces, as an important branch, can break the symmetry of sound wave propagation due to their asymmetric structure, achieving directional sound wave transmission and efficient impedance matching. Compared to traditional matching layers, spherical cap concave matching layers based on asymmetric metasurfaces exhibit unique advantages in achieving smaller thickness, wider frequency response, and lower sound wave reflection. Simultaneously, their asymmetric characteristics can also be used to achieve directional acoustic signal reception. By rotating the hydrophone, the system's ability to locate the sound source can be further enhanced, expanding its application potential in underwater target identification and sonar systems. Therefore, introducing asymmetric metasurface technology into the matching layer design of underwater acoustic transducers not only helps to solve the performance bottlenecks of traditional matching layers but also provides a technical path for the high performance, miniaturization, and intelligence of underwater acoustic devices.

[0005] Patent application number 202511084611.3 discloses a matching layer structure for a high-frequency underwater acoustic transducer used in deep-sea exploration. This structure includes an integrated gradient functional material layer whose acoustic impedance varies gradient along the sound wave propagation direction. The material layer also contains embedded pressure-adaptive microstructures. This invention addresses the issue of the comparative patent lacking a pressure-resistant structure, preventing acoustic performance drift caused by deformation under 60MPa high pressure. It utilizes a negative Poisson's ratio honeycomb matrix that expands rather than contracts under 60MPa hydrostatic pressure to resist compression deformation. The pressure-responsive silicone oil filling the microsphere system increases in density with increasing pressure, automatically compensating for changes in sound velocity and ensuring minimal acoustic impedance fluctuations. This supports long-term operation at depths of up to 6000 meters, solving the performance degradation problem of existing transducers in deep seas, while avoiding the micron-level displacement risk associated with the filled cone in the comparative patent. The aforementioned patent uses a gradient functional material layer (gradient material properties, not a structure) to achieve acoustic impedance gradient changes; and embedded microstructures (honeycomb matrix, silicone oil-filled microspheres) to achieve pressure adaptation, which presents significant manufacturing challenges. Summary of the Invention

[0006] This invention proposes a matching layer for a spherical cap concave underwater acoustic transducer based on an asymmetric metasurface, to address the problems of low transmission efficiency, limited bandwidth, and lack of direction recognition capability in traditional matching layers. Through a special geometric structure and material distribution design, the matching layer of this invention achieves high acoustic transmission coefficient, wide bandwidth response, and incident angle recognition functionality.

[0007] To achieve the above objectives, the technical solution of the present invention is as follows: a matching layer for a spherical cap concave underwater acoustic transducer based on an asymmetric metasurface, comprising a cylindrical structure with an inwardly concave spherical cap at the top, the cylindrical structure being composed of a rigid material layer and a flexible adhesive layer arranged alternately and periodically along the thickness direction. Both the rigid material layer and the flexible adhesive layer have the same concave curvature. Multiple hemispherical concave units are distributed on the spherical cap concave surface of the top rigid material layer, and the multiple concave units constitute an asymmetric metasurface.

[0008] Preferably, the rigid material layer is an aluminum layer, which provides strong acoustic contrast in the periodic structure and provides strong skeletal support for the entire concave structure; the flexible adhesive layer is made of epoxy resin composite material to achieve precise bonding and harmonize acoustic impedance.

[0009] Preferably, the concave hole units are arranged in an array along the hemispherical concave surface, and the size of the concave hole units gradually changes along the radial direction. The size of the concave hole units in the left and right regions is different, so that the unit sizes in the left and right regions are different, forming an orientation-dependent asymmetric structure.

[0010] Preferably, the lower part of the periodically alternating rigid material layer and flexible adhesive layer is provided with piezoelectric ceramic, and the cylindrical structure and piezoelectric ceramic are disposed inside the shell; The outer shell is made of a flexible polyacrylate material with a thickness of 2mm.

[0011] Preferably, the recessed hole units are evenly distributed in a ring, and the diameter of the recessed hole unit on one side of the ring is larger than the diameter of the recessed hole unit on the other side.

[0012] Preferably, the concave curvature of the rigid material layer and the flexible adhesive layer is 0.321π rad, and the radius of the spherical cap concave surface of the top layer is 12.59 mm.

[0013] Preferably, the cylindrical structure has eight layers from top to bottom, including four flexible adhesive layers with spherical cap concave surfaces of the same thickness located in the second, fourth, sixth, and eighth layers, with a thickness of d1, and four aluminum layers with spherical cap concave surfaces of the same thickness located in the first, third, fifth, and seventh layers, with a thickness of d2; the surface of the aluminum layer in the first layer has an asymmetric metasurface structure composed of multiple hemispherical concave hole units.

[0014] Preferably, the distribution rule of the hemispherical concave hole units on the spherical cap concave surface of the top aluminum layer is as follows: With the center of the spherical concave surface of the top aluminum layer as the base point, there is only one hemispherical concave hole b6, with the center of the sphere located at the base point and a radius of r6; Taking the center of the spherical cap concave surface of the top aluminum layer as the origin, rotating the base point counterclockwise by 10° gives the center of the hemispherical concave hole b5 with radius r5; rotating the base point counterclockwise by 20° gives the center of the hemispherical concave hole b4 with radius r4; rotating the base point counterclockwise by 30° gives the center of the hemispherical concave hole b3 with radius r3; rotating the base point counterclockwise by 40° gives the center of the hemispherical concave hole b2 with radius r2; rotating the base point counterclockwise by 50° gives the center of the hemispherical concave hole b1 with radius r2. The center of concave hole b1; rotating the base point clockwise by 10° gives the center of hemispherical concave hole b7 with radius r7; rotating the base point clockwise by 20° gives the center of hemispherical concave hole b8 with radius r8; rotating the base point clockwise by 30° gives the center of hemispherical concave hole b9 with radius r9; rotating the base point clockwise by 40° gives the center of hemispherical concave hole b10 with radius r10; rotating the base point clockwise by 50° gives the center of hemispherical concave hole b11 with radius r11. And r1=d1×1.6, r2=d1×1.5, r3=d1×1.4, r4=d1×1.3, r5=d1×1.2, r6=d1×1.1, r7=d1, r8=d1×0.9, r9=d1×0.8, r10=d1×0.7, r11=d1×0.6; Hemispherical concave hole b1, centered at the first origin, rotates by 18° to construct the concave hole array of the outermost first ring in the left region; hemispherical concave hole b2, centered at the second origin, rotates by 18° to construct the concave hole array at the left region position of the second ring; hemispherical concave hole b3, centered at the third origin, rotates by 26° to construct the concave hole array at the left region position of the third ring; hemispherical concave hole b4, centered at the fourth origin, rotates by 26° to construct the concave hole array at the left region position of the fourth ring; hemispherical concave hole b5, centered at the fifth origin, rotates by 60° to construct the concave hole array at the left region position of the fifth ring; hemispherical concave hole b7, centered at the fifth origin, rotates by 60° to construct the concave hole array at the right region position of the fifth ring; hemispherical concave hole b8, centered at the fourth origin, rotates by 26° to construct the concave hole array at the right region position of the fourth ring; Spherical concave hole b9, centered on the third origin, rotates by 26° to construct a concave hole array in the right region of the third ring; hemispherical concave hole b10, centered on the second origin, rotates by 18° to construct a concave hole array in the right region of the second ring; hemispherical concave hole b11, centered on the first origin, rotates by 18° to construct a concave hole array in the outermost first ring of the right region; the first origin, second origin, third origin, fourth origin, and the... All five origin points are located on a straight line between the origin and the base point. The first origin point is at the same height as the center of hemispherical concave holes b1 and b11, the second origin point is at the same height as the center of hemispherical concave holes b2 and b10, the third origin point is at the same height as the center of hemispherical concave holes b3 and b9, the fourth origin point is at the same height as the center of hemispherical concave holes b4 and b8, and the fifth origin point is at the same height as the center of hemispherical concave holes b5 and b7.

[0015] Preferably, the method for identifying the incident direction of sound waves includes the following steps: S1: The piezoelectric ceramic in the cylindrical structure of the underwater acoustic transducer receives the sound wave at the initial orientation to obtain the first response value; S2: After rotating the cylindrical structure 180° around its central axis, the sound wave is received to obtain the second response value; S3: Input the feature vector formed by the first response value and the second response value into the trained deep learning model, and output the predicted sound wave incident angle by the deep learning model.

[0016] Preferably, the deep learning model is trained by using the acoustic response features under different incident angles as training samples and the corresponding incident angles as labels; the deep learning model is a multilayer perceptron, with two neurons in the input layer, two neurons in the hidden layer and ReLU activation function, and one neuron in the output layer; During the training of the deep learning model, mean squared error is used as the loss function; the Adam optimizer is selected to independently maintain the exponential moving average of the first and second moments of each parameter in the deep learning model, and the adaptive parameter update step size is dynamically calculated based on the second moment.

[0017] The beneficial effects of this invention: The matching layer is composed of four aluminum layers and four epoxy resin bonding layers arranged alternately and periodically with the same curvature, forming an overall cylindrical structure with a concave top, consisting of eight layers from top to bottom. The surface of the first aluminum layer has an array of hemispherical concave holes with gradually varying sizes along the meridian direction, forming an asymmetric metasurface. This invention's structure significantly improves acoustic wave transmission efficiency and operating bandwidth through periodic impedance transition and geometric focusing effects, exhibiting a unique angle-frequency dependent transmission response in the 69–79 kHz frequency band. Simultaneously, its asymmetric characteristics allow the underwater acoustic transducer to generate differentiated acoustic signals under different incident directions. Combined with a back-end deep learning model for processing, high-precision identification of the acoustic wave incident direction can be achieved without mechanical scanning. This invention overcomes the problems of low transmission efficiency, limited bandwidth, and lack of directional identification capability in traditional matching layers, making it suitable for underwater acoustic communication, detection, and positioning systems. The specific advantages of this invention are: 1. The structure design employs a periodic arrangement of aluminum layers and adhesive layers, and improves sound energy transmission efficiency through periodic impedance matching; 2. The design of the spherical cap concave surface structure achieves a high transmission coefficient through geometric focusing effect and gradual matching of acoustic impedance; 3. Asymmetric metasurface structures further enhance the transmission coefficient through subwavelength-scale unit design and precise control over acoustic waves; 4. Asymmetric metasurface structures, through their orientation-dependent acoustic response, enable a single underwater acoustic transducer to identify the direction of sound wave incidence without the need for angle-by-angle mechanical scanning. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a three-dimensional schematic diagram of the present invention.

[0020] Figure 2 This is a top view of the present invention.

[0021] Figure 3 This is a cross-sectional view of the present invention.

[0022] Figure 4 This is a schematic diagram of the metasurface unit structure of the present invention.

[0023] Figure 5 This is a comparison of the transmission spectra of the present invention with those of a planar underwater acoustic transducer matching layer and a underwater acoustic transducer matching layer with a spherical concave surface without metasurface.

[0024] Figure 6 The transmission spectrum of this invention is shown at 69-79 kHz.

[0025] Figure 7 This is a schematic diagram of the acoustic analysis of X-rays at 0° incident angle according to the present invention.

[0026] Figure 8 This is a schematic diagram illustrating the principle of incident sound wave angle recognition in this invention.

[0027] Reference numerals in the attached figures: 1 is the first aluminum layer (with metasurface structure), 2 is the first adhesive layer, 3 is the second aluminum layer, 4 is the second adhesive layer, 5 is the third aluminum layer, 6 is the third adhesive layer, 7 is the fourth aluminum layer, 8 is the fourth adhesive layer, 9 is the piezoelectric ceramic, and 10 is the outer shell. Detailed Implementation

[0028] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention. Example

[0029] like Figure 1 As shown, a matching layer for a spherical cap-shaped concave underwater acoustic transducer based on an asymmetric metasurface is presented. The overall structure is a cylindrical form with an inwardly concave spherical cap at the top, composed of alternating rigid material layers and flexible adhesive layers arranged periodically along the thickness direction; the top forms the inwardly concave spherical cap. Compared to planar structures, the spherical cap-shaped concave surface has a larger acoustic wave receiving area, better adapting to incident sound waves at different angles and improving transmission efficiency. Furthermore, it can alter the propagation path and phase distribution of sound waves within the matching layer, helping to broaden the operating bandwidth. Simultaneously, the spherical cap-shaped concave surface structure provides better overall mechanical stability, enabling it to withstand higher hydrostatic pressure.

[0030] Both the rigid material layer and the flexible adhesive layer have the same concave curvature, ensuring that the wavefront phase of the sound wave is precisely, consistently, and continuously modulated in three-dimensional space as it passes through the periodic structure composed of materials with different acoustic impedances in the matching layer. This is akin to passing through a carefully calibrated curved lens, avoiding phase disturbances, scattering, and energy loss caused by interlayer geometric mismatch. This forms an optimal pressure-bearing structure similar to a concentric arch, allowing each layer to evenly distribute hydrostatic pressure, greatly improving resistance to buckling and delamination.

[0031] Multiple hemispherical concave units are distributed on the spherical cap concave surface of the top rigid material layer, and these concave units constitute an asymmetric metasurface.

[0032] The concave surface of the spherical cap contains multiple hemispherical recessed units distributed according to a predetermined rule, which are used to form an asymmetric metasurface structure. The recessed units are arranged in an array along the hemispherical concave surface. The size of the recessed units gradually changes along the meridian direction, and the size of the recessed units in the left and right regions differs, resulting in different unit sizes in the left and right regions, forming a direction-dependent asymmetric structure, which can achieve a direction-dependent acoustic response.

[0033] The rigid material layer is an aluminum layer, providing strong acoustic contrast in the periodic structure and serving as the physical basis for asymmetric metasurface acoustic wave modulation. Simultaneously, it provides strong skeletal support for the entire concave structure, effectively resisting deformation and hydrostatic pressure. The flexible bonding layer is made of epoxy resin composite material, achieving precise bonding and harmonizing acoustic impedance. The spherical cap concave structure, as the pre-contact structure of the underwater acoustic transducer, requires acoustic bonding with the transducer material. Epoxy resin composite material is used as the acoustic adhesive to ensure good coupling between the aluminum layers, as well as between the matching layer and the piezoelectric ceramic 9, reducing sound wave loss at the bonding interface. It is encapsulated with epoxy resin composite material. The periodically alternating aluminum layers, flexible bonding layer, and piezoelectric ceramic 9 are disposed within the housing 10. The piezoelectric ceramic 9 converts acoustic signals into electrical signals. This invention uses a 2mm thick flexible polyacrylate material to fabricate the housing 10, encapsulating the matching layer and piezoelectric ceramic 9 in the transducer assembly.

[0034] like Figure 2 As shown, the top of the matching layer is a spherical concave structure, and the hemispherical concave hole units on its surface are arranged in a specific pattern to form an asymmetric metasurface, which is used to control the direction of sound wave propagation. Figure 2The diagram illustrates the arrangement of the concave hole units, which gradually change in diameter along the radial direction. The concave hole units on the left side are larger in diameter than those on the right side, creating a direction-dependent asymmetric structure. The outer ring line represents the edge of the matching layer's outer shell 10, and the radius of the first aluminum layer is marked as 12.59 mm, a dimension that determines the effective acoustic aperture of the metasurface region. This asymmetric structure, by designing metasurface units with different acoustic properties on the left and right sides of the concave surface, causes differentiated modulation of the local resonances and propagation phases excited within the structure when sound waves are incident from different directions. This difference in spatial modulation ultimately manifests as an asymmetric "acoustic fingerprint" in the transmission spectrum or received signal that uniquely corresponds to the incident direction, enabling differentiated responses to sound waves incident from different directions and further enhancing its direction recognition capability. Figure 2 The structural layout of asymmetric metasurfaces in underwater acoustic transducers is clearly demonstrated. High-precision machining of complex curved surface structures and internal microstructures can be achieved using 3D printing or additive manufacturing technologies.

[0035] The recessed hole units are evenly distributed in a ring, and the diameter of the recessed hole unit on one side of the ring is larger than the diameter of the recessed hole unit on the other side.

[0036] The distribution pattern of hemispherical concave holes is as follows: Figure 4 As shown, taking the center position of the surface of the spherical cap concave structure of the uppermost first aluminum layer 1 as the base point, as... Figure 2 As shown, there is only one hemispherical concave hole b6 here, with the center of the sphere located at the base point and a radius of r6.

[0037] Taking the center of the spherical concave structure of the first aluminum layer 1 as the origin, rotating the base point counterclockwise by 10° gives the center of the hemispherical concave hole b5 with radius r5; rotating the base point counterclockwise by 20° gives the center of the hemispherical concave hole b4 with radius r4; rotating the base point counterclockwise by 30° gives the center of the hemispherical concave hole b3 with radius r3; rotating the base point counterclockwise by 40° gives the center of the hemispherical concave hole b2 with radius r2; and rotating the base point counterclockwise by 50° gives the center of the hemispherical concave hole b1 with radius r1. Rotating the base point clockwise by 10° gives the center of a hemispherical concave hole b7 with radius r7; rotating the base point clockwise by 20° gives the center of a hemispherical concave hole b8 with radius r8; rotating the base point clockwise by 30° gives the center of a hemispherical concave hole b9 with radius r9; rotating the base point clockwise by 40° gives the center of a hemispherical concave hole b10 with radius r10; rotating the base point clockwise by 50° gives the center of a hemispherical concave hole b11 with radius r11.

[0038] And r1=d1×1.6, r2=d1×1.5, r3=d1×1.4, r4=d1×1.3, r5=d1×1.2, r6=d1×1.1, r7=d1, r8=d1×0.9, r9=d1×0.8, r10=d1×0.7, r11=d1×0.6.

[0039] Hemispherical concave hole b1, centered at origin 1 and rotated by 18°, constructs the outermost concave hole array of the left region (ring 1); hemispherical concave hole b2, centered at origin 2 and rotated by 18°, constructs the concave hole array at the left region position of ring 2; hemispherical concave hole b3, centered at origin 3 and rotated by 26°, constructs the concave hole array at the left region position of ring 3; hemispherical concave hole b4, centered at origin 4 and rotated by 26°, constructs the concave hole array at the left region position of ring 4; hemispherical concave hole b5, centered at origin 5 and rotated by 60°, constructs the concave hole array at the left region position of ring 5; Spherical concave hole b7 is centered at origin 5 and rotated by an angle of 60° to construct a concave hole array in the right region of ring 5; hemispherical concave hole b8 is centered at origin 4 and rotated by an angle of 26° to construct a concave hole array in the right region of ring 4; hemispherical concave hole b9 is centered at origin 3 and rotated by an angle of 26° to construct a concave hole array in the right region of ring 3; hemispherical concave hole b10 is centered at origin 2 and rotated by an angle of 18° to construct a concave hole array in the right region of ring 2; hemispherical concave hole b11 is centered at origin 1 and rotated by an angle of 18° to construct a concave hole array in the outermost part of the right region (ring 1). Origin 1, Origin 2, Origin 3, Origin 4 and Origin 5 are all located on the straight line between the origin and the base point. Origin 1 is at the same height as the center of hemispherical concave hole b1 and hemispherical concave hole b11, Origin 2 is at the same height as the center of hemispherical concave hole b2 and hemispherical concave hole b10, Origin 3 is at the same height as the center of hemispherical concave hole b3 and hemispherical concave hole b9, Origin 4 is at the same height as the center of hemispherical concave hole b4 and hemispherical concave hole b8, and Origin 5 is at the same height as the center of hemispherical concave hole b5 and hemispherical concave hole b7. Example

[0040] like Figure 3 As shown, a matching layer for a spherical cap concave underwater acoustic transducer based on an asymmetric metasurface is composed of multiple layers of aluminum and flexible adhesive layers arranged in a periodic alternation pattern. The concave curvature of both the rigid material layer and the flexible adhesive layer is 0.321π rad. The overall height of the underwater acoustic transducer is 25 mm, the diameter is 29.18 mm, and the thickness of the outer shell 10 on the sidewalls and bottom is 2 mm.

[0041] Preferably, the matching layer consists of eight layers from top to bottom. Among them, four flexible adhesive layers with spherical cap concave surfaces of equal thickness are located in the second, fourth, sixth, and eighth layers, with a thickness of d1 (0.37 mm) and a concave surface curvature of 0.321π rad. Four aluminum layers with spherical cap concave surfaces of equal thickness are located in the first, third, fifth, and seventh layers, with a thickness of d2 (1.01 mm) and a concave surface curvature of 0.321π rad. The surface of the first aluminum layer has an asymmetric metasurface structure composed of multiple hemispherical concave hole units.

[0042] Specifically, from top to bottom, the layers are: a first aluminum layer 1 (1.01 mm), a first adhesive layer 2 (0.37 mm), a second aluminum layer 3 (1.01 mm), a second adhesive layer 4 (0.37 mm), a third aluminum layer 5 (1.01 mm), a third adhesive layer 6 (0.37 mm), a fourth aluminum layer 7 (1.01 mm), a fourth adhesive layer 8 (0.37 mm), and a piezoelectric ceramic 9 (2 mm). The first adhesive layer is used to bond the first aluminum layer 1 and the second aluminum layer 3; the second adhesive layer 4 is used to bond the second aluminum layer 3 and the third aluminum layer 5; the third adhesive layer 6 is used to bond the third aluminum layer 5 and the fourth aluminum layer 7; and the fourth adhesive layer 8 is used to bond the piezoelectric ceramic 9 and the fourth filter layer.

[0043] The aluminum layer, as a rigid material, possesses high acoustic impedance and excellent acoustic reflection characteristics, while the flexible bonding layer, typically a flexible material, exhibits low acoustic impedance and good transmission performance. By periodically alternating these layers, a periodic impedance transition is formed, optimizing the sound wave propagation path and energy loss within the underwater acoustic transducer matching layer. This overcomes the performance limitations of traditional matching layers, achieving performance improvements in acoustic energy transmission efficiency, directivity control, frequency response range, and structural lightweighting. The periodically alternating arrangement of the aluminum and epoxy resin layers constitutes a phononic crystal with strong acoustic characteristic contrast. By precisely designing the thickness of each layer, this structure significantly broadens the operating bandwidth through multi-layer distributed resonance effects. Furthermore, by controlling the layer thickness ratio, a gradually changing equivalent acoustic impedance is achieved from the transducer to the water body, minimizing interface reflection loss. External sound waves undergo broadband impedance matching and wavefront renormalization via the periodic structure formed by the aluminum and flexible bonding layers, maximizing the transmission of acoustic energy to the piezoelectric ceramic 9. The piezoelectric ceramic 9 acts as a sensor, converting the received acoustic signal into an electrical signal.

[0044] like Figure 4 As shown, the left side displays a lateral view of the hemispherical concave structure, while the right side is a sectional view of the same structure. Figure 4 Multiple hemispherical concave hole units are arranged radially from the center outward along the concave surface of the spherical cap, with their sizes gradually changing, namely r1(d1×1.6), r2(d1×1.5), r3(d1×1.4), r4(d1×1.3), r5(d1×1.2), r6(d1×1.1), r7(d1), r8(d1×0.9), r9(d1×0.8), r10(d1×0.7), and r11(d1×0.6), forming an asymmetric metasurface structure. Figure 4In the diagram, r6 is the critical radius, indicating the key location of the sound wave focusing region. The lower right image is a magnified view of the area near the critical radius r6, where r6 = d1 × 1.1. This design of unit dimensions varying according to a specific gradient, combined with the spherical cap concave surface structure, enables the formation of continuously varying acoustic resonance and phase delay distributions in the radial direction, thereby allowing for precise shaping and control of the incident sound wavefront. This metasurface structure, through its asymmetrical arrangement, results in different sound wave response characteristics in different directions, thus enhancing the transducer's sensitivity to sound waves from specific directions.

[0045] The transmission spectra of the planar underwater acoustic transducer matching layer, the concave underwater acoustic transducer matching layer without metasurface structure, and the spherical cap concave underwater acoustic transducer matching layer based on asymmetric metasurface of the present invention are as follows: Figure 5 As shown, a hemispherical concave structure was simulated using finite element analysis. The transmission coefficient and sound field distribution at different frequencies and incident angles were analyzed, clearly demonstrating its focusing and transmission enhancement effects on sound waves. The horizontal axis represents the sound wave incident angle, and the vertical axis represents the transmission coefficient. Figure 5 The diagram shows the transmission coefficient curves of three different structures as a function of incident angle: The planar structure (gray curve) exhibits significant angle sensitivity, with a maximum transmission coefficient close to 1 around 68° (-68°). At other angles, the transmission coefficient drops rapidly, particularly in the -16° to 16° range, where it falls to around 0.2, indicating poor reception of small-angle incident sound waves. The spherical cap concave structure (dark gray curve) performs excellently, maintaining a transmission coefficient above 0.8 consistently in the 35° to 77° (-35° to -77°) range, with an average transmittance approximately 17% higher than the planar structure. In the -20° to 20° range, the transmission coefficient increases to around 0.3, with an average transmittance approximately 6.5% higher than the planar structure, demonstrating better sound wave reception at low to medium angles. The asymmetric metasurface spherical cap concave structure (black curve) performs the best, maintaining excellent transmission performance over a wide angle range of ±60° to ±90°. Within the core range of -60° to 60°, by breaking acoustic symmetry, the transmission coefficients of incident sound waves at positive and negative angles no longer exhibit symmetry. A transmission valley forms at the negative angle of -38°, and a transmission peak forms at -36°; simultaneously, a transmission peak appears at the positive angle of 38°, and a transmission valley appears at 39°. This asymmetrical acoustic response, together with the monotonically increasing characteristic in the range of -19° to 19°, constitutes a unique "acoustic fingerprint."

[0046] like Figure 6As shown, within the frequency range of 69 kHz to 79 kHz, the spherical cap concave structure with an asymmetric metasurface continues its advantage of wide-angle high transmission and exhibits unique angle-frequency dependent characteristics. At 69 kHz incident light, the structure forms distinct transmission valleys and peaks near -38° (38°), exhibiting a typical asymmetric acoustic fingerprint. As the frequency gradually increases from 69 kHz to 79 kHz, the characteristic peak and valley positions of the transmission coefficient curve shift towards smaller angles, and the amplitude increases. The transmission oscillations in the small-angle range are further enhanced, with not only more peaks and valleys but also larger fluctuation amplitudes (a deep valley approaching 0 appears). The transmission response gradually evolves from a "single set of peaks and valleys" to "multiple sets of dense oscillations." At 79 kHz, the characteristic peak and valley positions shift to near -20° (20°), and a double peak appears. This pattern not only expands the structure's wideband operating capability but also provides richer "frequency + angle" dual-dimensional information for underwater acoustic detection through frequency-differentiated transmission responses.

[0047] Figure 7 A schematic diagram of the X-ray acoustic analysis of the matching layer of the underwater acoustic transducer with a spherical concave surface of the present invention at 0° incident angle is shown. Figure 7 The lines in the diagram represent the propagation path of the sound wave, showing the reflection of the sound wave as it passes through the matching layer. When the sound wave enters the concave structure, its wavefront gradually converges during propagation, enhancing the concentration of the sound wave on the transducer surface, thereby improving the receiving efficiency of the underwater acoustic transducer. This focusing effect can effectively compensate for the sound energy attenuation caused by the mismatch of the acoustic impedance of the medium.

[0048] The localization mechanism of the matching layer of a spherical concave underwater acoustic transducer based on an asymmetric metasurface is as follows: Figure 8 As shown, a "physical encoding + intelligent decoding" paradigm is adopted. In practical applications, the asymmetric metasurface and the underwater acoustic transducer constitute a direction-sensitive acoustic front-end system. Its direction recognition capability requires further extraction and analysis through back-end signal processing algorithms. A feature database of the transducer output signal under different incident directions is established through experimental measurements or numerical simulations. Using machine learning methods, a model is trained to identify signal characteristics under different incident directions. The real-time received signal characteristics are matched with the database to directly estimate the incident direction, achieving intelligent direction determination. This process eliminates the need for angle-by-angle scanning, significantly improving recognition efficiency and system response speed.

[0049] A method for identifying the incident direction of acoustic waves in an underwater acoustic transducer includes the following steps: S1: Receive sound waves at the initial azimuth using an underwater acoustic transducer to obtain the first response value; the initial azimuth is as follows: Figure 8 As shown at 0°. External sound waves pass through the matching layer structure and are sensed by the piezoelectric ceramic 9 (equivalent to a sensor). The underwater acoustic transducer outputs a signal, namely the first response value (x).

[0050] S2: After rotating the underwater acoustic transducer 180° around its central axis, it receives sound waves and obtains the second response value; the 180° rotation causes the left and right regions of the asymmetric metasurface to be interchanged in space. Since the size and distribution of the hemispherical concave holes in the left and right regions are different, the output signal of the underwater acoustic transducer after rotation is the second response value (y).

[0051] S3: Input the feature vector formed by the first response value and the second response value into the trained deep learning model, and output the predicted sound wave incident angle by the deep learning model.

[0052] The deep learning model is trained using acoustic response features at different incident angles as training samples, labeled with the corresponding incident angles. The deep learning model is a multilayer perceptron, with two neurons in the input layer, two neurons in the hidden layer using the ReLU activation function, and one neuron in the output layer. This deep learning model acts as an "intelligent decoder," learning the complex mapping relationship between the unique "acoustic fingerprint" generated by the asymmetric metasurface and the incident angle. It can output continuous incident angle estimates in real-time and with high precision based solely on the acoustic response values ​​of the transducer in two specific directions.

[0053] Specifically, the underwater acoustic transducer acquires a sound wave response value when incident at 0°, and then acquires the response value again after rotating 180° around the z-axis, forming a set of two-dimensional input features (x, y) to characterize the incident direction of the sound wave. To construct an accurate mapping relationship from the feature vector (x, y) to the incident angle, a deep learning model based on a multilayer perceptron (MLP) is designed. This network adopts a classic hierarchical architecture: the input layer contains two neurons, corresponding to receiving sound wave response features from two directions; the hidden layer has two neurons, using the ReLU (Rectified Linear Unit) activation function to introduce nonlinear transformation capabilities, enhancing the deep learning model's ability to express complex relationships between features; the output layer is designed with one neuron, directly outputting the predicted continuous value of the incident angle. During the training of the deep learning model, the mean squared error (MSE) is used as the loss function, which provides a clear gradient direction for model optimization by calculating the squared difference between the predicted angle and the true label. The optimization algorithm employs the Adaptive Moment Estimation (Adam) optimizer. By calculating the adaptive learning rate for each parameter, it significantly improves the convergence speed and training stability of the deep learning model, ultimately achieving high-precision prediction of the incident angle. The Adam optimizer maintains an exponential moving average of the first moment (gradient direction) and second moment (gradient magnitude) for each parameter in the model independently, and dynamically calculates the adaptive parameter update step size based on the latter. This method combines deep learning technology with underwater acoustic transducer design, providing an intelligent solution for underwater acoustic direction recognition.

[0054] The embodiments described above are preferred embodiments of the present invention, but the present invention is not limited to the above embodiments. Any obvious improvements, substitutions or modifications that can be made by those skilled in the art without departing from the essence of the present invention shall fall within the protection scope of the present invention.

[0055] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A matching layer for a spherical cap concave underwater acoustic transducer based on an asymmetric metasurface, characterized in that, It includes a cylindrical structure with an inwardly recessed spherical cap at the top, the cylindrical structure being composed of rigid material layers and flexible adhesive layers arranged alternately and periodically along the thickness direction; Both the rigid material layer and the flexible adhesive layer have the same concave curvature. Multiple hemispherical concave units are distributed on the spherical cap concave surface of the top rigid material layer, and the multiple concave units constitute an asymmetric metasurface.

2. The matching layer of the spherical cap concave underwater acoustic transducer based on an asymmetric metasurface according to claim 1, characterized in that, The rigid material layer is an aluminum layer, which provides strong acoustic contrast in the periodic structure and provides strong skeletal support for the entire concave structure; the flexible adhesive layer is made of epoxy resin composite material, which achieves precise bonding and harmonization of acoustic impedance.

3. The matching layer of the spherical cap concave underwater acoustic transducer based on an asymmetric metasurface according to claim 1 or 2, characterized in that, The concave hole units are arranged in an array along the hemispherical concave surface. The size of the concave hole units gradually changes along the radial direction. The size of the concave hole units in the left and right regions is different, resulting in different unit sizes in the left and right regions, forming an orientation-dependent asymmetric structure.

4. The matching layer for the spherical cap concave underwater acoustic transducer based on an asymmetric metasurface according to claim 3, characterized in that, The lower part of the periodically alternating rigid material layer and flexible adhesive layer is provided with piezoelectric ceramic (9), and the cylindrical structure and piezoelectric ceramic (9) are disposed inside the outer shell (10); The outer shell (10) is made of a flexible polyacrylate material with a thickness of 2 mm.

5. The matching layer for the spherical cap concave underwater acoustic transducer based on an asymmetric metasurface according to claim 3, characterized in that, The recessed hole units are evenly distributed in a ring, with the diameter of the recessed hole unit on one side of the ring being larger than the diameter of the recessed hole unit on the other side.

6. The matching layer for a spherical cap concave underwater acoustic transducer based on an asymmetric metasurface according to claim 4 or 5, characterized in that, The concave curvature of both the rigid material layer and the flexible adhesive layer is 0.321π rad, and the radius of the spherical cap concave surface of the top layer is 12.59 mm.

7. The matching layer for the spherical cap concave underwater acoustic transducer based on an asymmetric metasurface according to claim 6, characterized in that, The cylindrical structure consists of eight layers from top to bottom, including four flexible adhesive layers with spherical cap concave surfaces of equal thickness located in the second, fourth, sixth, and eighth layers, with a thickness of d1; and four aluminum layers with spherical cap concave surfaces of equal thickness located in the first, third, fifth, and seventh layers, with a thickness of d2; the surface of the aluminum layer in the first layer has an asymmetric metasurface structure composed of multiple hemispherical concave hole units.

8. The matching layer for the spherical cap concave underwater acoustic transducer based on an asymmetric metasurface according to claim 7, characterized in that, The distribution rule of the hemispherical concave hole units on the spherical cap concave surface of the top aluminum layer is as follows: With the center of the spherical concave surface of the top aluminum layer as the base point, there is only one hemispherical concave hole b6, with the center of the sphere located at the base point and a radius of r6; Taking the center of the spherical cap concave surface of the top aluminum layer as the origin, rotating the base point counterclockwise by 10° gives the center of the hemispherical concave hole b5 with radius r5; rotating the base point counterclockwise by 20° gives the center of the hemispherical concave hole b4 with radius r4; rotating the base point counterclockwise by 30° gives the center of the hemispherical concave hole b3 with radius r3; rotating the base point counterclockwise by 40° gives the center of the hemispherical concave hole b2 with radius r2; rotating the base point counterclockwise by 50° gives the center of the hemispherical concave hole b1 with radius r2. The center of concave hole b1; rotating the base point clockwise by 10° gives the center of hemispherical concave hole b7 with radius r7; rotating the base point clockwise by 20° gives the center of hemispherical concave hole b8 with radius r8; rotating the base point clockwise by 30° gives the center of hemispherical concave hole b9 with radius r9; rotating the base point clockwise by 40° gives the center of hemispherical concave hole b10 with radius r10; rotating the base point clockwise by 50° gives the center of hemispherical concave hole b11 with radius r11. And r1=d1×1.6, r2=d1×1.5, r3=d1×1.4, r4=d1×1.3, r5=d1×1.2, r6=d1×1.1, r7=d1, r8=d1×0.9, r9=d1×0.8, r10=d1×0.7, r11=d1×0.6; Hemispherical concave hole b1, centered at the first origin, rotates by 18° to construct the concave hole array of the outermost first ring in the left region; hemispherical concave hole b2, centered at the second origin, rotates by 18° to construct the concave hole array at the left region position of the second ring; hemispherical concave hole b3, centered at the third origin, rotates by 26° to construct the concave hole array at the left region position of the third ring; hemispherical concave hole b4, centered at the fourth origin, rotates by 26° to construct the concave hole array at the left region position of the fourth ring; hemispherical concave hole b5, centered at the fifth origin, rotates by 60° to construct the concave hole array at the left region position of the fifth ring; hemispherical concave hole b7, centered at the fifth origin, rotates by 60° to construct the concave hole array at the right region position of the fifth ring; hemispherical concave hole b8, centered at the fourth origin, rotates by 26° to construct the concave hole array at the right region position of the fourth ring; Spherical concave hole b9, centered on the third origin, rotates by 26° to construct a concave hole array in the right region of the third ring; hemispherical concave hole b10, centered on the second origin, rotates by 18° to construct a concave hole array in the right region of the second ring; hemispherical concave hole b11, centered on the first origin, rotates by 18° to construct a concave hole array in the outermost first ring of the right region; the first origin, second origin, third origin, fourth origin, and the... All five origin points are located on a straight line between the origin and the base point. The first origin point is at the same height as the center of hemispherical concave holes b1 and b11, the second origin point is at the same height as the center of hemispherical concave holes b2 and b10, the third origin point is at the same height as the center of hemispherical concave holes b3 and b9, the fourth origin point is at the same height as the center of hemispherical concave holes b4 and b8, and the fifth origin point is at the same height as the center of hemispherical concave holes b5 and b7.

9. The matching layer for a spherical cap concave underwater acoustic transducer based on an asymmetric metasurface according to claim 8, characterized in that, The method for identifying the incident direction of sound waves includes the following steps: S1: The piezoelectric ceramic in the cylindrical structure of the underwater acoustic transducer receives the sound wave at the initial orientation to obtain the first response value; S2: After rotating the cylindrical structure 180° around its central axis, the sound wave is received to obtain the second response value; S3: Input the feature vector formed by the first response value and the second response value into the trained deep learning model, and output the predicted sound wave incident angle by the deep learning model.

10. The matching layer of the spherical cap concave underwater acoustic transducer based on an asymmetric metasurface according to claim 9, characterized in that, The deep learning model is trained by using the acoustic response features under different incident angles as training samples and the corresponding incident angles as labels; the deep learning model is a multilayer perceptron, with two neurons in the input layer, two neurons in the hidden layer and ReLU activation function, and one neuron in the output layer. During the training of the deep learning model, mean squared error is used as the loss function; the Adam optimizer is selected to independently maintain the exponential moving average of the first and second moments of each parameter in the deep learning model, and the adaptive parameter update step size is dynamically calculated based on the second moment.

Images