Acoustic angle encoder, direction finding array and detection method based on geometric waveguide

Abstract

The present application relates to the field of physical acoustics and sound waveguide propagation technology, and discloses an acoustic angle encoder, a direction-finding array and a method based on a geometric waveguide; the acoustic angle encoder comprises at least one sound waveguide structure with a curved propagation path, the inner side wall of which is a smooth continuous curved surface, and the outer side wall of which is provided with a plurality of acoustic impedance discontinuous structures; after external sound waves are obliquely incident into the waveguide interior, under the synergistic effect of the centrifugal guidance generated by the curved geometry and the slow wave delay effect of the outer side wall, the local equivalent sound speed and the effective propagation path of sound waves in different directions produce significant differences; the present application breaks the physical aperture limitation of miniaturized equipment, amplifies the weak spatial angle difference into a terminal macroscopic phase difference at the physical level, and improves the positioning accuracy and robustness in a complex sound field environment.

Description

Technical Field

[0001] This invention relates to the fields of physical acoustics, spatial sound field manipulation and acoustic waveguide propagation, and particularly to an acoustic angle encoder, direction finding array and detection method based on geometric waveguides. Background Technology

[0002] With the development of intelligent voice interaction, robot perception, spatial audio, and miniaturized acoustic devices, how to achieve high-precision sound source direction detection under limited physical aperture conditions has become an important technical problem in the field of acoustic perception.

[0003] Most existing sound source localization systems rely on the time difference or phase difference of free-field propagation between multiple microphones to calculate direction. However, in miniaturized devices, due to the limited spacing between microphones, the natural phase difference in free space is usually weak, resulting in insufficient sensitivity of the system to small angular changes.

[0004] Some existing acoustic system solutions (such as simple biomimetic curved surfaces or conventional smooth cylindrical resonator tubes) rely solely on macroscopic geometric reflections to accumulate sound wave path differences. Due to the lack of intervention on the underlying physical property of sound wave propagation speed, such smooth cavities require long physical dimensions to accumulate observable phase differences, and cannot escape the size constraints of miniaturized devices.

[0005] Furthermore, existing technologies generally lack a spatial acoustic field coding architecture that can first convert directional information in a free field into directionally related initial propagation boundary conditions, and then further physically amplify the difference during subsequent waveguide propagation.

[0006] Therefore, a new spatial acoustic control scheme is urgently needed to achieve highly sensitive physical encoding and stable solution for minute spatial directional differences. Summary of the Invention

[0007] This invention proposes an acoustic angle encoder, direction finding array, and detection method based on geometric waveguides, aiming to solve the technical problems of weak free field direction information, difficulty in stable extraction of phase difference, and difficulty in effectively amplifying spatial direction features in existing miniaturized acoustic systems.

[0008] This invention utilizes a sound wave guiding structure with a curved propagation path. Under the combined effect of curved geometric guidance and the slow wave delay effect formed by the discontinuous acoustic impedance structure on the outer wall, the propagation path difference and propagation speed difference between sound waves in different directions are physically amplified in depth, thereby forming a phase difference and / or time delay feature with significant directional correlation at the sound output end.

[0009] As a further extension of the application scenarios of the present invention, in the external physical space at the front end of the acoustic waveguide structure, a pre-intervention structure such as an asymmetric step-reflection step, a phase gradient acoustic metasurface, or a high-low folded acoustic barrier can be independently combined to pre-establish a direction-related initial propagation path difference or phase change in the external free field, and then cooperate with the nonlinear slow wave effect in the waveguide structure to achieve cascade amplification under more extreme conditions.

[0010] This invention is an engineering abstraction and enhancement of the physical mechanism of "spatial direction precoding - waveguide propagation amplification" in biological hearing systems, thereby achieving high-sensitivity spatial direction physical coding.

[0011] In a first aspect, the present invention provides an acoustic angle encoder based on a geometric waveguide, comprising at least one acoustic waveguide structure; The acoustic waveguide structure is a hollow cavity with an acoustic injection end and an acoustic output end; The acoustic wave guiding structure is curved along the direction of acoustic wave propagation, and defines an inner wall near the center of curvature and an outer wall away from the center of curvature. The aspect ratio of the acoustic waveguide structure from the acoustic injection end to the acoustic output end is 5:1 to 20:1, wherein the aspect ratio is the ratio of the unfolded length of the acoustic waveguide structure along the central axis to the equivalent hydrodynamic diameter of the hollow cavity; when the cross-sectional area of ​​the hollow cavity is not uniform, the equivalent hydrodynamic diameter is the average equivalent inner diameter along the direction of acoustic wave propagation. The inner wall is a smooth, continuous curved surface; Multiple acoustic impedance discontinuities are provided on the outer side wall; The acoustic waveguide structure is configured as follows: Establish direction-dependent propagation path differences and / or propagation speed differences for sound waves incident from different spatial directions, and form phase difference and / or time delay characteristics corresponding to the incident direction at the sound output end.

[0012] Furthermore, the acoustic waveguide structure is either a three-dimensional conical spiral waveguide or a two-dimensional asymmetric curvature gradient shape: when it is a three-dimensional conical spiral waveguide, the hollow cavity extends continuously along the conical surface in three-dimensional space, and the cross-sectional area of ​​the hollow cavity gradually decreases from the acoustic injection end to the acoustic output end; when it is a two-dimensional asymmetric curvature gradient shape, the central axis of the hollow cavity is asymmetrically spirally bent and unfolded in a single plane; when the cross-sectional area of ​​the hollow cavity is not uniform, the equivalent hydrodynamic diameter is the average equivalent inner diameter along the direction of acoustic wave propagation.

[0013] Furthermore, the acoustic impedance discontinuity structure on the outer wall is a mesoscopic geometric structure, and its characteristic dimensions vary in gradient along the direction of sound wave propagation in order to maintain the continuity of equivalent sound velocity modulation in different frequency ranges and suppress high-frequency phase winding and local acoustic distortion.

[0014] Furthermore, the acoustic impedance discontinuity structure includes one or more of the following: groove, thread, protrusion, bypass tube, and local resonant cavity, and its characteristic size is less than 1 / 10 of the lowest operating wavelength of the target operating frequency band.

[0015] Secondly, the present invention provides a multi-dimensional acoustic direction finding array, employing at least two of the above-mentioned acoustic angle encoders and at least two sound sensors coupled thereto; The acoustic injection ends of each acoustic angle encoder are arranged according to a preset spatial topology to form an array input surface; Each of the acoustic angle encoders is configured to establish spatial phase coding features with different spatial directions.

[0016] Thirdly, the present invention provides a sound wave direction detection system, which employs the above-mentioned multi-dimensional acoustic direction finding array and a signal processing module electrically connected to each of the sound sensors; The signal processing module is configured to extract phase difference and / or time difference features amplified by the front-end physical structure, and to reconstruct the direction of the sound source through a spatial angle mapping model.

[0017] Fourthly, the present invention provides a method for detecting the direction of sound waves, comprising the following steps: Step 1: Receive external incident sound waves and guide them into each of the sound wave guiding structures for propagation, forcing sound waves with slight differences in incident angle to generate an effective propagation path length difference between the inner wall near the center of curvature and the outer wall far from the center of curvature. Step 2: During the propagation of sound waves, the acoustic impedance discontinuity structure set on the outer wall is used to apply a slow wave delay effect to the sound waves deflected towards the outer wall, so as to output an amplified terminal sound wave signal at the sound output end. The terminal sound wave signal carries a phase difference sequence corresponding to the spatial direction. Step 3: The phase difference sequence is picked up by the sound sensor and then reverse-decoded by the signal processing module to reconstruct the spatial direction information of the sound source.

[0018] Compared with the prior art, the present invention has the following advantages: 1) Overcoming physical aperture limitations to achieve physical amplification of the core phase: This invention utilizes a synergistic physical mechanism of "bending geometry guidance" and "mesoscopic slow wave delay" within a confined cavity to transform the weak spatial angle difference in a free field into a significantly amplified characteristic phase difference at the terminal, thus breaking the physical receiving aperture limitations faced by miniaturized devices.

[0019] 2) Breaking through the bottleneck of pure geometric reflection, introducing mesoscopic slow waves to achieve two-dimensional intervention of "sound speed-path": This invention abandons the inefficient approach of traditional acoustic waveguides (such as conventional resonant tubes) that rely solely on the back-and-forth reflection of sound waves between smooth tube walls to lengthen the geometric path. It innovatively introduces a discontinuous acoustic impedance structure on the outer wall of the waveguide, which, while lengthening the centrifugal geometric path, also forcibly reduces the local group velocity of the sound waves in that region through the "slow wave effect." This multiplier effect of "lengthened geometric path + slower propagation speed" amplifies weak phase differences over short distances.

[0020] 3) Deep hardware and software coupling enables end-to-end decoding of complex sound fields: This invention utilizes a front-end purely physical structure to construct a phase sequence with spatial direction correlation and configures a spatial angle mapping model for reverse unwinding. This architecture of "front-end physical amplification features + back-end intelligent algorithm reverse decoding" significantly improves the robustness of direction calculation in complex sound field environments.

[0021] 4) Front-end physical compatibility and scalability: The core physical amplification mechanism of this invention operates independently within the waveguide cavity. Furthermore, its acoustic injection end possesses extremely high openness, seamlessly compatible with pre-intervention structures such as step-type structures, acoustic metasurfaces, or high-low folded labyrinths. This decoupled architecture not only addresses the varying complexity of product deployment requirements but also, when used in combination, stimulates a multiplier-level secondary amplification effect, further enhancing detection sensitivity under extreme conditions. Attached Figure Description

[0022] The above and other objects, features, and advantages of exemplary embodiments of the present invention will become readily apparent upon reading the following detailed description with reference to the accompanying drawings. In the drawings, several embodiments of the invention are illustrated by way of example and not limitation, with the same or corresponding reference numerals denoteing the same or corresponding parts, wherein: Figure 1 This is a schematic diagram of the curved geometric waveguide in Embodiment 1 of the present invention; Figure 2 This is a schematic diagram of the three-dimensional conical spiral worm tube in Embodiment 2 of the present invention; Figure 3 The diagram shows a comparison of phase delay results under different frequencies and incident directions in Embodiment 1 of the present invention, where sub-graphs (a) to (f) correspond to test frequencies from 1 kHz to 3.5 kHz, respectively. Figure 4 The diagram shows a comparison of phase delay results under different frequencies and incident directions in Embodiment 2 of the present invention, where sub-graphs (a) to (h) correspond to test frequencies from 1 kHz to 10 kHz, respectively.

[0023] Figure 5 This is a schematic diagram of the acoustic comparison model of the straight cylindrical guide tube used in this invention to illustrate the principle of sound wave propagation distance difference.

[0024] Figure 6 This is a schematic diagram (front view) of a sound field precoding structure according to the present invention.

[0025] Figure 7 This is a schematic diagram (side view) of a sound field precoding structure in this invention.

[0026] Figure 8 This is a schematic diagram used in this invention to illustrate the principle of generating direction-dependent sound propagation path differences in the sound field precoding structure.

[0027] Explanation of icon numbers: A. Sound injection end; B. Sound output end; C1. Inner wall; D1. Outer wall; E. Sound wave acquisition point near the sound wave inlet; F. Sound wave acquisition point far from the sound wave inlet; G. First primary reflection surface of the sound field precoding structure; H. Second primary reflection surface of the sound field precoding structure; L1. Schematic diagram of a sound ray with a relatively long propagation path; L2. Schematic diagram of a sound ray with a relatively short propagation path. Detailed Implementation

[0028] The exemplary embodiments disclosed in this application will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of this application are shown in the drawings, it should be understood that this application can be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of this application and to fully convey the scope of this application to those skilled in the art. Unless otherwise specified, the technical means used in the embodiments are conventional means well known to those skilled in the art.

[0029] It should be noted that the accompanying drawings of this application (especially) Figure 1 , Figure 2The accompanying drawings (including those depicting microscopic features) are merely schematic diagrams illustrating the internal structure and physical principles, and are not drawn to strict engineering scale. To visually clearly demonstrate the "acoustic impedance discontinuity structure" (such as rectangular microgrooves) on the outer wall of the waveguide, its relative physical dimensions, thickness, and protrusion have been deliberately exaggerated in the drawings. In the actual physical prototype, as described in this specification, this structure is "mesoscopic" (i.e., its feature size is strictly less than 1 / 10 of the lowest operating wavelength of the target frequency band), and its actual geometric scale is much smaller than the visually presented scale in the drawings. Therefore, the exaggerated depiction in the drawings does not constitute a limitation on the actual structural dimensions or relative scale of the present invention.

[0030] In a first aspect, the present invention provides an acoustic angle encoder based on a geometric waveguide, comprising at least one acoustic waveguide structure, the acoustic waveguide structure being a hollow cavity, the acoustic waveguide structure having an acoustic injection end and an acoustic output end; the acoustic waveguide structure extends in a curved manner along the direction of acoustic wave propagation, and defines an inner wall C1 near the center of curvature and an outer wall D1 away from the center of curvature; the aspect ratio of the acoustic waveguide structure from the acoustic injection end A to the acoustic output end B is 5:1 to 20:1, wherein the aspect ratio is the ratio of the unfolded length of the acoustic waveguide structure along the central axis to the equivalent hydrodynamic diameter of the hollow cavity; when the cross-sectional area of ​​the hollow cavity is non-uniform, the equivalent hydrodynamic diameter is the average equivalent inner diameter along the direction of acoustic wave propagation; the inner wall C1 of the acoustic waveguide structure near the center of curvature is a smooth continuous curved surface; multiple acoustic impedance discontinuities are provided on the outer wall D1 of the acoustic waveguide structure away from the center of curvature.

[0031] Specifically, this invention provides an acoustic angle encoder based on a geometric waveguide, comprising at least one acoustic waveguide structure. The acoustic waveguide structure has an acoustic injection end A and an acoustic output end B, and its internal cavity defines a confined space for acoustic wave propagation. The inner wall of the acoustic waveguide structure has a curvature along the acoustic wave propagation direction, and the aspect ratio of the acoustic waveguide structure from the acoustic injection end A to the acoustic output end B is 5:1 to 20:1. An acoustic impedance discontinuity structure is continuously or discontinuously arranged on the outer wall D1 of the acoustic waveguide structure, the acoustic impedance discontinuity structure being used to modify... The equivalent acoustic impedance varies locally; the curvature is configured to generate a radial centrifugal force, forcing obliquely incident sound waves with a deflection angle to deflect towards the outer wall D1 of the confined space and couple with the acoustic discontinuities (i.e., acoustic impedance discontinuities) on the outer wall D1, thereby exciting a slow-wave delay effect that reduces the effective sound velocity; the sound wave guiding structure utilizes the physical path amplification caused by the curvature geometry and the slow-wave delay effect to physically encode the incident angle difference into a sound wave signal with a significant characteristic phase and / or time difference at the sound output end B.

[0032] In one specific embodiment, the sound wave guiding structure has two shapes: a horn-shaped structure and a three-dimensional collimated spiral shape, wherein the cavity is configured as a three-dimensional spiral channel winding along a cylindrical surface or a hyperboloid; in the above two shapes, when the cross-sectional area of ​​the cavity is not uniform, the equivalent hydrodynamic diameter is the average equivalent inner diameter along the direction of sound wave propagation.

[0033] In one specific embodiment, the acoustic wave guiding structure is in the form of a three-dimensional spatial collimated spiral, wherein the cavity is configured as a three-dimensional spiral channel wound along a cylindrical surface or a hyperboloid; in this way, when the incident acoustic wave propagates inside, it not only generates radial centrifugal offset, but also maintains continuous axial advancement, so as to break away from the energy convergence mode of the two-dimensional spiral closed cavity and ensure high fidelity of phase characteristics during transmission.

[0034] As a further extension of the application scenarios of this invention, a spatial acoustic field precoding structure can be integrally or spliced ​​onto the outer side of the acoustic injection end A of the acoustic waveguide structure. The spatial acoustic field precoding structure is configured to construct direction-dependent initial physical boundary conditions between the free field and the confined waveguide, and can specifically adopt one of the following three asymmetric front-end physical intervention configurations: One is the asymmetric step-reflection step: This structure utilizes the multipath reflection mechanism of ray acoustics. By deploying stepped surfaces with discrete physical size differences outside the sound injection end, it generates discrete initial path length differences for incident sound waves with different spatial vectors and guides them to cut into the main waveguide at differentiated grazing incidence angles. The second is the phase gradient acoustic metasurface: by arranging local resonant microstructures (such as variable cross-section Helmholtz resonant cavities or spatial coiled slits) in an array around the acoustic injection end, macroscopic path accumulation is abandoned, and the instantaneous wavefront anomalous deflection (generalized Snell's law) is excited by local acoustic resonance, and a strong initial phase perturbation is established within the physical thickness at the deep subwavelength scale. The third type is the high-low folded staggered acoustic barrier: by vertically protruding rigid staggered barriers arranged in an asymmetrical array outside the entrance, a Z-shaped meandering channel with an asymmetrical folding rate is defined, transforming the limited physical depth into a significantly extended effective propagation path.

[0035] Any of the above precoding structures can work independently in conjunction with the bending centrifugal guidance and slow wave effect within the waveguide to achieve a multiplier-level second-order cascade amplification of the external initial phase difference and the internal nonlinear slow wave effect.

[0036] In one specific embodiment, the characteristic dimensions of multiple acoustic impedance discontinuities provided on the outer wall D1 of the acoustic wave guiding structure are gradually varied along the direction of acoustic wave propagation, so as to avoid high-frequency distortion at the end where the cavity diameter becomes thinner and maintain stable modulation of the equivalent sound velocity in a wide frequency band.

[0037] In one specific embodiment, the acoustic impedance discontinuity structure includes one or more of the following: groove, thread, protrusion, bypass pipe communicating with the main cavity, and resonant cavity. The acoustic impedance discontinuity structure, without obstructing the axial flow of the main sound wave energy, causes the obliquely incident sound wave to produce the slow wave delay effect or resonant delay, thereby amplifying the characteristic phase and / or time difference.

[0038] In one specific embodiment, the local cross-section of the outer wall D1 of the acoustic wave guiding structure exhibits alternating undulations to construct a periodic acoustic impedance gradient within the cavity, and to maintain the determinism of the phase mapping law while avoiding cumulative nonlinear compression of acoustic wave energy.

[0039] Secondly, the present invention provides a multi-dimensional acoustic direction finding array, employing at least two of the above-mentioned acoustic angle encoders and at least two sound sensors coupled thereto. The acoustic injection terminals A of each acoustic angle encoder are arranged in a preset spatial topology to form an array input surface; the acoustic output terminals B of each acoustic angle encoder are respectively coupled to the independent sound sensor.

[0040] In one specific embodiment, at least two of the acoustic angle encoders constitute a spatial phase coding array: At least two of the acoustic angle encoders have different physical parameters so that the encoders at different positions have differentiated path amplification factors; for the same spatial sound source, at least two of the acoustic angle encoders output a set of phase difference sequences containing mutually orthogonal angle-phase mapping features at the same frequency; the multiple electrical signals output by the sound sensor use the phase difference sequence to form mutually orthogonal angle-phase mapping matrices in space to achieve robust azimuth and pitch angle calculation.

[0041] In one specific implementation, the array is configured as a frequency-domain multiplexing architecture, and at least two of the acoustic angle encoders are divided into at least two subgroups: The acoustic angle encoder of the first subgroup has a first geometric parameter feature, and its effective path length offset pattern is configured to be optimized for encoding the wavelength of the first frequency band sound wave; the acoustic angle encoder of the second subgroup has a second geometric parameter feature, and its effective path length offset pattern is configured to be optimized for encoding the wavelength of the second frequency band sound wave, wherein the first frequency band is different from the second frequency band.

[0042] Thirdly, the present invention provides a sound wave direction detection system, which employs the above-mentioned multidimensional acoustic direction finding array; and a signal processing module electrically connected to each of the sound sensors; The signal processing module is configured to: extract the relative phase difference and / or absolute time offset between each acoustic wave signal that is physically amplified by the front end to form a phase difference sequence; and input the phase difference sequence into a pre-stored spatial angle mapping model to output the three-dimensional spatial vector direction information of the incident acoustic wave. The spatial angle mapping model is either a deep neural network model trained with prior sound field sample data, or a nonlinear lookup table constructed based on a generalized cross-correlation algorithm, used to solve the nonlinear mapping matrix containing the increase in effective path length and the slow wave effect in reverse.

[0043] Fourthly, the present invention provides a method for sound wave direction detection, implemented using the above-mentioned system, comprising the following steps: Step 1: Use the spatial sound field precoding structure to establish the direction-dependent initial propagation path difference, initial phase perturbation and / or initial incident angle offset for the external incident sound wave; Step 2: The external incident sound wave modulated by the spatial sound field precoding structure is introduced into the sound wave guiding structure for propagation, forcing the sound wave to generate an effective propagation path length difference between the inner wall C1 near the center of curvature and the outer wall D1 far from the center of curvature. Step 3: During the propagation of sound waves, the acoustic impedance discontinuity structure set on the outer wall D1 is used to apply a slow wave delay effect to the sound waves that are biased towards the outer wall D1, so as to output an amplified terminal sound wave signal at the sound output end B. The terminal sound wave signal carries a phase difference characteristic corresponding to the spatial direction. Step 4: The phase difference sequence is picked up by the sound sensor and then reverse-decoded by the signal processing module to reconstruct the spatial direction information of the sound source.

[0044] It should be specifically noted that the "mesoscopic scale" or "mesoscopic geometry" (e.g., the aforementioned acoustic impedance discontinuity structure) mentioned in this invention refers to a structure whose physical size is on the deep subwavelength scale (specifically, a characteristic size less than 1 / 10 of the lowest operating wavelength of the target operating frequency band). The structural characteristics of this scale are that it neither causes strong acoustic wave diffraction and disordered scattering at the macroscopic scale, nor does it generate effective physical perturbations to the local acoustic impedance boundaries, thereby stimulating a significant slow-wave delay effect without blocking the macroscopic transmission of sound waves.

[0045] To further elucidate the physical encoding mechanism of the acoustic wave guiding structure for the direction of incoming spatial waves in this invention, a basic straight-cylinder guide tube will be used as an example for comparison and explanation. For example... Figure 5As shown, assuming the inner wall of the straight pipe is a smooth boundary, only the geometric reflection of sound waves is considered. When an external sound wave is incident from the left-side injection end A and radiated from the right-side output end B, sound waves with different spatial incident angles will inevitably experience different geometric paths with varying numbers of reflections and return trajectories within the pipe. This objective physical phenomenon causes the spatial incident direction to be deterministically encoded as a propagation path difference, which then manifests as a macroscopic time difference and phase difference at the output end.

[0046] However, Figure 5 This also reveals the physical limitations of this mechanism: the physical amplification effect of a regular and smooth straight tube on weak phase differences is extremely insignificant. If incident sound waves from different directions establish only a weak initial phase difference in a free field (or via an external pre-coding device), a smooth straight tube requires an extremely long physical dimension to accumulate it to a level suitable for robust decoding. Given this underlying physical bottleneck, this invention innovatively introduces asymmetric curved geometry and superimposes an acoustic impedance discontinuity structure with mesoscopic slow-wave effects (as described in subsequent embodiments) to significantly amplify phase differences within a limited physical aperture, providing more effective and robust physical cues for high-precision spatial positioning in complex sound fields.

[0047] Example 1 Basic acoustic angle encoder unit; Objective: This embodiment provides a basic two-dimensional acoustic waveguide structure, aiming to clearly demonstrate the core physical mechanism of "bending geometry + mesoscopic slow wave structure" synergistic amplification of small incident angle differences, and to elaborate in detail the scientific derivation process of the core acoustic parameters.

[0048] Structure: The acoustic waveguide structure in this embodiment is a waveguide with a constant cross-section but a curved arrangement; the physical length L of the waveguide along its central axis is 60 mm, the equivalent hydrodynamic diameter D of the internal cavity is 8 mm, and the length-to-diameter ratio L / D = 7.5:1. This waveguide has a two-dimensional waveguide shape, that is, the central axis of the hollow cavity is curved only in a single plane (unfolding in a bull's horn shape), without spiral ascent along the Z-axis.

[0049] Asymmetric acoustic impedance wall design: The inner wall C1 (i.e. the inner surface of the hollow cavity near the center of the curve) is a smooth, continuous curved surface that has been polished.

[0050] The outer wall D1 (i.e., the inner surface of the hollow cavity away from the center of curvature) is continuously machined with 25 periodically arranged rectangular micro-grooves (i.e., a discontinuous acoustic impedance structure). Specific parameters are: groove depth h = 1.5 mm, groove width w = 1.0 mm, and the ridge spacing between adjacent grooves d = 1.0 mm.

[0051] Scientific Derivation and Selection of the Acoustic Impedance Discontinuity Structure of the Outer Wall D1: In order to achieve a high signal-to-noise ratio slow-wave delay in broadband speech (e.g., 500Hz-4kHz), this embodiment preferably uses a rectangular microgroove array. Its key parameters (taking a target center frequency f=2000Hz and wavelength λ≈171.5mm as an example) are configured according to the following physical principles: Anti-diffraction criterion (periodic design): The structural period p must satisfy the subwavelength condition (p < λ / 10 ≈ 17.1 mm). In this embodiment, the period p is set to 2.0 mm, which effectively satisfies the deep subwavelength condition and ensures the smooth transmission of sound wave energy.

[0052] Loss prevention criteria (width design): At 2000Hz, the thickness of the air viscous layer is approximately 0.05mm. The width w=1.0mm set in this embodiment is much larger than this thickness, which greatly reduces the probability of sound waves getting trapped in the viscous layer and being converted into heat energy, thus ensuring a high signal-to-noise ratio for signal output.

[0053] Non-resonant delay criterion (depth design): To avoid the acoustic bandgap at λ / 4 that blocks sound waves, this embodiment sets the depth h=1.5mm. This non-resonant perturbation depth provides stable equivalent sound velocity attenuation without causing transmission dead zones.

[0054] Selection Comparison: Compared with the Helmholtz resonator with strong dispersion or the spatial labyrinth structure with high heat dissipation, the rectangular microgroove has better engineering applicability in terms of broadband fidelity and energy conservation.

[0055] Manufacturing process: Material selection: Acoustic hard boundary materials (such as photosensitive resin or metal) are used to ensure that the pipe wall has good sound wave reflection performance.

[0056] Molding and post-processing: SLA (stereolithography) 3D printing technology is used for one-piece molding. After printing, the inner wall C1 is fluid polished with polishing fluid, while the microgroove array of the outer wall D1 maintains the right-angle edge morphology to establish a sudden change in local acoustic impedance.

[0057] Working Process and Physical Mechanism: When a 2kHz sound wave enters directly opposite the injection end axis (0°), the main energy of the sound wave propagates along the central region of the cavity, with minimal influence from the groove group on the outer wall D1, and has a reference transmission time t0. When a sound wave with an angle (e.g., 10° angle) enters obliquely, the wavefront is guided by the bending geometry at the bend, generating a radial centrifugal force. The sound wave energy is deflected and propagates close to the outer wall D1. The sound wave propagating close to the outer wall D1 sweeps over 25 microgrooves, and the change in equivalent acoustic impedance excites a slow wave effect, causing the sound wave group velocity in this region to decrease. At the sound output end B, the transmission time of this oblique sound wave changes to t1. Based on the synergistic effect of "longer geometric path and slower transmission speed," the phase difference is significantly amplified.

[0058] Combination Figure 3 The phase delay results comparison diagram shown illustrates the specific acoustic performance of Example 1 under broadband response. Figure 3 Figures (a) to (f) show the phase delay at frequencies of 1 kHz, 1.5 kHz, 2 kHz, 2.5 kHz, 3 kHz, and 3.5 kHz, respectively. The physical structure of Example 1 is relatively simple, with a small aspect ratio, making it easy to fabricate. In the low-frequency range, it can effectively convert small angular differences into significant phase differences; however, Figure 3 (e) and Figure 3 (f) reveals the limitations of Example 1 in the high-frequency wide-angle region: as the frequency increases, phase winding becomes more pronounced. Specifically, at a frequency of 3 kHz, when the incident angle of the acoustic wave reaches approximately 75 degrees, the accumulated phase difference exceeds the unambiguous boundary of 2π, resulting in winding distortion. This indicates that the basic two-dimensional asymmetric waveguide is more suitable for detection tasks in the mid-to-low frequency or small-angle range.

[0059] Example 2 Three-dimensional cochlear tube morphology with impedance transformation Objective: This embodiment provides a three-dimensional spiral morphology with high engineering application value, which deeply integrates macroscopic impedance transformation with mesoscopic gradient slow wave structure.

[0060] Structural design: The acoustic waveguide structure has a three-dimensional waveguide shape, meaning that the central axis of the hollow cavity is both radially curved and spirally ascending along the Z-axis in three-dimensional space (forming a conical spiral wound around a worm shaft in three dimensions), with an effective unfolded length L = 150 mm. The cross-sectional area gradually decreases smoothly from the acoustic injection end A (inner diameter 10 mm) to the acoustic output end B (inner diameter 1.5 mm) to achieve effective acoustic energy focusing and impedance matching. The inner wall C1 of the waveguide structure remains smooth.

[0061] Gradient microgroove design: To match the gradually decreasing cross-section, the acoustic discontinuity of the outer wall D1 of the waveguide structure is preferably a gradient-gradient rectangular microgroove array; the depth h of the microgroove decreases proportionally along the direction of sound wave propagation as the diameter of the volute decreases (e.g., gradually transitioning from 2.0 mm at the inlet to 0.2 mm at the outlet). This design can avoid high-frequency distortion at the tapered end of the tube and maintain stable modulation of the equivalent sound velocity over a wide frequency band.

[0062] Manufacturing and assembly: Machining and Assembly: The worm gear is machined in parts from high-rigidity metal or alloy using a multi-axis CNC machine tool, and the inner wall C1 is polished. At least one high signal-to-noise ratio microphone is integrated via a flexible printed circuit board (FPC) at least at the worm head hole and cured using acoustic sealant. After the upper and lower housings are aligned, acoustic airtightness is achieved through methods such as laser spot welding.

[0063] Comparing the results from Example 1 and Example 2 further verifies the broadband superiority of the three-dimensional conical spiral worm tube in this embodiment. Figure 4 As shown in (a) to (h), the system underwent frequency sweep verification in the 1kHz to 10kHz frequency band. Figure 3 and Figure 4 The comparison shows that Example 2, with its more complex geometric gradient and three-dimensional slow-wave coupling mechanism, exhibits greater flexibility in phase calculation. Notably, Example 2 significantly delays the occurrence frequency of phase winding: winding appears in Example 1 at 3 kHz, while in Example 2, phase winding does not appear until 5 kHz. Meanwhile, Figure 4 It also reveals the inherent laws of high-frequency wave acoustics, namely that the higher the frequency, the lower the starting angle of phase winding (for example, the winding at 5 kHz starts at about 90 degrees, while the winding at 9 kHz starts at about 80 degrees).

[0064] Figure 3 and Figure 4 The revealed "frequency-angle-phase winding" mapping law provides a core physical guiding principle for the spatial arrangement of signal acquisition points (i.e., sound sensors) within the detection system: Firstly, anti-winding detection in the high-frequency band (near the incident end region, such as...) Figure 1 , Figure 2 (E, the sound wave acquisition point near the sound wave inlet): High-frequency sound waves have short physical wavelengths, and even a small path difference can cause a significant phase deflection. This is also... Figure 4The reason why mid-to-high frequency components are prone to entanglement at large angles is that if high-frequency waves are allowed to propagate over long distances within the waveguide, excessive path difference will cause the accumulated phase difference to exceed the 2π boundary. Therefore, the phase difference acquisition points for high-frequency components need to be deployed in the region near the acoustic wave injection end, that is, at the acoustic wave acquisition point E near the acoustic wave inlet, in order to limit the path difference and ensure the monotonicity and determinism of the high-frequency phase response.

[0065] Secondly, the maximum amplification in the low-frequency band (far incident region, such as...) Figure 1 , Figure 2 (Sound acquisition point F, far from the sound wave inlet): Low-frequency sound waves are limited by their long wavelength characteristics, resulting in a high physical threshold for preventing phase entanglement. If acquired near the inlet, it is difficult to establish a phase difference with a sufficient signal-to-noise ratio. Deploying the signal acquisition point deep within the vortex tube, far from the inlet, i.e., at sound acquisition point F far from the sound wave inlet, can obtain the maximum increase in centrifugal path and slow wave delay effect.

[0066] In summary, in conjunction with the aforementioned frequency space unwinding mechanism, multiple signal acquisition points are reserved along the propagation path on the waveguide wall in this embodiment. High-frequency micro-microphones are arranged near the entrance to capture the unwound high-frequency transient phase; the main microphone (responsible for full-frequency / low-frequency) is further deployed away from the entrance to obtain low-frequency phase data amplified by maximizing the path.

[0067] Example 3 Cascaded encoder combination with spatial sound field precoding Objective: This embodiment aims to illustrate how a basic acoustic waveguide structure can be compatiblely combined with an external front structure when challenging the limit of small angle detection resolution.

[0068] Extended design scheme: On the outside of the acoustic injection end A of the basic waveguide structure (such as a three-dimensional conical spiral tube) in Embodiment 1 or Embodiment 2, a spatial acoustic field precoding structure can be integrally formed or mechanically spliced ​​according to the frequency band and volume constraints of different application scenarios. To construct direction-dependent initial physical boundary conditions between the free field and the confined waveguide, the spatial acoustic field precoding structure preferably adopts one of the following three asymmetric front-end physical intervention configurations: Configuration 1: High and low folding staggered acoustic baffles (space folding maze architecture) The configuration includes a substrate extending circumferentially around the acoustic injection end A, and a plurality of rigid staggered baffles protruding vertically from the surface of the substrate. These rigid staggered baffles are arranged in a spatially asymmetrical array along the acoustic injection end, dividing the substrate surface into at least one high-fold region and at least one low-fold region.

[0069] Specifically, in the high-folding region, multiple rigid staggered baffles are arranged in a U-shaped or L-shaped nested configuration, with the ends of adjacent baffles overlapping each other, defining a Z-shaped meandering channel that forces the incident sound wave to undergo at least two spatial deflections. In the low-folding region, multiple rigid staggered baffles are arranged in a straight parallel or radial configuration, defining a straight or micro-folding channel without overlapping shielding. For incident sound waves from different spatial directions, this structure utilizes the asymmetry in the structural arrangement of the Z-shaped meandering channel in the high-folding region and the channel in the low-folding region. This allows sound waves with different incident angles to undergo physical propagation paths with significantly different numbers of reflections and lateral folding depths before entering the sound injection end A, thereby transforming the limited physical depth into a significantly extended effective propagation path and establishing an amplified initial propagation path difference.

[0070] Configuration 2: Phase gradient acoustic metasurface (microscopic local resonance architecture) In this configuration, multiple local resonant microstructures at deep subwavelength scales are arrayed in the external physical space directly opposite and obliquely surrounding the acoustic injection end A. The acoustic geometry parameters of each local resonant microstructure exhibit an asymmetric gradient variation along space.

[0071] This metasurface structure abandons the accumulation of macroscopic geometric paths and excites instantaneous phase abrupt changes in acoustic waves through internal local acoustic resonance, constructing an asymmetric equivalent phase interface around the acoustic injection end A. This equivalent phase interface utilizes a generalized Snell refraction / reflection anomaly mechanism to generate direction-dependent initial phase perturbations and anomalous wavefront deflections on long-wavelength incident acoustic waves from different spatial directions, introducing them into the waveguide interior with physically amplified differentiated grazing incidence angles.

[0072] Configuration 3: Asymmetric acoustic Fresnel reflector (anti-diffraction wavefront reshaping architecture) This configuration is made of solid material that satisfies the acoustic rigid reflection boundary conditions. Within the external space directly opposite and obliquely surrounding the sound injection end A, it is divided into multiple discrete phase modulation zones (step-like steps). Each phase modulation zone is asymmetrically distributed along the space, and there is a gradual change in geometric height between adjacent phase modulation zones.

[0073] To suppress parasitic diffraction of sound waves at the step edge, the geometric height transition region between adjacent phase modulation regions is configured as a smooth, continuous surface, thereby forming a continuous "acoustic blaze profile" on the asymmetric acoustic Fresnel reflector. This structure utilizes the acoustic blaze profile to asymmetricly reshape the wavefront of incident sound waves from different spatial directions, generating direction-dependent initial phase perturbations and initial incident angle shifts, allowing sound waves from different directions to enter the waveguide structure at differentiated grazing incident angles.

[0074] To further illustrate the physical form and engineering design concept of the spatial sound field precoding structure, combined with Figures 6 to 8As shown, this embodiment provides a specific example of an asymmetric multipath reflection configuration.

[0075] Compared to traditional acoustic receiver structures that are highly biomimetic or contain complex continuous folds and multiple micro-surfaces, complex surfaces, while increasing the coding precision of spatial angles, inevitably lead to extremely complex acoustic path distortion and disordered scattering. This excessively complex wavefront variation places stringent demands on the computational power and robustness of the backend acoustic modeling and reverse decoding algorithms.

[0076] Given the pain points of this project, such as Figure 6 (Front view) and Figure 7 As shown in the side view, the precoding structure of this embodiment adopts a macroscopically simplified asymmetric step-type reflection step design. This configuration eliminates redundant and complex folds, defining a first primary reflection surface G with a clear reflection surface and simple geometric contours, and a second primary reflection surface H near the center, within the physical space surrounding the acoustic injection end A. The first primary reflection surface G and the second primary reflection surface H form a significant high-low step difference in spatial depth.

[0077] like Figure 8 As shown, when external sound waves are incident, utilizing the multipath reflection mechanism of ray acoustics, part of the sound waves are reflected by the first primary reflecting surface G (as shown in sound ray L1), experiencing a relatively long return path; while another part of the sound waves are reflected by the second primary reflecting surface H (as shown in sound ray L2), experiencing a relatively short path. Through the asymmetrical arrangement of the first primary reflecting surface G and the second primary reflecting surface H, the system forces the incident sound waves to generate a definite initial path difference (i.e., the path difference between L1 and L2).

[0078] Furthermore, addressing the potential ambiguity in pitch angle spatial orientation caused by the geometric symmetry of the basic structure in the vertical direction, a preferred structural modification aimed at improving three-dimensional spatial resolution is proposed: referencing the asymmetric physical mechanism of the auricle in the sagittal plane, the upper and lower regions of the first primary reflector G are configured with an asymmetric shape. Specifically, the region of the first primary reflector G above the sound injection end A and the region below it are configured with different effective reflective areas, and their relative tilt angles toward the second primary reflector H are also different. This spatial geometric difference in the vertical dimension causes incident sound waves from different pitch angles to produce different multipath reflection trajectories, thereby effectively breaking down the ambiguity in sound source localization in the vertical direction.

[0079] This structural design not only successfully established initial phase perturbations related to the horizontal and pitch directions, but more importantly, its simple and clear physical reflection interface greatly reduces the difficulty of modeling acoustic boundary conditions. While ensuring the effectiveness of front-end precoding, it significantly reduces the computational complexity and decoding burden of the back-end signal processing module, providing an excellent balance solution for the engineering implementation of three-dimensional spatial direction finding for miniaturized acoustic detection equipment.

[0080] Workflow and cascade physical mechanism: When a sound source with a slight angular difference in the external space comes into contact with this system, the physical modulation is divided into two cascaded stages: Phase 1 (Free-field initial precoding): The incident sound wave is first incident on the spatial sound field precoding structure of any of the above configurations. Using spatial folding, resonant abrupt change, or blazing reflection mechanisms, weak external spatial differences are transformed into significant initial reflected beams or anomalous deflection wavefronts.

[0081] Phase Two (Secondary Cascaded Amplification Inside the Waveguide): Acoustic waves carrying differentiated initial boundary conditions enter the waveguide from the acoustic injection end A. Since the initial angles have been pre-differentiated, the centrifugal tangential forces experienced by the acoustic waves at the bends of the waveguide are differentiated, resulting in significant differences in the depth of interaction and the number of passes over the acoustic impedance discontinuities (gradient microgrooves) on the outer wall D1.

[0082] Conclusion: When the basic waveguide structure is combined with the above-mentioned spatial acoustic field precoding structure, deep physical coupling of "peripheral wavefront reshaping / folding + internal slow wave delay" is achieved, which enhances the system's ability to detect and resolve extremely small angles.

[0083] Example 4 Frequency-domain multiplexed multidimensional acoustic direction-finding array Objective: This embodiment demonstrates how to combine multiple worm gear encoders into a hardware array to achieve omnidirectional sound source detection in complex broadband environments.

[0084] Array topology arrangement: On a ring or polygonal substrate (e.g., 120mm in diameter), four worm gear encoders from Embodiment 2 are arranged radially at preset angles (e.g., 90° intervals). The acoustic injection end A of each encoder faces the external space, and the acoustic output end B is concentrated in the internal central region.

[0085] Frequency domain multiplexing group design: To improve the recognition accuracy of broadband speech, the four tubes are configured into two frequency domain multiplexing subgroups: The first subgroup (responsible for lower frequency bands): for example, positioned at 0° and 180°. The internal channels are relatively wide, and the initial depth of the D1 gradient microgrooves on the outer wall is relatively deep (e.g., 2.5 mm), which is suitable for exciting the slow-wave delay response of longer wavelength sound waves.

[0086] The second subgroup (responsible for higher frequency bands): for example, positioned at 90° and 270°. The internal channels are relatively compact, and the outer wall D1 has a shallowly distributed micro-tooth structure (e.g., 0.3mm). This design reduces the scattering and attenuation of high-frequency sound waves in the deep cavity, ensuring the continuity of high-frequency phase modulation.

[0087] Collaborative operation: When broadband voice impacts the array, different frequency band components acquire corresponding eccentric offsets and slow-wave delays in their respective matched subgroups. The multi-channel analog-to-digital converter (ADC) at the center of the array uses a unified clock for synchronous sampling, converting the physically amplified multi-channel sound waves into a digital audio stream containing spatially orthogonal characteristics.

[0088] Example 5 Acoustic signal angle encoding and detection method (hardware-software co-processing system) Objective: To provide a complete end-to-end system working method from physical front-end capture to final algorithm output of the three-dimensional coordinates of the sound source.

[0089] Offline Prior Stage (Constructing the Spatial Angle Mapping Model): Given the highly nonlinear phase amplification mechanism introduced by the front-end cochlear tube, this embodiment employs a deep neural network (such as MLP or 1D-CNN) to construct the mapping model. The multi-channel distortion phase difference feature vectors measured by multi-angle frequency sweeps are used as network inputs, and the physical spatial vectors (X, Y, Z) of the actual sound source are used as network labels for supervised training. The trained model implicitly learns the complex nonlinear spatial mapping matrix of the hardware front end and deploys it in the system processing chip (such as DSP or NPU).

[0090] Online detection phase (actual operation process): Step 1 (Spatial Feature Capture and Physical Centrifugal Guidance): The target acoustic wave contacts the array (if the front end is preferably configured with a spatial acoustic field pre-coding structure, then the initial boundary offset related to the direction is established by it). Subsequently, the external acoustic wave is guided into each acoustic waveguide structure (i.e., the spiral tube). Each acoustic waveguide structure converges acoustic energy through a gradually decreasing cross-section, while the curved spiral geometry forces the acoustic wave with a small difference in incident angle to produce an effective propagation path length difference (radial centrifugal offset) between the inner wall C1 and the outer wall D1.

[0091] Step 2 (Mesoscopic Slow Wave and Feature Physical Amplification): Centrifugal acoustic waves biased towards the outer wall D1 excite the slow wave effect of the microgroove. Through this delayed intervention, the initial minute spatial difference is purely physically amplified (or cascaded amplification if combined with a pre-coding structure) into a significant characteristic phase difference at the end.

[0092] Step 3 (Extraction and Reverse Decoding Reconstruction): The signal processing module performs frequency domain transformation (such as STFT) on the digital signal picked up by the sensor in real time to extract the phase difference sequence. Then, it calls the spatial angle mapping model to perform forward inference calculation, performs orthogonal reverse decoding on the nonlinear redundant sequence, and then outputs the three-dimensional spatial vector direction of the sound source in real time.

[0093] In this application, the term "physical amplification" or "phase amplification" refers to: using the aspect ratio, curvature geometry, and mesoscopic acoustic interface characteristics of the sound waveguide structure to physically intervene in obliquely incident sound waves from different directions, so that the difference in the propagation path length and the reduction in propagation speed before reaching the sound sensor produce a multiplier effect, thereby making the phase difference of the sound waves that finally reach each sensor significantly greater than the initial spatial phase difference of the sound waves in the free field due to different propagation distances.

[0094] In this application, the term "encoding" refers to the system not actively emitting a specific sequence of sound waves, but rather using a purely physical structure to differentiate the propagation paths and local sound speeds of sound waves from different incident directions, thereby "imprinting" or "mapping" the three-dimensional spatial direction information of the external sound field into a specific phase difference characteristic sequence between multiple output signals.

[0095] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. An acoustic angle encoder based on geometric waveguides, characterized in that, Includes at least one acoustic waveguide structure; The acoustic waveguide structure is a hollow cavity and has an acoustic injection end and an acoustic output end; The acoustic wave guiding structure extends in a curved manner along the direction of acoustic wave propagation, defining an inner wall near the center of curvature and an outer wall away from the center of curvature. The aspect ratio of the acoustic waveguide structure from the acoustic injection end to the acoustic output end is 5:1 to 20:1, wherein the aspect ratio is the ratio of the unfolded length of the acoustic waveguide structure along the central axis to the equivalent hydrodynamic diameter of the hollow cavity. The inner wall is a smooth, continuous curved surface; Multiple acoustic impedance discontinuities are provided on the outer side wall; The acoustic waveguide structure is configured as follows: Establish direction-dependent propagation path differences and / or propagation speed differences for sound waves incident from different spatial directions, and form phase difference and / or time delay characteristics corresponding to the incident direction at the sound output end.

2. The acoustic angle encoder based on geometric waveguide according to claim 1, characterized in that: The acoustic waveguide structure is either a three-dimensional conical spiral waveguide or a two-dimensional asymmetric curvature gradient structure. When it is in the form of a three-dimensional conical spiral waveguide, the hollow cavity extends continuously along the conical surface in three-dimensional space, and the cross-sectional area of ​​the hollow cavity gradually decreases from the sound injection end to the sound output end; when it is in the form of a two-dimensional asymmetric curvature gradient, the central axis of the hollow cavity bends and unfolds asymmetrically in a spiral shape in a single plane. When the cross-sectional area of ​​the hollow cavity is not uniform, the equivalent hydrodynamic diameter is the average equivalent inner diameter along the direction of sound wave propagation.

3. The acoustic angle encoder based on geometric waveguide according to claim 1, characterized in that: The characteristic dimensions of the multiple acoustic impedance discontinuities vary in gradient along the direction of sound wave propagation to maintain the continuity of equivalent sound velocity modulation in different frequency ranges and suppress high-frequency phase winding and local acoustic distortion.

4. An acoustic angle encoder based on a geometric waveguide according to claim 1, characterized in that: The acoustic impedance discontinuity structure includes one or more of the following: groove, thread, protrusion, bypass tube, and local resonant cavity, and its characteristic size is less than 1 / 10 of the lowest operating wavelength of the target operating frequency band.

5. A multidimensional acoustic direction-finding array, characterized in that, It includes at least two acoustic angle encoders as described in any one of claims 1 to 4, and at least two sound sensors coupled thereto; The acoustic injection ends of each acoustic angle encoder are arranged according to a preset spatial topology to form an array input surface; Each of the aforementioned acoustic angle encoders is configured as follows: Establish spatial phase coding features with differences for different spatial directions.

6. The multidimensional acoustic direction-finding array according to claim 5, characterized in that: At least two of the acoustic angle encoders have different geometric parameters or acoustic modulation parameters, so that the encoders at different positions form an angle-phase mapping relationship that can be distinguished from each other; For the same spatial sound source, at least two of the acoustic angle encoders output phase difference sequences corresponding to different directional features; The multiple electrical signals output by the sound sensor are used to construct a spatial angle mapping matrix to calculate the azimuth and pitch angles.

7. The multidimensional acoustic direction-finding array according to claim 5, characterized in that: The array is configured as a frequency domain multiplexing architecture; At least two of the acoustic angle encoders are divided into at least two subgroups; Different subgroups correspond to spatial phase coding within different wavelength ranges.

8. A sound wave direction detection system, characterized in that, It includes a multi-dimensional acoustic direction-finding array as described in any one of claims 5 to 7, and a signal processing module electrically connected to each of the sound sensors; The signal processing module is configured as follows: Extract the relative phase difference and / or time offset between each acoustic signal to form a phase difference sequence; The phase difference sequence is then input into a spatial mapping model trained based on the front-end physical propagation characteristics to output the spatial direction information of the incident sound wave.

9. A method for detecting the direction of sound waves, characterized in that, The acoustic direction detection system as described in claim 8 includes the following steps: Step 1: Receive external incident sound waves and guide them into each of the sound wave guiding structures for propagation, forcing sound waves with slight differences in incident angle to generate an effective propagation path length difference between the inner wall near the center of curvature and the outer wall far from the center of curvature. Step 2: During the propagation of sound waves, the acoustic impedance discontinuity structure set on the outer wall is used to apply a slow wave delay effect to the sound waves deflected towards the outer wall, so as to output an amplified terminal sound wave signal at the sound output end. The terminal sound wave signal carries a phase difference sequence corresponding to the spatial direction. Step 3: The phase difference sequence is picked up by the sound sensor and then reverse-decoded by the signal processing module to reconstruct the spatial direction information of the sound source.

Images