Composite superstructure compressed sensing acoustic imaging device and method

Abstract

The invention relates to a composite superstructure compressed sensing acoustic imaging device and method, the device comprises a disc-shaped superstructure formed by combining two fish-shaped bionic psittaclus superstructures, a plurality of internal cavities are distributed on the fish-shaped bionic psittaclus superstructures, eight folding channel units form an octagonal superstructure, and the octagonal superstructure is provided with a plurality of hollow cavities. Orthogonal sound channels are distributed on the folding channel units. The method comprises the following steps: according to a target sound source detection frequency band, determining geometric parameters of an acoustic unit, according to target sound information and an expected position, dividing a detection direction into a plurality of grid points, carrying out predictive detection on each target sound source at each grid point, and according to complex sound signals measured by internal sensors of two fish-shaped bionic psittaclus superstructures, determining a target sound source according to the complex sound signals. A compressed sensing observation matrix is jointly constructed, and system testing and training are completed; and according to complex sound signals measured by the two sensors, in combination with the compressed sensing observation matrix, reversely deducing the azimuth relation between the sound source and the two sensors, and according to the distance between the two sensors and the two azimuth angles, obtaining the sound source direction and the relative distance.

Description

Technical Field

[0001] The present invention belongs to the technical field of underwater acoustic imaging, and in particular relates to a composite superstructure compressed sensing acoustic imaging device and method. Background Art

[0002] The ocean not only brings resources and convenience to people's production and life, but is also an important part of the military strategic map. In underwater warfare, detection is a crucial link. Electromagnetic waves are difficult to propagate underwater, while sound waves have unique characteristics when propagating underwater, such as omnidirectionality and multi-media propagation, and can adapt to complex underwater environments. Sound positioning is widely used in various engineering fields, including structural health monitoring, target identification, underwater acoustic communication, etc. The development of small, low-cost, high-performance detection devices and technologies is of great strategic significance for establishing a strong marine strategic scientific and technological force. Underwater imaging plays an important role in sonar detection, which is widely used in the marine military field and underwater resource detection and development. Traditional underwater acoustic imaging is generally achieved by phased sensor arrays, but it is greatly affected by algorithms, the number of sensors, and the spatial array method. It has the disadvantages of high complexity, low space utilization, and high energy consumption. Therefore, improving the existing underwater acoustic detection devices has always been a hot topic in the field of acoustics. In recent years, people have begun to use acoustic superstructures with subwavelength size characteristics to achieve complex acoustic imaging functions. Acoustic superstructures are artificial materials that manipulate sound waves through structural design. They show superior performance compared to traditional materials. Combined with the theory of compressed sensing, by designing acoustic superstructures with high spatial anisotropy, complex acoustic imaging functions including multi-source identification, high-resolution perception and tracking can be achieved. The combination of compressed sensing and superstructures has proposed new ideas for establishing small-volume, high-precision underwater acoustic detection systems.

[0003] Although the underwater acoustic sensing device proposed in the past can detect weak underwater signals, it is only sensitive to sound waves in a single direction and resonant frequency, and cannot perceive complex sound sources in space. Although the current acoustic imaging device based on anisotropic superstructures can perceive and track multiple sound sources, it only stays in the air environment and is difficult to perceive sound sources in the underwater environment. It only detects the direction of the sound source and cannot determine the distance. The identification accuracy and resolution are not high enough to meet the actual engineering application. It is necessary to take into account the underwater sound transmission parameters in the design of the anisotropic acoustic superstructure, appropriately increase the number of sensors and consider multiple sound source information to achieve spatial sound source identification and positioning tracking functions within a wider frequency band underwater. The frequency response can be adjusted according to the actual application scenario to design an imaging device that meets the actual environment application needs. Summary of the invention

[0004] The purpose of the present invention is to overcome the shortcomings of existing acoustic imaging devices and provide a composite compressed sensing acoustic imaging device and method for realizing near- and far-field high-precision sound source identification, precise ranging and positioning tracking functions for multiple sound sources in air or underwater space, and the frequency response can be adjusted according to the actual application scenario.

[0005] In order to achieve the above object, the present invention adopts the following technical solution:

[0006] A composite superstructure compressed sensing acoustic imaging device comprises a disc-shaped structure composed of two fish-shaped bionic nautilus superstructures, on which a plurality of internal cavities are distributed, and an octagonal superstructure composed of eight folded channel units, on which orthogonal acoustic channels are distributed.

[0007] A further improvement of the present invention is that the outer spiral line of the fish-shaped bionic nautilus superstructure satisfies the equation: t =rcos(t h ), y t =rsin(t h ), z t = 0, where t h =360nt, a is a parameter that controls the degree of divergence of the spiral, n is the number of turns of the spiral, and t is an independent variable ranging from 0 to 1.

[0008] A further improvement of the present invention is that the device can realize acoustic imaging in a gas medium represented by air, or realize acoustic imaging in a liquid medium represented by water, and when working in different media, the cavity is filled with corresponding working media.

[0009] A further improvement of the present invention is that when the sound waves emitted by the sound source enter the disc-shaped superstructure from the outside, the sound waves in different directions pass through different orthogonal sound channels and cavities, and finally generate different phases and amplitudes of the sound signals at the two detection points, that is, the direction of the sound source is determined by the anisotropy presented by the artificial structure.

[0010] A further improvement of the present invention is that the frequency response can be directly adjusted by adjusting the geometric parameters of the acoustic unit, including the length of the orthogonal acoustic channel and the volume of the cavity.

[0011] A composite superstructure compressed sensing acoustic imaging method, based on the composite superstructure compressed sensing acoustic imaging device, comprises:

[0012] 1) According to the target sound source detection frequency band, determine the geometric parameters of the acoustic unit, including the length of the orthogonal sound channel and the shape and volume of the cavity;

[0013] 2) According to the target sound information and expected position, the detection direction is divided into several grid points, and the target sound source is predicted and detected at each grid point. According to the complex sound signals measured by the internal sensors of the two fish-shaped bionic nautilus superstructures, the compressed sensing observation matrix is ​​jointly constructed to complete the system testing and training;

[0014] 3) In actual measurement, based on the complex sound signals measured by the two sensors and combined with the compressed sensing observation matrix, the azimuth relationship between the sound source and the two sensors is reversely deduced, and the direction and relative distance of the sound source are obtained based on the distance between the two sensors and the two azimuths.

[0015] A further improvement of the present invention is that the outer spiral line of the fish-shaped bionic nautilus superstructure satisfies the equation: t =rcos(t h ), y t =rsin(t h ), z t = 0, where t h =360nt, a is a parameter that controls the degree of divergence of the spiral, n is the number of turns of the spiral, and t is an independent variable ranging from 0 to 1. A further improvement of the present invention is that the method can realize acoustic imaging in a gas medium represented by air, or realize acoustic imaging in a liquid medium represented by water, and when working in different media, the cavity is filled with corresponding working media. A further improvement of the present invention is that when the sound waves emitted by the sound source enter the disc-shaped superstructure from the outside, the sound waves in different directions pass through different folded sound channels and cavities, and finally generate different phases and amplitudes of the sound signals at the two detection points, that is, the direction of the sound source is determined by the anisotropy presented by the artificial structure. A further improvement of the present invention is that the frequency response can be directly adjusted by adjusting the geometric parameters of the acoustic unit, including the length of the orthogonal acoustic channel and the volume of the cavity. Compared with the prior art, the present invention has at least the following beneficial technical effects:

[0016] In summary, according to the above characteristics of a composite compressed sensing acoustic imaging device provided by the present invention, the acoustic imaging device provided by the present invention can realize single sound source positioning and ranging in air or underwater three-dimensional space. The structure of this device is simple and regular, unique and beautiful, and has a variety of designable factors such as: unit channel length and width, cavity shape and volume, number and shape of units, unit arrangement and opening direction, structural wall thickness and overall height, sensor position. By designing the nautilus superstructure and the folded channel superstructure, it is possible to achieve customized regulation of the incident sound wave, reduce the structural size and system complexity, and adjust the structural strength and perception range to adapt to different working environments. At the same time, the structural features of the two different Pisces Nautilus make the imaging process of the two sensors independent of each other, so that the sound source distance can be determined by two independent sound source orientation identifications. The use of the acoustic superstructure realizes sound wave capture through the physical structure, which is more stable and reliable. The resonance between the structural unit and the sound wave enhances the sound wave in the special frequency band, weakens the underwater transmission loss, and protects the information integrity. The use of multi-information fusion compressed sensing method and dual-sensor joint compressed sensing method reduces the computational complexity, improves the reconstruction accuracy, overcomes the defects of spatial sampling theorem and cone ambiguity of the perpendicular position, reduces the sampling rate, and realizes the resolution, positioning and tracking of multiple underwater sound sources. In summary, compared with the existing sensor array imaging system, this acoustic imaging device has a smaller size and more stable structure, reduces the number of sensors, and the imaging performance can be flexibly adjusted according to the actual working conditions. Compared with the existing compressed sensing super-structure imaging device, it integrates multiple information, improves accuracy, increases the number of sound sources that can be distinguished at the same time, and can realize the distance measurement of the sound source. It can be widely used in acoustic imaging of various devices and environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 a is a schematic diagram of the composite bionic superstructure;

[0018] Figure 1 b is a schematic diagram of superstructure acoustic transmission characteristics;

[0019] Figure 1 c is a schematic diagram of the superstructure acoustic field distribution;

[0020] Figure 2 a is the training process, and the schematic diagram of the measurement matrix is ​​obtained;

[0021] Figure 2 b is the test process, and the schematic diagram of the observation vector is obtained;

[0022] Figure 2 c is a schematic diagram of the joint compressed sensing imaging method;

[0023] Figure 2 d is a schematic diagram of the target vector obtained by compressed sensing reconstruction;

[0024] Figure 3 a is a schematic diagram of the transmission characteristics of the random folded superstructure;

[0025] Figure 3 b is a schematic diagram of the transmission characteristics of the sensor on the disk-shaped superstructure;

[0026] Figure 3 c is a schematic diagram of the sensor transmission characteristics under the disk-shaped superstructure;

[0027] Figure 3 d is a comparative schematic diagram of the transmission characteristics under 0° incident sound waves;

[0028] Figure 4 a is the diagram of the experimental setup;

[0029] Figure 4 b is an enlarged image of the bionic nautilus superstructure;

[0030] Figure 4 c is a schematic diagram of the sound position of the speaker;

[0031] Figure 5 a is the spectrum of sound source 1;

[0032] Figure 5 b is the spectrum of sound source 2;

[0033] Figure 5 c is the spectrum of sound source 3;

[0034] Figure 5 d is the spectrum of sound source 4;

[0035] Figure 6 a is a schematic diagram of the frequency response of sound sources incident from different directions;

[0036] Figure 6 b is the correlation coefficient graph;

[0037] Figure 6 c is a schematic diagram of the single sound source localization results using sensor 1, sensor 2, and dual sensors combined with compressed sensing;

[0038] Figure 6 d is a schematic diagram of the dual sound source identification results using sensor 1, sensor 2, and dual sensors combined with compressed sensing;

[0039] Figure 6 e is a schematic diagram of the difference between the dual sound source identification results of the dual sensor joint method and the single sensor compressed sensing method;

[0040] Figure 7 a is the test image of the sound localization and identification experiment with a resolution of 5°;

[0041] Figure 7 b is a schematic diagram of single sound source localization results;

[0042] Figure 7 c is a schematic diagram of the dual sound source identification result;

[0043] Figure 7 d is a schematic diagram of the three sound source identification results;

[0044] Figure 8 a is a schematic diagram of the detailed results of the sixth experiment with a single sound source;

[0045] Figure 8 b is a schematic diagram of the detailed results of the 8th experiment with dual sound sources;

[0046] Figure 8 c is a schematic diagram of the detailed results of the 13th experiment with three sound sources; Fig. 9 a is a schematic diagram of all the results of four sound source identification; Fig. 9 b is a schematic diagram of the detailed results of the fourth experiment of four-sound source identification; Fig. 9 c is a schematic diagram of the detailed results of the 14th experiment of four-sound source identification; Fig. 9 d is a schematic diagram of the detailed results of the 19th experiment of four-sound source identification; Fig.10 a is a schematic diagram of the experiment for determining the distance of the sound source; Fig.10 b is the principle diagram for determining the distance to the sound source; Fig.10 c is a schematic diagram of the position and distance of a single sound source relative to sensor 1 and sensor 2; Fig.10 d is a schematic diagram of the success rate of multi-sound source orientation and distance identification; Fig.11 a is a schematic diagram of the relative rotation method; Fig.11 b is the actual operation diagram of the relative rotation method; Fig.12 a is a schematic diagram of the experimental system and device; Fig.12 b is a schematic diagram of multiple sound source identification results; Fig.12 c is a schematic diagram of the detailed results of single sound source azimuth identification and distance positioning; Fig.12 d is a schematic diagram of the results of determining the distances of multiple sound sources; Fig.13 a is a schematic diagram of the far-field experimental system and device; Fig.13b is a comparison chart of the multi-source azimuth recognition performance of the acoustic imaging device at a height of 10 mm and a height of 200 mm; Fig.13 c is a comparison chart of the multi-source distance determination performance of the acoustic imaging device at a height of 10 mm and a height of 200 mm. DETAILED DESCRIPTION The technical scheme in the implementation case of the present invention is described below in conjunction with the relevant calculation and measurement results of the implementation case of the present invention in the accompanying drawings. Obviously, the implementation case described is only a part of the implementation case of the present invention, not all the implementation cases. Based on the implementation case in the present invention, all other implementation cases obtained by those skilled in the art without making creative work belong to the scope of protection of the present invention. Example 1 The present invention provides a composite superstructure compressed sensing acoustic imaging device, comprising a disc-shaped superstructure 2 composed of two fish-shaped bionic nautilus superstructures 1, a plurality of internal cavities 3 distributed on the fish-shaped bionic nautilus superstructure, an octagonal superstructure 5 composed of eight folded channel units 4, and orthogonal acoustic channels 6 distributed on the folded channel units. In this embodiment, the outer spiral line of the fish-shaped bionic nautilus superstructure 1 satisfies the equation: t =rcos(t h ), y t =rsin(t h ), z t = 0, where t h =360nt, a is a parameter that controls the degree of divergence of the spiral, n is the number of turns of the spiral, and t is an independent variable ranging from 0 to 1. In this embodiment, the device can realize acoustic imaging in a gas medium represented by air, or realize acoustic imaging in a liquid medium represented by water. When working in different media, the cavity 3 is filled with corresponding working media. In this embodiment, when the sound waves emitted by the sound source enter the disc-shaped superstructure 2 from the outside, the sound waves in different directions pass through different orthogonal sound channels 6 and cavity 3, and finally the phases and amplitudes of the sound signals generated at the two detection points are different, that is, the direction of the sound source is determined by the anisotropy presented by the artificial structure. In this embodiment, the frequency response can be directly adjusted by adjusting the geometric parameters of the acoustic unit, including the length of the orthogonal acoustic channel 6 and the volume of the cavity 3 . Among them, the superstructure designed by the device using the bionics design concept has simple and regular, unique and beautiful structural features, which reduces acoustic loss and ensures ease of production. The designed superstructure acoustic imaging device with the anisotropic characteristics of the Pisces spiral and the anisotropic properties of the curled folding provides a new design idea for the design of spatial coding superstructures and has good acoustic wave control performance. The device can perform multi-element joint observation based on frequency, phase, amplitude, and complex sound pressure information, thus improving the accuracy and performance of target recognition, differentiation, and detection. The device can achieve high-resolution and high-recognition rate recognition of multiple sound sources, enhance compressed sensing imaging performance, and simultaneously determine the direction and distance of the sound source. The device is not only capable of near-field imaging, but also has the advantage of excellent far-field free-space imaging performance similar to the actual application environment. It can identify the orientation of multiple far-field sound sources in a free-space environment and determine the distance of multiple sound sources. Example 2 The present invention provides a composite superstructure compressed sensing acoustic imaging method, which is based on the composite superstructure compressed sensing acoustic imaging device, and includes: 1) According to the target sound source detection frequency band, determine the geometric parameters of the acoustic unit, including the length of the orthogonal sound channel 6 and the shape and volume of the cavity 3; 2) According to the target sound information and expected position, the detection direction is divided into several grid points, and the target sound source is predicted and detected at each grid point. According to the complex sound signals measured by the internal sensors of the two fish-shaped bionic nautilus superstructures 1, the compressed sensing observation matrix is ​​jointly constructed to complete the system testing and training; 3) In actual measurement, based on the complex sound signals measured by the two sensors and combined with the compressed sensing observation matrix, the azimuth relationship between the sound source and the two sensors is reversely deduced, and the direction and relative distance of the sound source are obtained based on the distance between the two sensors and the two azimuths. Example 3 (I) Design of the structure of the composite superstructure compressed sensing acoustic imaging device like Figure 1 As shown in (a), the inside of the disc-shaped superstructure has two fish-shaped nautiluses in a head-to-tail embrace posture, and the two are cleverly combined together to make the internal structure have dual anisotropy. The outside of the disc-shaped superstructure has eight different folded channel units, which change the sound transmission through orthogonal modulation, so that the external structure has multi-directional anisotropy. The double fish disc-shaped superstructure and the folded channel octagonal superstructure work together and couple with each other to jointly realize the composite structural modulation of sound waves, wherein the double fish nautilus superstructure has different parameters, and the combination of the two makes the combination of dual detectors possible. Inspired by the anisotropic superstructure composed of a resonant gain cavity and a curled folded orthogonal modulation superstructure, as well as the gradient gradual change characteristics and self-similar characteristics of the nautilus structure in nature, the present invention designs a composite bionic acoustic superstructure, wherein the inner circle diameter D = 30cm, and the outer circle side length L = 40cm. The superstructure consists of two parts, the inner and outer parts. The inner part is a disc-shaped superstructure composed of two fish-shaped bionic nautilus superstructures, and the outer part is an octagonal superstructure composed of eight folded channel units. The combination of the inner and outer structures makes the superstructure have both spatial folding and spiral gradient characteristics. The inner disc-shaped superstructure is composed of two bionic nautilus superstructures. The two fish-shaped bionic nautilus superstructures have different spatial layers and opening positions, ensuring that the two parts of the disc-shaped superstructure have different acoustic transmission characteristics. The parametric equation for determining the outer spiral line of the nautilus structure is x t =rcos(t h ), y t =rsin(t h ), z t = 0, where t h =360nt, a is a parameter that controls the divergence of the spiral, n is the number of turns of the spiral, and t is an independent variable ranging from 0 to 1. The bionic nautilus superstructure contains multiple chambers, each of which gradually decreases in volume from the outside to the inside and from the beginning to the end of the spiral. The middle chambers are connected by a through body tube. At the same time, in order to enable the superstructure to receive omnidirectional sound signals, holes of different sizes are opened in sequence on each layer of the spiral shell. The opening expression of the outermost shell and the middle through body tube is d n =d1-(n-1)Δd, where d1=10mm (left-hand spiral) or d1=11mm (right-hand spiral) is the aperture of the outermost unit cell, and Δd=1mm is the unit cell aperture reduction factor. Considering the difficulty of manufacturing, the minimum aperture d is fixed. min =1mm. The expression of the openings of other layers of the shell is the same as that of the outermost shell, except that the aperture of the outermost unit cell of the inner shell is smaller, and the aperture decreases by 1mm. The external octagonal superstructure is composed of eight different folded channel units, each of which contains two cavities formed by three strips and slots opened at different positions on the strips, where the length of the strip is l=146mm, the width is w=2mm, the interval between the strips is s=10mm, and the length of the opening slot is g=24mm. The eight different folded channel units are orthogonal to the eight directions, so that sound waves incident from different directions have unique propagation paths after different reflections. The height of the superstructure is designed to be h=10mm, corresponding to λ / 11.3 at 3000Hz. The composite bionic superstructure can be regarded as a combination of different acoustic channels in multiple directions, which is equivalent to a combination of multiple filters. Figure 1In (a), each acoustic channel contains two layers of folded channels on the outside and two or three layers of spiral cavities on the inside. The outer single-layer folded channel and the inner single cavity each correspond to an acoustic unit. The geometric parameters of the acoustic unit, including the folded channel length, the volume of the cavity, and the position distribution of the holes, will directly affect the frequency response. The combination of cells with different geometric parameters constitutes an acoustic channel with different transmission characteristics, and the combination of acoustic channels with different transmission characteristics constitutes a superstructure with spatial frequency modulation performance. The finite element method is used in COMSOL multi-physics simulation software to calculate the frequency response of the superstructure model under the action of sound sources in 10 different incident directions, such as Figure 1 (b) As shown in the figure, although the frequency responses of sound sources incident from different directions appear similar in the narrowband frequency band, the frequency response changes dramatically with direction in the broadband frequency range, which indicates that sound sources incident from different directions have different frequency responses. At this point, the frequency response diagram of the superstructure can be used. Figure 1 (c) The upper right corner shows a graph in which different units have different colors, and different colors represent different frequency responses. The geometric size of the superstructure determines the operating frequency range in practical applications, and the length of the outer ring of the superstructure is defined as the characteristic size. The characteristic size corresponds to approximately 7 to 8 times the minimum wavelength of the effective working frequency band, and approximately 3 to 4 times the maximum wavelength of the effective working frequency band. Further, the sound field distribution of the superstructure under the action of incident sound sources from different directions is calculated, as shown in Figure 1 (c) It can be seen that even if the positions of the two sound sources differ by only 10°, the sound waves incident from different directions still have different sound pressure amplitudes after being modulated by the superstructure, and have different sound field distributions at different frequencies. The above results show that the superstructure has a direction-sensitive acoustic frequency response and a high resolution. 2. Joint Compressed Sensing Acoustic Imaging Based on Composite Superstructure The compressed sensing imaging process includes training and testing. Figure 2 As shown in (a), the sound source is played in different positions outside the superstructure in turn. After a series of preprocessing, the measurement matrix A = [A1, A2..., A c ]. The number of columns c of the matrix is ​​determined by the number of sound source positions p and the number of sound source types t, c = p × t. The test process is as follows Figure 2 As shown in (b), multiple sound signals are randomly selected from the audio library, and the sound signals are played at multiple positions outside the super structure to obtain the observation vector Y = S i (ω)°H j (w). Among them, S i (ω) is the frequency domain vector of the ith sound source, H j (w) is the frequency response vector at the jth position, and ° is the Hadamard product. Figure 2As shown in (c), since the two parts of the disc-shaped superstructure contain two different bionic nautilus superstructures, they can be regarded as two different sparse blocks. In the process of sound source identification, in order to simplify the problem, the plane area around the superstructure is divided into m grids, and a single sound source occupies one grid. Assume that there are two sound sources in the plane space, and the two sound sources are randomly distributed in different position blocks. In the actual test process, n frequency points are selected at each grid point, and the response data of the n frequency points are formed into a one-dimensional column vector. The sound source frequency responses obtained by the two sensors are combined to form a joint observation matrix. This process can be expressed as: Among them, Y1, A1 are the observation vector and measurement matrix obtained by sensor 1; Y2, A2 are the observation vector and measurement matrix obtained by sensor 2; S is the target vector obtained by the joint compressed sensing imaging method. Considering that using the complex sound pressure information of the sound source for compressed sensing imaging can overcome the high correlation defects of the regular structure itself, reduce the dependence on anisotropy, and have anti-noise interference performance, the joint compressed sensing imaging method also uses complex sound pressure information for identification and reconstruction. After obtaining the measurement matrix and observation vector, the two-step iterative shrinkage threshold method (TwIST) is used to solve the underdetermined equation system Y=AS to determine the target vector. This is an L0 norm minimization problem, which is an NP-complete problem. There is no corresponding fast solution, so it is often converted into an L1 norm minimization solution: in is the solution value of the target vector, ||S||1 represents the minimization of the L1 norm, corresponding to ∑ i |AS i -y i |Theoretically, the position with a sound source corresponds to a non-zero value, and the position without a sound source corresponds to a zero value. However, due to the existence of measurement errors, in reality, the maximum data point represents the actual position of the sound source, such as Figure 2 (d) as shown. (III) Single and dual sound source identification In order to prove that the composite superstructure can indeed reduce the loss, the loss of the randomly designed folded structure and the regularly designed anisotropic structure are compared by numerical calculation methods, such as Figure 3 As shown in the figure, compared with the random design structure, the regular design structure has a larger transmission coefficient in the frequency band of 3000-7000Hz, which proves that the regular structure has a smaller acoustic loss. The single sound source localization and dual sound source identification experiments were carried out using the composite bionic superstructure and the joint compressed sensing imaging method. Figure 4 The experimental test platform is shown in Figure 4(a) shows the composite bionic superstructure (enlarged view as shown in Figure 4 (b) is fixed in the middle of the optical platform and is made of PLA plastic through photocuring 3D printing technology. Two sensors are fixed in the center of the two parts of the superstructure. The model is B&K4958-A and is used to pick up sound signals. Two speakers are fixed on the outside of the superstructure to emit sound signals. The sound generation diagram is shown in Figure 4 (c) As shown. 10 acoustic waveguides are connected to the speaker, with a 10° interval between each two, for transmitting sound signals. The Motu 16A sound card is connected to the PC and the speaker to receive the original sound signal from the PC and drive the speaker. The data collector is connected to the sensor and the PC to receive the sound signal detected by the sensor and then input the signal into the PC for data processing. In order to study the sound modulation performance of the superstructure system, the sound signal generated by the chirp function in MATALB (such as Figure 5 As shown in Figure 2, the superstructure system is continuously incident from 10 different directions to obtain the superstructure frequency response characteristics, as shown in Figure 2. Figure 6 As shown in (a), different colors in the figure represent responses of different intensities. It can be seen from the figure that, on the one hand, the sound signals received by a single sensor from different directions have different spectral characteristics, and the frequency response is related to the direction; on the other hand, the sound signals received by two sensors have different frequency responses, and there are differences between the sensors. Define two responses μ i and μ j The coherence between them is μ ij = <μ i ,μ j > / |μ i ||μ j | is used to measure the overlap between two frequency responses. Further, the correlation coefficient of the response signals of the incident sound sources from different directions is calculated, such as Figure 6 (b) As shown in the figure, the average correlation coefficient of sensor 1 is about 0.73, and the average correlation coefficient of sensor 2 is about 0.63, which proves that the superstructure has certain different sound modulation performance. However, the two sensors have a large correlation coefficient, which makes the accuracy of sound source localization directly using the sound pressure amplitude signal low. Therefore, a complex sound pressure signal is used to reduce the influence of correlation, and the complex signal is used to locate and identify the sound source. A sound source signal is played at a random position to test the single sound source localization performance of the sound source monitoring system based on the superstructure, as shown in Figure 6 As shown in (c), it can be seen from the figure that the position of a single sound source can be determined by using sensor 1 and sensor 2 separately, and the estimated position of the sound source is consistent with the actual position. However, the joint observation of dual sensors can make the single sound source localization result clearer and more distinguishable. Furthermore, the dual sound source identification performance of the composite bionic superstructure combined with compressed sensing imaging system was tested. Two different test signals were randomly selected from the audio library and emitted simultaneously by two speakers. The mixed signal picked up by the sensors at the center of the two fish-shaped nautilus superstructures contains the location information and audio content information of the two sound sources. The reconstruction algorithm is used to restore the location and content information of the sound source. The results are shown in Figure 2. Figure 6 As shown in (d), the darker the color of the square, the higher the reconstruction strength at the corresponding sound source position. The correct recognition of the audio signal is defined as the sound source position and content corresponding to the highest reconstruction strength are consistent with the actual position and content. The recognition rate is defined as the ratio of the number of recognized audio to the total number of audio. Figure 6 (d) It can be seen that the success rate of dual sound source identification using the complex sound pressure information obtained by sensor 1 is 95%, and the success rate of dual sound source identification using the complex sound pressure information obtained by sensor 2 is also 95%. The recognition success rate of the dual sensor joint compressed sensing reconstruction method is higher than that of using only one sensor, which is 100%. The difference between the dual sound source identification results of the dual sensor joint observation compressed sensing imaging method and the single pixel compressed sensing imaging method is shown in the figure below. Figure 6 As shown in (e). As can be seen from the figure, whether relative to sensor 1 or sensor 2, the dual-sensor combined compressed sensing imaging method has a higher reconstruction strength. The results of single sound source localization and dual sound source identification prove the effectiveness of the combined compressed sensing imaging method and confirm that the method has the characteristics of high accuracy. 4. High-resolution multi-sound source localization and identification At present, the maximum resolution of devices using superstructures for compressed sensing acoustic imaging is only 10°, and the recognition success rate drops sharply as the number of sound sources increases, which limits the development of superstructure acoustic imaging devices. The introduction of the joint compressed sensing imaging method expands the measurement matrix in the compressed sensing process, making high-resolution and high-success-rate sound source localization and identification possible. Experiments using superstructures for high-resolution multi-sound source localization and identification, such as Figure 7 shown. Figure 7 (a) is a test image of the 5° resolution multi-sound source localization and identification experiment. 20 waveguides are arranged at 5° intervals outside the superstructure. The dual sensor combined with the compressed sensing imaging method is used to test the multi-sound source identification performance at 5° resolution. The single sound source localization results are shown in Figure 2. Figure 7 As shown in (b), it can be seen from the figure that the estimated position of a single sound source is consistent with the actual position. The dual sound source identification results are shown in Figure 7 As shown in (c), it can be seen from the figure that the success rate of dual sound source identification is 97%. The above results show that the imaging device based on the composite bionic superstructure can achieve single sound source localization and dual sound source identification with higher resolution. In addition, the dual sensor combined with compressed sensing method also makes it possible to identify more sound source targets at the same time with a high success rate. The three-source and four-source identification experiments were carried out using the superstructure. Figure 7 As shown in (d), it can be seen from the figure that the imaging device can identify the positions and contents of three sound sources, and the recognition success rate is 95%. Figure 8 The four sound source identification results are shown in Fig. 9 As shown in the figure, it can be seen that the imaging device can identify the location and content of four sound sources, with an identification success rate of 80%. The above results show that the imaging device based on the composite bionic superstructure can identify the location of the sound source with a resolution of 5° and simultaneously separate the audio content from multiple mixed signals. In general, this method can be used to improve the resolution and recognition rate of the acoustic imaging device, increase the number of sound source identifications, and has broad application prospects. (V) Determination of distance between multiple sound sources The single-pixel acoustic imaging method can only determine the direction of the sound source, but cannot determine the distance of the sound source. Although the binaural bionic coupling method can determine the distance of the sound source, it has the defect of cone fuzziness, making it difficult to locate the sound source at the midpoint position, and there is a complex coupling form between the two ears. The bionic dual-sensor combined compressed sensing imaging device proposed in this paper comprehensively utilizes related methods and theories such as composite superstructure, compressed sensing and dual-sensor combination to overcome the above defects. The bionic dual-sensor combined acoustic imaging device was used to conduct experiments to determine the distance of the sound source, such as Fig.10 The experimental test schematic diagram is shown in Fig.10 As shown in (a), after the sound wave enters the structure from free space, it is divided into two different parts, which are modulated by the inner and outer parts of the composite bionic superstructure respectively, and then received by the two sensors inside the structure. The two sensors are used to locate the sound source and then determine the distance of the sound source. The principle is as follows Fig.10 (b) As shown. The sound waves received by the two sensors are processed to determine the geometric orientation θ1 and θ2 of the sound source relative to sensor 1 and sensor 2. Two relative orientation lines are drawn using the relative position information of the sound source. The intersection of the two orientation lines is the actual position of the sound source. According to the exterior angle theorem of a triangle, an exterior angle of a triangle is equal to the sum of the two non-adjacent interior angles. Therefore, the angle θ3 between the sound source and the two sensors is θ1-θ2. According to the sine theorem of a triangle: Among them, l1 is the distance of the sound source from sensor 1, l2 is the distance of the sound source from sensor 2, d = 92 mm is the distance of the line between the two sensors, and θ1′ = π-θ1 is the complementary angle of the angle between the sound source and sensor 1. According to formula (3), the distances of the sound source from sensor 1 and sensor 2 can be calculated as l1 = (sinθ2 / sinθ3)*d and l2 = (sinθ1 / sinθ3)*d. Fig.10 (c) is the result of the position and distance of a single sound source relative to sensor 1 and sensor 2 determined by the composite bionic superstructure acoustic imaging device. As can be seen from the figure, both sensor 1 and sensor 2 can accurately locate the position of the single sound source and ultimately determine the distance of the single sound source. It is worth noting that it can also be seen from the figure that even if the distance of the sound source near the midpoint relative to the two sensors is the same, the distance of the sound source at that position can still be accurately determined. Fig.10 (d) is a summary result of the success rate of multi-sound source position identification and distance determination. As can be seen from the figure, the designed imaging device can simultaneously locate the positions of multiple sound sources and determine the distances of multiple sound sources. Among them, the success rate of single sound source distance determination is 100%, the success rate of dual sound source distance determination is 89%, the success rate of three sound source distance determination is 85%, and the success rate of four sound source distance determination is 49%. As the number of sound sources increases, the success rate of distance determination gradually decreases. Nevertheless, when the number of sound sources does not exceed 4, the average recognition rate is as high as 80%, which proves that the imaging device can successfully determine the distances of multiple sound sources while accurately identifying the positions of the sound sources. However, although the designed acoustic imaging device solves the problem of "conical ambiguity" of the median, it is difficult to determine the distance of the sound source at the position where the two sensors are connected because the two azimuth lines are parallel to each other and have no intersection. To address this problem, this paper draws inspiration from the behavior of owls turning their ears in the dark to catch prey and proposes a relative rotation method, such as Fig.11 As shown. When the target sound source is located at the position of the dual sensor connection, the acoustic imaging device is rotated by an angle, and the sound source will no longer be at the position of the dual sensor connection, so that the distance can be determined. In this process, the present invention cleverly utilizes the idea of ​​a relative rotation coordinate system. The present invention assumes that the acoustic imaging device is fixed, and the sound source is rotated by a certain angle in the opposite direction. On the one hand, since the omnidirectional sound propagation data of the sound source relative to the superstructure has been obtained, the azimuth of the sound source relative to the structure after rotation can be obtained according to the previous positioning data. On the other hand, since the sound source is no longer at the position of the dual sensor connection at this time, the two sound source azimuth determinations of the two sensors can be used to determine the distance of the sound source at the connection position. Therefore, this method can successfully solve the problem of connection positioning ambiguity. The above results show that the designed imaging device can identify the position of omnidirectional multi-sound sources and simultaneously measure the distance of omnidirectional multi-sound sources, which has great practical application value. (VI) Free space far-field sound source identification and distance determination Considering that the practical use value of the sound source identification and distance positioning performance of the acoustic imaging device under near-field waveguide conditions is limited, and the positioning performance is easily affected by the disturbance of the sound source position, research on far-field sound source identification and distance positioning in free space is carried out, such as Fig.12 As shown. The far field distance calculation formula is: D=2*R 2 / λ min (4) Where D is the far field distance, R is the array aperture, and λ min is the minimum sound wave length. Set R = 92mm, λ min =48mm is substituted into formula (4), and D = 348mm is calculated. Therefore, in order to test the far-field experimental effect, the distance from the sound source to the center of the array should be greater than 348mm. Experimental test system such as Fig.12 As shown in (a), an experimental test system is built in a superstructure anechoic chamber to test the performance of the superstructure acoustic imaging device. Eight sound sources are distributed at different locations around the superstructure to simulate the sound sources coming from all directions in the actual environment. The multi-sound source identification results are shown in Fig.12 As shown in (b), it can be seen from the figure that the dual-sensor combined compressed sensing method proposed in this paper can improve the recognition success rate of single sound source, dual sound source, triple sound source and quad sound source compared with the single pixel imaging method. Among them, single sound source, dual sound source and triple sound source can be fully recognized, and the recognition success rate of quad sound source is as high as 84%. The detailed results of single sound source azimuth recognition and distance positioning are shown in Fig.12 As shown in (c), it can be seen from the figure that both sensor 1 and sensor 2 can accurately determine the direction of the sound source and ultimately determine the distance of the sound source. The results of determining the distance of multiple sound sources are shown in Fig.12 As shown in (d), it can be seen from the figure that the designed imaging device can completely determine the distance of single sound source and dual sound source. In addition, as the number of sound sources increases, the distance identification success rate gradually decreases, but when the number of sound sources is less than 4, the average success rate of sound source identification is still 85.5%. The above results show that the designed imaging device can identify the positions of multiple sound sources in the far field of free space and determine the distance of multiple sound sources at the same time. In order to further optimize the free-space far-field sound source identification and distance localization performance of the acoustic imaging device, the superstructure height is considered to be increased from 10 mm to 200 mm, and its performance is tested in a free-field environment that meets the requirements of the actual application environment. The experimental results are shown in Figure 13. The experimental test system is as follows Fig.13 (a) shown. Fig.13(b) is a comparison chart of the multi-source azimuth identification performance of acoustic imaging devices at different heights. It can be seen from the figure that compared with the acoustic imaging device at a height of 10mm, the acoustic imaging device at a height of 200mm has improved the success rate of multi-source identification. Among them, single sound source, dual sound source and triple sound source can be fully identified, and the success rate of four sound source identification is 90.6%. Fig.13 (c) is a comparison chart of the multi-sound source distance determination performance of acoustic imaging devices at different heights. As can be seen from the figure, the acoustic imaging device with a height of 200mm improves the success rate of multi-sound source distance determination. Among them, the distances of single and dual sound sources can be completely determined, the distance determination success rate of three sound sources is 87.8%, the distance determination success rate of four sound sources is 76.3%, and the average distance determination success rate is as high as 91%. The above results show that the designed imaging device can not only accurately identify the far-field sound source position content and determine the far-field sound source distance, but also the sound source identification and distance positioning performance can be optimized by increasing the superstructure height, which has broad application prospects. According to the above data, it can be seen that the technical effects that the present invention can achieve are as follows: 1. The device can realize multi-source acoustic imaging in a gas medium represented by air, or realize multi-source acoustic imaging in a liquid medium represented by water; 2. The acoustic unit structure is simple and regular, the overall structure is unique and beautiful, and there are many available design factors. By changing the unit channel length and width, the shape and volume of the cavity, the arrangement of the unit and the opening direction, anisotropy can be adjusted to customize the acoustic imaging performance; 3. When the sound waves emitted by the sound source enter the superstructure from the outside, the sound waves in different directions pass through different channels and chambers, and finally the phases and amplitudes of the sound signals generated at the two detection points are different. That is, the direction of the sound source is determined by the anisotropy presented by the artificial structure. 4. The size and arrangement design of the unit structure reduces the size of the space, improves space utilization, and can be widely used in various narrow spaces; 5. By considering the material, wall thickness and overall height of the structure, the accuracy and range of acoustic positioning can be improved while ensuring sufficient impact resistance, so that it can be flexibly applied to different scenarios; 6. The use of acoustic superstructures reduces the need for the number of sensors and uses physical structures to capture evanescent waves more stably and reliably. 7. The acoustic unit can resonate and anti-resonate with the evanescent wave carrying the sound source information to enhance the sound waves in the special frequency band, reduce the attenuation of the wave, suppress the transmission loss, and protect the integrity of the sound source information; 8. The signal processing method is simplified through the compressed sensing theory, which gets rid of the dependence on the time difference and signal difference between the two sensors, reduces the requirements of acoustic imaging on the sensor, overcomes the defects of the spatial sampling theorem and the cone ambiguity of the midpoint position, and reduces the sampling rate; 9. Through the fusion of multiple sound source information and compressed sensing, high resolution and high recognition rate of sound sources are achieved, the number of identifiable sound sources is increased, and the compressed sensing imaging performance is enhanced; 10. The dual judgment of dual sensors improves the precision and accuracy of imaging, increases the number of identifiable sound sources, and realizes the measurement of sound source distance based on the perception of sound source orientation; In summary, according to the above characteristics of a composite superstructure compressed sensing acoustic imaging device provided by the present invention, the acoustic imaging device provided by the present invention can realize single sound source positioning and ranging in air or underwater three-dimensional space. The structure of this device is simple and regular, unique and beautiful, and has a variety of designable factors such as: unit channel length and width, cavity shape and volume, number and shape of units, unit arrangement and opening direction, structural wall thickness and overall height, sensor position. By designing the nautilus superstructure and the curled folded channel superstructure, it is possible to achieve customized regulation of the incident sound wave, reduce the structural size and system complexity, and adjust the structural strength and perception range to adapt to different working environments. At the same time, the structural features of the two different fish-shaped nautilus parts make the imaging processes of the two sensors independent of each other, so that the sound source distance can be determined by two independent sound source orientation identifications. The use of the acoustic superstructure realizes sound wave capture through the physical structure, which is more stable and reliable. The resonance between the structural unit and the sound wave enhances the sound wave in the special frequency band, weakens the underwater transmission loss, and protects the information integrity. The use of multi-information fusion compressed sensing method and dual-sensor joint compressed sensing method reduces the computational complexity, improves the reconstruction accuracy, overcomes the defects of spatial sampling theorem and cone ambiguity of the perpendicular position, reduces the sampling rate, and realizes the resolution, positioning and tracking of multiple underwater sound sources. In summary, compared with the existing sensor array imaging system, this acoustic imaging device has a smaller size and more stable structure, reduces the number of sensors, and the imaging performance can be flexibly adjusted according to the actual working conditions. Compared with the existing compressed sensing super-structure imaging device, it integrates multiple information, improves accuracy, increases the number of sound sources that can be distinguished at the same time, and can realize the distance measurement of the sound source. It can be widely used in acoustic imaging of various devices and environments. The above shows and describes the basic principles and main features of the present invention and the advantages of the present invention. It is obvious to those skilled in the art that the present invention is not limited to the details of the above exemplary embodiments, and the present invention can be implemented in other specific forms without departing from the spirit or basic features of the present invention. Therefore, no matter from which point of view, the embodiments should be regarded as exemplary and non-restrictive. The scope of the present invention is defined by the attached claims rather than the above description, and it is intended that all changes falling within the meaning and scope of the equivalent elements of the claims are included in the present invention. Any figure mark in the claims should not be regarded as limiting the claims involved. In addition, it should be understood that although this specification is described in accordance with the implementation modes, not every implementation mode contains only one independent technical solution. This description of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment can also be appropriately combined to form other implementation modes that can be understood by those skilled in the art. The above content is only to illustrate the technical idea of ​​the present invention, and cannot be used to limit the protection scope of the present invention. Any changes made on the basis of the technical solution according to the technical idea proposed by the present invention shall fall within the protection scope of the claims of the present invention.

Claims

1. A composite superstructure compressed sensing acoustic imaging device, characterized in that: The invention comprises a disc-shaped superstructure (2) formed by combining two fish-shaped bionic nautilus superstructures (1), a plurality of internal cavities (3) distributed on the fish-shaped bionic nautilus superstructure (1), an octagonal superstructure (5) composed of eight folded channel units (4), and orthogonal sound channels (6) distributed on the folded channel units (4).

2. The composite superstructure compressed sensing acoustic imaging device according to claim 1, characterized in that: Fish-shaped bionic nautilus superstructure (1) The outer spiral line satisfies the equation: t =rcos(t h ), y t =rsin(t h ), z t = 0, where t h =360nt, a is a parameter that controls the degree of divergence of the spiral, n is the number of turns of the spiral, and t is an independent variable ranging from 0 to 1.

3. The composite superstructure compressed sensing acoustic imaging device according to claim 1, characterized in that: The device can realize acoustic imaging in a gas medium represented by air, or in a liquid medium represented by water. When working in different media, the cavity (3) is filled with corresponding working media.

4. The composite superstructure compressed sensing acoustic imaging device according to claim 1, characterized in that: When the sound waves emitted by the sound source enter the disk-shaped superstructure (2) from the outside, the sound waves in different directions pass through different orthogonal sound channels (6) and cavities (3), and finally generate different phases and amplitudes of the sound signals at the two detection points, that is, the direction of the sound source is determined by the anisotropy presented by the artificial structure.

5. The composite superstructure compressed sensing acoustic imaging device according to claim 1, characterized in that: The frequency response can be directly adjusted by adjusting the geometric parameters of the acoustic unit, including the length of the orthogonal acoustic channel (6) and the volume of the cavity (3).

6. A composite superstructure compressed sensing acoustic imaging method, characterized in that: The method is based on a superstructure compressed sensing acoustic imaging device according to claim 1, comprising: 1) determining the geometric parameters of the acoustic unit, including the length of the orthogonal acoustic channel (6) and the shape and volume of the cavity (3), according to the target sound source detection frequency band; 2) According to the target sound information and expected position, the detection direction is divided into several grid points, and the target sound source is predicted and detected at each grid point. According to the complex sound signals measured by the internal sensors of the two fish-shaped bionic nautilus superstructures (1), a compressed sensing observation matrix is ​​jointly constructed to complete the system testing and training; 3) In actual measurement, based on the complex sound signals measured by the two sensors and combined with the compressed sensing observation matrix, the azimuth relationship between the sound source and the two sensors is reversely deduced, and the direction and relative distance of the sound source are obtained based on the distance between the two sensors and the two azimuths.

7. The composite superstructure compressed sensing acoustic imaging method according to claim 6, characterized in that: Fish-shaped bionic nautilus superstructure (1) The outer spiral line satisfies the equation: t =rcos(t h ), y t =rsin(t h ), z t = 0, where t h =360nt, a is a parameter that controls the degree of divergence of the spiral, n is the number of turns of the spiral, and t is an independent variable ranging from 0 to 1.

8. The composite superstructure compressed sensing acoustic imaging method according to claim 6, characterized in that: The method can realize acoustic imaging in a gas medium represented by air, or in a liquid medium represented by water. When working in different media, the cavity (3) is filled with corresponding working media.

9. The composite superstructure compressed sensing acoustic imaging method according to claim 6, characterized in that: When the sound waves emitted by the sound source enter the disk-shaped superstructure (2) from the outside, the sound waves in different directions pass through different orthogonal sound channels (6) and cavities (3), and finally generate different phases and amplitudes of the sound signals at the two detection points, that is, the direction of the sound source is determined by the anisotropy presented by the artificial structure.

10. The composite superstructure compressed sensing acoustic imaging method according to claim 6, characterized in that: The frequency response can be directly adjusted by adjusting the geometric parameters of the acoustic unit, including the length of the orthogonal acoustic channel (6) and the volume of the cavity (3).

Images