Hollow drum structure, low-frequency sound signal detection device and method based on the structure

Abstract

The present invention relates to the fields of fiber optic sensing and deep learning, and particularly to an ethereal drum structure, a low-frequency acoustic signal detection device and method based on this structure. By combining the advantages of an F-P cavity formed by two fiber Bragg gratings, tapered optical fiber, and the ethereal drum structure, the simple structure, high sensitivity, and easy multiplexing of the optical F-P cavity are fully utilized. Considering the characteristics that the fiber Bragg grating only reflects specific wavelengths and the natural high sensitivity of the tapered optical fiber, and combining the narrowband filtering advantage of the circular diaphragm, circular diaphragms of different sizes are prepared, with corresponding resonant frequencies being different from each other, enabling the ethereal drum structure to detect sound waves of different frequencies. Without the need for additional Fourier transform and filtering operations, the frequency of the external sound wave can be directly obtained, realizing the intelligent and autonomous recognition of the target sound signal. It is more sensitive, compact, and efficient than traditional low-frequency acoustic signal sensing systems, and at the same time has advantages such as narrowband filtering, real-time spectrum extraction, and autonomous target recognition.

Description

Technical Field

[0001] The present invention relates to the fields of fiber optic sensing and deep learning, and particularly relates to an ethereal drum structure, a low-frequency acoustic signal detection device and method based on this structure. The device uses the ethereal drum structure as a basic element, is equipped with fiber Bragg gratings and circular diaphragms, and combines neural network algorithms, and is particularly suitable for the early warning detection and intelligent identification of low-frequency acoustic signals generated by low-altitude unmanned aerial vehicles and natural disasters. Background Art

[0002] With the development of unmanned aerial vehicle technology and the frequent occurrence of natural disasters, how to detect the presence of low-altitude unmanned aerial vehicles or the occurrence of natural disasters is an urgent problem to be solved at present. Traditional low-frequency detection technologies face many difficulties, so it is of great significance to develop an acoustic signal detection device that can be applied to the low-frequency band and has extremely high sensitivity and intelligent identification capabilities.

[0003] Compared with traditional electrical sensing technologies, fiber optic sensing has advantages such as anti-electromagnetic interference, corrosion resistance, small size, and light weight, and has gradually been widely used in the field of low-frequency acoustic signal detection. Currently, the main sensing technology used for detecting low frequencies is Distributed Acoustic Sensing (DAS). Representative achievements in the literature reports on DAS technology include: Chinese invention patent application "A Method and System for Real-time Detection and Positioning of Low-altitude Unmanned Aerial Vehicles" (application number CN202310503374.4, publication date 2023-10-31), "Distributed Fiber Optic Sensing Seismograph" proposed by Ran Zengling, Rao Yunjiang, Wang Ximing, etc. Distributed Fiber Optic Sensing Seismograph Distributed Fiber Optic Sensing Seismograph and Its Applications [J], Geophysical Prospecting for Petroleum, 2022, 61(1): 41-49), "DAS Distributed Fiber Optic Acoustic Sensing Geophone" proposed by Yu Yongshuang (Design of DAS Distributed Fiber Optic Acoustic Sensing Geophone [D], Jilin: Changchun, Changchun University of Science and Technology, 2022).

[0004] The above-mentioned existing technologies use the DAS technology path based on fiber optic sensing to achieve the detection of low-frequency acoustic signals, which can improve the acoustic wave detection efficiency to a certain extent, but there is still room for further improvement in terms of sensitivity improvement, demodulation algorithm optimization, preparation cost reduction, large-scale array multiplexing, and target autonomous recognition. Summary of the Invention

[0005] The present invention aims to solve the defects of the existing technology and proposes an ethereal drum structure, a low-frequency acoustic signal detection device and method based on this structure, so as to achieve acoustic target detection with a compact structure, extremely high sensitivity, and target autonomous recognition, and is particularly suitable for the detection and intelligent identification fields of low-frequency acoustic signals generated by low-altitude unmanned aerial vehicles and natural disasters.

[0006] The technical solution adopted by the present invention is as follows:

[0007] An ethereal drum structure, the ethereal drum structure is a flat drum-shaped cavity with a hollow interior, and the surface is composed of a plurality of tongues 601 with different sizes. The tongues 601 and the drum-shaped cavity form a resonance cavity to achieve acoustic and structural sensitization. Each tongue 601 is hollowed out at the center and embedded with a circular diaphragm 604. The tapered optical fiber 603 is combined with the circular diaphragm 604 by means of gluing. Fiber Bragg gratings 602 are distributed on both sides of the tapered optical fiber 603. The circular diaphragm 604 is used to pick up the incident acoustic wave signal, convert the intensity of the incident acoustic wave signal into the vibration amplitude of the diaphragm, and then change the length of the tapered optical fiber 603 on the diaphragm. The tapered optical fiber 603 is used to convert the vibration amplitude of the diaphragm into a phase change of the optical signal, and then change the intensity of the output interference light; there are N tongues in total, N is a positive integer, and the sizes of the N tongues 601 are different from each other, and the sizes of the corresponding N circular diaphragms 604 are also different from each other. The two fiber Bragg gratings 602 on both sides of the tapered optical fiber 603 have the same properties and form an F-P cavity. The reflection spectra of the fiber Bragg gratings 602 corresponding to different F-P cavities are different from each other, so that the reflection center wavelengths of each F-P cavity are different.

[0008] Preferably, the circular diaphragm 604 adopts non-linear amplification based on the resonant working mode. When the acoustic wave frequency is approximately equal to the resonant frequency of a certain diaphragm and acts on the ethereal drum structure, the vibration of this diaphragm is more obvious than that of other diaphragms, which will lead to a significant change in the length of its corresponding F-P cavity. Using this property can effectively obtain the acoustic wave frequency and then achieve high sound pressure sensitivity.

[0009] Preferably, the circular diaphragm 604 adopts narrowband filtering based on the control of the resonant working mode, that is, the resonant frequencies corresponding to the diaphragms with different areas in the ethereal drum structure are not the same, so that each diaphragm will only have a significant vibration effect when it is affected by the corresponding resonant frequency. This method realizes the recognition of the acoustic wave spectrum characteristics;

[0010] Preferably, the sensing structure formed by the combination of the tongues 601, the circular diaphragms 604 and the F-P cavities formed by the adjacent fiber Bragg gratings 602 and the tapered optical fiber 603, the number N of which can be prepared according to requirements, realizes the resonant absorption of multi-frequency acoustic waves.

[0011] Preferably, the F-P cavity formed by the two fiber Bragg gratings 602 and the tapered optical fiber 603 can be more affected by the circular diaphragm while reflecting the optical signal of a specific Bragg wavelength, further improving the sound pressure sensitivity;

[0012] The present invention also provides a low-frequency acoustic signal detection device based on the above-mentioned kongling drum structure, which includes a broadband laser 1, a first 1×2 coupler 21, a delay optical fiber 3, a second 1×2 coupler 22, a piezoelectric ceramic 4, a circulator 5, a kongling drum structure 6, an optical isolator 7, a demultiplexer group 8, a photodetector group 9, a data acquisition card 10, and a host computer 11;

[0013] The laser emitted from the broadband laser 1 first passes through the first 1×2 coupler 21 and is divided into two optical signals with equal intensities. One of the optical signals passes through the delay optical fiber 3, resulting in a phase difference between this optical signal and the other optical signal. The other optical signal passes through the piezoelectric ceramic 4. The piezoelectric ceramic 4 is controlled by the host computer 11 to keep the phase difference between the two optical signals stable. Then, the two optical signals are emitted through the second 1×2 coupler 22 and successively enter one port of the circulator 5 and are emitted from the second port of the circulator 5. Next, the two optical signals successively enter the kongling drum structure 6 and successively pass through N F-P cavities composed of tapered optical fibers and fiber Bragg gratings distributed on both sides of the tapered optical fiber in the kongling drum structure 6. There are two reflected optical beams inside each F-P cavity. One is reflected by the first fiber Bragg grating in the F-P cavity, and the other is reflected by the second fiber Bragg grating after passing through the first grating. Since there are two optical signals entering the kongling drum structure 6 successively, there are a total of four reflected optical beams. According to the matching interference theory, since the length of the delay optical fiber 3 is equal to the length of the F-P cavity (i.e., the sum of the grating region of a single fiber Bragg grating and the length of the tapered optical fiber), that is, the difference between the phase difference generated at the delay optical fiber 3 and the phase difference generated at the F-P cavity satisfies the coherence condition of light. Therefore, interference optical signals will finally be generated at each F-P cavity. The interference optical signals generated by each F-P cavity correspond to different wavelengths. Subsequently, all the interference optical signals return to the second port of the circulator 5 and are emitted from the third port. The remaining optical signals that are not reflected are emitted to the optical isolator 7 after passing through the Nth F-P cavity to prevent them from reflecting back to the original optical path; then, the interference optical signals emitted from different F-P cavities respectively enter the demultiplexer group 8 composed of N demultiplexers. The demultiplexer group 8 separates the interference optical signals with different wavelengths from the interference optical signals with other wavelengths. Then, the interference optical signals are converted into electrical signals by the photodetector group 9 composed of N photodetectors and enter the data acquisition card 10 to obtain the changes in the interference optical signals of each group of F-P cavities. Finally, the obtained data is transmitted to the host computer 11, and the PGC algorithm (Li Heping, Research on Fiber Optic Vector Hydrophone Based on FBG-FP Cavity [D], Sichuan: Chengdu, University of Electronic Science and Technology of China, 2017) is used to demodulate the data. The demodulated target signal data is sent into the trained neural network model, and the type of the external sound source can be successfully identified.

[0014] The present invention also provides a low-frequency acoustic signal detection method based on the above-mentioned low-frequency acoustic signal detection device. The method is divided into the following steps:

[0015] S1. Modulation and transmission of optical signals between two 1×2 couplers:

[0016] S1.1 The broadband laser 1 emits an initial optical signal as a light source and then enters the first 1×2 coupler 21 with a splitting ratio of 50:50, where it is divided into two optical signals: One optical signal passes through the delay fiber 3 to generate a phase difference with the other optical signal. This phase difference is used to alleviate the situation where interference cannot occur due to a large difference between the cavity length of the subsequent F-P cavity and the coherence length of light, and thus there is a sequential order when the two optical signals enter the F-P cavity. A piezoelectric ceramic 4 is added to the second optical signal to maintain the stability of this phase difference, and the piezoelectric ceramic 4 is controlled by the host computer 11;

[0017] S1.2 The two optical signals enter the second 1×2 coupler 22 with the same performance as the first 1×2 coupler 21 in sequence, then enter from one port of the circulator 5 and are output from the second port, and then enter the ethereal drum structure 6 in sequence; There are N series-connected F-P cavities composed of two fiber Bragg gratings and one tapered fiber in the ethereal drum structure 6, where N is a positive integer, and the central wavelengths of the optical signals that each F-P cavity can reflect are different. These F-P cavities are respectively close to N circular diaphragms with different areas. The vibration of the circular diaphragms is used to cause the cavity length of the F-P cavity to change. When the sound wave is at the resonance frequency of a certain circular diaphragm, the diaphragm will generate significant deformation, which will drive the cavity length of the F-P cavity to change significantly, and then cause the optical phase transmitted in the fiber Bragg grating to change simultaneously, affecting the intensity of the final interference light; In addition, the inside of the F-P cavity is composed of a tapered fiber, and the tapered fiber has a higher pressure sensitivity than ordinary fibers and is more affected by the action of the circular diaphragm, so the change in cavity length is more obvious.

[0018] S2. Propagation and matching interference of optical signals in the F-P cavity composed of two fiber Bragg gratings and one tapered fiber:

[0019] After the two optical signals pass through N FP cavities in sequence, each time the optical signals pass through an FP cavity, the optical signals that meet the Bragg wavelength condition of the fiber Bragg grating will be reflected inside the corresponding FP cavity, and two types of reflected light will be generated at this time: one is the optical signal that is directly reflected after passing through the first fiber Bragg grating in the FP cavity, and the other is the optical signal that passes through the first fiber Bragg grating and is reflected by the second fiber Bragg grating and then re-transmitted back to the first fiber Bragg grating to return to the original path. Since the distance between the two fiber Bragg gratings is relatively far (usually several decimeters), there is a large phase difference between the two types of reflected light signals; since there are two optical signals that enter the hollow drum structure one after another, there are a total of four types of reflected light between the two optical signals: the first is the reflected light that is reflected by the first fiber Bragg grating in the FP cavity after passing through the delayed optical fiber; the second is the reflected light that is reflected by the FP cavity after passing through the delayed optical fiber. The first type is the reflected light reflected by the second fiber Bragg grating in the FP cavity; the third type is the reflected light reflected by the first fiber Bragg grating in the FP cavity without passing through the delayed fiber; the fourth type is the reflected light reflected by the second fiber Bragg grating in the FP cavity without passing through the delayed fiber; according to the matched interference theory (Lin Huizu, Research on Key Technologies of Fiber Bragg Grating Hydrophone Array Based on Matched Interference [D], Hunan: Changsha, National University of Defense Technology, 2013), there is only one situation in which these four reflected lights will interfere with each other, that is, the second reflected light and the third reflected light interfere with each other, and the remaining optical signal that does not interfere but is reflected back by the fiber Bragg grating is retained in the output signal in the form of a direct current. Finally, N interference lights of different wavelengths are generated in the N FP cavities and return to the two ports of the circulator. The remaining optical signal that is not reflected is emitted to the optical isolator 7 after passing through the Nth FP cavity to prevent it from being reflected back to the original optical path;

[0020] S3. Detection of interference light signals and target recognition:

[0021] S3.1 Detection:

[0022] N interference light signals of different wavelengths are output from the third port after returning to the second port of the circulator. Considering that the wavelengths of the N interference lights are inconsistent, a de-wavelength division multiplexer group 8 composed of N de-wavelength division multiplexers of different wavelengths is set to separate the wavelengths of the interference light signals. Then, the N optical signals are respectively converted into electrical signals by a photoelectric detector group 9 composed of N photoelectric detectors, and then the signals are uniformly collected by an acquisition card 10. The acquired data can be demodulated by PGC to obtain the target acoustic wave signal.

[0023] S3.2 Identification

[0024] Inputting the obtained target acoustic wave signal into the trained neural network model can identify the type of external sound source. Since the areas of each circular diaphragm are different from each other, and the corresponding resonance frequencies are also different, when the resonance frequency of a certain diaphragm is consistent with the external acoustic wave frequency, its vibration amplitude will be more significant than that of other diaphragms, and the corresponding change in the cavity length of the F-P cavity is larger, which will greatly affect the output intensity of the interference light on this F-P cavity. Therefore, the acoustic wave frequency can be used as the actual label to classify these data results, so as to realize the early warning detection and intelligent identification of low-altitude UAVs and natural disasters;

[0025] Inputting the demodulated target signal data into the trained neural network model can be used to identify external sound sources. According to the existing neural network models, common models include BP neural network, convolutional neural network, generative adversarial network, recurrent neural network, deep belief network, deep residual network, long short-term memory network, etc. Among them, the long short-term memory network (abbreviated as LSTM) was first proposed by Hochreiter and Schmidhuber in 1997 (Long short-term memory[J], Neural Computation, 1997, 9(8): 1735-1780). This neural network model has the characteristics of being able to process long-term dependence relationships, having dynamic memory and forgetting capabilities, strong generalization capabilities, and wide applicability compared with other models. Therefore, the long short-term memory network is selected as the main body to construct the neural network model. The following are the steps to construct and train the LSTM neural network model:

[0026] S3.2.1 After using the acoustic feature data of known sound sources obtained in actual tests as the driving signal of the transducer, place the transducer near the structure of the ethereal drum, so that the circular diaphragm in the ethereal drum structure vibrates, thereby changing the cavity length of the F-P cavity. Then collect and demodulate the acoustic signal data generated by these known sound sources. The obtained demodulated data is divided into a training set and a validation set according to a quantity ratio of 7:3: Among them, the training set is used to train and initially establish the neural network model, and the validation set is used to evaluate the performance of the neural network model obtained in each training. Classify each demodulated data in the training set and the validation set according to the acoustic wave frequency it represents. Each category is used as the true label, that is, as the basis for judging the prediction effect of the model during subsequent training, validation, and testing. A total of N categories are set, representing the resonance frequencies of N diaphragms with different areas respectively;

[0027] S3.2.2 After receiving the training set data at the input layer of the neural network model, randomly initialize the weights and biases of the neurons in the input layer. At the same time, set the number of iteration cycles to C times, the learning rate to S, and the regularization parameter to σ. Among them, the learning rate S is the amplitude for updating the model parameters (such as weights or biases) in each iteration, and the regularization parameter σ is used to avoid overfitting of the model. Usually, set the number of iteration cycles C to 600 times, the learning rate S to 0.001, and the regularization parameter σ to 0.01. According to what is described in S1, there are N interference optical signals transmitted and returned, and a total of N neurons are set in the input layer, and each neuron corresponds to an interference optical signal at a wavelength.

[0028] S3.2.3 The training set data propagates forward in the network and enters the hidden layer. A total of M hidden layers are set, where M is a positive integer. For each layer passed through, weighted summation and activation function processing are required to enable the model to handle complex training tasks. Among them, the commonly used activation functions mainly include Sigmoid, ReLU, Softmax, etc. Compared with other types of activation functions, Softmax is more proficient in multi-object classification and recognition. Therefore, Softmax is preferably used as the activation function of the neural network model.

[0029] S3.2.4 When the training set data enters the output layer through the hidden layer, preferably use the Hinge loss function as the loss function to calculate the loss between the predicted output and the true label, and use the gradient information of this function combined with the backpropagation algorithm (Beyond regression: New tools for prediction and anal-ysis in the behavioralsciences[D], 1974, Harvard University) to calculate the gradients corresponding to the weights (i.e., the proportion coefficients of each output value in each layer of the model) and biases (i.e., a constant term added to the output of each layer of the model) of the model. According to the gradient feedback, use the Adam gradient descent algorithm (Adam: A Method for Stochastic Optimization[C], International Conference on Learning Representations, 2015, USA) to update the weights and biases in the model in real time.

[0030] S3.2.5 Repeat the above method for C iterations until the performance of the neural network model on the training set reaches the convergence condition or the predetermined number of iterations C is reached. During the process of constructing the neural network model, it is necessary to use the validation set to evaluate the performance of the model after each iteration, including accuracy, recall, and F1 score, and optimize the learning rate S and the regularization parameter σ according to the above metrics for the next iteration; after the iteration process is completed, the model with the best performance on the validation set will be selected as the final neural network model;

[0031] S3.2.6 Place the ethereal drum structure in an environment with unknown sound sources, and use the host computer 11 to collect the interference optical signals generated by the influence of the sound sources; after the host computer 11 demodulates these signal data, input them into the neural network model for prediction and classification to achieve intelligent recognition of low-frequency sound targets.

[0032] The present invention innovatively combines the advantages of the F-P cavity composed of double fiber Bragg gratings, tapered optical fibers, and the ethereal drum structure, and proposes a low-frequency sound signal detection device based on the ethereal drum structure, which fully utilizes the advantages of the optical F-P cavity structure such as simple structure, high sensitivity, and easy multiplexing. Considering that the fiber Bragg grating only reflects specific wavelengths and the natural high-sensitivity characteristics of the tapered optical fiber, and combining the narrowband filtering advantage of the circular diaphragm, circular diaphragms with different area sizes are prepared, and the corresponding resonant frequencies are different from each other, so that the ethereal drum structure can detect sound waves of different frequencies, and the frequency composition of the external sound waves can be directly obtained without additional Fourier transform and filtering operations, realizing intelligent and autonomous recognition of the target sound signal. This device is more sensitive, compact, and efficient than the traditional low-frequency sound signal sensing system, and has the advantages of narrowband filtering, real-time spectrum extraction, and autonomous target recognition, with broad application prospects.

[0033] A high-sensitivity detection and intelligent recognition device for low-frequency acoustic signals based on a multi-frequency resonance mechanism provided by the present invention has the following beneficial technical effects compared with the prior art: First, an ethereal drum structure is innovatively introduced, and an F-P cavity composed of two fiber Bragg gratings is introduced on the outer surface. The center of the tongue of the ethereal drum is hollowed out and a circular diaphragm is embedded. The sound wave inside the ethereal drum is continuously reflected on the diaphragm to realize high-sensitivity detection of low-frequency disturbance signals and multiplexing arrays under a compact structure. Second, when a single diaphragm is affected by a sound wave with a resonant frequency, it has a non-linear amplification effect on this signal. Therefore, compared with other diaphragms, the vibration amplitude of this diaphragm is more obvious, and extremely high-sensitivity detection can be achieved. Third, by controlling the area of the diaphragm, the optimization design and array combination of the resonant frequency can be realized, and then the spectral characteristics of the incident acoustic wave signal can be extracted in real time, providing a basis for target intelligent recognition. In addition, a tapered fiber is introduced as the F-P cavity, which can further enhance the effect of the outside world on the F-P cavity. Finally, each diaphragm can be simulated as a neuron cell, and the demodulated target acoustic signal is used as the input of the neural network to realize the intelligent recognition of low-frequency acoustic targets, and the detection and early warning of low-altitude drones and natural disasters can be achieved to the greatest extent. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] The present invention will be further described with reference to the accompanying drawings. However, the embodiments in the drawings do not constitute any limitation to the present invention. For those of ordinary skill in the art, other drawings can be obtained according to the following drawings without creative efforts. In the drawings:

[0035] Figure 1 is a schematic assembly structure diagram of a high-sensitivity detection and intelligent recognition device for low-frequency acoustic signals based on a multi-frequency resonance mechanism according to the present invention;

[0036] Figure 2 is a three-dimensional structure diagram of the ethereal drum structure 6 according to the present invention;

[0037] Figure 3 is a three-dimensional structural cross-sectional view of the ethereal drum structure 6 according to the present invention;

[0038] Figure 4 is a three-dimensional structure diagram when the F-P cavity composed of the tapered fiber 603 and the two fiber Bragg gratings 602 on both sides is mounted on the circular diaphragm 604 embedded in the middle of the tongue 601 according to the present invention;

[0039] Figure 5 is a deformation diagram of the circular diaphragm 604 when affected by a sound wave according to the present invention;

[0040] Figure 6 is a schematic diagram of collecting external acoustic wave signals and importing them into a neural network model for real-time target recognition according to the present invention;

[0041] Description of reference numerals:

[0042] 1: broadband laser; 2: 1×2 coupler; 3: time-delay optical fiber; 4: piezoelectric ceramics; 5: circulator; 6: hollow drum structure; 7: optical isolator; 8: demultiplexer group; 9: photodetector group; 10: acquisition card; 11: host computer.

[0043] 601: sound tongue; 602: fiber Bragg grating; 603: tapered optical fiber; 604: circular diaphragm. DETAILED DESCRIPTION

[0044] In order to enable those skilled in the art to better understand the technical solution of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.

[0045] Figure 1 is a structural schematic diagram of an embodiment of a low-frequency sound signal high-sensitivity detection and intelligent identification device based on a multi-frequency resonance mechanism according to the present invention, Figure 2 is a schematic diagram of the three-dimensional structure of the hollow drum structure 6 of the present invention, Figure 3 is a schematic cross-sectional view of the three-dimensional structure of the hollow drum structure 6 of the present invention, Figure 4 6 is a schematic diagram of a three-dimensional structure when the FP cavity composed of the tapered optical fiber 603 and the fiber Bragg gratings 602 on both sides of the present invention is mounted on the central circular diaphragm 604 embedded in the sound tongue 601, Figure 5 is a schematic diagram of the deformation of the circular diaphragm 604 of the present invention when it is acted upon by sound waves, Figure 6 It is a schematic diagram of collecting external sound wave signals and then importing them into a neural network model for real-time target recognition as described in the present invention.

[0046] The present invention discloses a low-frequency sound signal detection and intelligent identification device based on a hollow drum structure, the device structure comprising: a broadband laser 1, a 1×2 coupler 2, a time-delay optical fiber 3, a piezoelectric ceramic 4, a circulator 5, a hollow drum structure 6, an optical isolator 7, a demultiplexer group 8, a photodetector group 9, a collection card 10 and a host computer 11;

[0047] The interior of the ethereal drum structure 6 is a hollow structure, and its surface is composed of sound tongues 601. A circular diaphragm 604 is embedded in the middle of the sound tongue 601. A conical optical fiber 603 is covered above, and fiber Bragg gratings 602 are distributed on both sides of the optical fiber. There are four such structures in total. The sizes of the four sound tongues 601 are different from each other, and the sizes of the corresponding four circular diaphragms 604 are also different from each other. The fiber Bragg gratings 602 on both sides of a single conical optical fiber have the same properties and form an F-P cavity. The refractive index modulation depths and grating periods of the fiber Bragg gratings 602 of each group of F-P cavities are different from each other, and thus the reflection center wavelengths between each F-P cavity are different from each other. In order to achieve the effect of low-frequency acoustic signal sensing, nitrile rubber is preferably used as the raw material when designing the four circular diaphragms. Its Young's modulus is 2.3×10 9 Pa, the Poisson's ratio is 0.49, and the density is 1300 kg / m 3 . The designed radii are 0.05 m, 0.1 m, 0.3 m, and 0.5 m from small to large, and the corresponding thicknesses are all 80 μm. According to formula (4), it can be known that the resonance frequencies of the four diaphragms are 143.3489 Hz, 35.8372 Hz, 3.9819 Hz, and 1.4335 Hz respectively, which are in line with the frequency band of low-frequency acoustic signals and can be used for the detection and identification of this frequency band. When a laser with a power of 25 mW exits from the broadband laser 1 with a bandwidth of 1520 nm to 1580 nm, the light source first passes through a 1×2 coupler 2 and is divided into two optical signals with an intensity ratio of 50:50. One of the paths has a delay optical fiber 3, so that there is a phase difference between the optical signal of this path and the optical signal of the other path The other path has a piezoelectric ceramic 4 that can be controlled by the host computer 11 to make the phase difference between the two paths remain stable. Then the two optical signals enter another 1×2 coupler 2 in sequence and enter one port of the circulator 5. After exiting from the second port of the circulator 5, they enter the ethereal drum structure 6. Then the two optical signals successively pass through four F-P cavities composed of adjacent fiber Bragg gratings 602 and conical optical fibers 603. The center wavelengths that can be reflected between the four F-P cavities are 1535 nm, 1550 nm, 1565 nm, and 1580 nm respectively. The remaining optical signal after passing through the fourth F-P cavity exits the optical fiber and reaches the optical isolator 7. There are two reflected light beams inside each F-P cavity. One is obtained after being reflected by the first grating, and the other is obtained after being reflected by the second grating after passing through the first grating. There will be a phase difference between these two reflected light beams According to the matching interference theory, the phase difference Δφ generated at the delay optical fiber and the phase difference generated at the F-P cavity Satisfying the coherence conditions of light, interference light signals will ultimately be generated at each F-P cavity, and the corresponding wavelengths of the respective interference light signals are different from each other. Subsequently, all the interference light signals return to the second port of the circulator 5, are coupled into a total interference light signal, and are emitted from the third port. Then, the light signals of different wavelengths enter the demultiplexer group 8 corresponding to the respective wavelengths, are separated from the light signals of other wavelengths, and after the optical signals are converted into electrical signals by the photodetector group 9, they enter the acquisition card 10 to obtain the change in the optical signal of each wavelength. Finally, the obtained signal data is imported into the host computer 11 for demodulation. Among them, the normalized interference light intensity change of the F-P cavity formed by a pair of fiber gratings is between 0.4 and 0.7. Finally, the obtained demodulation data is input into the trained long short-term memory network model, and the type of the external target sound signal can be recognized.

[0048] The present invention is based on the following principles:

[0049] A phase difference is generated between the optical signal corresponding to the optical path of the delay fiber and the other optical signal. The process is as follows:

[0050] Assume that the wavelengths of the optical signals in the two optical paths output from the coupler are both λ, the length of the delay fiber is D, and the refractive index of the fiber is n. Then, a stable phase difference Δφ can be generated between the two optical signals, and its magnitude is:

[0051]

[0052] And the coherence length l of light can be expressed as:

[0053]

[0054] Where, λ is the central wavelength of the broadband light source, Δλ is the spectral width of the light source, and the length D of the delay fiber is much greater than the coherence length l. Therefore, the optical signals on the two optical paths have a sequential order when passing through the second coupler and will not interfere.

[0055] When the acoustic wave frequency is at the resonance frequency of a certain circular diaphragm, the circular diaphragm will generate a more significant deformation, thereby driving a change in the cavity length of the F-P cavity. The deformation process of the circular diaphragm is as follows:

[0056] For the circular diaphragm, the deformation amount Δh at its center can be expressed as:

[0057]

[0058] Where, μ is the Poisson's ratio, r is the radius of the diaphragm, t is the thickness of the diaphragm, E is the Young's modulus, P is the sound pressure, f is the acoustic wave frequency, f0 is the resonance frequency, and ξ is the damping coefficient. It can be seen from equation (3) that the closer the acoustic wave frequency is to the resonance frequency, the greater the deformation amount at the center of the diaphragm. Among them, the resonance frequency f0 of the diaphragm is:

[0059]

[0060] Among them, ρ is the density of the diaphragm. It can be seen from equation (3) that by changing the radius of the diaphragm, i.e., the area size, the resonant frequency of the diaphragm can be effectively changed, which helps to identify different external acoustic signals. At the same time, since the tapered optical fiber where the F-P cavity is located is combined with the diaphragm by an adhesive method, the distance between the two can be regarded as zero. At this time, the tapered optical fiber is stressed and strained due to the deformation of the circular diaphragm, and then the cavity length of the F-P cavity changes, thereby changing the intensity of the interference light.

[0061] The optical signal band that satisfies the Bragg wavelength condition of the fiber Bragg grating will generate reflection inside the corresponding F-P cavity, and the process is as follows:

[0062] According to the characteristics of the fiber grating, when a beam of light passes through the grating, if the wavelength of the light satisfies:

[0063] λ m =2nΛ (5)

[0064] Then the beam of light is called the Bragg wavelength and will be reflected by the grating. Among them, n is the refractive index of the optical fiber, Λ is the period of the fiber grating. Considering that the double fiber gratings need to form an F-P cavity structure, the used grating belongs to a weak reflection grating to ensure that the subsequent optical signals can interfere normally.

[0065] When the optical signal passes through each F-P cavity, mainly two types of reflected light are generated, and there are two optical signals, so there are four kinds of reflected light in total. According to the matching interference effect, interference exists among the four kinds of reflected light, and the process is as follows:

[0066] It is known that there is a phase difference of Δφ between the two optical signals, and the length of the delay optical fiber is D. Select one of the F-P cavities as the object of discussion, and divide the incident optical signal into two cases for discussion. The first case is the optical signal that does not pass through the delay optical fiber: Assume that the amplitude of the optical signal is A and the initial phase is δ; the length of the tapered optical fiber is d, and its refractive index is the same as that of the ordinary optical fiber, both are n; the reflectivities of the fiber gratings on both sides of the tapered optical fiber are both R, the transmittivities are both T, and the grating region lengths are both L. Then the obtained reflected light can be expressed respectively as:

[0067] The expression of the reflected light directly reflected on the surface of the first fiber grating:

[0068] E1=ARe iδ (6)

[0069] The expression of the reflected light that passes through the first fiber grating, is reflected on the surface of the second fiber grating, and then passes through the first fiber grating again and returns:

[0070]

[0071] Among them, is the phase difference generated when passing through the F-P cavity, and can be expressed as:

[0072]

[0073] The second case is the optical signal after passing through the delay optical fiber. The initial phase of this optical signal is δ + Δφ. Ignoring the attenuation generated during the propagation of the optical signal, the two reflected lights obtained are respectively expressed as:

[0074] The expression of the reflected light directly reflected on the surface of the first fiber grating is also:

[0075] E3 = AR e i(δ+Δφ) (9)

[0076] The expression of the reflected light that passes through the first fiber grating, is reflected on the surface of the second fiber grating, and then passes through the first fiber grating again and returns is also:

[0077]

[0078] By comparison, it is not difficult to find that there is a possibility of interference only in the phase difference between E2 and E3 among the four reflected lights, that is, the phase difference compensation can be satisfied by adjusting the length of the delay optical fiber Furthermore, the phase difference is made to satisfy the coherence condition, and finally an interference effect is generated. And through subsequent derivation, it can be known that the intensity I of the interference light in the F-P cavity is:

[0079]

[0080] It shows that when the length of the delay optical fiber is closer to the sum of the lengths of the grating region of a single fiber Bragg grating and the tapered optical fiber, the interference effect is the best.

Claims

1. An ethereal drum structure, characterized in that: The structure of the ethereal drum is a flat drum-shaped cavity with a hollow interior. The surface is composed of multiple sound tongues (601) of different sizes. The sound tongues (601) and the drum-shaped cavity form a resonance cavity to achieve acoustic and structural sensitization. Each sound tongue (601) is hollowed out at the center and a circular diaphragm (604) is embedded. The tapered optical fiber (603) is combined with the circular diaphragm (604). Fiber Bragg gratings (602) are distributed on both sides of the tapered optical fiber (603). The circular diaphragm (604) is used to pick up the incident sound wave signal, convert the intensity of the incident sound wave signal into the vibration amplitude of the diaphragm, and then change the length of the tapered optical fiber (603) on the diaphragm. The tapered optical fiber (603) is used to convert the vibration amplitude of the diaphragm into the phase change of the optical signal, and then change the intensity of the output interference light. There are N sound tongues in total, where N is a positive integer, and the sizes of the N sound tongues (601) are different from each other, and the sizes of the corresponding N circular diaphragms (604) are also different from each other. The two fiber Bragg gratings (602) on both sides of the tapered optical fiber (603) have the same properties and form an F-P cavity. The reflection spectra of the fiber Bragg gratings (602) corresponding to different F-P cavities are different from each other, so that the reflection center wavelengths of each F-P cavity are different.

2. The structure of the ethereal drum according to claim 1, characterized in that: The circular diaphragm (604) is made of nitrile rubber.

3. The ethereal drum structure according to claim 1 or 2, characterized in that: The circular diaphragm (604) adopts non-linear amplification based on the resonant working mode. When the sound wave frequency is approximately equal to the resonant frequency of a certain diaphragm and acts on the structure of the ethereal drum, this diaphragm can achieve more obvious vibration compared with other diaphragms, which will cause a significant change in the cavity length of its corresponding F-P cavity. Using this property, the sound wave frequency can be effectively obtained, and then high sound pressure sensitivity can be achieved.

4. The ethereal drum structure according to claim 1 or 2, characterized in that: The circular diaphragm (604) adopts narrowband filtering based on resonant working mode control, that is, the resonant frequencies corresponding to the diaphragms of different areas in the structure of the ethereal drum are not the same, so that each diaphragm will only have a significant vibration effect when it is affected by the corresponding resonant frequency. This method realizes the recognition of the acoustic spectrum characteristics.

5. The ethereal drum structure according to claim 1 or 2, characterized in that: The tapered optical fiber (603) is combined with the circular diaphragm (604) by gluing.

6. The structure of the ethereal drum according to claim 1, wherein: The sensing structure formed by the combination of the sound tongue (601), the circular diaphragm (604), and the F-P cavity formed by the adjacent fiber Bragg grating (602) and the tapered optical fiber (603), the number N of which can be prepared according to requirements, realizes the resonant absorption of multi-frequency sound waves.

7. The structure of the ethereal drum according to claim 1, characterized in that: The F-P cavity formed by the two fiber Bragg gratings (602) and the tapered optical fiber (603) can be more affected by the circular diaphragm while reflecting the optical signal of a specific Bragg wavelength, further improving the sound pressure sensitivity.

8. A low-frequency sound signal detection device based on the structure of the ethereal drum according to claim 1, characterized in that: It includes a broadband laser (1), a first 1×2 coupler (21), a delay optical fiber (3), a second 1×2 coupler (22), a piezoelectric ceramic (4), a circulator (5), an ethereal drum structure (6), an optical isolator (7), a demultiplexer group (8), a photodetector group (9), a data acquisition card (10), and a host computer (11); The laser emitted from the broadband laser (1) first passes through the first 1×2 coupler (21) and is divided into two optical signals with equal intensities. One of the optical signals passes through the delay fiber (3) so that there is a phase difference between this optical signal and the other optical signal. The other optical signal has a piezoelectric ceramic (4). The piezoelectric ceramic (4) is controlled by the host computer (11) to keep the phase difference between the two paths stable. Then, the two optical signals are emitted through the second 1×2 coupler (22) and enter one port of the circulator (5) successively and are emitted from the second port of the circulator (5). Then, the two optical signals enter the ethereal drum structure (6) successively and pass through N F-P cavities composed of tapered optical fibers and fiber Bragg gratings distributed on both sides of the tapered optical fiber in the ethereal drum structure (6). There are two reflected optical beams inside each F-P cavity. One is reflected by the first fiber Bragg grating in the F-P cavity, and the other is reflected by the second fiber Bragg grating after passing through the first grating. Since there are two optical signals entering the ethereal drum structure (6) successively, there are a total of four reflected optical beams. According to the matching interference theory, since the length of the delay fiber (3) is equal to the length of the F-P cavity, that is, the difference between the phase difference generated at the delay fiber (3) and the phase difference generated at the F-P cavity satisfies the coherence condition of light. Therefore, interference optical signals will be generated at each F-P cavity finally. The interference optical signals generated by each F-P cavity correspond to different wavelengths. Subsequently, all the interference optical signals return to the second port of the circulator (5) and are emitted from the third port. The remaining optical signals that are not reflected are emitted to the optical isolator (7) after passing through the Nth F-P cavity to prevent them from reflecting back to the original optical path. Then, the interference optical signals emitted from different F-P cavities enter the demultiplexer group (8) composed of N demultiplexers respectively. The demultiplexer group (8) separates the interference optical signals with different wavelengths from the interference optical signals with other wavelengths. Then, the interference optical signals are converted into electrical signals by the photodetector group (9) composed of N photodetectors and enter the acquisition card (10) to obtain the change of the interference optical signals of each group of F-P cavities. Finally, the obtained data is transmitted into the host computer (11) and the data is demodulated using the PGC algorithm. The demodulated target signal data is sent into the trained neural network model, and the type of the external sound source can be successfully identified.

9. A method for detecting low-frequency acoustic signals based on the low-frequency acoustic signal detection device according to claim 8, characterized in that, The method is divided into the following steps: S1. Modulation and transmission of optical signals between two 1×2 couplers: S1.1 The broadband laser (1) emits an initial optical signal as the light source and then enters the first 1×2 coupler (21) with a splitting ratio of 50:50 and is divided into two optical signals: One optical signal passes through the delay fiber (3) to generate a phase difference with the other optical signal. This phase difference is used to alleviate the situation that interference cannot occur due to the large difference between the cavity length of the subsequent F-P cavity and the coherence length of light, and this leads to a sequential order when the two optical signals enter the F-P cavity. A piezoelectric ceramic (4) is added to the second optical signal to maintain the stability of this phase difference. The piezoelectric ceramic (4) is controlled by the host computer (11). S1.2 Two optical signals enter the second 1×2 coupler (22) with the same performance as the first 1×2 coupler (21) successively, then enter from one port of the circulator (5) and are output from the second port, and successively enter the ethereal drum structure (6); there are N series-connected F-P cavities composed of two fiber Bragg gratings and one tapered fiber in the ethereal drum structure (6), where N is a positive integer, and the central wavelengths of the optical signals that each F-P cavity can reflect are different. These F-P cavities are respectively close to N circular diaphragms with different areas. The vibration of the circular diaphragms is used to cause the change of the F-P cavity length. When the sound wave is at the resonance frequency of a certain circular diaphragm, the diaphragm will produce significant deformation, thereby driving the obvious change of the F-P cavity length, and then causing the optical phase transmitted in the fiber Bragg grating to change simultaneously, affecting the light intensity of the final interference light; in addition, the inside of the F-P cavity is composed of a tapered fiber, and the tapered fiber has higher pressure sensitivity than ordinary fibers and is more affected by the action of the circular diaphragm, so the change of the cavity length is more obvious; S2. Propagation and matching interference of optical signals in the F-P cavity composed of two fiber Bragg gratings and one tapered fiber: After the two optical signals pass through N F-P cavities in sequence, every time an optical signal passes through an F-P cavity, the optical signal that satisfies the Bragg wavelength condition of the fiber Bragg grating will be reflected inside the corresponding F-P cavity. At this time, two types of reflected light will be generated: one is the optical signal directly reflected after passing through the first fiber Bragg grating in the F-P cavity, and the other is the optical signal that passes through the first fiber Bragg grating and is reflected by the second fiber Bragg grating and then re-transmitted back to the first fiber Bragg grating and returns along the original path. Since the distance between the two fiber Bragg gratings is several decimeters, there is a large phase difference between these two types of reflected optical signals; due to the existence of two optical signals that enter the ethereal drum structure successively, there are a total of four types of reflected light between these two optical signals: the first is the reflected light reflected by the first fiber Bragg grating in the F-P cavity after passing through the delay fiber; the second is the reflected light reflected by the second fiber Bragg grating in the F-P cavity after passing through the delay fiber; The third is the reflected light reflected by the first fiber Bragg grating in the F-P cavity without passing through the delay fiber; the fourth is the reflected light reflected by the second fiber Bragg grating in the F-P cavity without passing through the delay fiber; according to the matching interference theory, there is only one situation for interference to occur between these four types of reflected light, that is, the second reflected light interferes with the third reflected light. The remaining optical signals that are not interfered but are reflected by the fiber Bragg grating are retained in the output signal in the form of direct current. Finally, N interference lights with different wavelengths are generated in N F-P cavities and return to the second port of the circulator. The remaining optical signals that are not reflected are output to the optical isolator (7) after passing through the Nth F-P cavity to prevent them from reflecting back to the original optical path; S3. Detection and target recognition of interference optical signals: S3.1 Detection: After the interference optical signals of N different wavelengths return to the second port of the circulator, they are output from the third port. Considering that the wavelengths of the N interference optical signals are inconsistent, a demultiplexer group (8) composed of N demultiplexers with different wavelengths is set to separate the bands of the interference optical signals. Then, the N optical signals are respectively converted into electrical signals by a photodetector group (9) composed of N photodetectors, and then the signals are collected by a data acquisition card (10) uniformly. The acquired data is input to the host computer (11) and can obtain the target acoustic signal after PGC demodulation processing; S3.2 Identification Inputting the demodulated target signal data into the trained neural network model can be used to identify the external sound source. The long short-term memory network is selected as the main body to construct the neural network model. The following are the steps to construct and train the LSTM neural network model: S3.2.1 After using the acoustic feature data of the known sound source obtained in the actual test as the driving signal of the transducer, place the transducer near the structure of the guzheng drum, so that the circular diaphragm in the guzheng drum structure vibrates, which in turn changes the cavity length of the F-P cavity. Then collect and demodulate the acoustic signal data generated by the known sound source. The obtained demodulated data is divided into a training set and a validation set according to the quantity ratio of 7:3: Among them, the training set is used to train and initially establish the neural network model, and the validation set is used to evaluate the performance of the neural network model obtained by each training; Classify each demodulated data in the training set and the validation set according to the acoustic wave frequency it represents. Each category is used as the true label, that is, as the basis for judging the prediction effect of the model during subsequent training, validation, and testing. A total of N categories are set, which represent the resonant frequencies of diaphragms with N different areas respectively; S3.2.2 After the training set data is received by the input layer of the neural network model, randomly initialize the weights and biases of the neurons in the input layer. At the same time, set the number of iteration cycles to C times, the learning rate S, and the regularization parameter σ. Among them, the learning rate S is the amplitude of updating the model parameters during each iteration, and the regularization parameter σ is used to avoid overfitting of the model; According to S1, it is known that there are N transmitted and returned interference optical signals, and a total of N neurons are set in the input layer, and each neuron corresponds to an interference optical signal at a wavelength; S3.2.3 The training set data propagates forward in the network and enters the hidden layer. A total of M hidden layers are set, where M is a positive integer. After passing through each layer, weighted summation and activation function processing are required to enable the model to handle complex training tasks; The Softmax function is selected as the activation function of the neural network model; S3.2.4 When the training set data enters the output layer through the hidden layer, the Hinge loss function is selected as the loss function to calculate the loss between the predicted output and the true label, and the gradient information of this function is used to calculate the gradients corresponding to the weights and biases of the model in combination with the backpropagation algorithm. According to the gradient feedback, the Adam gradient descent algorithm is used to update the weights and biases in the model in real time; S3.2.5 Repeat the above method for C iterations until the performance of the neural network model on the training set reaches the convergence condition or the predetermined number of iterations C is reached; during the process of constructing the neural network model, it is necessary to use the validation set to evaluate the performance of the model after each round of iteration, including accuracy, recall rate, and F1 score, and optimize the learning rate S and the size of the regularization parameter σ with the above indicators for the next round of iteration; after the iteration process is completed, the model with the best performance on the validation set will be selected as the final neural network model; S3.2.6 Place the steel tongue drum structure in an environment with unknown sound sources, and use the host computer (11) to collect the interference optical signals generated by the influence of the sound sources; after the host computer (11) demodulates these signal data, it is input into the neural network model for prediction and classification to achieve intelligent recognition of low-frequency sound targets.

10. A low-frequency acoustic signal detection method according to claim 9, characterized in that: Set the iteration period C to 600 times, the learning rate S to 0.001, and the regularization parameter σ to 0.01.

Images