A device and method for locating a moving dipole source of sound in a small aperture array

Abstract

This invention relates to a device and method for locating a moving dipole sound source using a very small aperture array. The device includes a vector acoustic array and a processor. The vector acoustic array comprises at least three elements, each spaced 1 / 6λ apart, where λ is the wavelength. The vector acoustic array is used to measure sound pressure signals, and the processor is used to process the sound pressure signals and obtain the location result of the dipole source through a nonlinear optimization algorithm. The method includes: measuring the sound pressure signal within a preset time period using the vector acoustic array and converting the measured time-domain sound pressure signal into a frequency-domain sound pressure signal; calculating the vibration velocity signal based on the frequency-domain sound pressure signal and establishing a dipole sound source propagation model; and performing a nonlinear optimization solution on the dipole sound source propagation model to obtain the location result of the dipole source. Compared with the prior art, this invention can achieve high-precision and real-time positioning of moving dipole sources.

Description

Technical Field

[0001] This invention relates to the field of applied acoustics technology, and in particular to a moving dipole sound source localization device and method using a very small aperture array. Background Technology

[0002] A dipole is an acoustic or electromagnetic concept that describes two vibrating sources that are equal in size but in opposite directions. In acoustics, it is often used to describe vibration modes in which two sources vibrate in opposite directions, producing specific sound field modes. Acoustic dipoles can be applied to the design and analysis of devices such as antennas, loudspeakers, microphones, and sound sensors.

[0003] Because dipole sound sources are based on two equal but oppositely oriented charges, their sound pressure distribution in space exhibits strong directionality. The sound wave amplitude is strong in the same direction and weak in other directions. Therefore, dipole sound sources are typically used in applications requiring precise directionality. Dipole sound sources are divided into stationary dipoles and moving dipoles. Propellers (including air propellers and underwater propellers) are typical examples of moving dipole sound sources. When the blades rotate, they periodically strike the fluid, generating alternating thrust and drag acting on the medium, equivalent to a time-oscillating couple.

[0004] Localization of moving dipole sources is beneficial for reliably confirming their location, thus enabling strategic tracking and localization. Literature review reveals that acoustic dipole source localization technology has developed rapidly in recent years. Current mainstream localization algorithms mainly include Time Delay Oscillation (TDOA) methods, subspace algorithms, sparse reconstruction (compressed sensing) algorithms, and deep learning algorithms. Furthermore, the dipole localization method proposed by Bing Yang et al. upgrades the traditional TDOA / SRP-PHAT process to a deep learning version, using an array with more than 8 elements and an annular aperture of 0.5λ to 1λ. It utilizes CNN-RNN algorithms for inference and SRP peak search per frame. This method requires pre-training on large-scale labeled data under reverberation and noise conditions similar to the target scene. When dealing with moving sources, it generally suffers from the following problems: requiring a larger array aperture and more elements; and requiring a large amount of computation, making real-time updates difficult to achieve with limited computing power. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of the prior art by providing a moving dipole sound source positioning device and method with a very small aperture array, which can achieve high-precision and real-time positioning of the moving dipole source.

[0006] The objective of this invention can be achieved through the following technical solution: a mobile dipole sound source localization device with a very small aperture array, comprising a vector acoustic array and a processor, wherein the vector acoustic array comprises at least three array elements, and each array element is spaced 1 / 6λ apart, where λ is the wavelength. The vector acoustic array is used to measure sound pressure signals, and the processor is used to process the sound pressure signals and obtain the localization result of the dipole source by solving a nonlinear optimization algorithm.

[0007] Furthermore, the array element in the vector microphone array includes a support base, on which a vector microphone is mounted. The vector microphone includes four scalar microphones, including a central scalar microphone and three peripheral scalar microphones. The central scalar microphone is located at the center position, and the three peripheral scalar microphones are evenly arranged around the center position.

[0008] Furthermore, a limiting ring is provided on the support base, and the vector microphone is mounted and fixed on the limiting ring.

[0009] Furthermore, the spacing between the central scalar microphone and the peripheral scalar microphones is set according to the dipole source frequency.

[0010] A method for locating a moving dipole sound source using a very small aperture array includes the following steps:

[0011] S1. Measure the sound pressure signal within a preset time using a vector acoustic array, and convert the measured time-domain sound pressure signal into a frequency-domain sound pressure signal;

[0012] S2. Calculate the vibration velocity signal based on the frequency domain sound pressure signal and establish a dipole sound source propagation model;

[0013] S3. Perform nonlinear optimization on the dipole sound source propagation model to obtain the localization result of the dipole source.

[0014] Further, the specific process of step S1 is as follows: each element in the vector microphone array measures the sound pressure signal within a preset time period, and performs fast Fourier transform processing on the time-domain sound pressure signal measured by each element to obtain the frequency-domain sound pressure signal of each scalar microphone in the array element.

[0015] Further, step S2 includes the following steps:

[0016] S21. Calculate the measured value of the vibration velocity signal based on the frequency domain sound pressure signal of each scalar microphone in the array element;

[0017] S22. Based on the measured and theoretical values ​​of vibration velocity signals, construct a dipole sound source propagation model that includes an objective function and constraints.

[0018] Furthermore, the formula for calculating the vibration velocity signal measurement value in step S21 is as follows:

[0019]

[0020] Where v is the measured value of the vibration signal, i is the imaginary unit, ω is the angular frequency, ρ is the density, and Δ is the distance between the central scalar microphone and the peripheral scalar microphone.

[0021] Furthermore, the theoretical values ​​of the vibration velocity signal in step S22 include radial vibration velocity and angular vibration velocity:

[0022]

[0023] Among them, v r Radial vibration velocity, v θ For angular vibration velocity, D s Let θ be the dipole moment intensity, k be the wave number, r be the distance from the sound source to the array element, and θ be the distance from the source to the array element. sound It is a dipole moment;

[0024] The specific dipole sound source propagation model is as follows:

[0025]

[0026] stx min ≤x≤x max ,y min ≤y≤y max ,θ∈[0,2π)

[0027] Where x and y are coordinate data, It is the difference in the direction of vibration velocity. This is the measured value of the vibration velocity signal. x is the theoretical value of the vibration velocity signal. min x max and y min y max The coordinate range limit for the target area.

[0028] Furthermore, step S3 specifically involves using SQP (sequence quadratic program) for nonlinear optimization. x = [x, y, θ] T Ω represents all feasible regions that satisfy the boundary and nonlinear constraints. After solving, the initial position coordinates of the dipole source and the direction angle of the dipole moment (x, y, θ) are obtained. sound ).

[0029] Compared with the prior art, the present invention has the following advantages:

[0030] This invention designs a vector acoustic array and a processor. The vector acoustic array includes at least three array elements, each element being spaced 1 / 6λ apart (λ being the wavelength). The vector acoustic array is used to measure sound pressure signals, the processor processes the sound pressure signals, and a nonlinear optimization algorithm is used to obtain the localization result of the dipole source. Thus, by employing a vector sensor array and a nonlinear optimization algorithm in a collaborative design, high-precision, real-time localization of a moving dipole source can be achieved under conditions of extremely small array aperture (i.e., element spacing of 1 / 6λ).

[0031] This invention constructs a vector microphone array, wherein each array element contains four scalar microphones (one central scalar microphone and three peripheral scalar microphones). By utilizing the three-dimensional sound field perception capability of the vector microphones, an effective acoustic interference network can be constructed with only 3 array elements, significantly reducing the physical size of the vector microphone array.

[0032] This invention first uses a vector acoustic array to measure the sound pressure signal within a preset time period and converts the measured time-domain sound pressure signal into a frequency-domain sound pressure signal. Then, based on the frequency-domain sound pressure signal, the vibration velocity signal is calculated, and a dipole sound source propagation model is established. Finally, a nonlinear optimization solution is performed on the dipole sound source propagation model to obtain the localization result of the dipole source. This realizes a dipole source localization method that calculates the vibration velocity signal from the frequency-domain sound pressure signal, establishes a dipole propagation model, and uses vector vibration velocity joint feature modeling. It can maintain a localization accuracy of ±0.3λ even in complex sound field environments with a signal-to-noise ratio below 10dB. Combined with a nonlinear optimization algorithm, the sound source position and dipole moment direction can be quickly solved. Attached Figure Description

[0033] Figure 1 This is a schematic diagram of a single array element in this invention;

[0034] Figure 2 This is a schematic diagram of the method flow of the present invention;

[0035] Figure 3 This is a schematic diagram of the vector acoustic array layout in the embodiment;

[0036] The markings in the diagram are as follows: 1. Support base, 2. Limiting ring, 3. Scalar microphone. Detailed Implementation

[0037] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments.

[0038] Example

[0039] This scheme first proposes a mobile dipole sound source localization device with a minimal aperture array, including a vector acoustic array and a processor. The vector acoustic array includes at least three array elements, and each array element is spaced 1 / 6λ apart, where λ is the wavelength, thereby achieving the minimal aperture condition. The vector acoustic array is used to measure the sound pressure signal, and the processor is used to process the sound pressure signal and obtain the localization result of the dipole source through a nonlinear optimization algorithm.

[0040] like Figure 1 As shown, the array elements in the vector microphone array include a support base 1, on which a vector microphone is mounted. The vector microphone includes four scalar microphones 3, including a central scalar microphone and three peripheral scalar microphones. The central scalar microphone is located at the center position, and the three peripheral scalar microphones are evenly arranged around the center position. The distance between the central scalar microphone and the peripheral scalar microphones is set according to the dipole source frequency. Usually, this distance Δ must be much smaller than the wavelength λ of the maximum frequency being analyzed, generally less than 1 / 10λ.

[0041] This design incorporates a limiting ring 2 on the support base, and the vector microphone is mounted and fixed on the limiting ring 2.

[0042] Using the above-described device, a method for locating a moving dipole sound source with an extremely small aperture array is implemented, such as... Figure 2 As shown, it includes the following steps:

[0043] S1. Measure the sound pressure signal within a preset time using a vector acoustic array, and convert the measured time-domain sound pressure signal into a frequency-domain sound pressure signal;

[0044] Specifically, the sound pressure signal of each element in the vector microphone array is measured within a preset time period. The time-domain sound pressure signal measured by each element is processed by fast Fourier transform to obtain the frequency domain sound pressure signal of each scalar microphone in the array element.

[0045] S2. Calculate the vibration velocity signal based on the frequency domain sound pressure signal and establish a dipole sound source propagation model;

[0046] Specifically:

[0047] First, calculate the vibration velocity signal measurement value based on the frequency domain sound pressure signal of each scalar microphone in the array element;

[0048] Then, based on the measured and theoretical values ​​of the vibration velocity signal, a dipole sound source propagation model containing the objective function and constraints was constructed.

[0049] S3. Perform nonlinear optimization on the dipole sound source propagation model to obtain the localization result of the dipole source.

[0050] In this embodiment, the above scheme is applied. First, array elements are constructed, and four scalar microphones are correspondingly installed on the limiting rings 2 of the support base 1 and fixed by external threads, so that one scalar microphone is in the center position, and the other three scalar microphones are evenly distributed on the outer ring along the circumference. Then, as follows... Figure 3 As shown, the three array elements are arranged sequentially at intervals of 1 / 6λ, and then the dipole source is located according to the following process:

[0051] Step 1: Measure the sound pressure signal p over a period of time using the vector microphone within the array element, and perform Fast Fourier Transform (FFT) processing on the signal beats to obtain the frequency domain sound pressure signal P for each channel (i.e., each scalar microphone); that is, analyze the frequency based on the measured signal, perform time domain sampling, and then perform beat processing on the sampled signal of each channel to convert the time domain signal into a frequency domain signal.

[0052] Step 2: Calculate the vibration velocity signal v based on the frequency domain sound pressure signal P;

[0053] Step 3: Establish a dipole sound source propagation model. Based on the measured sound pressure amplitude and vibration velocity amplitude and direction, construct a nonlinear measurement equation that includes the source position coordinates and dipole moment direction. Define the cost function as the sum of squares of the second norm of the difference between the measured value and the theoretical value.

[0054] Step 4: Perform nonlinear optimization to obtain the position coordinates and dipole moment direction of the dipole source.

[0055] In step two, through The sound pressure signal of each vector microphone is processed to obtain the vibration velocity signal v, where i is the imaginary unit, ω is the angular frequency, ρ is the density, and Δ is the distance between the central scalar microphone and the peripheral scalar microphones.

[0056] In step three, considering that the signal measured by the microphone changes as the source moves, the amplitude and direction of the received sound pressure signal and vibration velocity signal of each microphone will affect the distance r from the sound source to the array element and the dipole moment θ. sound The influence of this leads to a series of measurement data.

[0057] Through the dipole velocity formula The theoretical vibration amplitude under each vector microphone can be calculated. and the direction of theoretical vibration velocity

[0058] Among them, v r Radial vibration velocity, v θ For angular vibration velocity, D s denoted as , where is the dipole moment intensity and k is the wave number.

[0059] To measure the deviation between measured and theoretical values, the objective function (cost function) is defined as the sum of squares of the second norm as follows:

[0060]

[0061] in, It is the difference in the direction of vibration velocity, and the periodicity of the angle needs to be considered. And mapped to [-π, π]. Regarding constraints, given: x min ≤x≤x max ,y min ≤y≤y max ,θ∈[0,2π), locate in the target area to ensure that the source location and physical quantity are within a reasonable range.

[0062] In step four, SQP is used for nonlinear optimization solution. Where x = [x, y, θ] T Ω represents all feasible regions that satisfy the boundary and nonlinear constraints;

[0063] After solving, the initial position coordinates of the dipole source and the direction angle of the dipole moment (x, y, θ) are obtained. sound ).

[0064] In summary, this scheme adopts a collaborative design of vector sensor array and nonlinear optimization algorithm, which can measure vibration velocity signal using vector microphone under the condition of extremely small array aperture (i.e., array element spacing 1 / 6λ), construct optimization objective function, and realize fast, real-time and high-precision positioning of dipole source through nonlinear optimization solution algorithm. It can be widely used in dynamic sound source positioning scenarios such as rotating machinery fault diagnosis and underwater vehicle tracking.

Claims

1. A method for locating a moving dipole sound source using a very small aperture array, applied to a device for locating a moving dipole sound source using a very small aperture array, characterized in that... The mobile dipole sound source localization device includes a vector microphone array and a processor. The vector microphone array includes at least three array elements, and each array element is spaced 1 / 6λ apart, where λ is the wavelength. The vector microphone array is used to measure the sound pressure signal. The processor is used to process the sound pressure signal and obtain the localization result of the dipole source through a nonlinear optimization algorithm. The array elements in the vector microphone array include a support base, and a vector microphone is installed on the support base. The vector microphone includes four scalar microphones, including a central scalar microphone and three peripheral scalar microphones. The central scalar microphone is located at the center position, and the three peripheral scalar microphones are evenly arranged around the center position. The method for locating a moving dipole sound source includes the following steps: S1. Measure the sound pressure signal within a preset time using a vector acoustic array, and convert the measured time-domain sound pressure signal into a frequency-domain sound pressure signal; S2. Calculate the vibration velocity signal based on the frequency domain sound pressure signal and establish a dipole sound source propagation model; S3. Perform nonlinear optimization on the dipole sound source propagation model to obtain the localization result of the dipole source; Step S2 includes the following steps: S21. Calculate the measured value of the vibration velocity signal based on the frequency domain sound pressure signal of each scalar microphone in the array element; S22. Based on the measured and theoretical values ​​of vibration velocity signals, construct a dipole sound source propagation model that includes an objective function and constraints. The theoretical values ​​of the vibration velocity signal include radial vibration velocity and angular vibration velocity: in, Radial vibration velocity, Angular vibration velocity, For dipole moment intensity, i The imaginary unit, k For wave number, r The distance from the sound source to the array element. θ sound It is a dipole moment; The specific dipole sound source propagation model is as follows: s.t. in, x , y For coordinate data, It is the difference in the direction of vibration velocity. This is the measured value of the vibration velocity signal. This is the theoretical value of the vibration velocity signal. , and , The coordinate range limit for the target area.

2. The method for locating a moving dipole sound source using a very small aperture array according to claim 1, characterized in that, A limiting ring is provided on the support base, and the vector microphone is mounted and fixed on the limiting ring.

3. The method for locating a moving dipole sound source using a very small aperture array according to claim 1, characterized in that, The spacing between the central scalar microphone and the peripheral scalar microphones is set according to the dipole source frequency.

4. The method for locating a moving dipole sound source using a very small aperture array according to claim 1, characterized in that, The specific process of step S1 is as follows: each element in the vector microphone array measures the sound pressure signal within a preset time period, and performs fast Fourier transform processing on the time-domain sound pressure signal measured by each element to obtain the frequency-domain sound pressure signal of each scalar microphone in the array element.

5. The method for locating a moving dipole sound source using a very small aperture array according to claim 1, characterized in that, The formula for calculating the vibration velocity signal measurement value in step S21 is as follows: in, v This is the measured value of the vibration velocity signal. i The imaginary unit, ω Angular frequency, ρ Let denot be the density, and Δ be the distance between the central scalar microphone and the peripheral scalar microphones.

6. The method for locating a moving dipole sound source using a very small aperture array according to claim 1, characterized in that, Step S3 specifically involves using SQP for nonlinear optimization solution. x = [ x, y, θ ] T Ω represents the feasible region satisfying the boundary and nonlinear constraints. After solving, the initial position coordinates of the dipole source and the direction angle of the dipole moment are obtained. .

Images