Method and device for acoustically imaging fan noise, electric vehicle charging pile

Abstract

The present disclosure relates to a method and device for acoustically imaging fan noise, an electric vehicle charging pile, and a storage medium. The method comprises: acquiring a noise time-domain signal generated by a fan when the fan is running; performing time-frequency conversion on the noise time-domain signal based on a short-time Fourier transform algorithm to obtain a noise frequency-domain signal; inputting the noise frequency-domain signal into a delay-and-sum beamformer to perform beamforming to obtain beam data; positioning principal component data of the beam data to obtain sound power data; and generating a sound image containing a sound source position according to the sound power data. In this embodiment, the resolution of the noise can be improved by positioning the principal component of the beam data, which achieves the effect of accurately positioning the sound source position and is conducive to obtaining a more accurate sound image.

Description

TECHNICAL FIELD

[0001] The present disclosure relates to the technical field of data processing, in particular to a method and device for acoustically imaging fan noise, an electric vehicle charging pile, and a storage medium. BACKGROUND

[0002] Large functional devices such as air conditioners, charging piles, etc. are provided with fans to meet their heat dissipation needs. During heat dissipation, the operation of the fan will form noise, so noise reduction of the sound source noise is needed.

[0003] However, the sound source noise is a block-shaped sound source, and the definition of the position of the noise in the related art is not accurate, which cannot meet the corresponding needs. SUMMARY

[0004] The present disclosure provides a method and device for acoustically imaging fan noise, an electric vehicle charging pile, and a storage medium to solve the deficiencies of the related art.

[0005] According to a first aspect of an embodiment of the present disclosure, a method for acoustically imaging fan noise is provided, the method comprising:

[0006] obtaining a noise time-domain signal generated by a fan when running;

[0007] performing time-frequency conversion on the noise time-domain signal based on a short-time Fourier transform algorithm to obtain a noise frequency-domain signal;

[0008] inputting the noise frequency-domain signal into a delay-and-sum beamformer to perform beamforming to obtain beam data;

[0009] positioning principal component data of the beam data to obtain sound power data;

[0010] generating an acoustic image containing a noise position according to the sound power data.

[0011] Optionally, positioning the principal component data of the beam data to obtain sound power data comprises:

[0012] obtaining sound source intensity distribution data of the principal component data in the beam data;

[0013] obtaining an actual intensity vector of a spatial point sound source according to the sound source intensity distribution;

[0014] obtaining a horizontal angle and a pitch angle in a three-dimensional coordinate system according to coordinate data of the spatial point sound source in a U-space;

[0015] obtaining coordinate data of the spatial point sound source in the three-dimensional coordinate system according to the horizontal angle and the pitch angle;

[0016] The coordinate data and the actual intensity vector constitute the sound power data.

[0017] Optionally, the sound source intensity distribution data of the principal component data in the beam data is obtained, comprising:

[0018] The square value of the beam data of each grid point in the region of interest is obtained to obtain the sound power of each grid point;

[0019] The sound power of all grid points in the region of interest constitutes the sound source intensity distribution data.

[0020] Optionally, the actual intensity vector of the spatial point sound source is obtained according to the sound source intensity distribution, comprising:

[0021] The auxiliary intensity vector, the auxiliary vector and the iteration step length are obtained;

[0022] The power data is obtained according to a preset gradient function and the auxiliary vector, and a projection of the power data in a non-negative quadrant of a preset coordinate system is obtained to obtain an intermediate vector;

[0023] The auxiliary intensity vector is updated according to a preset shrinkage operator and the intermediate vector,

[0024] The iteration step length is updated, and the auxiliary vector is updated according to the iteration step length and the auxiliary intensity vector;

[0025] The sound image is generated according to the auxiliary intensity vector, and a main lobe in the sound image is obtained;

[0026] When the main lobe does not satisfy a preset condition, the step of obtaining the power data according to the preset gradient function and the auxiliary vector is continuously executed until the main lobe satisfies the preset condition to stop iteration, and the auxiliary intensity vector is taken as the actual intensity vector of the spatial point sound source.

[0027] Optionally, the preset condition comprises that the main lobe area is equal to or greater than a preset area threshold, and / or the difference between the maximum value and the minimum value in the main lobe range after each iteration step is greater than or equal to a preset difference threshold.

[0028] Optionally, the projection of the power data in the non-negative quadrant of the preset coordinate system is obtained to obtain the intermediate vector, comprising:

[0029] The ratio vector of the gradient function and the Lipschitz constant is obtained;

[0030] The difference between the auxiliary vector and the ratio vector is obtained to obtain a proportional difference vector;

[0031] The projection in the non-negative quadrant of the preset coordinate system is obtained to obtain the intermediate variable.

[0032] Optionally, the value of the Lipschitz constant is the maximum of eigenvalues of a product of a transposed point spread function matrix and the point spread function matrix.

[0033] Optionally, the auxiliary intensity vector is updated according to the preset shrinkage operator and the intermediate vector, including:

[0034] A difference vector of an absolute value of the intermediate vector and the sparse regularization parameter is obtained.

[0035] A larger value of each element in the difference vector and a set constant is obtained to obtain a difference correction vector.

[0036] A sign operation is performed on the intermediate vector to obtain an intermediate correction vector.

[0037] A dot product of the difference correction vector and the intermediate correction vector is obtained to obtain the auxiliary intensity vector.

[0038] Optionally, the sparse regularization parameter is obtained by:

[0039] An infinite norm of the sound source intensity distribution data is obtained to obtain a candidate value.

[0040] A product of the candidate value and a preset weight is obtained.

[0041] A smaller value of the product and a preset empirical value is obtained to obtain the value of the sparse regularization parameter.

[0042] Optionally, the auxiliary vector is updated according to the iteration step and the auxiliary intensity vector, including:

[0043] An auxiliary difference vector of the auxiliary intensity vector and the updated auxiliary intensity vector is obtained.

[0044] A difference of the iteration step and 1 is obtained, and a ratio of the difference and the updated iteration step is obtained to obtain a step ratio.

[0045] A product of the auxiliary difference vector and the step ratio is obtained to obtain an auxiliary product vector.

[0046] A sum of the auxiliary product vector and the auxiliary intensity vector is obtained to obtain the updated auxiliary vector.

[0047] Optionally, the horizontal angle and the pitch angle in a three-dimensional coordinate system are obtained according to coordinate data of the spatial point sound source in the U-space space, including:

[0048] A sine value and a cosine value of the horizontal angle, and a sine value and a cosine value of the pitch angle are obtained.

[0049] obtaining a first product of a cosine value of the horizontal angle and a sine value of the pitch angle, and constructing a first equation according to the first product and an abscissa of coordinate data of the spatial point sound source in a U-space space;

[0050] obtaining a second product of a sine value of the horizontal angle and a cosine value of the pitch angle, and constructing a second equation according to the second product and an ordinate of the coordinate data of the spatial point sound source in the U-space space;

[0051] calculating the horizontal angle and the pitch under the three-dimensional coordinate system according to the first equation and the second equation.

[0052] According to a second aspect of the embodiments of the present disclosure, an apparatus for acoustic imaging of fan noise is provided, and the apparatus comprises:

[0053] a time domain signal obtaining module configured to obtain a time domain signal of noise generated by a fan when the fan is running;

[0054] a frequency domain signal obtaining module configured to perform time-frequency conversion on the time domain signal of noise based on a short-time Fourier transform algorithm to obtain a frequency domain signal of noise;

[0055] a beam data obtaining module configured to input the frequency domain signal of noise into a delay-and-sum beamformer to perform beamforming to obtain beam data;

[0056] a sound power obtaining module configured to locate principal component data of the beam data to obtain sound power data;

[0057] a sound image generating module configured to generate a sound image containing a noise position according to the sound power data.

[0058] Optionally, the sound power obtaining module comprises:

[0059] an intensity distribution obtaining sub-module configured to obtain sound source intensity distribution data of the principal component data in the beam data;

[0060] an actual intensity obtaining sub-module configured to obtain an actual intensity vector of a spatial point sound source according to the sound source intensity distribution;

[0061] an angle obtaining sub-module configured to obtain a horizontal angle and a pitch angle under a three-dimensional coordinate system according to coordinate data of the spatial point sound source in a U-space space;

[0062] a coordinate data obtaining sub-module configured to obtain coordinate data of the spatial point sound source under the three-dimensional coordinate system according to the horizontal angle and the pitch angle;

[0063] the coordinate data and the actual intensity vector constitute the sound power data.

[0064] Optionally, the intensity distribution acquisition submodule comprises:

[0065] an acoustic power acquisition unit configured to acquire a square value of the beam data of each grid point in the region of interest to obtain the acoustic power of each grid point;

[0066] an intensity distribution acquisition unit configured to determine the acoustic power of all grid points in the region of interest as the sound source intensity distribution data.

[0067] Optionally, the actual intensity acquisition submodule comprises:

[0068] an initial value acquisition unit configured to acquire the auxiliary intensity vector, the auxiliary vector and the iteration step length;

[0069] an intermediate vector acquisition unit configured to acquire power data according to a preset gradient function and the auxiliary vector, and acquire a projection of the power data in a non-negative quadrant of a preset coordinate system to obtain an intermediate vector;

[0070] an auxiliary intensity updating unit configured to update the auxiliary intensity vector according to a preset shrinkage operator and the intermediate vector,

[0071] an iteration step length updating unit configured to update the iteration step length;

[0072] an auxiliary vector updating unit configured to update the auxiliary vector according to the iteration step length and the auxiliary intensity vector;

[0073] an image main lobe acquisition unit configured to generate an acoustic image according to the auxiliary intensity vector and acquire a main lobe in the acoustic image;

[0074] an actual intensity acquisition unit configured to, when the main lobe does not satisfy a preset condition, continue to perform the step of acquiring power data according to a preset gradient function and the auxiliary vector until the main lobe satisfies the preset condition to stop iteration, and take the auxiliary intensity vector as the actual intensity vector of the spatial point sound source.

[0075] Optionally, the preset condition comprises that the main lobe area is at or equal to a preset area threshold value, and / or the difference between the maximum value and the minimum value in the main lobe range after each iteration step is greater than or equal to a preset difference threshold value.

[0076] Optionally, the intermediate vector acquisition unit comprises:

[0077] a ratio vector acquisition subunit configured to acquire a ratio vector of the gradient function and a Lipschitz constant;

[0078] a proportional difference acquisition subunit configured to acquire a difference between the auxiliary vector and the ratio vector to obtain a proportional difference vector;

[0079] An intermediate variable obtaining subunit is configured to obtain an intermediate variable by projecting in a non-negative quadrant of a preset coordinate system.

[0080] Optionally, the value of the Lipschitz constant is a maximum value of eigenvalues of a product of a transposed point spread function matrix and the point spread function matrix.

[0081] Optionally, the auxiliary intensity updating unit comprises:

[0082] A difference vector obtaining subunit is configured to obtain a difference vector between an absolute value of the intermediate vector and the sparse regularization parameter;

[0083] A difference correction vector obtaining subunit is configured to obtain a difference correction vector by taking a larger value between each element of the difference vector and a set constant;

[0084] An intermediate correction vector obtaining subunit is configured to obtain an intermediate correction vector by performing a sign operation on the intermediate vector;

[0085] An auxiliary intensity vector obtaining subunit is configured to obtain the auxiliary intensity vector by taking a dot product of the difference correction vector and the intermediate correction vector.

[0086] Optionally, the auxiliary intensity updating unit further comprises a regularization parameter obtaining subunit configured to obtain the sparse regularization parameter; the regularization parameter obtaining subunit comprises:

[0087] A candidate value obtaining subunit is configured to obtain an infinite norm of the sound source intensity distribution data to obtain a candidate value;

[0088] A product obtaining subunit is configured to obtain a product of the candidate value and a preset weight;

[0089] A regularization parameter obtaining subunit is configured to obtain a smaller value between the product and a preset experience value to obtain a value of the sparse regularization parameter.

[0090] Optionally, the auxiliary vector updating unit comprises:

[0091] An auxiliary difference obtaining subunit is configured to obtain an auxiliary difference vector between the auxiliary intensity vector and the updated auxiliary intensity vector;

[0092] A step ratio obtaining subunit is configured to obtain a difference between the iteration step and 1 and obtain a ratio between the difference and the updated iteration step to obtain a step ratio;

[0093] An auxiliary product obtaining subunit is configured to obtain an auxiliary product vector by taking a product of the auxiliary difference vector and the step ratio;

[0094] The auxiliary vector updating subunit is configured to obtain a sum of the auxiliary product vector and the auxiliary intensity vector to obtain an updated auxiliary vector.

[0095] Optionally, the angle obtaining sub-module comprises:

[0096] The sine-cosine obtaining unit is configured to obtain a sine value and a cosine value of the horizontal angle and a sine value and a cosine value of the pitch angle.

[0097] The first equation obtaining unit is configured to obtain a first product of the cosine value of the horizontal angle and the sine value of the pitch angle and construct a first equation according to the first product and an abscissa of the coordinate data of the spatial point sound source in the U-space.

[0098] The second equation obtaining unit is configured to obtain a second product of the sine value of the horizontal angle and the cosine value of the pitch angle and construct a second equation according to the second product and an ordinate of the coordinate data of the spatial point sound source in the U-space.

[0099] The angle obtaining unit is configured to calculate the horizontal angle and the pitch angle in the three-dimensional coordinate system according to the first equation and the second equation.

[0100] According to a third aspect of embodiments of the present disclosure, an electric vehicle charging pile is provided, comprising a charging assembly and a fan for dissipating heat of the charging assembly, and further comprising:

[0101] a memory and a processor;

[0102] The memory is configured to store a computer program executable by the processor.

[0103] The processor is configured to execute the computer program in the memory to implement the method according to any one of the first aspect.

[0104] According to a fourth aspect of embodiments of the present disclosure, a chip is provided, comprising:

[0105] a processor and an interface; the processor is configured to read instructions through the interface to implement the method according to any one of the first aspect.

[0106] According to a fifth aspect of embodiments of the present disclosure, a non-transitory computer readable storage medium is provided, when the executable computer program in the storage medium is executed by a processor, the method according to any one of the first aspect can be implemented.

[0107] The technical solutions provided by the embodiments of the present disclosure can include the following beneficial effects:

[0108] In the scheme provided by the embodiments of the present disclosure, a noise time-domain signal generated by a fan during operation can be acquired; a short-time Fourier transform algorithm is used to perform time-frequency conversion on the noise time-domain signal to obtain a noise frequency-domain signal; the noise frequency-domain signal is input into a delay-sum beamformer to perform beamforming to obtain beam data; then, principal component data of the beam data is positioned to obtain sound power data; finally, a sound image containing a noise position is generated according to the sound power data. In this way, by positioning the principal component of the beam data, the resolution of the noise can be improved, the sound source position can be accurately positioned, and a more accurate sound image can be obtained.

[0109] It should be understood that the foregoing general description and the following detailed description are only exemplary and explanatory, and cannot limit the present disclosure. BRIEF DESCRIPTION OF DRAWINGS

[0110] The accompanying drawings, which are incorporated into and form part of the specification, illustrate embodiments consistent with the present disclosure and, together with the description, serve to explain the principles of the present disclosure.

[0111] Figure 1 is a schematic diagram of noise positioning in the related art.

[0112] Figure 2 is a flowchart of a method of acoustic imaging of fan noise according to an exemplary embodiment.

[0113] Figure 3 is a schematic diagram of a grid for dividing a region of interest according to an exemplary embodiment.

[0114] Figure 4 is a flowchart of acquiring sound power data according to an exemplary embodiment.

[0115] Figure 5 is a flowchart of acquiring an actual intensity vector according to an exemplary embodiment.

[0116] Figure 6 is an effect diagram of noise positioning according to an exemplary embodiment.

[0117] Figure 7 is another effect diagram of noise positioning according to an exemplary embodiment.

[0118] Figure 8 is a flowchart of a method of acoustic imaging of fan noise according to an exemplary embodiment.

[0119] Figure 9 is a block diagram of an apparatus for acoustic imaging of fan noise according to an exemplary embodiment. DETAILED DESCRIPTION

[0120] The exemplary embodiments will be described in detail below with reference to the drawings. In the following description, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The following exemplary embodiments are not representative of all embodiments consistent with the present disclosure. Rather, they are merely examples of devices consistent with some aspects of the present disclosure as detailed in the appended claims. It is noted that features of the below-described embodiments and implementations can be combined with each other, to the extent not conflicting with each other.

[0121] In the related art, when acoustic imaging is performed, a device such as a fan is taken as a block sound source, and a positioning result as shown in FIG. 1 is generated. In the block sound source, two sound sources are combined into one sound source, and the imaging effect is poor, resulting in inaccurate position of noise, which cannot meet the corresponding requirements. Figure 1

[0122] To solve the above technical problems, the embodiments of the present disclosure provide a method and device for performing acoustic imaging on fan noise, an electric vehicle charging pile, and a storage medium. The method for performing acoustic imaging on fan noise can be applied to an electric vehicle charging pile, Figure 2 FIG. 2 is a flowchart of a method for performing acoustic imaging on fan noise according to an exemplary embodiment. Referring to FIG. 2, a method for performing acoustic imaging on fan noise includes steps 21-25. Figure 2

[0123] In step 21, a noise time-domain signal generated by a fan when the fan is running is acquired.

[0124] In this step, the voice acquisition array includes M array elements, and each array element can be composed of at least one microphone. The voice acquisition array is pre-set at a specified position and used to acquire noise of a sound source. In an example, the specified position refers to a front of the sound source, and a center point of the voice acquisition array is arranged opposite to a center point of the sound source.

[0125] In this step, the voice acquisition array can acquire a voice time-domain signal in real time or according to a set period. Since the fan is a sound source around the voice acquisition array and the fan noise is noise, the voice acquisition array can acquire the voice time-domain signal, which is referred to as a noise time-domain signal. In this step, a sampling frequency f s of the voice acquisition array can be 16 kHz, 44.1 kHz, or 48 kHz. In an example, the sampling frequency f s of the voice acquisition array is 16 kHz, so as to occupy less hardware and computing power.

[0126] ​​It can be understood that the voice acquisition array can be started synchronously with the sound source (such as a fan), and the voice acquisition array starts to work only after the sound source starts, so that the power consumption of the voice acquisition array can be reduced and the noise time domain signal can be acquired.

[0127] In this step, the processor of the electric vehicle charging pile can communicate with the voice acquisition array to obtain the noise time domain signal. The noise time domain signal is a column vector s(t).

[0128] In step 22, the noise time domain signal is subjected to time-frequency conversion based on a short-time Fourier transform algorithm to obtain a noise frequency domain signal.

[0129] In this step, the processor can perform frequency domain conversion on the noise time domain signal to obtain noise frequency domain data. For example, the processor can perform windowing and framing on the noise time domain signal, and perform Fourier transform (i.e., FFT) processing on each frame to obtain the lth frame, kth spectral component signal s(k, l), i.e., to obtain the noise frequency domain data. The windowing and framing and Fourier transform are also called short-time Fourier transform (STFT). In an example, the length of each frame is 64 ms, i.e., the frame length L = 1024. The window function w(n) adopts a Hanning window, the length of which is the same as the frame length, and the frame shift is half of the frame length, i.e., inc = 512.

[0130] In an example, the sound source is a device such as a fan, the operating frequency (or speed) of which is relatively fixed, and the frequency of the corresponding noise is also relatively fixed. At this time, it can be understood that the noise belongs to a narrowband signal; or the main frequency (principal component) of the sound source noise is a narrowband signal, such as a 120 Hz signal. For convenience of processing, the parameter k in the spectral component signal is omitted in subsequent embodiments, i.e., the lth frame, kth spectral component signal s(k, l) becomes the lth frame spectral component signal s(l).

[0131] For ease of understanding, in this example, the principal component is defined as noise data of low-frequency background noise superimposed with low-frequency single-frequency noise, or noise signal with low-frequency noise as background component and containing some order signals, wherein the low frequency refers to no more than 1000 Hz.

[0132] In step 23, the noise frequency domain signal is input into a delay-sum beamformer to perform beamforming to obtain beam data.

[0133] In this step, the processor can obtain the steering vector of the target sound source. It is assumed that the frequency is f k , and the spatial position (i.e., the spatial three-dimensional coordinates) of each element is pos m, m represents the mth array element, and m is 1 to M. It is assumed that the spatial position of the sound source is pos0. In the far field condition, the steering vector of the sound source is:

[0134]

[0135] In formula (1), ω k = 2f k represents the angular frequency of the noise signal, c = 340 m / s represents the speed of sound in air, represents the time difference between the arrival of the sound wave at the mth array element pos m and the first array element pos1, ||·||2 represents the two norm.

[0136] In this step, the processor can obtain the weight coefficient of the DSB (Delay and Sum Beamformer) beamformer according to the steering vector of the sound source:

[0137]

[0138] In formula (2), (·) H represents the conjugate transpose.

[0139] In this step, the processor can obtain the beam data according to the frequency domain data and the weight coefficient, that is:

[0140]

[0141] In formula (3), z(l) represents the beam data.

[0142] In an example, the processor can take the average value of multiple frames of beam data as the value of the current frame of beam data, at which time the frame parameter l can be omitted.

[0143] In this step, referring to Figure 3 , the processor can uniformly divide the region of interest in space (i.e. the plane 32 a certain distance in front of the voice collection array 31) into N grid points; then, the processor can perform DSB (Delay and Sum Beamformer) beamforming on each grid point to obtain beam data z.

[0144] In step 24, the principal component data of the beam data is positioned to obtain sound power data.

[0145] In this step, referring to Figure 4In step 41, the processor can obtain the sound source intensity distribution data of the principal component data in the beam data. In an example, the processor can square the beam data of each grid point in the region of interest to obtain the sound power of each grid point. Continuing to refer to FIG. 4, the processor can calculate the sound power of each grid point as Figure 3 The processor can determine that the sound powers of all grid points in the region of interest constitute the sound source intensity distribution data b. When the sound powers of all grid points are calculated, the sound source intensity distribution data b = [b1, b2…bN]Tcan be obtained. N T .

[0146] In step 42, the processor can obtain the actual intensity vector of the spatial point sound source according to the sound source intensity distribution. In this step, the sound source intensity distribution data formed by the DSB beam data is regarded as a linear combination of the point spread function matrix (PSF) and the actual intensity of the sound source, as shown in equation (4).

[0147] b = Py (4)

[0148] In equation (4), y represents the actual intensity column vector of the spatial point sound source, and P represents the point spread function matrix (PSF), which is a characteristic matrix corresponding to the setting of the voice acquisition array, and the value of the nth row and n’th column element is:

[0149]

[0150] In equation (5), the steering vector d(n) and d(n’) of the nth or n’th grid point of the point spread function matrix P are related.

[0151] It can be understood that the goal in this step is to obtain the actual intensity column vector y of the spatial point sound source in equation (4). Since the point spread function matrix P is usually not a full rank matrix, the actual intensity column vector y cannot be calculated by y = P -1 b. Therefore, the solving process of the actual intensity column vector y is converted into a least square solving problem in this step, and the deconvolution beam forming method such as the Fast Iterative Shrinkage Thresholding Algorithm (FISTA) can be used.

[0152] In an example, referring to FIG. 5, Figure 5 In step 51, the processor can obtain the auxiliary intensity vector, the auxiliary vector, and the iteration step length. In this step, the processor can perform initialization work, i.e., the auxiliary intensity vector y (0) = [0, 0…0]T, the auxiliary vector q (1) = y (0) , and the iteration step length λ (1) ​​= 1.

[0153] In step 52, the processor can obtain the power data according to the preset gradient function and the auxiliary vector, and obtain the projection of the power data in the non-negative quadrant of the preset coordinate system to obtain an intermediate vector.

[0154] In this step, the processor can obtain the preset gradient function Then, the processor can substitute the radiation vector and the point spread function matrix P into the above preset gradient function to obtain the power data b. Then, the processor can project the power data b in the non-negative quadrant (i.e., the first quadrant and the fourth quadrant) of the preset coordinate system to obtain an intermediate vector. Alternatively, the processor can obtain a ratio vector of the gradient function and the Lipschitz constant; then, obtain the difference between the auxiliary vector and the ratio vector to obtain a proportional difference vector; and then, project in the non-negative quadrant of the preset coordinate system to obtain the intermediate variable. The intermediate variable is shown in equation (6).

[0155]

[0156] In equation (6), represents the Euclidean projection on the non-negative quadrant. L represents the Lipschitz constant, which is taken as P T the maximum value of the eigenvalue of P.

[0157] In step 53, the processor can update the auxiliary intensity vector according to the preset shrinkage operator and the intermediate vector.

[0158] In this step, the processor can obtain the preset shrinkage operator and the intermediate vector to update the auxiliary intensity vector, including that the processor can obtain a difference vector of the absolute value of the intermediate vector and a sparse regularization parameter. In this step, the sparse regularization parameter is obtained as follows: the processor can obtain the infinity norm of the sound source intensity distribution data b to obtain a candidate value ||b|| ∞ . The processor can obtain the product of the candidate value and a preset weight, taking the preset weight 0.001 as an example, the product is 0.001||b|| ∞ . The processor can obtain the smaller value of the product and a preset experience value (such as 0.005) to obtain the value of the sparse regularization parameter. The value of the sparse regularization parameter ξ is shown in equation (7):

[0159] ξ = min{0.005, 0.001||b|| ∞}; (7)

[0160] In equation (7), min{·} represents the smaller value of the two, and ||·|| ∞ represents the infinity norm.

[0161] In this step, the value of the sparse regularization parameter ξ is obtained, which can limit the number of iterations, accelerate the convergence speed of the iteration process, and ensure the acoustic image effect.

[0162] Then, the processor can obtain the larger value of each element in the difference value vector and the set constant to obtain a difference correction vector. After that, the processor can perform a sign operation on the intermediate vector to obtain an intermediate correction vector. Finally, the processor can obtain the dot product of the difference correction vector and the intermediate correction vector to obtain the auxiliary intensity vector. The auxiliary intensity vector is obtained as shown in equation (8):

[0163]

[0164] In equation (8), represents the Hadamard product, sign(·) represents the sign operation, and the sparse regularization parameter ξ is an empirical parameter, which can be 0.005.

[0165] In step 54, the processor can update the iteration step and update the auxiliary vector according to the iteration step and the auxiliary intensity vector.

[0166] In this step, the processor can update the iteration step.

[0167]

[0168] In equation (9), a represents the number of iterations.

[0169] In this step, the processor can update the auxiliary vector according to the iteration step and the radiation intensity vector, including: obtaining an auxiliary difference vector of the auxiliary intensity vector and the updated auxiliary intensity vector; obtaining the difference between the iteration step and 1 and obtaining the ratio of the difference to the updated iteration step to obtain a step ratio; obtaining the product of the auxiliary difference vector and the step ratio to obtain an auxiliary product vector; obtaining the sum of the auxiliary product vector and the auxiliary intensity vector to obtain the updated auxiliary vector.

[0170] As shown in equation (10).

[0171]

[0172] In equation (10), y (a) -y (a-1) represents the auxiliary difference vector, represents the step ratio, represents the auxiliary product vector.

[0173] In step 55, the processor can generate an acoustic image according to the auxiliary intensity vector and obtain a main lobe in the acoustic image.

[0174] In this step, the processor can generate an acoustic image according to the auxiliary intensity vector, that is, sequentially sorting the auxiliary intensity vector according to the grid point order and generating an acoustic image to obtain the acoustic image. It should be noted that the processor can use color 1 to represent the acoustic power exceeding the preset power threshold (such as -40 dB), and color 2 to represent the acoustic power below the preset power threshold.

[0175] The voice acquisition array is a 3*3 array, the element spacing is 0.34 cm, the frequency of narrowband beamforming is 500 Hz, and the region of interest is divided into 100*100 grid points for imaging. Two sound sources u x and u y The coordinates are P1[0, 0] and P12[0.3, 0.3], respectively, and the sparse regularization parameter ξ = 0.005, and the simulation effect after 100 iterations is as shown in Figure 6 In this way, the processor can obtain the region 61 composed of color 1, that is, the noise main lobe.

[0176] In step 56, when the main lobe does not satisfy the preset condition, the processor can continue to perform the step of obtaining power data according to the preset gradient function and the auxiliary vector until the main lobe satisfies the preset condition to stop iteration, and the auxiliary intensity vector is taken as the actual intensity vector of the spatial point sound source.

[0177] In this step, a preset condition is provided in the electric vehicle charging pile. The preset condition includes that the main lobe area is equal to or less than a preset area threshold, and / or, the difference between the maximum value and the minimum value in the main lobe range after each iteration step is greater than or equal to a preset difference threshold. The preset area threshold can be an empirical value, for example, it can be adjusted according to the area of the region of interest, such as 10% to 20% of the area of the region of interest. Alternatively, the difference between the maximum value and the minimum value of the acoustic power in the main lobe range is greater than or equal to the preset difference threshold, that is, the difference between the maximum value and the minimum value is still greater than or equal to the preset difference threshold after each contraction of the main lobe, so as to ensure that the main lobe area does not appear excessive contraction. In this step, by using the preset condition to limit the number of iterations, the convergence speed of the iteration process can be accelerated, and the acoustic image effect can be ensured.

[0178] In this step, when the main lobe is obtained, the processor can continue to perform the step of obtaining power data according to the preset gradient function and the auxiliary vector when the main lobe obtained in the current iteration process does not satisfy the preset condition, that is, return to step 52. When the main lobe obtained in the current iteration process satisfies the preset condition, the processor can stop iteration, and at this time the processor can take the auxiliary intensity vector in the current iteration step as the actual intensity vector of the spatial point sound source.

[0179] In step 43, the processor can obtain the horizontal angle and the pitch angle in the three-dimensional coordinate system according to the coordinate data of the spatial point sound source in the U-Space space.

[0180] Considering that the point spread function matrix P has translational invariance in the U-Space space, the above processing process is to calculate in the U-Space space, that is, the acoustic image in the U-Space space can be obtained. At this time, the processor can obtain the horizontal angle and the pitch angle in the three-dimensional coordinate system according to the coordinate data of the spatial point sound source in the U-Space space. For example, the processor can obtain the sine value and the cosine value of the horizontal angle, and the sine value and the cosine value of the pitch angle; then, the processor can obtain a first product of the cosine value of the horizontal angle and the sine value of the pitch angle, and construct a first equation according to the first product and the horizontal coordinate of the coordinate data of the spatial point sound source in the U-space space; then, the processor can obtain a second product of the sine value of the horizontal angle and the cosine value of the pitch angle, and construct a second equation according to the second product and the vertical coordinate of the coordinate data of the spatial point sound source in the U-space space; finally, the processor can calculate the horizontal angle and the pitch in the three-dimensional coordinate system according to the first equation and the second equation.

[0181] Continuing to refer to Figure 3 , the relationship between the coordinate data in the U-Space space and the horizontal angle and the pitch angle in the three-dimensional coordinate system is shown in equation (11).

[0182]

[0183] In equation (11), u x ,u y respectively represent the coordinate data in the U-Space space, respectively represent the horizontal angle and the pitch in the three-dimensional coordinate system.

[0184] In step 44, the processor can obtain the coordinate data of the spatial point sound source in the three-dimensional coordinate system according to the horizontal angle and the pitch angle; the coordinate data and the actual intensity vector constitute the acoustic power data. In this step, continuing to refer to Figure 3 , the processor can obtain the coordinate data of the spatial point sound source (each grid point) in the three-dimensional coordinate system according to the horizontal angle and the pitch angle.

[0185] In step 25, the acoustic image containing the noise position is generated according to the acoustic power data.

[0186] In this step, the processor can generate the acoustic image containing the noise position according to the acoustic power data, and the effect is shown in Figure 7 . It should be noted that Figure 7 is the acoustic image in the U-Space space. Compared withFigure 6 and Figure 7 , Figure 7 the main lobe is relatively Figure 6 larger in area, which can cover the center positions (positions shown by the circles) of the two sound sources, thereby avoiding Figure 6 the problem that the main lobe shown in the prior art is too narrow to cover the center positions of the two sound sources.

[0187] In an example, referring to Figure 8 , the processor can perform STFT processing on the sound source noise, then perform DSB beamforming, and then perform sparse regularization estimation; thereafter, the FISTA deconvolution beamforming iteration processing is performed by using the sparse regularization parameter, and the sound image can be output when a preset condition is met.

[0188] So far, in the scheme provided by the embodiment of the present disclosure, the noise time domain signal can be acquired; then, the beam data corresponding to the noise time domain signal is acquired; thereafter, the principal component data of the beam data is positioned to obtain the sound power data; finally, the sound image containing the noise position is generated according to the sound power data. In this way, in the embodiment, by positioning the principal component of the beam data, the resolution of the noise can be improved, the effect of accurately positioning the sound source position is achieved, and a more accurate sound image is obtained.

[0189] On the basis of the method for performing acoustic imaging on fan noise provided by the embodiment of the present disclosure, the present disclosure further provides a device for performing acoustic imaging on fan noise, referring to Figure 9 , the device comprises:

[0190] a time domain signal acquisition module 91 configured to acquire a noise time domain signal;

[0191] a beam data acquisition module 92 configured to input the noise frequency domain signal into a delay-and-sum beamformer to perform beamforming and obtain beam data;

[0192] a sound power acquisition module 93 configured to position the principal component data of the beam data to obtain sound power data;

[0193] a sound image generation module 94 configured to generate a sound image containing a noise position according to the sound power data.

[0194] Optionally, the sound power acquisition module comprises:

[0195] an intensity distribution acquisition sub-module configured to acquire sound source intensity distribution data of the principal component data in the beam data;

[0196] an actual intensity acquisition sub-module configured to acquire an actual intensity vector of a spatial point sound source according to the sound source intensity distribution;

[0197] an angle obtaining submodule, configured to obtain a horizontal angle and a pitch angle in a three-dimensional coordinate system according to coordinate data of the spatial point sound source in a U-space;

[0198] a coordinate data obtaining submodule, configured to obtain coordinate data of the spatial point sound source in a three-dimensional coordinate system according to the horizontal angle and the pitch angle;

[0199] the coordinate data and the actual intensity vector constitute the sound power data.

[0200] Optionally, the intensity distribution obtaining submodule comprises:

[0201] a sound power obtaining unit, configured to obtain a square value of the beam data of each grid point in the region of interest, to obtain sound power of each grid point;

[0202] an intensity distribution obtaining unit, configured to determine that the sound power of all grid points in the region of interest constitutes the sound source intensity distribution data.

[0203] Optionally, the actual intensity obtaining submodule comprises:

[0204] an initial value obtaining unit, configured to obtain an auxiliary intensity vector, an auxiliary vector and an iteration step length;

[0205] an intermediate vector obtaining unit, configured to obtain power data according to a preset gradient function and the auxiliary vector, and obtain a projection of the power data in a non-negative quadrant of a preset coordinate system, to obtain an intermediate vector;

[0206] an auxiliary intensity updating unit, configured to update the auxiliary intensity vector according to a preset shrinkage operator and the intermediate vector,

[0207] an iteration step length updating unit, configured to update the iteration step length;

[0208] an auxiliary vector updating unit, configured to update the auxiliary vector according to the iteration step length and the auxiliary intensity vector;

[0209] an image main lobe obtaining unit, configured to generate an acoustic image according to the auxiliary intensity vector and obtain a main lobe in the acoustic image;

[0210] an actual intensity obtaining unit, configured to continue to perform the step of obtaining the power data according to the preset gradient function and the auxiliary vector when the main lobe does not satisfy a preset condition, until the main lobe satisfies the preset condition to stop iteration, and take the auxiliary intensity vector as the actual intensity vector of the spatial point sound source.

[0211] Optionally, the preset condition comprises that a main lobe area is equal to or less than a preset area threshold value, and / or a difference between a maximum value and a minimum value in the main lobe range after each iteration step is greater than or equal to a preset difference threshold value.

[0212] Optionally, the auxiliary intensity updating unit comprises:

[0213] a difference vector obtaining sub-unit, configured to obtain a difference vector between an absolute value of the intermediate vector and the sparse regularization parameter;

[0214] a difference correction vector obtaining sub-unit, configured to obtain a larger value between each element in the difference vector and a set constant, to obtain a difference correction vector;

[0215] an intermediate correction vector obtaining sub-unit, configured to perform a sign operation on the intermediate vector, to obtain an intermediate correction vector;

[0216] an auxiliary intensity vector obtaining sub-unit, configured to obtain a dot product of the difference correction vector and the intermediate correction vector, to obtain the auxiliary intensity vector.

[0217] Optionally, the auxiliary intensity updating unit further comprises a regularization parameter obtaining sub-unit, configured to obtain the sparse regularization parameter; the regularization parameter obtaining sub-unit comprises:

[0218] a candidate value obtaining sub-unit, configured to obtain an infinite norm of the sound source intensity distribution data, to obtain a candidate value;

[0219] a product obtaining sub-unit, configured to obtain a product of the candidate value and a preset weight;

[0220] a regularization parameter obtaining sub-unit, configured to obtain a smaller value between the product and a preset experience value, to obtain a value of the sparse regularization parameter.

[0221] It should be noted that the system embodiment shown in the embodiment matches the content of the above-mentioned method embodiment, and the content of the above-mentioned method embodiment can be referred to, and details are not described herein.

[0222] In an example embodiment, an electric vehicle charging pile is also provided, comprising a charging assembly and a fan for dissipating heat of the charging assembly, further comprising:

[0223] a memory and a processor;

[0224] the memory is configured to store a computer program executable by the processor;

[0225] the processor is configured to execute the computer program in the memory to implement the method as described above.

[0226] In an example embodiment, a non-transitory computer-readable storage medium, such as a memory including instructions executable by a processor, is also provided, in which the executable computer program can be executed by the processor. The readable storage medium can be a ROM, a random access memory (RAM), a CD-ROM, a magnetic tape, a floppy disc, and an optical data storage device, etc.

[0227] Embodiments of the present disclosure provide a chip for executing the above method. The chip can be a conventional CPU (central processing unit) chip, a GPU (graphics processing unit) chip, etc., or can be a special-purpose acceleration chip for artificial intelligence technology, such as an AI (Artificial Intelligence) accelerator, etc.

[0228] Other embodiments of the present disclosure will be apparent to those skilled in the art with the consideration of the specification and practice of the present disclosure disclosed herein. The present disclosure is intended to cover any variations, uses, or adaptations of the present disclosure following the general principles thereof and including the modifications and variations thereof that are obvious to those skilled in the art and which are included within the scope of the present disclosure. The specification and examples are to be regarded as illustrative only, and the true scope and spirit of the present disclosure are indicated by the following claims.

[0229] It should be understood that the present disclosure is not limited to the precise structures as herein described and illustrated in the drawings, and that various modifications and changes can be made without departing from its scope. The scope of the present disclosure is limited only by the claims that follow.

Claims

1. A method of acoustically imaging fan noise, characterized by, The method comprises: acquiring a noise time-domain signal generated by a fan during operation; performing time-frequency conversion on the noise time-domain signal based on a short-time Fourier transform algorithm to obtain a noise frequency-domain signal; inputting the noise frequency-domain signal into a delay-sum beamformer to perform beamforming to obtain beam data; positioning principal component data of the beam data to obtain sound power data; the principal component data refers to noise data of low-frequency background noise superimposed with a low-frequency single-frequency signal; generating a sound image containing a noise position according to the sound power data.

2. The method of claim 1, wherein, The positioning of the principal component data of the beam data to obtain the sound power data comprises: acquiring sound source intensity distribution data of the principal component data in the beam data; acquiring an actual intensity vector of a spatial point sound source according to the sound source intensity distribution; acquiring a horizontal angle and a pitch angle in a three-dimensional coordinate system according to coordinate data of the spatial point sound source in a U-space; acquiring coordinate data of the spatial point sound source in the three-dimensional coordinate system according to the horizontal angle and the pitch angle; the coordinate data and the actual intensity vector constitute the sound power data.

3. The method of claim 2, wherein, The acquisition of the sound source intensity distribution data of the principal component data in the beam data comprises: acquiring a square value of the beam data of each grid point in a region of interest to obtain sound power of each grid point; determining that the sound power of all grid points in the region of interest constitutes the sound source intensity distribution data.

4. The method of claim 2, wherein, The acquisition of the actual intensity vector of the spatial point sound source according to the sound source intensity distribution comprises: acquiring an auxiliary intensity vector, an auxiliary vector and an iteration step length; acquiring power data according to a preset gradient function and the auxiliary vector, and acquiring a projection of the power data in a non-negative quadrant of a preset coordinate system to obtain an intermediate vector; updating the auxiliary intensity vector according to a preset shrinkage operator and the intermediate vector; updating the iteration step length and updating the auxiliary vector according to the iteration step length and the auxiliary intensity vector; generating a sound image according to the auxiliary intensity vector and acquiring a main lobe in the sound image; when the main lobe does not satisfy a preset condition, continuing to perform the step of acquiring the power data according to the preset gradient function and the auxiliary vector until the main lobe satisfies the preset condition to stop iteration, and taking the auxiliary intensity vector as the actual intensity vector of the spatial point sound source.

5. The method of claim 4, wherein, The preset condition comprises that the main lobe area is less than or equal to a preset area threshold value, and / or the difference between the maximum value and the minimum value in the main lobe range after each iteration step is greater than or equal to a preset difference threshold value.

6. The method of claim 4, wherein, The acquisition of the projection of the power data in the non-negative quadrant of the preset coordinate system to obtain the intermediate vector comprises: acquiring a ratio vector of the gradient function to a Lipschitz constant; acquiring a difference value of the auxiliary vector and the ratio vector to obtain a proportional difference vector; projecting the proportional difference vector in the non-negative quadrant of the preset coordinate system to obtain the intermediate vector.

7. The method of claim 6, wherein, The value of the Lipschitz constant is the maximum value of the eigenvalues of the product of the rotation matrix of the point spread function matrix and the point spread function matrix.

8. The method of claim 4, wherein, The updating of the auxiliary intensity vector according to the preset shrinkage operator and the intermediate vector comprises: obtaining a difference vector of absolute values of the intermediate vector and the sparse regularization parameter; obtaining a larger value of each element in the difference vector and a set constant to obtain a difference correction vector; performing a sign operation on the intermediate vector to obtain an intermediate correction vector; obtaining a dot product of the difference correction vector and the intermediate correction vector to obtain the auxiliary intensity vector.

9. The method of claim 8, wherein, The sparse regularization parameter is obtained by the following method, comprising: obtaining an infinite norm of the sound source intensity distribution data to obtain a candidate value; obtaining a product of the candidate value and a preset weight; obtaining a smaller value of the product and a preset empirical value to obtain the value of the sparse regularization parameter.

10. The method of claim 4, wherein, Updating the auxiliary vector according to the iteration step and the auxiliary intensity vector, comprising: obtaining an auxiliary difference vector of the auxiliary intensity vector and the updated auxiliary intensity vector; obtaining a difference value of the iteration step and 1 and obtaining a ratio of the difference value and the updated iteration step to obtain a step ratio value; obtaining a product of the auxiliary difference vector and the step ratio value to obtain an auxiliary product vector; obtaining a sum of the auxiliary product vector and the auxiliary intensity vector to obtain the updated auxiliary vector.

11. The method of claim 2, wherein, According to the coordinate data of the spatial point sound source in the U-space space, the horizontal angle and the pitch angle in the three-dimensional coordinate system are obtained, comprising: obtaining the sine and cosine values of the horizontal angle, and the sine and cosine values of the pitch angle; obtaining a first product of the cosine value of the horizontal angle and the sine value of the pitch angle, and constructing a first equation according to the first product and the horizontal coordinate of the coordinate data of the spatial point sound source in the U-space space; obtaining a second product of the sine value of the horizontal angle and the cosine value of the pitch angle, and constructing a second equation according to the second product and the vertical coordinate of the coordinate data of the spatial point sound source in the U-space space; calculating the horizontal angle and the pitch angle in the three-dimensional coordinate system according to the first equation and the second equation.

12. An apparatus for acoustically imaging fan noise, the apparatus comprising: The device comprises: a time domain signal acquisition module for acquiring a noise time domain signal generated by a fan during operation; a frequency domain signal acquisition module for performing time-frequency conversion on the noise time domain signal based on a short-time Fourier transform algorithm to obtain a noise frequency domain signal; a beam data acquisition module for inputting the noise frequency domain signal into a delay-and-sum beamformer to perform beamforming to obtain beam data; a sound power acquisition module for positioning principal component data of the beam data to obtain sound power data; the principal component data refers to noise data of a low-frequency background noise superimposed with a low-frequency single-frequency signal; a sound image generation module for generating a sound image containing a noise position according to the sound power data.

13. An electric vehicle charging station, characterized in that, Comprising: a charging assembly and a fan for dissipating heat for the charging assembly, further comprising a memory and a processor; The memory is used to store a computer program executable by the processor; The processor is used to execute the computer program in the memory to implement the method of any one of claims 1-11.

14. A chip, characterized by Comprising: a processor and an interface; The processor is used to read instructions through the interface to execute the method of any one of claims 1-11.

15. A non-transitory computer-readable storage medium, comprising: The executable computer program in the storage medium, when executed by a processor, can implement the method of any one of claims 1-11.

Images