A positioning identification method and device for abnormal sound monitoring of power equipment

Abstract

The present application relates to a kind of positioning identification method and device for power equipment abnormal sound monitoring, the method first defines spherical coordinate system, determines the number of microphone in spherical microphone array and the coordinate position of each microphone thereof;Then construct the ideal sound signal model of spherical microphone array, and construct the transfer matrix of focusing surface and array surface based on this in turn obtain noise source intensity solving equation, since focusing surface division point number is much more than array microphone number, so the equation is underdetermined equation group;Finally, construct block sparse bayesian learning optimization framework and solve to obtain source intensity coefficient vector, finally achieve the reliable monitoring and accurate positioning of power equipment abnormal sound.The present application compared with prior art, overcome the deficiency that conventional planar array is difficult to realize flexible autonomous positioning of three-dimensional space sound field, can effectively improve the positioning accuracy of space sound field, and still can guarantee good anti-interference performance in low signal-to-noise ratio environment.

Description

Technical Field

[0001] This invention relates to the field of fault detection technology, and in particular to a method and apparatus for locating and identifying abnormal noises in power equipment. Background Technology

[0002] Urban substations are crucial locations in power systems for transforming electrical energy, and their operational stability and reliability directly impact the safe and stable operation of the entire power system. However, the actual environment within substations is complex, with numerous noise-generating devices of varying noise levels. This noise can severely affect the physical and mental health of nearby residents. Therefore, given the influence of multiple coupling factors and the complex acoustic environment of substations, accurately locating abnormal noises from electrical equipment within the substation is of great significance for clarifying the sound field distribution and implementing corresponding noise reduction measures.

[0003] Most mainstream microphone array noise monitoring systems currently use a planar array topology. However, this type of array is limited by its topology and can only meet the monitoring requirements in a single direction, making it difficult to adapt to the noise detection needs of the three-dimensional sound field environment in substations. Furthermore, when a large number of microphones need to be deployed, this array type becomes extremely large, which is not conducive to on-site testing and portability.

[0004] Beamforming technology based on spherical microphone arrays is a reliable three-dimensional spatial sound source localization technology. However, it is usually difficult to achieve good localization results in low signal-to-noise ratio environments. At the same time, due to the limitation of the Rayleigh criterion, the sound source localization accuracy is proportional to the number of sensors. Given a sphere radius, there is an upper limit to the number of microphones that can be placed on a single sphere, which is not conducive to improving the localization effect.

[0005] Chinese patent application CN202110529779.6 discloses a mechanical fault detection system and method for GIS equipment based on acoustic imaging. The system includes an acoustic signal acquisition module, an acoustic signal processing module, a sound field imaging module, a video acquisition module, a superimposed positioning output module, and a human-computer interaction terminal. The acoustic signal acquisition module includes a microphone array with a non-uniform spiral arrangement and an analog-to-digital conversion module for acquiring multi-channel sound signals. The detection method processes the acquired acoustic signals using spectral subtraction, constructs a sound field model based on the array, including a measurement point plane model and a sound source focal point model, then calculates the intermediate structure using a beamforming algorithm, performs iterative calculations using the deconvolution DAMAS algorithm, outputs a sound field cloud distribution map, and then superimposes and fuses the image data acquired synchronously with the acoustic signals to output a sound field effect map. This invention reduces the sidelobes of the array and enhances anti-interference capability by setting the microphone array with a non-uniform spiral structure. However, this invention requires a large number of microphones to form a microphone array to ensure the fault detection effect; when the number of microphones is small, the fault detection rate is low.

[0006] In summary, there is currently a lack of a location identification method, storage medium, and electronic device for monitoring abnormal noises in substation power equipment, which can achieve accurate abnormal noise location accuracy in complex environments with low signal-to-noise ratios and a limited number of microphones. Summary of the Invention

[0007] The purpose of this invention is to overcome the shortcomings of the prior art and provide a method and device for the location and identification of abnormal noises in power equipment that can achieve accurate location accuracy under conditions of low signal-to-noise ratio and limited number of microphones.

[0008] The objective of this invention can be achieved through the following technical solutions:

[0009] One aspect of the present invention provides a method for locating and identifying abnormal noises in power equipment. This method achieves location identification by establishing a spherical microphone array, comprising the following steps: establishing an ideal acoustic signal model of the microphone array based on its position information; obtaining a noise source strength equation based on the ideal acoustic signal model; and obtaining noise source information based on the noise source strength equation to achieve noise source location identification.

[0010] As a preferred technical solution, the steps of establishing the ideal acoustic signal model include: establishing a spherical coordinate system and obtaining the coordinate information of each microphone on the spherical coordinate system; and establishing the ideal acoustic signal model based on the coordinate information.

[0011] As a preferred technical solution, the step of obtaining the noise source strength solution equation includes: constructing the transfer matrix of the focusing surface and the array surface according to the ideal acoustic signal model, and obtaining the noise source strength solution equation.

[0012] As a preferred technical solution, the noise source strength solution equation is an underdetermined system of equations.

[0013] As a preferred technical solution, the steps of the noise source information acquisition process include: constructing a block sparse Bayesian learning optimization framework based on the noise source strength solution equation to obtain the optimized solution equation; and solving the source strength coefficient vector based on the optimized solution equation to obtain the noise source information.

[0014] In another aspect, the present invention provides a location identification device for monitoring abnormal noise in power equipment, comprising: a data sensing module including a spherical microphone array; a data acquisition module connected to the data sensing module for acquiring signals from the microphones and performing preprocessing; and a data analysis module for acquiring noise source information according to the location identification method for monitoring abnormal noise in power equipment described above.

[0015] As a preferred technical solution, the data analysis module acquires the pre-processed microphone signal via wireless communication.

[0016] As a preferred technical solution, the system also includes a result visualization module connected to the data analysis module, used to display the noise source information.

[0017] In another aspect, an electronic device is provided, comprising one or more processors and a memory, wherein the memory stores one or more programs, the one or more programs including instructions for executing the above-described location identification method for monitoring abnormal noises in power equipment.

[0018] In another aspect, the present invention provides a computer-readable storage medium comprising one or more programs executable by one or more processors of an electronic device, the one or more programs including instructions for performing the above-described location identification method for monitoring abnormal noises in power equipment.

[0019] Compared with the prior art, the present invention has the following advantages:

[0020] (1) Compared with the mainstream microphone array noise monitoring system that uses a planar array topology, the present invention establishes a spherical microphone array for noise localization and identification, which overcomes the shortcomings of conventional planar arrays in achieving flexible and autonomous localization of the three-dimensional spatial sound field. It can accurately reconstruct the acoustic parameters of the target sound source, thereby achieving effective localization of the sound source.

[0021] (2) The sound source localization accuracy of a spherical microphone array is proportional to the number of sensors. Under a given sphere radius, there is an upper limit to the number of microphones that can be arranged on a single sphere. In this invention, for environments with a small number of microphones and a low signal-to-noise ratio, the block sparse Bayesian learning method, which has good signal reconstruction performance and fast reconstruction speed in sparse sampling technology, is used to solve the established noise source strength equation, which significantly improves the signal reconstruction performance and reconstruction speed.

[0022] (3) The data analysis module obtains the microphone signal of the data acquisition module through wireless communication, which facilitates the detection of abnormal noises in power equipment in multiple data acquisition module usage scenarios. Attached Figure Description

[0023] Figure 1 This is a flowchart of the location identification method for monitoring abnormal noises in power equipment, as shown in the embodiment.

[0024] Figure 2 This is a schematic diagram of a positioning and identification device used for monitoring abnormal noises in power equipment in the embodiment;

[0025] Figure 3 This is a schematic diagram of the spherical array in the simulation;

[0026] Figure 4 This is a schematic diagram of the focusing area in the simulation;

[0027] Figure 5 This is a schematic diagram comparing the dual-source localization results of the conventional compressed sensing beamforming method and the method proposed in this invention. (a) is a schematic diagram of the sound source localization results obtained by the conventional compressed sensing beamforming method, and (b) is a schematic diagram of the sound source localization results obtained by the method of this embodiment.

[0028] Figure 6 The diagrams illustrate the five-source localization results of conventional compressed sensing beamforming and the method proposed in this invention. (a) shows the sound source localization result obtained using the conventional compressed sensing beamforming method, and (b) shows the sound source localization result obtained using the method of this embodiment.

[0029] The system includes: 1. Data sensing module; 101. Spherical microphone array; 2. Data acquisition module; 3. Data analysis module; 4. Result visualization module; and 5. Power supply module. Detailed Implementation

[0030] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0031] Example 1

[0032] like Figure 1 This embodiment provides a location identification method for monitoring abnormal noises in power equipment, including the following steps:

[0033] Step S1: Establish a spherical coordinate system, obtain the coordinate information of each microphone in the spherical coordinate system, and establish an ideal acoustic signal model. The specific steps are as follows: First, define the acoustic signal model of each microphone on the spherical array based on spherical coordinates:

[0034]

[0035] In the formula, r h N represents the radius of the spherical microphone array. s q represents the total number of real sound sources. i This represents the source intensity coefficient of the i-th sound source. This represents the transfer function between the m-th microphone and the i-th sound source. Let represent the spherical coordinates of the m-th microphone and the i-th sound source, respectively. Further, the above equation can be expressed in matrix form:

[0036] [p h ] m×1 =[G f ] m×n ·[q] n×1 (2)

[0037] In the formula, p h G represents the sound pressure signal vector of all microphones in an m×1 dimensional spherical array. f Let represent the transfer matrix between the m×n dimension microphone array surface and the sound source focusing surface, and q represent the n×1 dimension source intensity coefficient vector to be determined.

[0038] Step S2: Based on the ideal acoustic signal model, construct the transfer matrices of the focusing surface and the array surface, and obtain the equation for solving the noise source intensity. The specific steps are as follows: Transfer matrix G f This can be specifically represented as:

[0039]

[0040] transfer function It can be calculated using the following formula:

[0041]

[0042] In the formula, R a (·) represents the radial function, and k represents the wave number. Let l represent a class of l-th order a-type spherical harmonics, (·) * Indicates conjugate.

[0043] For sound source localization technology in spherical coordinates, the focal plane of the actual sound source is usually a spherical surface concentric with the spherical array but with a larger radius. Therefore, the inverted focal source coordinates range θ∈[0,180°], φ∈[0,360°]. To avoid discrepancies between the inverted focal source coordinates and the actual sound source, the number of focal plane nodes n is usually much larger than the number of microphones m. This means that solving for the unknown source strength becomes solving an underdetermined system of equations, and conventional methods often fail to yield satisfactory results.

[0044] Step S3: Based on the noise source strength solution equation, construct a block sparse Bayesian learning optimization framework to obtain the optimized solution equation. Then, solve for the source strength coefficient vector using the optimized solution equation to obtain the noise source information. The specific steps are as follows: The block sparse Bayesian learning method, which possesses good signal reconstruction performance and fast reconstruction speed among sparse sampling techniques, is used for solving the problem. The solution equation can be further expressed as:

[0045]

[0046] In the formula, Let be a certain constant.

[0047] Unlike other sparse sampling techniques, the block sparse Bayesian learning method first processes the sound pressure signal p h Transfer matrix G f By performing block partitioning, we have:

[0048]

[0049] The above-mentioned operation fully utilizes the correlation within the data block, and then uses the block sparse Bayesian learning method to solve equation (5), which can significantly improve the signal reconstruction performance and reconstruction speed.

[0050] To highlight the superiority of the method in this embodiment, calculations were performed using the method of this invention on a specific measured data. The simulation setup was as follows: the radius of the spherical array measurement surface was set to 0.5m, and the number of microphones to 100. The desired sound source focusing spherical region was set to be concentric with the array and have a radius of 2.5m. The focusing grid points were discretized into 1106 nodes at angular intervals of Δθ = 7.2° and Δφ = 7.35°. During the simulation calculation, both the method of this invention and conventional compressed sensing beamforming methods added Gaussian white noise with a signal-to-noise ratio of 20dB to simulate the actual test environment. A schematic diagram of the spherical array in the simulation is shown below. Figure 3 The schematic diagram of the focal region in the simulation is as follows: Figure 4 As stated above.

[0051] Simulation 1 Setup: Two point sound sources with different intensities are set up, with initial coordinates of (r, θ, φ) = (2.5m, 90°, 90°) and (r, θ, φ) = (2.5m, 270°, 90°) respectively, and the sound source frequency is 2000Hz. The results of Simulation 1 are as follows: Figure 5 In the above, (a) is a schematic diagram of the sound source localization result obtained by the conventional compressed sensing beamforming method, and (b) is a schematic diagram of the sound source localization result obtained by the method of this embodiment.

[0052] Simulation 2 settings: Five point sound sources with equal intensity were set, with initial coordinates of (4.5m, 90°, 45°), (4.5m, 135°, 67.5°), (4.5m, 180°, 90°), (4.5m, 225°, 112.5°), and (4.5m, 270°, 135°), and the sound source frequency was 3000Hz. Simulation results are as follows. Figure 6 In the above, (a) is a schematic diagram of the sound source localization result obtained by the conventional compressed sensing beamforming method, and (b) is a schematic diagram of the sound source localization result obtained by the method of this embodiment.

[0053] Depend on Figure 5, 6 The sound source localization results show that the method of this invention can accurately locate the sound source, demonstrating superior localization performance compared to conventional compressed sensing beamforming. This sound source localization and identification method overcomes the limitations of conventional planar arrays in achieving flexible and autonomous localization of three-dimensional spatial sound fields. It can accurately reconstruct the acoustic parameters of the target sound source, thereby achieving effective sound source localization; and it maintains good anti-interference performance even in low signal-to-noise ratio environments. In summary, this invention is simple, efficient, and accurate, and the resulting three-dimensional spatial sound field localization and identification method provides precise and reliable technical support for monitoring abnormal noises in power equipment within substations.

[0054] Example 2

[0055] like Figure 2 This embodiment provides a positioning device for monitoring abnormal noise in power equipment, comprising: a data sensing module 1, including a spherical microphone array 101; a data acquisition module 2, connected to the data sensing module 1, for acquiring raw sound pressure data from the data sensing module 1 and performing preprocessing including filtering, signal amplification, and analog-to-digital conversion to obtain preprocessed sound pressure information; a data analysis module 3, for acquiring noise source information according to the method in Embodiment 1 based on the acquired sound pressure information; a visualization module 4, connected to the data analysis module 3, for displaying the noise source information; and a power supply module 5, connected to the data sensing module 1. The data analysis module 3 acquires the preprocessed microphone sound pressure signal via wireless communication.

[0056] Data sensing module 1 is used to collect the acoustic signature signal of the target device. Data acquisition module 2 is used to process the raw acoustic signature data, specifically including collecting the acoustic signature signal obtained by the data sensing module and amplifying the signal, ADC acquisition and signal storage, and transmitting the data to the data analysis module. Data analysis module 3 is used to process and calculate the acoustic signature signal data. First, the acoustic signature model of the spherical microphone array 101 is constructed using the collected acoustic signature signal. Based on this, the transfer matrix of the focusing surface and the array surface is constructed to obtain the noise source strength solution equation. Then, the block sparse Bayesian learning method in sparse sampling technology is used to construct the solution optimization framework and solve for the source strength coefficient vector. By determining the peak value of the weight coefficients, the location information of the noise source is located and the intensity quantization result is obtained. At the same time, the image signal is processed to generate a monitoring background image and overlay it with the sound field cloud map to output a noise cloud map matching the acoustic signature and image. Visualization module 4 is used to output the sound field environment distribution cloud map of the power equipment in the monitoring area.

[0057] Example 3

[0058] This embodiment provides an electronic device, including one or more processors and a memory, wherein the memory stores one or more programs, and the one or more programs include instructions for executing the location identification method for monitoring abnormal noises in power equipment as described in Embodiment 1.

[0059] Example 4

[0060] This embodiment provides a computer-readable storage medium including one or more programs executable by one or more processors of an electronic device, the one or more programs including instructions for performing a location identification method for monitoring abnormal noises in power equipment as described in Embodiment 1.

[0061] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in the present invention, and these modifications or substitutions should all be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for locating and identifying abnormal noises in power equipment, characterized in that, Positioning and identification are achieved by establishing a spherical microphone array. The method includes the following steps: Acquire sound pressure information, and establish an ideal acoustic signal model of the microphone array based on the position information of the spherical microphone array and the sound pressure information; Based on the ideal acoustic signal model, obtain the equation for solving the noise source intensity; By solving the equation based on the noise source intensity, noise source information is obtained, enabling the localization and identification of the noise source. The steps in establishing the ideal acoustic signal model include: Establish a spherical coordinate system and obtain the coordinate information of each microphone in the spherical coordinate system; Based on the coordinate information, the following ideal acoustic signal model is established: In the formula, This represents the radius of the spherical microphone array. This represents the total number of real sound sources. Indicates the first The source intensity coefficient of each sound source Indicates the first The microphone and the first Transfer function between sound sources They represent the first The microphone, the first The spherical coordinates of a sound source The steps for obtaining the noise source strength equation include: Based on the ideal acoustic signal model, the transfer matrices of the focusing surface and the array surface are constructed, and the following equation for solving the noise source intensity is obtained: In the formula, It is a radial function. k Indicates wave number, express l Step a Spherical harmonic functions, Indicates conjugate. The equations for solving the noise source intensity are an underdetermined system of equations. The steps in the process of acquiring noise source information include: Based on the noise source strength equation, the optimized solution equation is obtained as follows. In the formula, Let be a certain constant. express The sound pressure signal vectors of all microphones in the spherical microphone array of the specified dimension. express The transfer matrix between the microphone array surface and the sound source focusing surface in dimensionality. Indicates pending A vector of source strength coefficients of dimension; Based on the optimized solution equation, the sound pressure signal Transfer matrix Perform block partitioning; The source strength coefficient vector is obtained by using the block sparse Bayesian learning method, and the noise source information is obtained.

2. A positioning and identification device for monitoring abnormal noises in power equipment, characterized in that, The apparatus for implementing the positioning and identification method as described in claim 1 includes: The data sensing module (1) includes a spherical microphone array (101). The data acquisition module (2) is connected to the data sensing module (1) and is used to acquire the original sound pressure data from the data sensing module (1) and perform preprocessing to obtain the preprocessed sound pressure information. The data analysis module (3) is used to obtain the preprocessed sound pressure information from the data acquisition module (2) and obtain noise source information according to the location identification method for monitoring abnormal noise of power equipment as described in claim 1.

3. A positioning and identification device for monitoring abnormal noises in power equipment according to claim 2, characterized in that, The data analysis module (3) acquires the pre-processed microphone signal through wireless communication.

4. A positioning and identification device for monitoring abnormal noises in power equipment according to claim 2, characterized in that, It also includes a result visualization module (4) connected to the data analysis module (3) for displaying the noise source information.

5. An electronic device, characterized in that, It includes one or more processors and a memory, wherein the memory stores one or more programs, the one or more programs including instructions for executing the location identification method for monitoring abnormal noises in power equipment as described in claim 1.

6. A computer-readable storage medium, characterized in that, It includes one or more programs that are executed by one or more processors of an electronic device, the one or more programs including instructions for performing the location identification method for monitoring abnormal noises in power equipment as described in claim 1.

Images