Electrical equipment fault positioning method, device and equipment

Abstract

The invention provides an electrical equipment fault positioning method, device and equipment. The method comprises the following steps: acquiring a real-time initial acoustic signal of electrical equipment and real-time initial thermal image data of the electrical equipment; performing space-time alignment processing on the real-time initial acoustic signal to obtain a target acoustic signal; performing space-time alignment processing on the real-time initial thermal image data to obtain target thermal image data; carrying out acoustic-thermal coupling feature fusion on the target acoustic signal and the target thermal image data to obtain an acoustic-thermal coupling feature set; obtaining a fault source parameter according to the acoustic-thermal coupling feature set and a relation model of the acoustic emission frequency and the temperature rise rate; and determining fault data according to the fault source parameters and the dynamic mapping three-dimensional model. According to the invention, the electrical equipment fault positioning efficiency, accuracy and decision-making intelligence degree can be improved.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a method, apparatus and equipment for locating faults in electrical equipment. Background Technology

[0002] With the advancement of digitalization in new power systems, the demands for "non-invasive, precise, and real-time" operation and maintenance of primary electrical equipment in high-voltage direct current (HVDC) transmission projects have significantly increased. According to the operation and maintenance requirements of HVDC transmission projects, fault diagnosis within enclosed equipment must achieve a positioning accuracy of ≤5mm, and opening the enclosure for maintenance must be avoided. Current mainstream solutions rely on data from a single sensor, which is insufficient to address the issues of metal casing shielding and multi-source signal aliasing in HVDC transmission projects in desert and Gobi regions. Existing operation and maintenance technologies suffer from drawbacks such as "difficulty in penetrating metal shielding, poor fusion of multi-source data, and weak fault visualization," specifically in the following aspects: 1. Low fault location accuracy: Single physical field analysis cannot penetrate metal shielding, and the internal fault location error exceeds 15mm. Especially in the operation and maintenance of gas-insulated switchgear (GIS) equipment for DC transmission projects in desert areas, it is necessary to open the cover for inspection and confirmation of faults, which increases the risk of equipment downtime and project delays. 2. Poor data fusion: The spatiotemporal references of multi-source data (acoustic, thermal imaging, point cloud) are inconsistent, resulting in feature distortion after fusion. It is impossible to establish acoustic-thermal coupling correlation, and the accuracy of fault feature extraction is less than 70% in dense equipment scenarios. 3. Weak visualization: Fault source parameters cannot be dynamically mapped to the 3D model. Maintenance personnel can only judge the fault location through numerical values, which lacks spatial intuitiveness. In the fault diagnosis of transformer windings in the DC transmission project in the desert area, the location took more than 1 hour. 4. Low decision-making efficiency: There is no quantitative matching of historical cases, relying on human experience, and the misjudgment rate of fault type exceeds 25%. Under high-load operation scenarios, the lag in decision-making can easily lead to secondary faults. Summary of the Invention

[0003] The technical problem to be solved by the present invention is to provide a method, apparatus, and equipment for locating faults in electrical equipment. This can improve the efficiency, accuracy, and intelligence of fault location in electrical equipment.

[0004] To solve the above-mentioned technical problems, the technical solution of the present invention is as follows: A method for locating faults in electrical equipment, comprising: Acquire real-time initial acoustic signals and real-time initial thermal image data of electrical equipment; The real-time initial acoustic signal is spatiotemporally aligned to obtain the target acoustic signal; The real-time initial thermal image data is spatiotemporally aligned to obtain the target thermal image data; The acoustic signal and thermal image data of the target are fused with acoustic-thermal coupling features to obtain an acoustic-thermal coupling feature set. Based on the aforementioned acoustic-thermal coupling feature set and the relationship model between acoustic emission frequency and temperature rise rate, the fault source parameters are obtained; Based on the fault source parameters and the dynamically mapped 3D model, the fault data is determined.

[0005] Optionally, the real-time initial acoustic signal is subjected to spatiotemporal alignment processing to obtain the target acoustic signal, including: The point cloud data of electrical equipment is filtered to obtain spatial anchor points; Based on the spatial anchor points, the physical coordinates are transformed to obtain the point cloud coordinates of the acoustic sensor; Add timestamp markers to the point cloud coordinates of the acoustic sensor to obtain the target acoustic signal; The real-time initial thermal image data is spatiotemporally aligned to obtain the target thermal image data, including: The point cloud data of electrical equipment is filtered to obtain spatial anchor points; Based on the spatial anchor points, the physical coordinates are transformed to obtain the point cloud coordinates of the thermal imaging unit; Add timestamp markers to the point cloud coordinates of the thermal imaging unit to obtain the target thermal image data.

[0006] Optionally, acoustic-thermal coupling feature fusion is performed on the target acoustic signal and target thermal image data to obtain an acoustic-thermal coupling feature set, including: The target acoustic signal is subjected to noise reduction processing to obtain noise reduction acoustic characteristics; Heat source inversion is performed on the target thermal image data to obtain heat source distribution data; The spatial correlation matrix is ​​obtained based on the point cloud coordinates of the acoustic sensor and the thermal imaging unit; Based on the spatial correlation matrix, the noise reduction features and heat source distribution data are weighted and fused to obtain the acoustic-thermal coupling feature set.

[0007] Optionally, based on the acoustic-thermal coupling feature set and the relationship model between acoustic emission frequency and temperature rise rate, fault source parameters are obtained, including: Based on the aforementioned acoustic-thermal coupling feature set, and combining the heat conduction equation and the acoustic wave equation, a coupled partial differential equation between the acoustic emission frequency and the temperature rise rate is established; By applying regularization constraints to the coupled partial differential equations and combining them with the objective function, the fault source parameters are obtained.

[0008] Optionally, based on the fault source parameters and the dynamically mapped 3D model, fault data is determined, including: Based on the fault source parameters, the three-dimensional coordinates of the fault source are obtained; Time-frequency analysis was performed on the noise reduction characteristics to obtain a time-frequency waterfall plot; Based on the heat source distribution data, a thermal gradient cloud map is obtained; The three-dimensional coordinates, time-frequency waterfall plot, thermal gradient cloud plot, and electrical parameters are superimposed onto the preset three-dimensional model to obtain a dynamically mapped three-dimensional model; Fault data are determined based on the dynamic mapping 3D model.

[0009] Optionally, fault data can be determined based on the dynamically mapped 3D model, including: Based on the fault source parameters, the three-dimensional coordinates, displacement, and temperature rise rate of the fault source are obtained. Based on the three-dimensional coordinates, displacement, and temperature rise rate, and combined with the similarity metric function, the fault similarity value is obtained; Fault data is obtained based on the fault similarity value.

[0010] Optionally, the similarity measurement function is:

[0011] Where Sim represents the similarity between the current fault and the i-th historical fault; x is the current fault source parameter; For the i-th historical fault; This represents the weighted Euclidean distance between the two in the M-dimensional feature space, where M is the parameter dimension. This is the distance attenuation factor; The thermo-mechanical coupling index of the current fault; Let be the thermo-mechanical coupling index of the i-th historical fault.

[0012] Embodiments of the present invention also provide an electrical equipment fault location device, comprising: The acquisition module is used to acquire the real-time initial acoustic signal and the real-time initial thermal image data of the electrical equipment. The processing module is used to perform spatiotemporal alignment processing on the real-time initial acoustic signal to obtain the target acoustic signal; perform spatiotemporal alignment processing on the real-time initial thermal image data to obtain the target thermal image data; perform acoustic-thermal coupling feature fusion on the target acoustic signal and the target thermal image data to obtain an acoustic-thermal coupling feature set; obtain fault source parameters based on the acoustic-thermal coupling feature set and the relationship model between acoustic emission frequency and temperature rise rate; and determine fault data based on the fault source parameters and the dynamic mapping three-dimensional model.

[0013] Embodiments of the present invention also provide a computing device, including: one or more processors; and a storage device for storing one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the electrical equipment fault location method of the present invention.

[0014] Embodiments of the present invention also provide a computer-readable storage medium storing a program that, when executed by a processor, implements the electrical equipment fault location method of the present invention.

[0015] The above-described technical solution of the present invention has at least the following technical effects: The electrical equipment fault location method of the present invention acquires real-time initial acoustic signals and real-time initial thermal image data of the electrical equipment; performs spatiotemporal alignment processing on the real-time initial acoustic signals to obtain target acoustic signals; performs spatiotemporal alignment processing on the real-time initial thermal image data to obtain target thermal image data; performs acoustic-thermal coupling feature fusion on the target acoustic signals and target thermal image data to obtain an acoustic-thermal coupling feature set; obtains fault source parameters based on the acoustic-thermal coupling feature set and the relationship model between acoustic emission frequency and temperature rise rate; and determines fault data based on the fault source parameters and a dynamic mapping three-dimensional model. This improves the efficiency, accuracy, and intelligence of electrical equipment fault location. Attached Figure Description

[0016] Figure 1 This is a flowchart illustrating the electrical equipment fault location method of the present invention; Figure 2 This is a schematic diagram of the electrical equipment fault location device of the present invention. Detailed Implementation

[0017] Exemplary embodiments of the invention will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the invention are shown in the drawings, it should be understood that the invention may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this invention will be thorough and complete, and will fully convey the scope of the invention to those skilled in the art.

[0018] like Figure 1 As shown, an embodiment of the present invention proposes a method for locating faults in electrical equipment, comprising: Step S1: Acquire the real-time initial acoustic signal and the real-time initial thermal image data of the electrical equipment; Step S2: Perform spatiotemporal alignment processing on the real-time initial acoustic signal to obtain the target acoustic signal; Step S3: Perform spatiotemporal alignment processing on the real-time initial thermal image data to obtain the target thermal image data; Step S4: Perform acoustic-thermal coupling feature fusion on the target acoustic signal and target thermal image data to obtain an acoustic-thermal coupling feature set; Step S5: Based on the acoustic-thermal coupling feature set and the relationship model between acoustic emission frequency and temperature rise rate, obtain the fault source parameters; Step S6: Determine the fault data based on the fault source parameters and the dynamic mapping 3D model.

[0019] In this embodiment, as Figure 1 As shown, in the electrical equipment fault location method, firstly, real-time status data of the electrical equipment is acquired, including initial acoustic signals and initial thermal imaging data. Initial acoustic signal acquisition involves uniformly arranging an 8-channel microphone array on the equipment casing, setting the sampling rate to 48kHz, and setting the acquisition duration to continuous 24 hours, generating one data segment every 5 minutes and storing it in WAV format. Simultaneously, data is collected under both no-load and rated-load conditions to ensure coverage of different operating states. Initial acoustic signal acquisition obtains raw acoustic data including internal mechanical vibrations (such as vibrations caused by loose connecting rods). The 48kHz sampling rate can accurately capture fault frequency band signals of 1-8kHz, providing sufficient data support for subsequent noise reduction and feature extraction. Acquiring data under two operating conditions avoids missing fault features due to a single operating condition. Initial thermal imaging data is obtained by fixing a 640×512 resolution infrared thermal imager 3 meters directly in front of the equipment, adjusting the lens focal length to fully include the equipment casing in the field of view, setting the sampling interval to 1 minute, collecting the surface temperature distribution data of the equipment, and storing it in TIFF format. Before data acquisition, the thermal imager is calibrated to ensure that the temperature measurement error is ≤ ±0.5℃. The initial thermal image data acquisition obtains the raw data of the heat distribution on the equipment surface. The 640×512 resolution can clearly show the temperature differences in different areas of the equipment (such as the flange and the area corresponding to the outer shell of the contact).

[0020] Then, the real-time initial acoustic signal is spatiotemporally aligned to obtain the target acoustic signal; the real-time initial thermal image data is spatiotemporally aligned to obtain the target thermal image data. Next, acoustic-thermal coupling feature fusion is performed on the target acoustic signal and target thermal image data to obtain an acoustic-thermal coupling feature set; Furthermore, based on the aforementioned acoustic-thermal coupling feature set and the relationship model between acoustic emission frequency and temperature rise rate, the fault source parameters are obtained; Finally, the fault data is determined based on the fault source parameters and the dynamic mapping 3D model.

[0021] The solution of this invention can solve the problems of existing technologies that rely solely on a single physical field (acoustic or vibration) for analysis, which cannot penetrate the metal shielding layer. This makes it difficult to locate micron-level mechanical faults inside enclosed electrical equipment (such as GIS and oil-immersed transformers). In scenarios with dense equipment and complex environments, such as DC transmission projects in desert and Gobi areas, maintenance requires frequent opening of covers for inspection, increasing the risk of outages. It can also solve the problems of inconsistent spatiotemporal references of multi-source data (acoustic signals, thermal imaging data, point cloud data), resulting in feature distortion after data fusion and an inability to provide a reliable basis for fault diagnosis. Furthermore, it can solve the problem that fault source parameters (three-dimensional coordinates, displacement, temperature rise rate) of concealed equipment cannot be dynamically mapped to a three-dimensional model, making it difficult for maintenance personnel to intuitively perceive the fault location and evolution trend. Finally, it can solve the problem that fault diagnosis lacks historical case correlation, and maintenance decisions rely on human experience, which can easily lead to decision lag in high-load operation scenarios of DC transmission projects in desert and Gobi areas.

[0022] In an optional embodiment of the present invention, step S2, performing spatiotemporal alignment processing on the real-time initial acoustic signal to obtain the target acoustic signal, includes: Step S21: Filter the point cloud data of electrical equipment to obtain spatial anchor points; Step S22: Based on the spatial anchor point, transform the physical coordinates to obtain the acoustic sensor point cloud coordinates; Step S23: Add a timestamp to the point cloud coordinates of the acoustic sensor to obtain the target acoustic signal; In step S3, the real-time initial thermal image data is subjected to spatiotemporal alignment processing to obtain target thermal image data, including: Step S31: Filter the point cloud data of electrical equipment to obtain spatial anchor points; Step S32: Based on the spatial anchor point, transform the physical coordinates to obtain the point cloud coordinates of the thermal imaging unit; Step S33: Add a timestamp to the point cloud coordinates of the thermal imaging unit to obtain the target thermal image data.

[0023] In this embodiment, firstly, laser point cloud data of the equipment is acquired using a laser scanner with an accuracy of 0.1mm. Starting from one end of the equipment, a full-size scan of the equipment casing is performed following a spiral scanning path (moving 5cm for every 15° rotation). The scan range covers the entire equipment and the surrounding environment within 1 meter. The output point cloud data format is PLY. During the scanning process, three reflective markers are evenly placed around the equipment for subsequent point cloud stitching calibration. A high-precision 3D point cloud model of the equipment casing is generated using the laser point cloud data. The 0.1mm accuracy accurately reproduces the shape of the equipment casing, the position of flange bolts, and other detailed features. The spiral scanning path avoids scanning blind spots, and the reflective marker calibration ensures that the point cloud data stitching is error-free, laying the foundation for spatial anchor point extraction.

[0024] The collected PLY format point cloud data is preprocessed by removing noisy points according to the rule of 50 neighboring points and a standard deviation multiple of 2. Then, the data is downsampled to reduce the amount of data. Finally, the point cloud of the main device and the point cloud of the surrounding environment are segmented, and the point cloud of the main device is retained to eliminate interference for subsequent spatial anchor point key point detection and improve the accuracy of spatial anchor point key point extraction.

[0025] For the preprocessed point cloud of the main equipment body, local feature values ​​of the point cloud are calculated. Combined with the geometric features of the flange bolts (e.g., the bolt head is hexagonal with a side length of approximately 15mm), candidate key points are further screened. Finally, the flange bolt corner points are determined as spatial anchor points, and the three-dimensional coordinates (x, y, z) of the anchor points in the point cloud coordinate system are output. Using the unique and stable flange bolt corner points as spatial anchor points, with a coordinate error ≤0.05mm in the point cloud coordinate system, provides a unified spatial reference for subsequent coordinate registration of acoustic sensors and thermal imaging units.

[0026] A temporary coordinate system is established with the device base as the origin. The position of each acoustic sensor in the physical space is measured to obtain its physical coordinates (x1, y1, z1). Through the correspondence between the point cloud coordinates (x, y, z) of the spatial anchor point and the temporary coordinate system, the physical coordinates of the acoustic sensor are transformed to the point cloud coordinate system to obtain the coordinates (x1', y1', z1') of the acoustic sensor in the point cloud coordinate system.

[0027] For thermal image data, the intrinsic parameters (focal length f, principal point coordinates (u0, v0)) and extrinsic parameters (translation (tx, ty, tz) and rotation matrix R of the thermal imager relative to the point cloud coordinate system) are first determined. Using the camera projection model u=(fx+txf / z+u0), v=(fy+tyf / z+v0), where (u, v) are the pixel coordinates of the thermal image unit and (x, y, z) are the coordinates in the point cloud coordinate system, the mapping relationship between the pixel coordinates of the thermal image unit and the point cloud coordinate system is established. The coordinates of each thermal image unit are registered to the point cloud coordinate system to obtain the coordinates (x2', y2', z2') of the thermal image unit in the point cloud coordinate system.

[0028] By using unified point cloud coordinates, the coordinates of acoustic sensors and thermal imaging units are accurately registered with the point cloud coordinate system, ensuring spatial consistency of different types of data and eliminating spatial deviations for subsequent multi-source data fusion.

[0029] Then, timestamps are added to the point cloud coordinates of the acoustic sensor and the point cloud coordinates of the thermal imaging unit to obtain the target acoustic signal and the target thermal image data. Clock modules are installed on the acoustic signal acquisition equipment, infrared thermal imager, laser scanner, and data processing server. The clock module of the data processing server serves as the master clock, and the clock modules of the other devices serve as slave clocks. The master clock acquires a standard time signal (time accuracy ≤10ns) through a GPS receiver module to calibrate its own clock. Then, the master clock sends time synchronization messages to each slave clock at a 1-second interval. After receiving the messages, the slave clocks adjust their own clocks according to the message transmission delay to achieve synchronization with the master clock.

[0030] When each device collects data, the synchronized clock module adds a timestamp to each data point (each sampling point of the acoustic signal, each frame of the thermal image data, and each point of the point cloud data) in the format YYYY-MM-DDHH:MM:SS.ssssss, accurate to microseconds. Then, the timestamped acoustic data (including point cloud coordinates), thermal image data (including point cloud coordinates), and point cloud data are transmitted to the data processing server. The server sorts and integrates the three types of data according to the timestamps, deletes data with mismatched timestamps (time difference exceeding 100μs), and finally outputs spatiotemporally unified target acoustic signals and target thermal image data.

[0031] In an optional embodiment of the present invention, step S4 involves performing acoustic-thermal coupling feature fusion on the target acoustic signal and the target thermal image data to obtain an acoustic-thermal coupling feature set, including: Step S41: Perform noise reduction processing on the target acoustic signal to obtain noise reduction acoustic features; Step S42: Perform heat source inversion on the target thermal image data to obtain heat source distribution data; Step S43: Obtain the spatial correlation matrix based on the point cloud coordinates of the acoustic sensor and the thermal imaging unit; Step S44: Based on the spatial correlation matrix, the noise reduction features and heat source distribution data are weighted and fused to obtain the acoustic-thermal coupling feature set.

[0032] In this embodiment, firstly, the target acoustic signal is denoised to obtain denoised acoustic features; then, the target acoustic signal in the spatiotemporally unified target state data is decomposed into 5-level wavelet packet decomposition; during the decomposition process, the frequency range of the acoustic signal (0-24kHz) is divided into 32 frequency bands, each with a frequency width of 24kHz / 32=0.75kHz, and the coefficients of each decomposed frequency band are stored to form a wavelet packet decomposition coefficient matrix; Then, the acoustic signal under normal operating conditions is acquired as a background noise sample, and the same 5-level wavelet packet decomposition is performed on it to calculate the energy value of each frequency band and determine the frequency bands where the background noise is mainly distributed (usually 0-1kHz and 8-24kHz). Then, the energy values ​​of each frequency band of the acoustic signal under fault conditions are compared with the energy values ​​of each frequency band of the background noise. The frequency bands with an energy difference of less than 30% of the background noise energy are identified as noise frequency bands. The wavelet packet decomposition coefficients of these frequency bands are thresholded and suppressed, with the preferred threshold set at 1.5 times the energy value of the background noise in this frequency band. Finally, the wavelet packet decomposition coefficients after noise suppression are reconstructed to obtain the noise-reduced acoustic signal, and the energy value of the 1-8kHz fault frequency band is calculated. This frequency band is the main distribution frequency band of the vibration signal of internal mechanical faults (such as loose connecting rods) in the equipment, and the noise-reduced acoustic characteristics (energy of the 1-8kHz fault frequency band) are output.

[0033] Noise reduction processing can effectively separate shell reflection noise (mainly distributed in 0-1kHz and 8-24kHz) from internal mechanical vibration characteristics (mainly distributed in 1-8kHz). The signal-to-noise ratio of the noise reduction signal is improved by ≥20dB, and the energy calculation accuracy of the 1-8kHz fault frequency band is ≥90%, providing pure acoustic characteristic data for subsequent acoustic-thermal coupling characteristic fusion.

[0034] Next, heat source inversion is performed on the target thermal image data to obtain heat source distribution data; Preprocessing is performed on the TIFF format target thermal image data in the spatiotemporal unified target state data. First, salt-and-pepper noise in the thermal image is removed. Then, using the radiometric calibration coefficient built into the thermal imager, the grayscale values ​​of the thermal image units are converted into temperature values. The surface temperature distribution image of the equipment is obtained through radiometric calibration (calibration error ≤ ±0.3℃). Next, based on the material parameters and operating environment parameters of the equipment, the boundary conditions of the heat conduction equation are determined. For example, the outer shell material is aluminum alloy with a thermal conductivity k = 200 W / (m•K), specific heat capacity c = 900 J / (kg•K), and density ρ = 2700 kg / m³. 3 The operating ambient temperature T0 = 25℃, and the convective heat transfer coefficient h = 10 W / (m²) 2 •K)), the convective heat transfer boundary condition between the equipment surface and the environment: -k∂T / ∂n=h(T-T0), where k is the thermal conductivity, n is the normal vector of the equipment surface, T is the equipment surface temperature, T0 is the ambient temperature, and h is the convective heat transfer coefficient.

[0035] The three-dimensional heat conduction equation inside the device is established as follows: ∂T(x,y,z,t) / ∂t=α(∂ 2 T(x, y, z, t) / ∂x 2 +∂ 2 T(x, y, z, t) / ∂y 2 +∂2 T(x, y, z, t) / ∂z 2 )+q(x,y,z,t) / (ρc) Where α=k / (ρc) is the thermal diffusivity, q(x,y,z,t) is the intensity of the internal heat source, T(x,y,z,t) is the temperature of point (x,y,z) inside the equipment at time t, and t is time.

[0036] The internal space of the equipment is discretized into a three-dimensional mesh, with a preferred mesh size of 1mm×1mm×1mm. The heat conduction equation is discretized into a system of algebraic equations. Using the equipment surface temperature data as observations, the relationship between the observations and the internal temperature field is established. By optimizing and solving the heat conduction equation, the heat source intensity q(x,y,z,t) of each mesh node inside the equipment is obtained, thereby determining the internal heat source distribution and outputting heat source distribution data. The heat source distribution data includes the three-dimensional coordinates and heat source intensity of the overheated area, such as the overheated area of ​​the contact, which is defined as a heat source intensity q>500W / m. 3 The area.

[0037] By accurately reversing the internal heat sources of the equipment, the heat source distribution corresponding to faults such as contact overheating can be accurately identified, providing reliable heat source distribution data for subsequent acoustic-thermal coupling feature fusion.

[0038] Next, the spatial correlation matrix is ​​obtained based on the point cloud coordinates of the acoustic sensor and the thermal imaging unit; Extract the coordinates (x1', y1', z1') of the acoustic sensors in the point cloud coordinate system (for a total of 128 acoustic sensors) and the coordinates (x2', y2', z2') of the thermal imaging units in the point cloud coordinate system (for a total of 128 thermal imaging units). For each acoustic sensor i (i=1,2,...,128) and each thermal imaging unit j (j=1,2,...,128), calculate the spatial distance between them according to the Euclidean distance formula:

[0039] The spatial distance weight w_ij between acoustic sensor i and thermal imaging unit j is calculated using a Gaussian weighting function, with the formula: w_ij=exp(-d_ij) 2 / (2σ 2 )), where σ is the standard deviation of the Gaussian function. Based on the size of the equipment and the arrangement density of the acoustic sensors and thermal imaging units, σ ​​is preferably set to 0.3m to ensure that the acoustic sensors and thermal imaging units that are closer together have higher weights and the ones that are farther apart have lower weights, so as to conform to the actual physical field propagation law. The calculated 128×128 weight values ​​w_ij are arranged in the order of acoustic sensor number as row and thermal imaging unit number as column to construct a 128×128-dimensional spatial correlation matrix W=[w_ij].

[0040] For noise reduction characteristics, the energy E_ac in the 1-8kHz fault frequency band consists of 128 values, with one energy value corresponding to each acoustic sensor. These values ​​are then standardized using the following formula: E_ac'_i=(E_ac_i-μ_ac) / σ_ac,; Where μ_ac is the mean of the energy values ​​of the 128 acoustic features, and σ_ac is the acoustic standard deviation; For the heat source distribution data, the heat source intensity is q_th, with a total of 128 values. Each thermal imaging unit corresponds to one heat source intensity value. The data is then standardized using the following formula: q_th'_j=(q_th_j-μ_th) / σ_th, Where μ_th is the mean of the 128 heat source intensity values, and σ_th is the standard deviation of the heat source.

[0041] After standardization, the mean of E_ac' and q_th' is 0, and the standard deviation is 1. This eliminates the influence of differences in dimensions and numerical ranges between different types of features (acoustic energy, heat source intensity), such as acoustic energy in dB and heat source intensity in W / m². 3 The numerical ranges may differ by several orders of magnitude, ensuring that the two types of features have equal weight in subsequent fusion calculations. This prevents one type of feature from dominating the fusion result due to excessively large values, thus laying the foundation for accurate fusion.

[0042] Finally, based on the spatial correlation matrix, the noise reduction features and heat source distribution data are weighted and fused to obtain the acoustic-thermal coupling feature set.

[0043] The standardized noise reduction features E_ac'=[E_ac'_1,E_ac'_2,...,E_ac'_128], the heat source feature vector q_th'=[q_th'_1,q_th'_2,...,q_th'_128], and the 128×128-dimensional spatial correlation matrix W are weighted and fused. Specifically, a fused feature matrix F is constructed, where F(i,j)=E_ac'_i×w_ij+q_th'_j×(1-w_ij), where (1-w_ij) represents the complementary weights of the heat source features, ensuring that the sum of the weights of the noise reduction features and the heat source features at different spatial locations is 1. The mean values ​​of rows and columns of the fusion feature matrix F are calculated to obtain a 128-dimensional acoustic-thermal coupling feature set F_final=[F_avg_1,F_avg_2,...,F_avg_128], where F_avg_k is the mean value of the k-th row and the k-th column of the fusion feature matrix (k=1,2,...,128).

[0044] The spatial adaptive fusion of acoustic features and heat source features is achieved by using a spatial correlation matrix. The fused feature set F_final can reflect both the acoustic vibration information at different locations of the equipment and the heat source distribution characteristics. The feature dimension is reduced from 256 dimensions (128 dimensions of acoustics + 128 dimensions of heat source) to 128 dimensions. While retaining key information, data redundancy is reduced, the recognizability of the fused features is improved, and high-value feature input is provided for fault source parameter inversion.

[0045] In an optional embodiment of the present invention, step S5, based on the acoustic-thermal coupling feature set and the relationship model between acoustic emission frequency and temperature rise rate, obtains fault source parameters, including: Step S51: Based on the aforementioned acoustic-thermal coupling feature set, and combining the heat conduction equation and the acoustic wave equation, establish a coupled partial differential equation between the acoustic emission frequency and the temperature rise rate; Step S52: Apply regularization constraints to the coupled partial differential equations and combine them with the objective function to obtain the fault source parameters.

[0046] In this embodiment, based on the acoustic-thermal coupling feature set F_final, the coupling relationship between the acoustic emission frequency f and the temperature rise rate r generated by internal mechanical faults (such as loose connecting rods) is analyzed—the higher the mechanical vibration frequency, the more intense the component friction, and the faster the temperature rise rate. Model variables are defined as follows: the independent variable is the eigenvalue F_final_k (corresponding to coupling characteristics at different spatial locations) in the acoustic-thermal coupling feature set; the dependent variables are the acoustic emission frequency f and the temperature rise rate r; and the intermediate variables are the fault source vibration displacement s and the frictional heat generation power P.

[0047] By combining the heat conduction coupling equation and the acoustic wave equation, a coupled partial differential equation for the acoustic emission frequency and the temperature rise rate is established: Acoustic wave equation: ∂ 2 p / ∂t 2 =c 2 (∂ 2 p / ∂x 2 +∂ 2 p / ∂y 2 +∂ 2 p / ∂z 2 )+K×F_final×(∂s / ∂t), where p is the sound pressure, c is the sound speed, K is the coupling coefficient, and ∂s / ∂t is the vibration velocity.

[0048] The heat conduction coupling equation is: ∂T / ∂t = α(∂ 2 T / ∂x 2 +∂ 2 T / ∂y 2 +∂ 2 T / ∂z 2)+η×F_final×P, where η is the heat conversion efficiency, and the relationship between P and the vibration displacement s is P=μ×m×ω 2 ×s, μ is the friction coefficient, m is the mass of the faulty component, and ω is the angular frequency = 2πf.

[0049] By combining the two equations and eliminating intermediate variables, we obtain a set of coupled partial differential equations with acoustic emission frequency f and temperature rise rate r (r=∂T / ∂t) as outputs.

[0050] By applying regularization constraints to the coupled partial differential equations and combining them with the objective function, the fault source parameters are obtained.

[0051] Apply regularization constraints and construct the objective function: J = ||Ax - b|| 2 +λ||Lx|| 2 , where: A is the coefficient matrix; x is the parameter vector of the fault source to be determined; b is the observation vector (the sound pressure and temperature observation values ​​obtained by converting the acoustic-thermal coupling feature set F_final); λ is the regularization parameter; L is the regularization matrix.

[0052] Applying regularization constraints can suppress noise interference in equation solving, ensure the stability and uniqueness of the solution, and avoid huge deviations in parameter solution results due to small fluctuations in data.

[0053] In an optional embodiment of the present invention, step S6, determining fault data based on the fault source parameters and the dynamically mapped three-dimensional model, includes: Step S61: Obtain the three-dimensional coordinates of the fault source based on the fault source parameters; Step S62: Perform time-frequency analysis on the noise reduction characteristics to obtain a time-frequency waterfall plot; Step S63: Obtain a thermal gradient cloud map based on the heat source distribution data; Step S64: The three-dimensional coordinates, time-frequency waterfall plot, thermal gradient cloud plot and electrical parameters are superimposed on the preset three-dimensional model to obtain a dynamically mapped three-dimensional model; Step S65: Determine the fault data based on the dynamic mapping 3D model.

[0054] In this embodiment, a pre-built parametric 3D model of the equipment is loaded, including the geometric models and material properties of key components such as the equipment housing, flanges, contacts, and connecting rods.

[0055] Based on the fault source parameters, the three-dimensional coordinates of the fault source are obtained; the fault source parameters contain the three-dimensional coordinates of the fault source. Then, time-frequency analysis was performed on the noise reduction acoustic characteristics to obtain a time-frequency waterfall plot; the time-frequency analysis results of the acoustic signal were converted into a time-frequency waterfall plot. Next, based on the heat source distribution data, a thermal gradient cloud map is obtained; the temperature values ​​in the heat source distribution data are converted into pseudo-color codes, with a temperature range of 25-80℃, blue representing low temperature and red representing high temperature, to generate a thermal gradient cloud map. Next, obtain the opening and closing current curves from the equipment monitoring system to obtain electrical parameters; Finally, the three-dimensional coordinates, time-frequency waterfall plot, thermal gradient cloud plot and electrical parameters are superimposed on the preset three-dimensional model to obtain the dynamic mapping three-dimensional model; In the dynamically mapped 3D model, when the mouse clicks on a fault area, the corresponding time-frequency waterfall plot fault frequency band (1-8kHz), thermal gradient cloud plot fault area (red high temperature zone), and abnormal current curve segment (such as the peak deviation segment of opening and closing current) are automatically highlighted. The interactive control supports adding zoom, pan, and rotate controls, allowing maintenance personnel to adjust the panel size, position, and 3D model perspective. It also provides data query functions, such as displaying specific values ​​when the mouse hovers over the cloud plot or curve, such as "Time: 5s, Frequency: 3kHz, Sound Pressure Level: 65dB".

[0056] By dynamically mapping the 3D model, it is possible to achieve deep integration of multi-physics information with the 3D model. Maintenance personnel can view the multi-dimensional characteristics of the fault, including sound, heat, and electricity, through interactive operations. This greatly improves the efficiency of information acquisition and avoids misjudgment of faults caused by a single data dimension.

[0057] In an optional embodiment of the present invention, step S65, determining fault data based on the dynamically mapped three-dimensional model, includes: Step S651: Based on the fault source parameters, obtain the three-dimensional coordinates, displacement, and temperature rise rate of the fault source. Step S652: Based on the three-dimensional coordinates, displacement, and temperature rise rate, and combined with the similarity metric function, obtain the fault similarity value; Step S653: Obtain fault data based on the fault similarity value.

[0058] In this embodiment, the three-dimensional coordinates, displacement, and temperature rise rate of the fault source are obtained from the fault source parameters. Then, based on the three-dimensional coordinates, displacement, and temperature rise rate, and combined with the similarity metric function, the fault similarity value is obtained; specifically, the similarity between the input features and historical fault database cases is calculated based on the similarity function. The similarity function is obtained based on the relationship between the current fault source parameters, the thermo-mechanical coupling index of the current fault, and the thermo-mechanical coupling index of historical faults. The thermo-mechanical coupling index is calculated from the temperature rise rate and displacement of the corresponding fault. Finally, based on the fault similarity value, the fault type (such as loose connecting rod, poor contact of contact, etc.) and confidence level are output, thereby obtaining the operation and maintenance decision instructions. The operation and maintenance decision instructions may include a list of required spare parts, such as connecting rods and bolts; torque parameters, such as bolt tightening torque values; operation trajectory, such as the steps and spatial paths for operation and maintenance personnel to disassemble and replace parts, etc.

[0059] The operation and maintenance decision-making instruction generation process uses quantitative matching based on historical cases to reduce the misjudgment rate to ≤5% and shorten the decision-making time from 2 hours to 30 minutes. This adapts to the efficient operation and maintenance needs of complex DC transmission projects and reduces decision-making errors caused by reliance on human experience. At the same time, the multi-physics overlay interface provides operation and maintenance personnel with comprehensive fault information, helping to formulate accurate operation and maintenance plans.

[0060] In an optional embodiment of the present invention, in step S652, the similarity measurement function is:

[0061] Where Sim represents the similarity between the current fault and the i-th historical fault; x is the current fault source parameter; For the i-th historical fault; This represents the weighted Euclidean distance between the two in the M-dimensional feature space, where M is the parameter dimension. This is the distance attenuation factor; The thermo-mechanical coupling index of the current fault; Let be the thermo-mechanical coupling index of the i-th historical fault.

[0062] In this embodiment, the similarity measurement function is:

[0063] Where Sim represents the similarity between the current fault and the i-th historical fault; x is the current fault source parameter; For the i-th historical fault; This represents the weighted Euclidean distance between the two in the M-dimensional feature space, where M is the parameter dimension. This is the distance attenuation factor; The thermo-mechanical coupling index of the current fault; Let be the thermo-mechanical coupling index of the i-th historical fault.

[0064] In this embodiment, preferably, M=5, which includes three-dimensional coordinates of x / y / z, displacement, and temperature rise rate, and the weights are set according to the degree of influence of each parameter on the fault type; The similarity index is used to adjust the impact of distance on similarity. The thermo-mechanical coupling index is calculated from the temperature rise rate and displacement of the corresponding fault. Specifically, the thermo-mechanical coupling index = temperature rise rate × displacement × 1000. Finally, historical cases with a similarity Sim≥0.85 are selected. If there are no cases with Sim≥0.85, the current fault characteristics are added to the unlabeled case library to remind maintenance experts to make manual judgments.

[0065] Application examples of this invention: A 220kV transformer at a substation experienced a localized overheating fault. The fault location method described in this solution was adopted for operation and maintenance. A three-dimensional parametric model of the transformer was constructed, and the fault source parameters (coordinates x=850mm, y=620mm, z=1200mm, temperature rise rate 2.1℃ / min) were mapped onto the model. By rendering the fault area and overlaying thermal gradient cloud maps and electrical parameter curves, a multiphysics fusion interface was generated.

[0066] Maintenance personnel can accurately locate faults within 12 minutes (compared to 36 minutes for traditional static numerical displays), improving efficiency by 3 times; through a visual interface, they can clearly understand the thermal distribution and electrical characteristics of faults, formulate precise repair plans, and prevent faults from escalating.

[0067] The present invention overcomes the limitations of metal shielding by coupling partial differential equations and regularization constraints, achieving non-invasive localization of internal faults with a positioning accuracy of ≤5mm; based on feature space anchor points and time synchronization, it achieves spatiotemporal unification of acoustic, thermal imaging, and point cloud data, with a fusion accuracy of ≥95%; by mapping fault source parameters to a three-dimensional model in real time, it enhances spatial intuitiveness; and based on the similarity analysis of historical fault thermo-mechanical coupling index, it achieves accurate matching of fault types, improving decision-making efficiency by 60%.

[0068] The advantages of this invention are mainly reflected in the following aspects: Regarding fault location accuracy, existing technologies using single physical field analysis have a location error exceeding 15mm, requiring opening the cover for maintenance; this invention solves the problem through acoustic-thermal coupling inverse problem, with an error ≤5mm, eliminating the need for opening the cover. In the operation and maintenance of DC transmission equipment in desert areas, this avoids the risk of equipment downtime and shortens the maintenance period by 80%.

[0069] Regarding data fusion, existing technologies suffer from chaotic spatiotemporal benchmarks for multi-source data, resulting in feature extraction accuracy of less than 70%. This invention improves fusion accuracy to 95% through spatial anchor registration and time synchronization. In scenarios with dense equipment, fault features remain undistorted, providing a reliable basis for diagnosis.

[0070] In terms of visualization, existing technologies can only display static values ​​without spatial correlation; this invention achieves dynamic mapping between fault source parameters and three-dimensional models, and superimposes multi-physics field information, enabling maintenance personnel to locate fault locations within 15 minutes, improving efficiency by 3 times in the operation and maintenance of transformers in DC transmission projects in desert areas.

[0071] In terms of intelligent decision-making, existing technologies rely on human experience, with a misjudgment rate of over 25%. This invention uses quantitative matching of historical cases to reduce the misjudgment rate to ≤5% and shorten the decision-making time from 2 hours to 30 minutes, thus meeting the high-efficiency operation and maintenance needs of complex DC transmission projects.

[0072] like Figure 2 As shown, an embodiment of the present invention also provides an electrical equipment fault location device 20, comprising: The acquisition module 21 is used to acquire the real-time initial acoustic signal and the real-time initial thermal image data of the electrical equipment. Processing module 22 is used to perform spatiotemporal alignment processing on the real-time initial acoustic signal to obtain the target acoustic signal; perform spatiotemporal alignment processing on the real-time initial thermal image data to obtain the target thermal image data; perform acoustic-thermal coupling feature fusion on the target acoustic signal and the target thermal image data to obtain an acoustic-thermal coupling feature set; obtain fault source parameters based on the acoustic-thermal coupling feature set and the relationship model between acoustic emission frequency and temperature rise rate; and determine fault data based on the fault source parameters and the dynamic mapping three-dimensional model.

[0073] Optionally, the real-time initial acoustic signal is subjected to spatiotemporal alignment processing to obtain the target acoustic signal, including: The point cloud data of electrical equipment is filtered to obtain spatial anchor points; Based on the spatial anchor points, the physical coordinates are transformed to obtain the point cloud coordinates of the acoustic sensor; Add timestamp markers to the point cloud coordinates of the acoustic sensor to obtain the target acoustic signal; The real-time initial thermal image data is spatiotemporally aligned to obtain the target thermal image data, including: The point cloud data of electrical equipment is filtered to obtain spatial anchor points; Based on the spatial anchor points, the physical coordinates are transformed to obtain the point cloud coordinates of the thermal imaging unit; Add timestamp markers to the point cloud coordinates of the thermal imaging unit to obtain the target thermal image data.

[0074] Optionally, acoustic-thermal coupling feature fusion is performed on the target acoustic signal and target thermal image data to obtain an acoustic-thermal coupling feature set, including: The target acoustic signal is subjected to noise reduction processing to obtain noise reduction acoustic characteristics; Heat source inversion is performed on the target thermal image data to obtain heat source distribution data; The spatial correlation matrix is ​​obtained based on the point cloud coordinates of the acoustic sensor and the thermal imaging unit; Based on the spatial correlation matrix, the noise reduction features and heat source distribution data are weighted and fused to obtain the acoustic-thermal coupling feature set.

[0075] Optionally, based on the acoustic-thermal coupling feature set and the relationship model between acoustic emission frequency and temperature rise rate, fault source parameters are obtained, including: Based on the aforementioned acoustic-thermal coupling feature set, and combining the heat conduction equation and the acoustic wave equation, a coupled partial differential equation between the acoustic emission frequency and the temperature rise rate is established; By applying regularization constraints to the coupled partial differential equations and combining them with the objective function, the fault source parameters are obtained.

[0076] Optionally, based on the fault source parameters and the dynamically mapped 3D model, fault data is determined, including: Based on the fault source parameters, the three-dimensional coordinates of the fault source are obtained; Time-frequency analysis was performed on the noise reduction characteristics to obtain a time-frequency waterfall plot; Based on the heat source distribution data, a thermal gradient cloud map is obtained; The three-dimensional coordinates, time-frequency waterfall plot, thermal gradient cloud plot, and electrical parameters are superimposed onto the preset three-dimensional model to obtain a dynamically mapped three-dimensional model; Fault data are determined based on the dynamic mapping 3D model.

[0077] Optionally, fault data can be determined based on the dynamically mapped 3D model, including: Based on the fault source parameters, the three-dimensional coordinates, displacement, and temperature rise rate of the fault source are obtained. Based on the three-dimensional coordinates, displacement, and temperature rise rate, and combined with the similarity metric function, the fault similarity value is obtained; Fault data is obtained based on the fault similarity value.

[0078] Optionally, the similarity measurement function is:

[0079] Where Sim represents the similarity between the current fault and the i-th historical fault; x is the current fault source parameter; For the i-th historical fault; This represents the weighted Euclidean distance between the two in the M-dimensional feature space, where M is the parameter dimension. This is the distance attenuation factor; The thermo-mechanical coupling index of the current fault; Let be the thermo-mechanical coupling index of the i-th historical fault.

[0080] It should be noted that all implementation methods in the above method embodiments are applicable to the embodiments of this device and can achieve the same technical effect.

[0081] Embodiments of the present invention also provide a computing device, including: one or more processors; and a storage device for storing one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the electrical equipment fault location method of the present invention. All implementations in the above method embodiments are applicable to the embodiments of this computing device and can achieve the same technical effects.

[0082] Embodiments of the present invention also provide a computer-readable storage medium storing a program that, when executed by a processor, implements the electrical equipment fault location method of the present invention. All implementations in the above method embodiments are applicable to embodiments of this computing device and can achieve the same technical effects.

[0083] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0084] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0085] In the embodiments provided by this invention, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0086] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0087] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0088] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, ROM, RAM, magnetic disks, or optical disks.

[0089] Furthermore, it should be noted that in the apparatus and method of the present invention, it is obvious that the components or steps can be decomposed and / or recombined. These decompositions and / or recombinations should be considered equivalent solutions of the present invention. Moreover, the steps performing the above series of processes can naturally be executed in the order described, but are not necessarily required to be executed in chronological order; some steps can be executed in parallel or independently of each other. Those skilled in the art will understand that all or any step or component of the method and apparatus of the present invention can be implemented in any computing device (including processors, storage media, etc.) or network of computing devices, in hardware, firmware, software, or a combination thereof. This is something that those skilled in the art can achieve by using their basic programming skills after reading the description of the present invention.

[0090] Therefore, the object of the present invention can also be achieved by running a program or a set of programs on any computing device. The computing device can be a known general-purpose device. Therefore, the object of the present invention can also be achieved simply by providing a program product containing program code for implementing the method or apparatus. That is, such a program product also constitutes the present invention, and the storage medium storing such a program product also constitutes the present invention. Obviously, the storage medium can be any known storage medium or any storage medium developed in the future. It should also be noted that in the apparatus and method of the present invention, it is obvious that the components or steps can be decomposed and / or recombined. These decompositions and / or recombinations should be considered equivalent to the present invention. Furthermore, the steps for performing the above series of processes can naturally be performed in the order described, but are not necessarily required to be performed in chronological order. Some steps can be performed in parallel or independently of each other.

[0091] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for locating faults in electrical equipment, characterized in that, include: Acquire real-time initial acoustic signals and real-time initial thermal image data of electrical equipment; The real-time initial acoustic signal is spatiotemporally aligned to obtain the target acoustic signal; The real-time initial thermal image data is spatiotemporally aligned to obtain the target thermal image data; The acoustic signal and thermal image data of the target are fused with acoustic-thermal coupling features to obtain an acoustic-thermal coupling feature set. Based on the aforementioned acoustic-thermal coupling feature set and the relationship model between acoustic emission frequency and temperature rise rate, the fault source parameters are obtained; Based on the fault source parameters and the dynamically mapped 3D model, the fault data is determined.

2. The electrical equipment fault location method according to claim 1, characterized in that, The real-time initial acoustic signal is spatiotemporally aligned to obtain the target acoustic signal, including: The point cloud data of electrical equipment is filtered to obtain spatial anchor points; Based on the spatial anchor points, the physical coordinates are transformed to obtain the point cloud coordinates of the acoustic sensor; Add timestamp markers to the point cloud coordinates of the acoustic sensor to obtain the target acoustic signal; The real-time initial thermal image data is spatiotemporally aligned to obtain the target thermal image data, including: The point cloud data of electrical equipment is filtered to obtain spatial anchor points; Based on the spatial anchor points, the physical coordinates are transformed to obtain the point cloud coordinates of the thermal imaging unit; Add timestamp markers to the point cloud coordinates of the thermal imaging unit to obtain the target thermal image data.

3. The electrical equipment fault location method according to claim 1, characterized in that, The acoustic signal and thermal image data of the target are fused using acoustic-thermal coupling features to obtain an acoustic-thermal coupling feature set, including: The target acoustic signal is subjected to noise reduction processing to obtain noise reduction acoustic characteristics; Heat source inversion is performed on the target thermal image data to obtain heat source distribution data; The spatial correlation matrix is ​​obtained based on the point cloud coordinates of the acoustic sensor and the thermal imaging unit; Based on the spatial correlation matrix, the noise reduction features and heat source distribution data are weighted and fused to obtain the acoustic-thermal coupling feature set.

4. The electrical equipment fault location method according to claim 1, characterized in that, Based on the aforementioned acoustic-thermal coupling feature set and the relationship model between acoustic emission frequency and temperature rise rate, the fault source parameters are obtained, including: Based on the aforementioned acoustic-thermal coupling feature set, and combining the heat conduction equation and the acoustic wave equation, a coupled partial differential equation between the acoustic emission frequency and the temperature rise rate is established; By applying regularization constraints to the coupled partial differential equations and combining them with the objective function, the fault source parameters are obtained.

5. The electrical equipment fault location method according to claim 3, characterized in that, Based on the fault source parameters and the dynamically mapped 3D model, fault data is determined, including: Based on the fault source parameters, the three-dimensional coordinates of the fault source are obtained; Time-frequency analysis was performed on the noise reduction characteristics to obtain a time-frequency waterfall plot; Based on the heat source distribution data, a thermal gradient cloud map is obtained; The three-dimensional coordinates, time-frequency waterfall plot, thermal gradient cloud plot, and electrical parameters are superimposed on the preset three-dimensional model to obtain a dynamically mapped three-dimensional model; Fault data are determined based on the dynamic mapping 3D model.

6. The electrical equipment fault location method according to claim 5, characterized in that, Based on the dynamically mapped 3D model, fault data is determined, including: Based on the fault source parameters, the three-dimensional coordinates, displacement, and temperature rise rate of the fault source are obtained. Based on the three-dimensional coordinates, displacement, and temperature rise rate, and combined with the similarity metric function, the fault similarity value is obtained; Based on the fault similarity value, fault data is obtained.

7. The electrical equipment fault location method according to claim 6, characterized in that, The similarity measurement function is: Where Sim represents the similarity between the current fault and the i-th historical fault; x is the current fault source parameter; For the i-th historical fault; This represents the weighted Euclidean distance between the two in the M-dimensional feature space, where M is the parameter dimension. This is the distance attenuation factor; The thermo-mechanical coupling index of the current fault; Let be the thermo-mechanical coupling index of the i-th historical fault.

8. A fault location device for electrical equipment, characterized in that, include: The acquisition module is used to acquire the real-time initial acoustic signal and the real-time initial thermal image data of the electrical equipment. The processing module is used to perform spatiotemporal alignment processing on the real-time initial acoustic signal to obtain the target acoustic signal; The real-time initial thermal image data is spatiotemporally aligned to obtain the target thermal image data; The acoustic signal and thermal image data of the target are fused by acoustic-thermal coupling features to obtain an acoustic-thermal coupling feature set; based on the acoustic-thermal coupling feature set and the relationship model between acoustic emission frequency and temperature rise rate, the fault source parameters are obtained; based on the fault source parameters and the dynamic mapping three-dimensional model, the fault data is determined.

9. A computing device, characterized in that, include: One or more processors; A storage device for storing one or more programs, which, when executed by one or more processors, cause the one or more processors to implement the method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a program that, when executed by a processor, implements the method as described in any one of claims 1 to 7.

Images