Switch cabinet partial discharge positioning method based on multi-source collaborative modeling and dynamic neighborhood optimization

Abstract

The invention discloses a switch cabinet partial discharge positioning method based on multi-source collaborative modeling and dynamic neighborhood optimization, and relates to the technical field of power equipment insulation state monitoring, and the method comprises the steps: building a three-dimensional geometric model according to the physical structure of a switch cabinet, applying a partial discharge equivalent source at a selected position, and obtaining a partial discharge equivalent source; and cooperative response data of the electromagnetic field, the sound field and the transient ground pressure field are solved, and a switch cabinet cavity physical field cooperative response model is constructed. And acquiring a multi-source data matrix and extracting a multi-dimensional feature vector, inputting the multi-source data matrix and the multi-dimensional feature vector into the collaborative response model, performing inversion by adopting a multi-source data collaborative residual minimum fitting method to acquire a three-dimensional coordinate of a discharge source point, calculating a credible score based on a residual value, and performing dynamic correction in combination with the credible score to generate a final positioning result. The method has the advantages of multi-domain response fusion, accurate initial solution, real-time correction in the positioning process and strong anti-interference capability, and improves the precision and reliability of partial discharge positioning detection.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of power equipment insulation state monitoring, in particular to a switch cabinet partial discharge positioning method based on multi-source collaborative modeling and dynamic neighborhood optimization. BACKGROUND

[0002] Early switch cabinet partial discharge detection relies on ultra-high frequency (UHF) or acoustic emission (AE) single technology, and the positioning accuracy is limited; in recent years, researchers have gradually explored the combination of acoustic, ultra-high frequency and ground voltage and other multi-source signals, and built a cavity physical model by cooperating with finite element simulation to realize the spatial mapping of the discharge source point.

[0003] Although the above-mentioned technology has achieved certain results in improving the discharge recognition rate, there are still significant deficiencies, first, the traditional multi-source data fusion mostly uses weighted average or static fitting, lacks dynamic correction strategy based on error evaluation, and cannot respond to the credibility change of the discharge source point in real time, so the positioning stability is insufficient; second, the existing method usually assumes that the sensor node layout is ideal, ignores the spatial coverage blind area and inconsistency of multi-source response caused by improper layout, and it is difficult to balance the anti-interference ability and spatial resolution. SUMMARY

[0004] In view of the above-mentioned existing problems, the present application is proposed.

[0005] Therefore, the present application provides a switch cabinet partial discharge positioning method based on multi-source collaborative modeling and dynamic neighborhood optimization to solve the problems of insufficient multi-source signal cooperation, lack of dynamic correction ability and sensor node layout blind area in the prior art.

[0006] To solve the above technical problems, the present application provides the following technical scheme: In a first aspect, the present application provides a switch cabinet partial discharge positioning method based on multi-source collaborative modeling and dynamic neighborhood optimization, which comprises: establishing a three-dimensional geometric model of a switch cabinet cavity and a ground network model of the switch cabinet according to the physical structure of the switch cabinet and the ground network of the switch cabinet; Based on the three-dimensional geometric model of the switch cabinet cavity, a partial discharge equivalent source is applied to the selected spatial position, and the collaborative response data of the electromagnetic field, the acoustic field and the transient ground pressure field of the switch cabinet cavity are solved; According to the collaborative response data of the electromagnetic field, the acoustic field and the transient ground pressure field of the switch cabinet cavity, a switch cabinet cavity physical field collaborative response model is established; A multi-source data matrix is obtained, a multi-dimensional feature vector is extracted, the multi-dimensional feature vector is input into the switch cabinet cavity physical field collaborative response model, and the three-dimensional spatial coordinates of the partial discharge source point inside the switch cabinet are obtained by a multi-source data collaborative residual error minimum fitting method, and the credibility score of the three-dimensional spatial coordinates is calculated based on the residual error value; According to the reliable score of the three-dimensional space coordinates, the three-dimensional space coordinates of the partial discharge source point inside the switch cabinet are dynamically corrected to generate a final positioning result.

[0007] As a preferred scheme of the switch cabinet partial discharge positioning method of the multi-source collaborative modeling and dynamic neighborhood optimization of the application, wherein: according to the physical structure of the switch cabinet and the grounding network of the switch cabinet, a three-dimensional geometric model of the switch cabinet cavity and a grounding network model of the switch cabinet are established, and the specific steps are, According to the actual physical structure of the switch cabinet, a three-dimensional geometric model of the switch cabinet cavity is constructed, and the material properties of the switch cabinet cavity are defined, and the material properties of the switch cabinet cavity are assigned to the three-dimensional geometric model; According to the grounding network of the switch cabinet, the nodes and branches of the electrical network are defined, and the parameters of the electrical elements in the electrical network are determined, and the grounding network model of the switch cabinet is constructed.

[0008] As a preferred scheme of the switch cabinet partial discharge positioning method of the multi-source collaborative modeling and dynamic neighborhood optimization of the application, wherein: based on the three-dimensional geometric model of the switch cabinet cavity, the selected spatial position is subjected to partial discharge equivalent source, and the collaborative response data of the electromagnetic field, the sound field and the transient ground pressure field of the switch cabinet cavity are solved, and the specific steps are, In the three-dimensional geometric model after assigning the material properties, the spatial position and discharge parameters of the partial discharge equivalent source are set; According to the spatial position and discharge parameters of the partial discharge equivalent source, the selected spatial position is subjected to partial discharge equivalent source, and the time domain response data of the electromagnetic field, the sound field propagation response data and the transient ground pressure field voltage fluctuation time domain response data under the action of the partial discharge equivalent source are solved; The electromagnetic field time domain response data, the sound field propagation response data and the transient ground pressure field voltage response data are integrated to obtain the collaborative response data of the electromagnetic field, the sound field and the transient ground pressure field of the switch cabinet cavity.

[0009] As a preferred scheme of the switch cabinet partial discharge positioning method of the multi-source collaborative modeling and dynamic neighborhood optimization of the application, wherein: according to the collaborative response data of the electromagnetic field, the sound field and the transient ground pressure field of the switch cabinet cavity, a physical field collaborative response model of the switch cabinet cavity is established, and the specific steps are, According to the collaborative response data of the electromagnetic field, the sound field and the transient ground pressure field of the switch cabinet cavity, the multi-source response characteristics of different spatial sensing nodes in the switch cabinet cavity are extracted; The extracted multi-source response characteristics and the spatial coordinate information of the partial discharge equivalent source points in the three-dimensional geometric model of the switch cabinet cavity are correspondingly related to form a mapping data sample set of the discharge source points and the multi-source response, the mathematical mapping relationship between the discharge source points and the multi-source response in the mapping data sample set is fitted, and the physical field collaborative response model of the switch cabinet cavity is constructed.

[0010] As a preferred scheme of the switch cabinet partial discharge positioning method of the multi-source collaborative modeling and dynamic neighborhood optimization of the application, wherein: the multi-source response characteristics include electromagnetic field response characteristics, sound field response characteristics and transient ground pressure field response characteristics.

[0011] As a preferred scheme of the switch cabinet partial discharge positioning method of the multi-source collaborative modeling and dynamic neighborhood optimization of the application, wherein: the specific steps of obtaining the multi-source data matrix and extracting the multi-dimensional feature vector are, Based on the switch cabinet cavity physical field collaborative response model, determine the spatial coordinate position suitable for arranging the multi-source sensing node inside the switch cabinet; According to the spatial coordinate position suitable for arranging the multi-source sensing node inside the switch cabinet, arrange the multi-source sensing node inside the switch cabinet; Based on the multi-source sensing node arranged inside the switch cabinet, collect the electromagnetic field signal, acoustic signal and transient ground pressure field signal of each multi-source sensing node; According to the electromagnetic field signal, acoustic signal and transient ground pressure field signal of the multi-source sensing node, merge and arrange according to the unified time axis and spatial coordinate to form the multi-source data matrix; According to the multi-source data matrix, extract the multi-dimensional feature vector of each multi-source sensing node.

[0012] As a preferred scheme of the switch cabinet partial discharge positioning method of the multi-source collaborative modeling and dynamic neighborhood optimization of the application, wherein: the multi-source sensing node includes an electric field sensing node, a magnetic field sensing node, an acoustic sensing node and a transient ground pressure field sensing node.

[0013] As a preferred scheme of the switch cabinet partial discharge positioning method of the multi-source collaborative modeling and dynamic neighborhood optimization of the application, wherein: the multi-dimensional feature vector is input into the switch cabinet cavity physical field collaborative response model, and the three-dimensional spatial coordinates of the local discharge source point inside the switch cabinet are obtained by the multi-source data collaborative residual error minimum fitting method, and the credibility score of the three-dimensional spatial coordinates is calculated based on the residual value, and the specific steps are, Input the multi-dimensional feature vector into the switch cabinet cavity physical field collaborative response model, calculate the theoretical characteristics of the multi-source response under different three-dimensional spatial coordinates; According to the comparison between the theoretical characteristics of the multi-source response under each three-dimensional spatial coordinate and the multi-dimensional feature vector, the three-dimensional spatial coordinates of the local discharge source corresponding to the minimum residual value are solved by the multi-source data collaborative residual error minimum fitting method; Based on the minimum residual value, calculate the credibility score of the three-dimensional spatial coordinates of the local discharge source point inside the switch cabinet according to the residual score function.

[0014] As a preferred scheme of the switch cabinet partial discharge positioning method of the multi-source collaborative modeling and dynamic neighborhood optimization of the application, wherein: the dynamic correction of the three-dimensional space coordinates of the partial discharge source point in the switch cabinet is based on the reliable score of the three-dimensional space coordinates, and a final positioning result is generated, and the specific steps are, The reliable score of the three-dimensional space coordinates of the partial discharge source point in the switch cabinet is compared with the reliable score threshold value. If the reliable score of the three-dimensional space coordinates is lower than the reliable score threshold value, dynamic correction is performed to generate a set of alternative coordinate points. According to the multi-source response theoretical characteristics of each alternative coordinate point, the multi-source response theoretical characteristics of each alternative coordinate point are recalculated by using the switch cabinet cavity physical field collaborative response model, and the multi-source response theoretical characteristics of each alternative coordinate point are obtained. According to the multi-source response theoretical characteristics of each alternative coordinate point and the multi-dimensional feature vector, a residual error is calculated, and the alternative coordinate point with the smallest residual error is selected as the optimal neighborhood coordinate point. According to the three-dimensional space coordinates of the partial discharge source point in the switch cabinet, the optimal neighborhood coordinate point, and the reliable score of the three-dimensional space coordinates of the partial discharge source point in the switch cabinet, a residual error convergence factor is calculated, and a nonlinear fusion function containing the residual error convergence factor is constructed. The final positioning result is generated by the nonlinear fusion function containing the residual error convergence factor.

[0015] As a preferred scheme of the switch cabinet partial discharge positioning method of the multi-source collaborative modeling and dynamic neighborhood optimization of the application, wherein: the dynamic correction refers to constructing a dynamic correction search neighborhood based on the three-dimensional space coordinates of the partial discharge source point in the switch cabinet, and generating a set of alternative coordinate points in the dynamic correction search field through neighborhood search.

[0016] The application has the following beneficial effects: by establishing a switch cabinet cavity physical field collaborative response model, using the multi-source response characteristics of electromagnetic fields, acoustic fields and transient ground pressure fields, the theoretical response of the discharge source point based on multi-domain data is quickly deduced, the accuracy and calculation efficiency of the positioning initial solution are improved; through the dynamic correction search neighborhood mechanism, the positioning deviation caused by interference signals is effectively suppressed; through sensitivity analysis and optimization selection of sensor layout positions, real-time monitoring requirements in the operation of power equipment can be met. BRIEF DESCRIPTION OF DRAWINGS

[0017] In order to more clearly illustrate the technical solutions of the embodiments of the application, the following will briefly introduce the drawings needed to be used in the embodiment description. Obviously, the drawings in the following description are only some embodiments of the application, and other drawings can be obtained by those skilled in the art without creating laborious work.

[0018] Figure 1Flow chart for partial discharge positioning method of switch cabinet based on multi-source collaborative modeling and dynamic neighborhood optimization.

[0019] Figure 2 Flow chart for switch cabinet cavity physical field collaborative response modeling.

[0020] Figure 3 Flow chart for multi-source data collaborative residual minimum fitting and credible score calculation.

[0021] Figure 4 Flow chart for dynamic correction search neighborhood.

[0022] Figure 5 Computer device diagram DETAILED DESCRIPTION In order to make the above objectives, features and advantages of the present application more apparent, specific embodiments of the present application will be described in detail below with reference to the accompanying drawings.

[0023] In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present application. However, it will be apparent to one skilled in the art that the present application can be practiced without the specific details given herein, that the present application can be practiced with other than the described embodiments, and that the present application can be practiced with different or additional components, elements, acts, or steps. Thus, the present application is not limited to the embodiments described herein but is instead broad in scope.

[0024] Secondly, the "one embodiment" or "embodiment" referred to herein means that the specific features, structures or characteristics can be included in at least one implementation of the present application. "In one embodiment" appearing in different places in the specification does not mean the same embodiment, nor is it an independent or alternative embodiment that excludes other embodiments.

[0025] Reference Signs List Figures 1-5 For one embodiment of the present application, the embodiment provides a partial discharge positioning method of switch cabinet based on multi-source collaborative modeling and dynamic neighborhood optimization, comprising the following steps: S1: According to the physical structure of the switch cabinet and the grounding network of the switch cabinet, a three-dimensional geometric model of the switch cabinet cavity and a grounding network model of the switch cabinet are established.

[0026] According to the actual physical structure of the switch cabinet, a three-dimensional geometric model of the switch cabinet cavity is constructed by a finite element simulation software, and the material properties of the switch cabinet cavity are defined. The material properties of the switch cabinet cavity are assigned to the three-dimensional geometric model.

[0027] Further, the three-dimensional geometric model of the switch cabinet cavity usually includes three-dimensional structures such as metal cabinet, internal busbar, wire, insulating part, door body, observation window, etc.

[0028] According to the physical structure of the switch cabinet, the types and attribute parameters of the materials constituting the switch cabinet cavity are determined, wherein the material types of the switch cabinet generally include steel plate material parts, busbar copper material parts, and internal insulation parts.

[0029] The material attribute parameters include electrical conductivity, magnetic permeability, density, elastic model, and acoustic impedance.

[0030] After determining the types and attribute parameters of the materials constituting the switch cabinet cavity, in the three-dimensional geometric model of the switch cabinet cavity, different structure regions are divided according to the actual materials of the switch cabinet, and the material attributes of the actual physical structure of the switch cabinet are assigned, to ensure that accurate material attributes are used for calculation in electromagnetic field simulation, acoustic simulation, and transient ground pressure field simulation.

[0031] According to the grounding network of the switch cabinet, the nodes and branches of the electrical network are defined, and the parameters of the electrical elements in the electrical network are determined, to construct a grounding network model of the switch cabinet.

[0032] Further, key grounding nodes in the grounding network structure of the switch cabinet, such as grounding terminals, cabinet welding points, and grounding buses, are selected as nodes of the electrical network, and the conductive paths are divided according to the actual connection mode, and each path is built as a branch, wherein each branch contains series or parallel RLC electrical elements, and the electrical parameters of the branch electrical elements are obtained, wherein the electrical element parameters include resistance, inductance, and capacitance, etc.; each path from the discharge source to the ground is constructed by a finite element simulation software, wherein each ground voltage sampling point establishes an independent branch, forming an equivalent transfer network of the switch cabinet discharge current injection ground potential response, i.e. the grounding network model of the switch cabinet.

[0033] S2: Based on the three-dimensional geometric model of the switch cabinet cavity, the equivalent source of local discharge is applied to the selected spatial position, and the collaborative response data of the electromagnetic field, acoustic field, and transient ground pressure field of the switch cabinet cavity are solved.

[0034] In the three-dimensional geometric model after assigning the material attributes, the spatial position and discharge parameters of the equivalent source of local discharge are set.

[0035] Further, according to the three-dimensional spatial coordinate information in the three-dimensional geometric model of the switch cabinet cavity, a typical local discharge prone area is selected as the spatial position of the equivalent source of local discharge; wherein the determination of the typical local discharge prone area is specifically, according to the historical failure data of the switch cabinet, combined with engineering experience to determine the high-frequency area of local discharge, such as the end of the busbar, the short-circuit connector, the insulation sleeve connection area, etc.

[0036] The discharge parameters of the equivalent source of local discharge are set at the selected spatial position, wherein the discharge parameters include equivalent discharge energy, equivalent discharge frequency, discharge waveform, and discharge duration.

[0037] For example, in the three-dimensional geometric model of the switch cabinet cavity, a spatial position (0.52m, 0.28m, 1.36m) located in the corner area of the busbar is selected as the loading point of the partial discharge equivalent source, and the discharge parameters are set as follows: the equivalent discharge energy is 20 microjoules, the equivalent discharge efficiency is set to 50 kHz considering the repetition frequency characteristics of partial discharge in a typical industrial environment, a double exponential pulse waveform is used as the discharge waveform, and the discharge duration is set to 500 nanoseconds.

[0038] According to the spatial position of the partial discharge equivalent source and the discharge parameters, the partial discharge equivalent source is applied to the selected spatial position, and the time-domain response data of the electromagnetic field, the sound field propagation response data, and the transient ground pressure field voltage fluctuation time-domain response data of the switch cabinet cavity under the action of the partial discharge equivalent source are solved.

[0039] Further, in the three-dimensional geometric model of the switch cabinet cavity, each sensing node is set, and the discharge parameters of the partial discharge equivalent source are applied to the selected spatial position coordinates of the switch cabinet cavity in the form of transient current excitation for electromagnetic field, sound field, and transient ground pressure field simulation.

[0040] By solving the Maxwell equations using the finite element method, the time-varying electromagnetic field time-domain response data at each sensing node is obtained, forming an electromagnetic field time-domain response data set.

[0041] By solving the sound wave propagation equation using the finite element method, the propagation process of the sound wave excited by the partial discharge equivalent source in the switch cabinet cavity is calculated, the sound pressure curve at each simulation node is obtained, and a sound field propagation response data set is formed.

[0042] By arranging each voltage sensing node in the grounding network, the transient ground voltage curve over time is obtained, forming a transient ground pressure field voltage fluctuation time-domain response data set.

[0043] It should be noted that the response data obtained by the electromagnetic field, sound field, and transient ground pressure field in the partial discharge detection are all time-varying fluctuation curve data.

[0044] The electromagnetic field time-domain response data, sound field propagation response data, and transient ground pressure field response data are integrated to obtain the electromagnetic field, sound field, and transient ground pressure field collaborative response data of the switch cabinet cavity.

[0045] Further, the electromagnetic field time-domain response data, sound field propagation response data, and transient ground pressure field response data of the switch cabinet cavity are uniformly integrated according to the node coordinates and time axis, forming the electromagnetic field, sound field, and transient ground pressure field collaborative response data of the switch cabinet cavity.

[0046] S3: Establish a switch cabinet cavity physical field collaborative response model according to the electromagnetic field, acoustic field, and transient ground pressure field collaborative response data of the switch cabinet cavity.

[0047] According to the electromagnetic field, acoustic field, and transient ground pressure field collaborative response data of the switch cabinet cavity, the multi-source response characteristics of different spatial sensing nodes in the switch cabinet cavity are extracted.

[0048] Further, the electromagnetic field, acoustic field, and transient ground pressure field collaborative response data of the switch cabinet cavity are normalized to obtain the normalized electromagnetic field, acoustic field, and transient ground pressure field collaborative response data of the switch cabinet cavity.

[0049] The multi-source response characteristics of different spatial sensing nodes in the switch cabinet cavity are extracted from the normalized electromagnetic field, acoustic field, and transient ground pressure field collaborative response data, wherein the multi-source response characteristics include electromagnetic field response characteristics, acoustic field response characteristics, and transient ground pressure field response characteristics.

[0050] Further, the electromagnetic field, acoustic field, and transient ground pressure field collaborative response data are respectively subjected to fast Fourier transform to extract the main frequency characteristics of each electromagnetic field, acoustic field, and transient ground pressure field.

[0051] The electromagnetic field, acoustic field, and transient ground pressure field collaborative response data are subjected to Fourier transform to calculate the spectral energy center frequency of each electromagnetic field, acoustic field, and transient ground pressure field.

[0052] The electromagnetic field, acoustic field, and transient ground pressure field collaborative response data are respectively subjected to Hilbert transform to extract the time-domain envelope of each electromagnetic field, acoustic field, and transient ground pressure field. The time-domain envelope is fitted by least squares method to solve the attenuation coefficient of each electromagnetic field, acoustic field, and transient ground pressure field.

[0053] The maximum amplitude in the electromagnetic field time-domain response data, acoustic field propagation response data, and transient ground pressure field response data of each sensing node is taken as the spatial distribution characteristics of each electromagnetic field, acoustic field, and transient ground pressure field.

[0054] The main frequency characteristics, spectral energy center frequency, attenuation coefficient, and maximum amplitude of the electromagnetic field, acoustic field, and transient ground pressure field are output as electromagnetic field response characteristics, acoustic field response characteristics, and transient ground pressure field response characteristics.

[0055] According to the multi-source response characteristics of different sensing nodes in the switch cabinet cavity, a switch cabinet cavity physical field collaborative response model is established.

[0056] Furthermore, the three-dimensional spatial coordinates of each sensing node are mapped one-to-one with the corresponding multi-source response features to form a mapping data sample set of discharge point and multi-source response. By fitting the mathematical mapping relationship between discharge point and multi-source response in the mapping data sample set, a physical field collaborative response model of the switch cabinet cavity is constructed.

[0057] It should be noted that the physical field cooperative response model of the switch cabinet cavity is a digital model. The input is the three-dimensional spatial coordinates of the partial discharge power source inside the switch cabinet cavity, and the output is the multi-source response characteristics of the corresponding three-dimensional spatial coordinates. Among them, the three-dimensional spatial coordinates of the partial discharge power source inside the switch cabinet cavity and the multi-source response characteristics of the corresponding three-dimensional spatial coordinates establish a functional numerical mapping relationship between spatial position and response characteristics.

[0058] The mathematical mapping relationship between the discharge point and the multi-source response in the physical field cooperative response model of the switchgear cavity is expressed as follows: ; in, Indicates the first The theoretical maximum amplitude of the electric field at each sensing node. Indicates the first The theoretical maximum amplitude of the magnetic field at each sensing node Indicates the first The theoretical maximum amplitude of the sound field at each sensor node. Indicates the first The theoretical maximum amplitude of the ground voltage response of each sensing node. Represents the electric field propagation proportionality coefficient. This represents the proportionality coefficient of magnetic field propagation. This represents the sound field propagation scaling factor. This represents the proportionality coefficient of the ground voltage response propagation. This represents the discharge energy of the equivalent source of partial discharge. Indicates the direction of discharge relative to the first The angle between the observation directions of each sensor node, Indicates the equivalent source of partial discharge and the first Spatial distance between sensor nodes Represents the relative permittivity. This represents the dominant frequency component excited by the equivalent source of partial discharge. Represents a time variable. This indicates the phase shift angle of the electric field propagation. Indicates the phase shift angle of magnetic field propagation. Indicates the phase shift angle of sound field propagation. Indicates the phase shift angle of the ground voltage response. This represents the energy attenuation coefficient of the medium during electric field propagation. This represents the medium absorption attenuation factor during magnetic field propagation. represents the attenuation coefficient of the sound wave medium absorption, represents the spatial propagation loss coefficient, represents the acoustic impedance of the propagation medium, represents the equivalent grounding resistance of the partial discharge equivalent source to the first sensor node, represents the equivalent grounding capacitance on the grounding network path.

[0059] It should be noted that, , , , The value range of is [0, 1], and =1.

[0060] The switch cabinet cavity physical field cooperative response model solves , , , , , , , by fitting simulation data, forming a multi-source physical field cooperative response database.

[0061] The maximum amplitude of the theoretical electric field, magnetic field, acoustic field and transient ground pressure field response of each sensor node refers to the maximum time domain amplitude reached at the sensor node position, which can be determined by the peak value in the curve waveform of each physical field domain with time.

[0062] The propagation phase shift angle of each physical field domain refers to the phase change angle of the local discharge source signal on the characteristic main frequency component, which can be determined by the main frequency component.

[0063] Further, the spatial distance between the local discharge source point and the first sensor node is calculated as: ; wherein, , , represents the three-dimensional space coordinates of the local discharge source point in the switch cabinet cavity, , , represents the three-dimensional space coordinates of the first sensor node in the switch cabinet cavity.

[0064] S4: Obtain a multi-source data matrix, extract a multi-dimensional feature vector, input the multi-dimensional feature vector into a switch cabinet cavity physical field collaborative response model, and obtain the three-dimensional space coordinates of the partial discharge source point inside the switch cabinet through a multi-source data collaborative residual error minimum fitting method, and calculate the credibility score of the three-dimensional space coordinates based on the residual error value.

[0065] Based on the switch cabinet cavity physical field collaborative response model, the spatial coordinate position suitable for arranging the multi-source sensing node inside the switch cabinet is determined.

[0066] Further, the shell wall and internal structure of the switch cabinet cavity are discretized in three-dimensional space grid to form a candidate sensing node coordinate set, wherein each coordinate point represents a space position where a sensor can be arranged. Assuming the three-dimensional space coordinates of the discharge source point, and taking each assumed discharge source point as an excitation source, the multi-source response characteristics of the electromagnetic field, acoustic field, and transient ground pressure field of each candidate sensing node are calculated according to the switch cabinet cavity physical field collaborative response model, to obtain the multi-source response characteristics of the electromagnetic field, acoustic field, and transient ground pressure field corresponding to each candidate sensing node. According to each candidate sensing node and the multi-source response characteristics of the electromagnetic field, acoustic field, and transient ground pressure field corresponding to each candidate sensing node, a sensitivity Jacobian matrix is constructed. According to the sensitivity Jacobian matrix, a sensing arrangement point that satisfies the maximum Fisher information optimization formula is selected from the candidate sensing node coordinate set to form a multi-source sensing node arrangement subset.

[0067] According to the spatial coordinate position suitable for arranging the multi-source sensing node inside the switch cabinet, the multi-source sensing node is arranged inside the switch cabinet.

[0068] Further, based on the multi-source sensing node arrangement subset, a multi-source sensor is installed inside or on the surface of the switch cabinet cavity. An electric field sensor, a magnetic field sensor, and an acoustic sensor are arranged at each multi-source sensing node coordinate, and a low-voltage sensor is arranged in the switch cabinet node network to form a completed multi-source sensing node set. Subsequently, according to a high-precision clock synchronization protocol, all completed multi-source sensing nodes are time-synchronized and calibrated.

[0069] Based on the completed multi-source sensing node inside the switch cabinet, synchronous data acquisition is performed to acquire electromagnetic field signals, acoustic signals, and transient ground pressure field signals of each multi-source sensing node.

[0070] According to the electromagnetic field signals, acoustic signals, and transient ground pressure field signals of the multi-source sensing node, the signals are merged and sorted according to a unified time axis and spatial coordinates to form a multi-source data matrix.

[0071] Further, after the switch cabinet triggers a partial discharge event, each multi-source sensing node collects electromagnetic field signals, acoustic signals and transient ground pressure field signals at a unified sampling frequency, and arranges them in order according to timestamps to form a multi-source data matrix indexed according to the multi-source sensing node number.

[0072] According to the multi-source data matrix, a multi-dimensional feature vector of each multi-source sensing node is extracted.

[0073] Based on the formed multi-source data matrix, the electromagnetic field signals, acoustic signals and transient ground pressure field signals of each multi-source sensing node are respectively subjected to feature extraction operations.

[0074] The feature extraction operation specifically includes calculating the maximum amplitude, dominant frequency feature, spectral energy barycenter frequency, attenuation coefficient and spatial distribution feature of each signal, and constructing a multi-dimensional feature vector of each multi-source sensing node.

[0075] The multi-dimensional feature vector is input into a switch cabinet cavity physical field cooperative response model, and theoretical features of multi-source responses under different three-dimensional spatial coordinates are calculated based on the switch cabinet cavity physical field cooperative response model.

[0076] Further, the multi-dimensional feature vector is input into the switch cabinet cavity physical field cooperative response model, and the mathematical mapping relationship between the three-dimensional spatial coordinates of the discharge source point and the multi-source response features in the switch cabinet cavity physical field cooperative response model is used to output the theoretical multi-source response features under different assumed three-dimensional spatial coordinates.

[0077] According to the comparison between the theoretical multi-source response features under the three-dimensional spatial coordinates of each discharge source point and the multi-dimensional feature vector, the three-dimensional spatial coordinates of the local discharge source corresponding to the minimum residual error are solved by a multi-source data cooperative residual error minimum fitting method.

[0078] Further, based on the switch cabinet cavity physical field cooperative response model, a residual error objective function is constructed, and the fitting residual error values of the multi-dimensional feature vectors actually obtained by each multi-source sensing node and the theoretical multi-source response features calculated by the switch cabinet cavity physical field cooperative response model are used as evaluation indexes, wherein the constructed residual error objective function is represented as: ; Wherein, represents the fitting residual error value of the local discharge source point at the three-dimensional spatial coordinates, represents the total number of multi-source sensing nodes, represents the maximum amplitude of the electric field signal actually measured by the i th sensing node, represents the maximum amplitude of the magnetic field signal actually measured by the i th sensing node, represents the maximum amplitude of the acoustic signal actually measured by the i th sensing node, represents the maximum amplitude of the transient ground pressure field signal actually measured by the i th sensing node, represents the maximum amplitude of the transient ground pressure field signal actually measured by the i th sensing node, The maximum amplitude of the acoustic signal measured at each sensing node. Indicates the first The maximum amplitude of the transient ground pressure signal measured at each sensing node.

[0079] The quasi-Newton method is used to optimize the fitting residual value by minimizing the residual value. The minimum residual value is obtained, and the three-dimensional spatial coordinates of the local discharge source corresponding to the minimum residual value are obtained by inversion. The three-dimensional spatial coordinates corresponding to the minimum residual value are used as the three-dimensional coordinates of the local discharge source point.

[0080] The minimum residual value is obtained by using the multi-source data collaborative residual minimum fitting method, and the credibility score of the three-dimensional spatial coordinates of the local discharge point inside the switch cabinet is calculated based on the residual scoring function.

[0081] Furthermore, based on the solved minimum residual value, the reliability score of the three-dimensional spatial coordinates of the partial discharge point inside the switchgear is calculated using the residual scoring function. The formula for the residual scoring function is as follows: ; in, A reliable score representing three-dimensional spatial coordinates. This represents the residual value obtained by solving the residual objective function. This represents an empirical coefficient, with a value range of []. ].

[0082] It should be noted that, Less than Normally, this is not meaningful; a value greater than 1 can lead to an excessively large search step size or abnormally amplify the fitting error.

[0083] S5: Based on the reliability score of the three-dimensional spatial coordinates, the three-dimensional spatial coordinates of the local discharge point inside the switch cabinet are dynamically corrected to generate the final positioning result.

[0084] The credibility score of the three-dimensional spatial coordinates of the local discharge point inside the switch cabinet is compared with the credibility score threshold.

[0085] Furthermore, based on the reliable scoring data of the three-dimensional spatial coordinates of historical partial discharge points inside the switchgear, a reliable scoring threshold is set. The reliability score threshold has the same calculation range as the reliability score of the partial discharge point in the three-dimensional spatial coordinates inside the switchgear, and its value range is []. [1] ; Reliability score of the three-dimensional spatial coordinates of the local discharge point inside the switchgear. With the set credibility scoring threshold Compare; if ≥ If the three-dimensional space coordinate of the partial discharge source point in the switch cabinet is considered to be in the three-dimensional space coordinate inside the switch cabinet, the three-dimensional space coordinate is directly output as the final positioning result without correction.

[0086] If the three-dimensional space coordinate of the partial discharge source point in the switch cabinet is considered to be in the three-dimensional space coordinate inside the switch cabinet, the three-dimensional space coordinate is directly output as the final positioning result without correction.

[0087] Further, if < , a space neighborhood is constructed around the three-dimensional space coordinate of the partial discharge source point in the switch cabinet, wherein the neighborhood search radius is set according to the actual space inside the switch cabinet, and the value range is , ], the neighborhood search radius under the current confidence score is calculated as follows: ; wherein, represents the neighborhood search radius under the current confidence score, represents the minimum neighborhood search radius allowed when =1, represents the maximum neighborhood search radius allowed when =0.

[0088] The three-dimensional space coordinates of the multi-source sensing nodes searched in the neighborhood are taken as the candidate coordinate points in the neighborhood, forming a candidate coordinate point set.

[0089] According to the candidate coordinate point set, the multi-source response characteristics of each candidate coordinate point are recalculated using the switch cabinet cavity physical field cooperative response model, and the multi-source response theoretical characteristics of each candidate coordinate point are obtained.

[0090] According to the multi-source response theoretical characteristics of each candidate coordinate point and the multi-dimensional feature vector, the residual error is calculated, and the candidate coordinate point with the smallest residual error is selected as the optimal coordinate point in the neighborhood .

[0091] According to the three-dimensional space coordinate of the partial discharge source point in the switch cabinet and the optimal coordinate point in the neighborhood , the residual error convergence factor is calculated, and a nonlinear fusion function containing the residual error convergence factor is constructed.

[0092] Further, the residual error convergence factor formula is calculated as follows: ; wherein, represents the residual error convergence factor, and the value range is , a fitting residual value representing a partial discharge source point at a neighborhood optimal coordinate point, a minimum constant for preventing the denominator from being zero.

[0093] wherein, when the neighborhood optimal coordinate point is better, > 0, otherwise negative, indicating that the three-dimensional space coordinate of the partial discharge source point inside the switch cabinet is more reliable.

[0094] Further, a hyperbolic tangent function is used to construct a nonlinear fusion function containing a residual convergence factor, represented as: ; wherein, a nonlinear fusion coefficient between the three-dimensional space coordinate of the partial discharge source point inside the switch cabinet and the field coordinate.

[0095] Through the nonlinear fusion function containing the residual convergence factor, the final positioning result is generated.

[0096] Further, a weighted fusion calculation formula for generating the final positioning result is represented as: ; wherein, , , a three-dimensional space coordinate of the final corrected partial discharge source point inside the switch cabinet cavity, , , a three-dimensional space coordinate of the partial discharge source point inside the switch cabinet, , , a three-dimensional space coordinate of the neighborhood optimal coordinate point.

[0097] The embodiment also provides a computer device suitable for the switch cabinet partial discharge positioning method of multi-source collaborative modeling and dynamic neighborhood optimization, comprising: a memory and a processor; the memory is used to store computer executable instructions, and the processor is used to execute the computer executable instructions to realize the switch cabinet partial discharge positioning method of multi-source collaborative modeling and dynamic neighborhood optimization proposed in the above embodiment.

[0098] The computer device can be a terminal, and the computer device includes a processor, a memory, a communication interface, a display screen and an input device connected by a system bus. The processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for running the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is configured to perform wired or wireless communication with an external terminal. The wireless communication can be achieved by WIFI, an operator network, NFC (Near Field Communication) or other technologies. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen. The input device of the computer device can be a touch layer overlaid on the display screen, or a key, trackball or touchpad arranged on the shell of the computer device, or an external keyboard, touchpad or mouse, etc.

[0099] The embodiment also provides a storage medium having a computer program stored thereon, the program being executed by a processor to implement the switch cabinet partial discharge positioning method for realizing multi-source collaborative modeling and dynamic neighborhood optimization as proposed in the above embodiment; and the storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as a static random access memory (SRAM), an electrically erasable programmable read-only memory (EEPROM), an erasable programmable read-only memory (EPROM), a programmable read-only memory (PROM), a read-only memory (ROM), a magnetic memory, a flash memory, a magnetic disk or an optical disk.

[0100] To sum up, the application achieves the rapid deduction of the theoretical response of the discharge source point based on multi-domain data by establishing a switch cabinet cavity physical field collaborative response model and using the multi-source response characteristics of the electromagnetic field, the acoustic field and the transient ground pressure field, thereby improving the accuracy and the calculation efficiency of the positioning initial solution; the positioning deviation caused by the interference signal is effectively inhibited through the dynamic correction search neighborhood mechanism; the sensor layout position is optimized through the sensitivity analysis, and the real-time monitoring demand in the operation of the power equipment can be met.

[0101] It should be noted that the above examples are only used to illustrate the technical solutions of the present application but not limit the present application. Although the present application is described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or equivalently replaced, without departing from the spirit and scope of the technical solutions of the present application, which should be covered in the scope of the claims of the present application.

Claims

1. A switch cabinet partial discharge positioning method based on multi-source collaborative modeling and dynamic neighborhood optimization, characterized in that: The application relates to a local discharge source positioning method for a switch cabinet. According to the physical structure of the switch cabinet and the grounding network of the switch cabinet, a three-dimensional geometric model of a switch cabinet cavity and a grounding network model of the switch cabinet are established; Based on the three-dimensional geometric model of the switch cabinet cavity, local discharge equivalent sources are applied to selected spatial positions, and electromagnetic field, sound field and transient ground pressure field collaborative response data of the switch cabinet cavity are solved; According to the electromagnetic field, sound field and transient ground pressure field collaborative response data of the switch cabinet cavity, a switch cabinet cavity physical field collaborative response model is established; A multi-source data matrix is obtained, a multi-dimensional feature vector is extracted, the multi-dimensional feature vector is input into the switch cabinet cavity physical field collaborative response model, and three-dimensional spatial coordinates of a local discharge source point in the switch cabinet are obtained through a multi-source data collaborative residual minimum fitting method, and a credibility score of the three-dimensional spatial coordinates is calculated based on a residual value; According to the credibility score of the three-dimensional spatial coordinates, the three-dimensional spatial coordinates of the local discharge source point in the switch cabinet are dynamically corrected, and a final positioning result is generated.

2. The multi-source collaborative modeling and dynamic neighborhood optimization switchgear partial discharge positioning method of claim 1, wherein: The three-dimensional geometric model of the switch cabinet cavity is constructed according to the actual physical structure of the switch cabinet, and the material properties of the switch cabinet cavity are defined. According to the grounding network of the switch cabinet, the nodes and branches of the electrical network are defined, and the parameters of the electrical elements in the electrical network are determined, and the grounding network model of the switch cabinet is constructed. In the three-dimensional geometric model with the material properties, the spatial positions and discharge parameters of the local discharge equivalent sources are set.

3. The multi-source collaborative modeling and dynamic neighborhood optimization switchgear partial discharge positioning method of claim 2, wherein: According to the spatial positions and discharge parameters of the local discharge equivalent sources, the local discharge equivalent sources are applied to the selected spatial positions, and electromagnetic field time domain response data, sound field propagation response data and transient ground pressure field voltage fluctuation time domain response data generated by the local discharge equivalent sources are solved. The electromagnetic field time domain response data, sound field propagation response data and transient ground pressure field voltage response data are integrated, and the electromagnetic field, sound field and transient ground pressure field collaborative response data of the switch cabinet cavity are obtained. According to the electromagnetic field, sound field and transient ground pressure field collaborative response data of the switch cabinet cavity, the multi-source response characteristics of different spatial sensing nodes in the switch cabinet cavity are extracted. The extracted multi-source response characteristics and the spatial coordinate information of the local discharge equivalent source points in the three-dimensional geometric model of the switch cabinet cavity are correspondingly related, mapping data sample sets of the discharge source points and the multi-source responses are formed, a mathematical mapping relationship between the discharge source points and the multi-source responses in the mapping data sample sets is fitted, and the switch cabinet cavity physical field collaborative response model is constructed.

4. The multi-source collaborative modeling and dynamic neighborhood optimization switchgear partial discharge positioning method of claim 3, wherein: The multi-source response characteristics include electromagnetic field response characteristics, sound field response characteristics and transient ground pressure field response characteristics. ​ ​ 5. The multi-source collaborative modeling and dynamic neighborhood optimization switchgear partial discharge positioning method of claim 4, wherein: ​ 6. The multi-source co-modeling and dynamic neighborhood optimization switchgear partial discharge locating method of claim 5, wherein: The method comprises the following steps of: Based on the switch cabinet cavity physical field collaborative response model, the space coordinate position suitable for arranging the multi-source sensing node inside the switch cabinet is determined. According to the space coordinate position suitable for arranging the multi-source sensing node inside the switch cabinet, the multi-source sensing node is arranged inside the switch cabinet. Based on the multi-source sensing node arranged inside the switch cabinet, the electromagnetic field signal, the acoustic signal and the transient ground pressure field signal of each multi-source sensing node are collected. According to the electromagnetic field signal, the acoustic signal and the transient ground pressure field signal of the multi-source sensing node, the multi-source data matrix is formed by merging and arranging according to the unified time axis and the space coordinate. According to the multi-source data matrix, the multi-dimensional feature vector of each multi-source sensing node is extracted.

7. The multi-source co-modeling and dynamic neighborhood optimization switchgear partial discharge locating method of claim 6, wherein: The multi-source sensing node comprises an electric field sensing node, a magnetic field sensing node, an acoustic sensing node and a transient ground pressure field sensing node.

8. The multi-source co-modeling and dynamic neighborhood optimization switchgear partial discharge locating method of claim 7, wherein: The method comprises the following steps of: The multi-dimensional feature vector is input into the switch cabinet cavity physical field collaborative response model, and the three-dimensional space coordinate of the partial discharge source point inside the switch cabinet is obtained by the multi-source data collaborative residual error minimum fitting method. The multi-dimensional feature vector is input into the switch cabinet cavity physical field collaborative response model, and the theoretical feature of the multi-source response under different three-dimensional space coordinates is calculated. The three-dimensional space coordinate of the partial discharge source is solved by the multi-source data collaborative residual error minimum fitting method according to the comparison between the theoretical feature of the multi-source response under each three-dimensional space coordinate and the multi-dimensional feature vector.

9. The multi-source co-modeling and dynamic neighborhood optimization switchgear partial discharge locating method of claim 8, wherein: Based on the minimum residual error value, the credible score of the three-dimensional space coordinate of the partial discharge source point inside the switch cabinet is calculated according to the residual error score function. The three-dimensional space coordinate of the partial discharge source point inside the switch cabinet is dynamically corrected according to the credible score of the three-dimensional space coordinate, and the final positioning result is generated. The credible score of the three-dimensional space coordinate of the partial discharge source point inside the switch cabinet is compared with the credible score threshold. If the credible score of the three-dimensional space coordinate is lower than the credible score threshold, the dynamic correction is performed to generate a set of alternative coordinate points. According to the set of alternative coordinate points, the multi-source response theoretical feature of each alternative coordinate point is recalculated by using the switch cabinet cavity physical field collaborative response model, and the multi-source response theoretical feature of each alternative coordinate point is obtained. According to the multi-source response theoretical feature of each alternative coordinate point and the multi-dimensional feature vector, the residual error is calculated, and the alternative coordinate point with the minimum residual error is selected as the optimal coordinate point in the neighborhood. According to the three-dimensional space coordinate of the partial discharge source point inside the switch cabinet, the optimal coordinate point in the neighborhood and the credible score of the three-dimensional space coordinate of the partial discharge source point inside the switch cabinet, the residual error convergence factor is calculated, and a nonlinear fusion function containing the residual error convergence factor is constructed.

10. The multi-source co-modeling and dynamic neighborhood optimization switchgear partial discharge locating method of claim 9, wherein: The final positioning result is generated by using the nonlinear fusion function containing the residual error convergence factor. The dynamic correction refers to constructing a dynamic correction search neighborhood based on the three-dimensional space coordinate of the partial discharge source point inside the switch cabinet, and generating a set of alternative coordinate points in the dynamic correction search field by neighborhood search.

Images