Insulation defect discharge identification method and device, electronic equipment and storage medium

Abstract

The invention discloses an insulation defect discharge identification method and device, electronic equipment and a storage medium, which are used for solving the technical problem of coexistence of early warning information overload and effective alarm omission caused by the fact that real insulation defect discharge and interference events such as line operation transient state and lightning pulse cannot be effectively distinguished in the prior art. The method comprises the following steps: when insulation defect discharge occurs, synchronously acquiring an electrical sequence signal of each node in a medium-voltage complex network, and generating a frequency-energy spectrogram of each node by adopting the electrical sequence signal; performing spectrogram reconstruction on the frequency-energy spectrogram to obtain a reconstructed frequency-energy spectrogram; extracting insulation defect features from the reconstructed frequency-energy spectrogram; taking the insulation defect characteristics of each node as input, and adopting a random forest algorithm to construct an insulation defect discharge identification model; and when the real-time characteristic data of the medium-voltage complex network are collected, inputting the real-time characteristic data of the medium-voltage complex network into the insulation defect discharge identification model, and carrying out insulation defect discharge identification.

Description

Technical Field

[0001] The present invention relates to the technical field of insulation defects, and in particular, to a method, device, electronic device, and storage medium for identifying insulation defect discharges. Background Art

[0002] Currently, the methods for identifying insulation defect discharges mainly rely on the detection of basic parameters such as the amplitude, frequency, and phase of partial discharge signals, and judge typical insulation deterioration problems through threshold triggering or pulse waveform matching means. This method can achieve basic discharge monitoring functions in medium-voltage distribution lines with simple structures, stable loads, and controllable interference.

[0003] However, with the evolution of medium-voltage distribution networks towards a complex network form with multiple branches, multiple connections, and multiple device accesses, the existing technologies have exposed significant limitations: traditional single-point detection is prone to feature confusion in cable-overhead line hybrid complex network lines; the composite electromagnetic environment formed by dynamic load fluctuations and high-frequency interference of power equipment causes distortion of the time-frequency domain characteristics of partial discharge signals, and the misjudgment rate of conventional frequency band filtering and waveform matching algorithms has increased significantly; existing systems generally lack the ability of multi-node synchronous monitoring and spatio-temporal correlation analysis, and cannot effectively distinguish real insulation defect discharges from interference events such as line operation transients and lightning pulses, resulting in the coexistence of overloaded warning information and missed effective alarms. Summary of the Invention

[0004] The present invention provides a method, device, electronic device, and storage medium for identifying insulation defect discharges, which are used to solve the technical problem that the existing technologies cannot effectively distinguish real insulation defect discharges from interference events such as line operation transients and lightning pulses, resulting in the coexistence of overloaded warning information and missed effective alarms.

[0005] The present invention provides a method for identifying insulation defect discharges, including:

[0006] When an insulation defect discharge occurs, synchronously collect the electrical timing signals of each node in the medium-voltage complex network, and generate a frequency-energy spectrogram for each node by using the electrical timing signals; the nodes include overhead line nodes and overhead-cable connection nodes;

[0007] Perform spectrogram reconstruction on the frequency-energy spectrogram to obtain a reconstructed frequency-energy spectrogram;

[0008] Extract insulation defect features from the reconstructed frequency-energy spectrogram;

[0009] Use the insulation defect features of each node as input, and construct an insulation defect discharge identification model by using the random forest algorithm;

[0010] When real-time characteristic data of a medium-voltage complex network is collected, the real-time characteristic data of the medium-voltage complex network is input into the insulation defect discharge identification model for insulation defect discharge identification.

[0011] Optionally, the step of synchronously collecting electrical timing signals of each node in the medium-voltage complex network and generating a frequency-energy spectrogram of each node by using the electrical timing signals includes:

[0012] When insulation defect discharge occurs, electrical timing signals of each node in the medium-voltage complex network at each frequency point are collected with a preset frequency interval as a step size;

[0013] Calculate the energy value of each electrical timing signal within a preset time window;

[0014] Arrange the energy values in ascending order of frequency to obtain the frequency-energy spectrogram of each node.

[0015] Optionally, the step of performing spectrogram reconstruction on the frequency-energy spectrogram to obtain a reconstructed frequency-energy spectrogram includes:

[0016] Perform normalization processing on the frequency-energy spectrogram to obtain a normalized spectrogram;

[0017] Perform filtering processing on the normalized spectrogram to obtain a filtered spectrogram;

[0018] Perform modal decomposition on the filtered spectrogram to obtain modal components;

[0019] Remove noise in the modal components to obtain denoised modal components;

[0020] Perform spectrogram reconstruction by using all the denoised modal components to generate a reconstructed frequency-energy spectrogram.

[0021] Optionally, the step of extracting insulation defect features from the reconstructed frequency-energy spectrogram includes:

[0022] According to the energy distribution of the reconstructed frequency-energy spectrogram, determine the peak frequency point corresponding to the energy peak;

[0023] Divide key frequency bands according to the peak frequency point;

[0024] Calculate the mean, variance, kurtosis and skewness of the energy within the key frequency bands as the insulation defect features in the reconstructed frequency-energy spectrogram.

[0025] The present invention also provides an insulation defect discharge identification device, including:

[0026] A frequency-energy spectrogram generation module, which is used to synchronously collect the electrical time series signals of each node in a medium-voltage complex network when an insulation defect discharge occurs, and generate a frequency-energy spectrogram of each node by using the electrical time series signals; the nodes include overhead line nodes and overhead-cable connection nodes;

[0027] A spectrogram reconstruction module, which is used to reconstruct the frequency-energy spectrogram to obtain a reconstructed frequency-energy spectrogram;

[0028] An insulation defect feature extraction module, which is used to extract insulation defect features from the reconstructed frequency-energy spectrogram;

[0029] An insulation defect discharge identification model construction module, which is used to construct an insulation defect discharge identification model by using the insulation defect features of each node as input and adopting a random forest algorithm;

[0030] An insulation defect discharge identification module, which is used to input the real-time feature data of the medium-voltage complex network into the insulation defect discharge identification model for insulation defect discharge identification when the real-time feature data of the medium-voltage complex network is collected.

[0031] Optionally, the frequency-energy spectrogram generation module includes:

[0032] An electrical time series signal acquisition sub-module, which is used to collect the electrical time series signals of each node in the medium-voltage complex network at each frequency point at a preset frequency interval as a step when an insulation defect discharge occurs;

[0033] An energy value calculation sub-module, which is used to calculate the energy value of each electrical time series signal within a preset time window;

[0034] A frequency-energy spectrogram generation sub-module, which is used to arrange the energy values in ascending order of frequency to obtain the frequency-energy spectrogram of each node.

[0035] Optionally, the spectrogram reconstruction module includes:

[0036] A normalization sub-module, which is used to perform normalization processing on the frequency-energy spectrogram to obtain a normalized spectrogram;

[0037] A filtering processing sub-module, which is used to perform filtering processing on the normalized spectrogram to obtain a filtered spectrogram;

[0038] A modal decomposition sub-module, which is used to perform modal decomposition on the filtered spectrogram to obtain modal components;

[0039] A denoising sub-module, which is used to remove the noise in the modal components to obtain denoised modal components;

[0040] A spectrogram reconstruction sub-module, which is used to reconstruct a spectrogram by using all the denoised modal components to generate a reconstructed frequency-energy spectrogram.

[0041] Optionally, the insulation defect feature extraction module includes:

[0042] A peak frequency point determination sub-module, which is used to determine the peak frequency point corresponding to the energy peak according to the energy distribution of the reconstructed frequency-energy spectrogram;

[0043] A key frequency band division sub-module, which is used to divide a key frequency band according to the peak frequency point;

[0044] An insulation defect feature extraction sub-module, which is used to calculate the mean, variance, kurtosis and skewness of the energy within the key frequency band as the insulation defect features in the reconstructed frequency-energy spectrogram.

[0045] The present invention also provides an electronic device, which includes a processor and a memory:

[0046] The memory is used to store program codes and transmit the program codes to the processor;

[0047] The processor is used to execute the insulation defect discharge identification method described in any one of the above according to the instructions in the program codes.

[0048] The present invention also provides a computer-readable storage medium, which is used to store program codes, and the program codes are used to execute the insulation defect discharge identification method described in any one of the above.

[0049] It can be seen from the above technical solutions that the present invention has the following advantages: The embodiments of the present invention provide an insulation defect discharge identification method, and specifically disclose that when insulation defect discharge occurs, electrical timing signals of each node in a medium-voltage complex network are synchronously collected, and frequency-energy spectrograms of each node are generated by using the electrical timing signals; the nodes include overhead line nodes and overhead-cable connection nodes; the frequency-energy spectrograms are subjected to spectrogram reconstruction to obtain reconstructed frequency-energy spectrograms; insulation defect features are extracted from the reconstructed frequency-energy spectrograms; an insulation defect discharge identification model is constructed by using a random forest algorithm with the insulation defect features of each node as inputs; when real-time feature data of the medium-voltage complex network are collected, the real-time feature data of the medium-voltage complex network are input into the insulation defect discharge identification model for insulation defect discharge identification.

[0050] By synchronously collecting the electrical timing signals of the overhead line nodes and the overhead-cable connection nodes, the present invention breaks through the problems existing in the prior art, namely, the difficulty in accurately locating the discharge source by single-point detection due to the impedance characteristic differences of the cable-overhead line hybrid line, and the problem of characteristic confusion caused by the mutual coupling of discharge signals in different branches. The present invention also constructs the frequency-energy spectrograms of each node and uses an algorithm to reconstruct the spectrograms, realizing spectrogram decoupling and noise reduction, and effectively reducing the broadband composite electromagnetic noise interference formed by dynamic load fluctuations and high-frequency interference of power electronic devices. Thereby, it can effectively distinguish the true insulation defect discharge from interference events such as line operation transients and lightning pulses, avoiding overloading of warning information and omission of effective alarms. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0052] Figure 1 It is a flowchart of the steps of a method for identifying insulation defect discharge provided by an embodiment of the present invention;

[0053] Figure 2 It is a flowchart of the steps of a method for identifying insulation defect discharge provided by another embodiment of the present invention;

[0054] Figure 3 It is a structural block diagram of a device for identifying insulation defect discharge provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0055] The embodiments of the present invention provide a method, a device, an electronic device, and a storage medium for identifying insulation defect discharge, which are used to solve the technical problem that the prior art cannot effectively distinguish the true insulation defect discharge from interference events such as line operation transients and lightning pulses, resulting in the coexistence of overloading of warning information and omission of effective alarms.

[0056] In order to make the object, features, and advantages of the present invention more obvious and understandable, the following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the embodiments described below are only some embodiments of the present invention, rather than all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0057] Please refer to Figure 1 , Figure 1It is a flowchart of the steps of a method for identifying insulation defect discharges provided by an embodiment of the present invention.

[0058] A method for identifying insulation defect discharges provided by the present invention may specifically include the following steps:

[0059] Step 101, when an insulation defect discharge occurs, synchronously collect the electrical timing signals of each node in the medium-voltage complex network, and generate a frequency-energy spectrum diagram for each node using the electrical timing signals; the nodes include overhead line nodes and overhead-cable connection nodes.

[0060] In an embodiment of the present invention, when an insulation defect discharge occurs, the electrical timing signals of each node of the overhead lines and overhead-cable connection nodes in the medium-voltage complex network can be realized through broadband transformers and high-precision GPSs to generate a dedicated frequency-energy spectrum diagram for each node.

[0061] In specific implementation, the broadband transformers and high-precision GPSs can be deployed at nodes such as the poles and branch points of the overhead lines, as well as the overhead-cable connection nodes.

[0062] Step 102, perform spectrum reconstruction on the frequency-energy spectrum diagram to obtain a reconstructed frequency-energy spectrum diagram.

[0063] Next, perform spectrum reconstruction on the frequency-energy spectrum to obtain a reconstructed frequency-energy spectrum diagram to achieve decoupling and noise reduction of the frequency-energy spectrum diagram.

[0064] Step 103, extract insulation defect features from the reconstructed frequency-energy spectrum diagram.

[0065] Step 104, use the insulation defect features of each node as input, and build an insulation defect discharge identification model using the random forest algorithm.

[0066] Random Forest is an ensemble learning algorithm and belongs to the supervised learning method, which is widely used in classification and regression tasks. It improves the accuracy and stability of the model by constructing multiple decision trees and combining them.

[0067] After completing the spectrum reconstruction, insulation defect features can be extracted from the reconstructed frequency-energy spectrum diagram, and an insulation defect discharge identification model can be built using the random forest algorithm with the insulation defect features as input.

[0068] Step 105, when real-time characteristic data of the medium-voltage complex network is collected, input the real-time characteristic data of the medium-voltage complex network into the insulation defect discharge identification model for insulation defect discharge identification.

[0069] When real-time characteristic data of a medium-voltage complex network is collected, the real-time characteristic data of the medium-voltage complex network is input into a trained insulation defect discharge identification model, and it can be determined whether insulation defect discharge occurs in the medium-voltage complex network currently.

[0070] By synchronously collecting the electrical timing signals of the overhead line nodes and the overhead-cable connection nodes, the present invention breaks through the problems existing in the prior art that it is difficult to accurately locate the discharge source by single-point detection due to the impedance characteristic differences of the cable-overhead line hybrid line, and the characteristic confusion caused by the mutual coupling of the discharge signals of different branches; the present invention also constructs the frequency-energy spectrograms of each node and uses an algorithm to reconstruct the spectrograms to achieve spectrogram decoupling and noise reduction, effectively reducing the broadband composite electromagnetic noise interference formed by the dynamic load fluctuation and the high-frequency interference of power electronic devices. Thus, it can effectively distinguish the real insulation defect discharge from the interference events such as line operation transients and lightning pulses, avoiding the overloading of warning information and the omission of effective alarms.

[0071] Please refer to Figure 2 , Figure 2 which is the step flow chart of the insulation defect discharge identification method provided by another embodiment of the present invention. Specifically, it may include the following steps:

[0072] Step 201, when insulation defect discharge occurs, at a preset frequency interval as the step size, collect the electrical timing signals of each node in the medium-voltage complex network at each frequency point; the nodes include overhead line nodes and overhead-cable connection nodes;

[0073] Step 202, calculate the energy values of each electrical timing signal within a preset time window;

[0074] Step 203, arrange the energy values in ascending order of frequency to obtain the frequency-energy spectrogram of each node;

[0075] In the embodiment of the present invention, broadband current transformers and high-precision GPS can be deployed at the key nodes (towers, branch points) of the overhead line and the overhead-cable connection points.

[0076] The frequency band of the broadband current transformer is , with a frequency interval as the step size, calculate the energy values of the corresponding electrical timing signals of each node at frequency points ( ) within the time window :

[0077]

[0078] With a frequency interval as the step size, sort in ascending order of frequency to generate the exclusive frequency-energy spectrogram of each node , the horizontal axis is the frequency , the vertical axis represents the energy values corresponding to each frequency point.

[0079] In this process, the high-precision GPS receives satellite signals to obtain the reference time, which serves as the time reference for the entire system. The master node sends the reference time information to each slave node through the wireless network. After receiving the reference time information, each slave node adopts the NTP or PTP network time synchronization protocol to achieve efficient time synchronization.

[0080] Step 204, perform spectrogram reconstruction on the frequency-energy spectrogram to obtain a reconstructed frequency-energy spectrogram;

[0081] Next, perform spectrogram reconstruction on the frequency-energy spectrum to obtain a reconstructed frequency-energy spectrogram, so as to achieve decoupling and noise reduction of the frequency-energy spectrogram.

[0082] In an example, the step of performing spectrogram reconstruction on the frequency-energy spectrogram to obtain a reconstructed frequency-energy spectrogram may include the following sub-steps:

[0083] S41, perform normalization processing on the frequency-energy spectrogram to obtain a normalized spectrogram;

[0084] S42, perform filtering processing on the normalized spectrogram to obtain a filtered spectrogram;

[0085] S43, perform modal decomposition on the filtered spectrogram to obtain modal components;

[0086] S44, remove the noise in the modal components to obtain denoised modal components;

[0087] S45, use all the denoised modal components to perform spectrogram reconstruction to generate a reconstructed frequency-energy spectrogram.

[0088] In a specific implementation, the spectrogram reconstruction of the frequency-energy spectrum includes the following links: spectrogram preprocessing, improved variational mode decomposition, sparse regularization denoising, and energy conservation reconstruction.

[0089] 1) Spectrogram preprocessing:

[0090] a. Normalization processing: Perform logarithmic compression on the original energy values to enhance the visibility of low-energy components:

[0091]

[0092] Among them, is the normalized spectrogram.

[0093] b. Trend term elimination: Use the Savitzky-Golay filter to eliminate baseline drift:

[0094]

[0095] Among them, is the filtering spectrogram, w is the window length, that is, the number of sampling points used in sliding polynomial fitting.

[0096] 2) Improved variational mode decomposition:

[0097] a. Based on the objective function in the existing VMD algorithm, introduce the frequency band energy weight factor , to achieve adaptive bandwidth allocation:

[0098]

[0099]

[0100] Among them, K represents the number of modes; f represents the frequency variable; Frequency band energy weight factor; is the derivative; is the Dirac function; is the k-th mode component; represents the central frequency of the k-th mode, is the imaginary unit.

[0101] b. Dynamically determine the number of modes K through spectral kurtosis analysis:

[0102]

[0103] Among them, is the kurtosis calculation function.

[0104] The calculation process is as follows:

[0105] ① Select the sliding window width 2M + 1, and M can be taken as 5;

[0106] ② For each center point Find the second-order central moment , the fourth-order central moment ;

[0107] ③ Calculate the sample kurtosis at the i-th position:

[0108]

[0109] Among them, is a constant, which can be taken as 10 -8 ;

[0110] ④ is the summation; is the maximum value.

[0111] 3) Sparse regularization denoising:

[0112] a. Suppress the noise in the modal components using the truncated L1 norm:

[0113]

[0114] The threshold function is:

[0115]

[0116] Where is the instantaneous amplitude of a certain modal component at a certain frequency point, is the k-th modal component, is the regularization parameter, which is used to balance the smoothness of denoising and the sparsity of retaining signal features, is the truncation threshold, and different modal components correspond to different truncation thresholds.

[0117] b. Dynamically adjust based on the modal energy ratio:

[0118]

[0119] Where is the truncation threshold of the k-th modal component.

[0120] 4) Reconstruction with energy conservation:

[0121] Ensure the reconstruction energy conservation through the projection gradient method:

[0122]

[0123] Where is the support domain constraint of the original frequency-energy spectrogram.

[0124] Step 205. Extract the insulation defect features from the reconstructed frequency-energy spectrogram;

[0125] In the embodiment of the present invention, step 205 may include the following sub-steps:

[0126] S51. Determine the peak frequency point corresponding to the energy peak according to the energy distribution of the reconstructed frequency-energy spectrogram;

[0127] S52. Divide the key frequency bands according to the peak frequency point;

[0128] S53. Calculate the mean, variance, kurtosis and skewness of the energy in the key frequency bands as the insulation defect features in the reconstructed frequency-energy spectrogram.

[0129] In a specific implementation, the key frequency bands may be divided according to the energy distribution of the reconstructed frequency-energy spectrogram, and the specific process is:

[0130] Calculate the energy distribution of the frequency-energy spectrogram and determine the peak frequency point corresponding to the energy peak:

[0131]

[0132] Divide the bandwidth of the key frequency band according to the peak frequency point:

[0133]

[0134]

[0135] Calculate the mean, variance, kurtosis, and skewness of the energy within the key frequency band to obtain the statistical features within the key frequency band, which are used as the insulation defect features for reconstructing the frequency-energy spectrogram.

[0136] Step 206: Use the insulation defect features of each node as input and construct an insulation defect discharge identification model using the random forest algorithm;

[0137] After completing the spectrogram reconstruction, the insulation defect features can be extracted from the reconstructed frequency-energy spectrogram, and an insulation defect discharge identification model can be constructed using the random forest algorithm with the insulation defect features as input.

[0138] Among them, the random forest model is:

[0139]

[0140] Among them, represents the prediction output of the b-th tree for the input sample x;

[0141] represents the set of random parameters used when training the b-th tree;

[0142] is the total number of trees in the forest.

[0143] Its loss function is:

[0144]

[0145] Among them, N is the total number of samples, that is, the number of observation points (or training samples) used to calculate the loss; is the true class label of the i-th sample; is the probability output of the model predicting the i-th sample as the positive class.

[0146] Furthermore, in the embodiments of the present invention, 5-fold cross-validation can be used to optimize the hyperparameters, and indicators such as accuracy, F1-Score, and ROC-AUC can be used to evaluate the model effect. The specific evaluation process can refer to the conventional evaluation methods and will not be elaborated here.

[0147] Step 207: When the real-time characteristic data of the medium-voltage complex network is collected, input the real-time characteristic data of the medium-voltage complex network into the insulation defect discharge identification model for insulation defect discharge identification.

[0148] After training the insulation defect discharge identification model, it can be deployed in the target environment for real-time insulation defect discharge identification.

[0149] Furthermore, in the embodiment of the present invention, real-time data can also be collected online and stored in the time series database InfluxDB according to the time stamp. When the proportion of new data exceeds 20% of the historical data, retraining is triggered.

[0150] The present invention synchronously collects the electrical time series signals of the overhead line nodes and the overhead-cable connection nodes, breaking through the problems existing in the prior art that it is difficult to accurately locate the discharge source by single-point detection due to the impedance characteristic differences of the cable-overhead line hybrid line, and the characteristic confusion caused by the mutual coupling of the discharge signals of different branches; the present invention also constructs the frequency-energy spectrogram of each node and uses an algorithm to reconstruct the spectrogram to achieve spectrogram decoupling and noise reduction, effectively reducing the broadband composite electromagnetic noise interference formed by the dynamic load fluctuation and the high-frequency interference of power electronic devices. Thus, it can effectively distinguish the real insulation defect discharge from the interference events such as line operation transients and lightning pulses, avoiding the overloading of warning information and the omission of effective alarms.

[0151] Please refer to Figure 3 , Figure 3 which is the structural block diagram of an insulation defect discharge identification device provided by the embodiment of the present invention.

[0152] The embodiment of the present invention provides an insulation defect discharge identification device, including:

[0153] A frequency-energy spectrogram generation module 301, configured to synchronously collect the electrical time series signals of each node in the medium-voltage complex network when an insulation defect discharge occurs, and generate the frequency-energy spectrogram of each node by using the electrical time series signals; the nodes include overhead line nodes and overhead-cable connection nodes;

[0154] A spectrogram reconstruction module 302, configured to perform spectrogram reconstruction on the frequency-energy spectrogram to obtain a reconstructed frequency-energy spectrogram;

[0155] An insulation defect feature extraction module 303, configured to extract insulation defect features from the reconstructed frequency-energy spectrogram;

[0156] An insulation defect discharge identification model construction module 304, configured to use the insulation defect features of each node as input and construct an insulation defect discharge identification model by using the random forest algorithm;

[0157] The insulation defect discharge identification module 305 is used to input the real-time characteristic data of the medium-voltage complex network into the insulation defect discharge identification model for insulation defect discharge identification when the real-time characteristic data of the medium-voltage complex network is collected.

[0158] In an embodiment of the present invention, the frequency-energy spectrogram generation module 301 includes:

[0159] The electrical time-series signal acquisition sub-module is used to acquire the electrical time-series signals of each node in the medium-voltage complex network at each frequency point at a preset frequency interval as a step when insulation defect discharge occurs;

[0160] The energy value calculation sub-module is used to calculate the energy values of each electrical time-series signal within a preset time window;

[0161] The frequency-energy spectrogram generation sub-module is used to arrange the energy values in ascending order of frequency to obtain the frequency-energy spectrogram of each node.

[0162] In an embodiment of the present invention, the spectrogram reconstruction module 302 includes:

[0163] The normalization sub-module is used to perform normalization processing on the frequency-energy spectrogram to obtain a normalized spectrogram;

[0164] The filtering processing sub-module is used to perform filtering processing on the normalized spectrogram to obtain a filtered spectrogram;

[0165] The modal decomposition sub-module is used to perform modal decomposition on the filtered spectrogram to obtain modal components;

[0166] The denoising sub-module is used to remove the noise in the modal components to obtain denoised modal components;

[0167] The spectrogram reconstruction sub-module is used to perform spectrogram reconstruction using all denoised modal components to generate a reconstructed frequency-energy spectrogram.

[0168] In an embodiment of the present invention, the insulation defect feature extraction module 303 includes:

[0169] The peak frequency point determination sub-module is used to determine the peak frequency point corresponding to the energy peak according to the energy distribution of the reconstructed frequency-energy spectrogram;

[0170] The key frequency band division sub-module is used to divide the key frequency band according to the peak frequency point;

[0171] The insulation defect feature extraction sub-module is used to calculate the mean, variance, kurtosis and skewness of the energy within the key frequency band as the insulation defect features in the reconstructed frequency-energy spectrogram.

[0172] An embodiment of the present invention also provides an electronic device, which includes a processor and a memory:

[0173] The memory is used to store program code and transmit the program code to the processor;

[0174] The processor is used to execute the insulation defect discharge identification method of the embodiment of the present invention according to the instructions in the program code.

[0175] The embodiment of the present invention also provides a computer-readable storage medium, which is used to store program code, and the program code is used to execute the insulation defect discharge identification method of the embodiment of the present invention.

[0176] Those skilled in the art can clearly understand that for the convenience and conciseness of description, the specific working processes of the above-described systems, devices, and units can refer to the corresponding processes in the foregoing method embodiments, and will not be described herein again.

[0177] Each embodiment in this specification is described in a progressive manner. Each embodiment focuses on the differences from other embodiments. The same and similar parts among the various embodiments can be referred to each other.

[0178] Those skilled in the art should understand that the embodiments of the present invention can be provided as a method, a device, or a computer program product. Therefore, the embodiments of the present invention can take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Moreover, the embodiments of the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk memories, CD-ROMs, optical memories, etc.) containing computer-usable program code.

[0179] The embodiments of the present invention are described with reference to the flowcharts and / or block diagrams of methods, terminal devices (systems), and computer program products according to the embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram can be implemented by computer program instructions, and the combination of processes and / or blocks in the flowchart and / or block diagram can also be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing terminal devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing terminal devices generate means for realizing the functions specified in one Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks.

[0180] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing terminal device to work in a specific manner, so that the instructions stored in the computer-readable memory generate a manufactured article including an instruction device, and the instruction device realizes the functions in the process Figure 1One or more processes and / or blocks Figure 1 The functions specified in one or more blocks.

[0181] These computer program instructions can also be loaded onto a computer or other programmable data processing terminal device, so that a series of operation steps are executed on the computer or other programmable terminal device to generate a computer-implemented process. Thus, the instructions executed on the computer or other programmable terminal device provide steps for implementing the functions specified in one Figure 1 One or more processes and / or blocks Figure 1 The steps of the functions specified in one or more blocks.

[0182] Although the preferred embodiments of the embodiments of the present invention have been described, those skilled in the art can make additional changes and modifications to these embodiments once they know the basic creative concept. Therefore, the appended claims are intended to be construed to include the preferred embodiments and all changes and modifications falling within the scope of the embodiments of the present invention.

[0183] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties. And the collection, use, and processing of relevant data need to comply with the relevant laws, regulations, and standards of relevant countries and regions, and corresponding operation entrances are provided for users to choose to authorize or refuse.

[0184] Finally, it should also be noted that in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or terminal device comprising a series of elements not only includes those elements, but also includes other elements not expressly listed, or also includes elements inherent to such process, method, article or terminal device. Without further limitation, an element defined by the statement "comprising a..." does not exclude the presence of additional identical elements in the process, method, article or terminal device comprising the said element.

[0185] As described above, the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. An insulation defect discharge identification method, characterized in that Including: When an insulation defect discharge occurs, synchronously collect the electrical timing signals of each node in the medium-voltage complex network, and generate a frequency-energy spectrogram of each node using the electrical timing signals; the nodes include overhead line nodes and overhead-cable connection nodes; Perform spectrogram reconstruction on the frequency-energy spectrogram to obtain a reconstructed frequency-energy spectrogram; Extract insulation defect features from the reconstructed frequency-energy spectrogram; Using the insulation defect features of each node as input, construct an insulation defect discharge identification model using the random forest algorithm; When real-time characteristic data of the medium-voltage complex network is collected, input the real-time characteristic data of the medium-voltage complex network into the insulation defect discharge identification model for insulation defect discharge identification.

2. The method according to claim 1, wherein The step of synchronously collecting the electrical timing signals of each node in the medium-voltage complex network and generating a frequency-energy spectrogram of each node using the electrical timing signals includes: When an insulation defect discharge occurs, take the preset frequency interval as the step size, and collect the electrical timing signals of each node in the medium-voltage complex network at each frequency point; Calculate the energy values of each electrical timing signal within a preset time window; Arrange the energy values in ascending order of frequency to obtain the frequency-energy spectrogram of each node.

3. The method according to claim 1, characterized in that, The step of performing spectrogram reconstruction on the frequency-energy spectrogram to obtain a reconstructed frequency-energy spectrogram includes: Perform normalization processing on the frequency-energy spectrogram to obtain a normalized spectrogram; Perform filtering processing on the normalized spectrogram to obtain a filtered spectrogram; Perform modal decomposition on the filtered spectrogram to obtain modal components; Remove the noise in the modal components to obtain denoised modal components; Use all the denoised modal components for spectrogram reconstruction to generate a reconstructed frequency-energy spectrogram.

4. The method according to claim 1, wherein The step of extracting insulation defect features from the reconstructed frequency-energy spectrogram includes: According to the energy distribution of the reconstructed frequency-energy spectrogram, determine the peak frequency points corresponding to the energy peaks; Divide the key frequency bands according to the peak frequency points; Calculate the mean, variance, kurtosis, and skewness of the energy within the key frequency bands as the insulation defect features in the reconstructed frequency-energy spectrogram.

5. An insulation defect discharge identification device, characterized in that, Including: A frequency-energy spectrogram generation module for synchronously collecting the electrical timing signals of each node in the medium-voltage complex network when an insulation defect discharge occurs, and generating a frequency-energy spectrogram of each node using the electrical timing signals; the nodes include overhead line nodes and overhead-cable connection nodes; A spectrogram reconstruction module for performing spectrogram reconstruction on the frequency-energy spectrogram to obtain a reconstructed frequency-energy spectrogram; An insulation defect feature extraction module for extracting insulation defect features from the reconstructed frequency-energy spectrogram; An insulation defect discharge identification model construction module for constructing an insulation defect discharge identification model using the random forest algorithm with the insulation defect features of each node as input; An insulation defect discharge identification module for inputting the real-time characteristic data of the medium-voltage complex network into the insulation defect discharge identification model for insulation defect discharge identification when the real-time characteristic data of the medium-voltage complex network is collected.

6. The device according to claim 5, characterized in that The frequency-energy spectrogram generation module includes: The electrical timing signal acquisition sub-module is used to collect the electrical timing signals of each node in the medium-voltage complex network at each frequency point with a preset frequency interval as the step when an insulation defect discharge occurs; The energy value calculation sub-module is used to calculate the energy values of each electrical timing signal within a preset time window; The frequency-energy spectrogram generation sub-module is used to arrange the energy values in ascending order of frequency to obtain the frequency-energy spectrogram of each node.

7. The device according to claim 5, characterized in that, The spectrogram reconstruction module includes: The normalization sub-module is used to perform normalization processing on the frequency-energy spectrogram to obtain a normalized spectrogram; The filtering processing sub-module is used to perform filtering processing on the normalized spectrogram to obtain a filtered spectrogram; The modal decomposition sub-module is used to perform modal decomposition on the filtered spectrogram to obtain modal components; The denoising sub-module is used to remove the noise in the modal components to obtain denoised modal components; The spectrogram reconstruction sub-module is used to perform spectrogram reconstruction using all the denoised modal components to generate a reconstructed frequency-energy spectrogram.

8. The device according to claim 5, characterized in that, The insulation defect feature extraction module includes: The peak frequency point determination sub-module is used to determine the peak frequency point corresponding to the energy peak according to the energy distribution of the reconstructed frequency-energy spectrogram; The key frequency band division sub-module is used to divide the key frequency band according to the peak frequency point; The insulation defect feature extraction sub-module is used to calculate the mean, variance, kurtosis, and skewness of the energy within the key frequency band as the insulation defect features in the reconstructed frequency-energy spectrogram.

9. An electronic device, characterized in that, The device includes a processor and a memory: The memory is used to store program code and transmit the program code to the processor; The processor is used to execute the insulation defect discharge identification method according to any one of claims 1-4 according to the instructions in the program code.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium is used to store program code, and the program code is used to execute the insulation defect discharge identification method according to any one of claims 1-4.

Images