A method and system for classifying and identifying partial discharge of power cables

Abstract

The present invention provides a method for classifying and identifying partial discharge in power cables, comprising obtaining a discharge pulse waveform; and extracting waveform features from the discharge pulse waveform. The waveform feature extraction includes analyzing and classifying the partial discharge pulse data using an equivalent time-frequency transformation method. After extracting the waveform features from the discharge pulse waveform, the extracted waveform features are compressed into a feature space as feature vectors of the pulse waveform. The present invention can identify the signal features of different types of partial discharges and determine the type of partial discharge.

Description

Technical Field

[0001] The present invention relates to the technical field of partial discharge, and in particular to a method and system for classifying and identifying partial discharge of power cables. Background Art

[0002] The statements in this section merely provide background information related to the present invention and do not necessarily constitute prior art.

[0003] Power cable failures severely impact the safe operation of power grids, and the reliability of power cables largely depends on their insulation properties. Cables are prone to defects such as air gaps, impurities, protrusions, and burrs during production, transportation, installation, and operation. These defects trigger localized electric field concentrations, leading to partial discharge (PD). Therefore, partial discharge (PD) is the most direct indicator of the various insulation conditions in power cables. Furthermore, the development of PD can ultimately lead to insulation breakdown, causing equipment damage. Therefore, it is necessary to accurately and promptly identify the type of PD. However, current PD identification methods are inaccurate, unable to fully capture the characteristics that characterize the pulse waveform, and the identification process is cumbersome and inconvenient. Summary of the Invention

[0004] In order to solve the above problems, the present invention proposes a method and system for classifying and identifying partial discharge of power cables. The present invention can realize the identification of different types of partial discharge signal characteristics.

[0005] According to some embodiments, the present invention adopts the following technical solutions:

[0006] A method for classifying and identifying partial discharge of power cables, comprising:

[0007] Obtaining discharge pulse waveform;

[0008] Extract waveform features of discharge pulse waveform;

[0009] The waveform feature extraction includes analyzing and classifying the partial discharge pulse data by using an equivalent time-frequency transformation method.

[0010] Furthermore, after extracting the waveform features of the discharge pulse waveform, the extracted waveform features are compressed into a feature space as feature vectors of the pulse waveform.

[0011] Furthermore, the equivalent time-frequency transformation method includes calculating the time center of gravity and the frequency center of gravity of the partial discharge pulse based on the time domain signal of the partial discharge pulse and the signal after the frequency domain Fourier transformation.

[0012] Furthermore, the equivalent time-frequency transformation method further includes obtaining an equivalent time width and an equivalent frequency width of a pulse waveform based on the time center of gravity and the frequency center of gravity of the partial discharge pulse.

[0013] Furthermore, the equivalent time-frequency transformation method further includes extracting the characteristic value of each discharge pulse based on the equivalent time width and equivalent frequency width of the pulse waveform to obtain the characteristic value of the entire partial discharge pulse and its corresponding time-frequency signal point.

[0014] Furthermore, the waveform feature extraction also includes, under the premise of the equivalent time-frequency transform method, increasing the wavelet packet energy feature dimension.

[0015] Furthermore, the wavelet packet energy feature is used to increase the resolution of different pulse waveforms to facilitate cluster analysis.

[0016] A power cable partial discharge classification and identification system, comprising:

[0017] The data acquisition module is configured to acquire a discharge pulse waveform;

[0018] The feature extraction module is configured to extract waveform features of the discharge pulse waveform;

[0019] The waveform feature extraction includes analyzing and classifying the partial discharge pulse data by using an equivalent time-frequency transformation method.

[0020] A computer-readable storage medium stores a plurality of instructions, wherein the instructions are suitable for being loaded and executed by a processor of a terminal device, for a method for classifying and identifying partial discharge of a power cable.

[0021] A terminal device includes a processor and a computer-readable storage medium, wherein the processor is used to implement various instructions; the computer-readable storage medium is used to store multiple instructions, wherein the instructions are suitable for being loaded and executed by the processor to implement a method for classifying and identifying partial discharge of power cables.

[0022] Compared with the prior art, the present invention has the following beneficial effects:

[0023] The present invention can realize the recognition of different types of partial discharge signal features, and thus realize the judgment of the partial discharge type.

[0024] The waveform feature extraction process maximizes the characterization of the pulse waveform and contains the most information, thereby clustering highly similar pulses together in the feature space. This maximizes the characterization of the pulse waveform and accurately determines the type of discharge. High accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] The drawings in the specification, which constitute a part of this application, are used to provide further understanding of this application. The illustrative embodiments of this application and their descriptions are used to explain this application and do not constitute improper limitations on this application.

[0026] Figure 1 This is a distribution diagram of only calculating the equivalent time and equivalent frequency on the two-dimensional spectrum in Example 1;

[0027] Figure 2 It is a three-dimensional spectrum of wavelet packet energy in the third dimension added on the basis of equivalent time-frequency transform in the first embodiment;

[0028] Figure 3 This is the data processing flow of this embodiment 1. DETAILED DESCRIPTION

[0029] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0030] It should be noted that the following detailed descriptions are illustrative and intended to provide further explanation of the present application. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which the present application belongs.

[0031] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present application. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprise" and / or "include" are used in this specification, they indicate the presence of features, steps, operations, devices, components and / or combinations thereof.

[0032] Example 1

[0033] like Figure 1 As shown, a method for classifying and identifying partial discharge of power cables includes:

[0034] Obtaining discharge pulse waveform;

[0035] Extract waveform features of discharge pulse waveform;

[0036] The waveform feature extraction includes analyzing and classifying the partial discharge pulse data by using an equivalent time-frequency transformation method.

[0037] Specifically,

[0038] Waveform features are extracted for each pulse in the pulse group and compressed into a feature space as the characteristic vector of the pulse waveform. The waveform feature extraction process must maximize the characterization of the pulse waveform while retaining the most information, clustering highly similar pulses together in the feature space. To maximize the characterization of the pulse waveform, the equivalent time-frequency transform method is often used to analyze and classify partial discharge pulse data.

[0039] The equivalent time-frequency transform method is as follows:

[0040] Based on the time domain signal of the partial discharge pulse group and the signal after Fourier transformation in the frequency domain, the time center of gravity and frequency center of gravity of the partial discharge pulse are calculated respectively, and the equivalent time width and equivalent frequency width of the pulse waveform are obtained;

[0041] Based on the equivalent time width and equivalent frequency width of the pulse waveform, the characteristic quantity of each discharge pulse is extracted, and the characteristic quantity of the entire partial discharge pulse group and its corresponding time-frequency signal points are obtained.

[0042] Specifically, the acquired partial discharge pulse group based on clock synchronization is processed as follows:

[0043]

[0044] Where: j is the jth pulse; a i is the time domain waveform value of the i-th point; n is the number of pulse points; Δt is the sampling time interval; Δt(i-1) is the time corresponding to the i-th point. Similarly, the pulse waveform is fast Fourier transformed to obtain:

[0045]

[0046] Where A i is the spectrum amplitude of the i-th point; Δf(i-1) is the frequency value of the i-th point.

[0047] When using the equivalent time-frequency method to extract the characteristic quantity of the pulse waveform, it is necessary to j Do the following processing: Calculate the time center of gravity T0 of the pulse waveform through formula (3) j and frequency center F0 j , and then use formula (4) to obtain the equivalent time width T of the pulse waveform j and equivalent bandwidth F j .

[0048]

[0049]

[0050] Extract the characteristic value of each discharge pulse (eTj ,eF j ), then the characteristic vector of the entire pulse group (eT j ,eF j ),j=1,2,...,N, where N is the number of pulses.

[0051] The equivalent time-frequency method effectively extracts features from discharge pulse waveforms. However, when the classification algorithm requires more waveform-characteristic features, the equivalent time-frequency method becomes ineffective in classifying discharge pulse groups. In this case, it is necessary to extract more waveform-characteristic features from the discharge pulses to complete the classification. Based on the equivalent time-frequency transform method, the wavelet packet energy feature dimension is added. Clustering using only the equivalent time and equivalent frequency two-dimensional space often results in poor differentiation and ineffective clustering. The addition of the wavelet packet energy feature increases the resolution of different pulse waveforms and facilitates cluster analysis.

[0052] Wavelet analysis of a signal is a complete tree structure, requiring relatively high computational effort. The advantage of wavelet packet analysis lies in its ability to decompose not only the low-frequency portion of a signal, but also the high-frequency portion. While signal decomposition using wavelet packets results in relatively little information loss, excessive decomposition can negatively impact signal integrity. This invention utilizes the db4 wavelet basis to perform a three-layer wavelet packet decomposition of the acquired signal, preserving signal integrity to the greatest extent possible.

[0053] For each scale signal decomposed by wavelet packet, the energy corresponding to the wavelet packet coefficients S(i,0), S(I,1),…S(I,2i-1) of the signal after the first layer of decomposition is:

[0054] Ei=||S(i,j) 2 ;

[0055] The total energy formula of the signal wavelet packet is:

[0056]

[0057] Processing flow:

[0058] The PD data acquisition frequency is 100 MHz, the sampling interval is 10 ns, the acquisition method is threshold triggered acquisition, the data acquisition period is 5000 ns, and the acquired data is first preprocessed.

[0059] The data preprocessing method is to set a threshold, and the data of the thirty points before the threshold, i.e., 300ns, and the data of the 170 points after the threshold, i.e., 1700ns, are used as feature extraction sample data.

[0060] The purpose of data preprocessing is to reduce interference from invalid portions of the sampled data, maximizing the extraction of valid discharge data for classification. 200 points are selected because, after extensive analysis of measured data, they can fully capture the data of a single discharge pulse and eliminate interfering signals while preserving the discharge pulse to the greatest extent possible. The preprocessed data is then subjected to an equivalent time-frequency transform and wavelet packet energy calculation to obtain three characteristic features. Finally, clustering calculations are performed on this three-dimensional data to enable classification and identification of partial discharges.

[0061] Figure 1 For this data, we calculate the distribution of equivalent time and equivalent frequency on a two-dimensional spectrum, where the horizontal axis represents equivalent time and the vertical axis represents equivalent frequency. Each point in the spectrum represents a set of continuous sampled data, that is, a complete pulse waveform. After data preprocessing, the corresponding values of equivalent time and equivalent frequency are calculated (the results of Equation 4).

[0062] Figure 2 The three-dimensional spectrum of wavelet packet energy is added to the equivalent time-frequency transform. The equivalent time dimension and equivalent frequency dimension have the same meaning as Figure 1 Similarly, the actual meaning of the wavelet packet energy dimension is the value obtained by calculating the wavelet packet energy of the pulse data after data preprocessing. Each group of 200 data pulse waveforms corresponds to a wavelet packet energy. (That is, a group of 200 preprocessed pulse waveform data can be used to calculate the equivalent time, equivalent frequency, and wavelet packet energy using formulas 1-6. Corresponding Figure 2 A point in the.

[0063] like Figure 3The data processing process uses a PD data acquisition frequency of 100 MHz, a sampling interval of 10 ns, and a threshold-triggered acquisition method with a data acquisition period of 5000 ns. The collected data first undergoes data preprocessing. This preprocessing method sets a threshold, denotes the position of the first point exceeding the threshold as n, and then uses the data thirty points preceding position n, or 300 ns before, as the starting point. 200 points of data are continuously collected as feature extraction samples. The purpose of data preprocessing is to reduce interference from invalid portions of the sampled data and maximize the extraction of valid information from the discharge data for classification. The reason for extracting 200 points is that, after analyzing a large amount of measured data, 200 points can fully encompass the data of a single discharge pulse and can eliminate interference signals while maximally retaining the discharge pulse. The preprocessed data is then subjected to equivalent time-frequency transform and wavelet packet energy calculation to obtain the three characteristic data. Finally, clustering calculation is performed on the three-dimensional data. Since the collected partial discharge data has no regularity and the number of categories is not fixed, the mean shift clustering algorithm (clustering in three-dimensional space) is used. It does not need to specify the number of cluster categories to achieve the classification and identification of partial discharge.

[0064] Example 2

[0065] A power cable partial discharge classification and identification system, comprising:

[0066] The data acquisition module is configured to acquire a discharge pulse waveform;

[0067] The feature extraction module is configured to extract waveform features of the discharge pulse waveform;

[0068] The waveform feature extraction includes analyzing and classifying the partial discharge pulse data by using an equivalent time-frequency transformation method.

[0069] Example 3

[0070] A computer-readable storage medium stores a plurality of instructions, wherein the instructions are suitable for being loaded and executed by a processor of a terminal device to provide a method for classifying and identifying partial discharge of a power cable provided in this embodiment 1.

[0071] Example 4

[0072] A terminal device includes a processor and a computer-readable storage medium, the processor is used to implement various instructions; the computer-readable storage medium is used to store multiple instructions, and the instructions are suitable for being loaded and executed by the processor to provide a method for classifying and identifying partial discharge of power cables provided in this embodiment 1.

[0073] Those skilled in the art will appreciate that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the present application can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0074] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the steps in the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0075] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0076] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0077] The above description is merely a preferred embodiment of the present application and is not intended to limit the present application. Various modifications and variations are possible for those skilled in the art. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present application shall be included within the scope of protection of the present application.

[0078] Although the above describes the specific embodiments of the present invention in conjunction with the accompanying drawings, it is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art on the basis of the technical solution of the present invention without any creative work are still within the scope of protection of the present invention.

Claims

1. A method for classifying and identifying partial discharge of power cables, characterized in that: include: Obtaining discharge pulse waveform; Extracting waveform features of the discharge pulse waveform; and compressing the extracted waveform features into a feature space as a feature vector of the pulse waveform. The waveform feature extraction includes analyzing and classifying the partial discharge pulse data by using an equivalent time-frequency transformation method; The equivalent time-frequency transformation method includes: calculating the time center of gravity and the frequency center of gravity of the partial discharge pulse based on the time domain signal and the signal after the frequency domain Fourier transformation; obtaining the equivalent time width and equivalent frequency width of the pulse waveform based on the time center of gravity and the frequency center of gravity of the partial discharge pulse; The waveform feature extraction also includes, under the premise of the equivalent time-frequency transform method, increasing the wavelet packet energy feature dimension; the wavelet packet energy feature is used to increase the resolution of different pulse waveforms to facilitate cluster analysis, and the specific steps are: selecting the db4 wavelet basis to perform three-layer wavelet packet decomposition on the acquired signal, for each scale signal after the wavelet packet decomposition, the energy corresponding to the wavelet packet coefficients S(i,0), S(i,1), ... S(i,2i-1) after the third layer of decomposition of the signal is: ； The total energy formula of the signal wavelet packet is: ； After equivalent time-frequency transform and wavelet packet energy calculation, the characteristic data of three dimensions are obtained, and clustering calculation is performed on the characteristic data of three dimensions. Specifically, the mean shift clustering algorithm is used to realize the classification and identification of partial discharge.

2. A method for classifying and identifying partial discharge of power cables according to claim 1, characterized in that: The equivalent time-frequency transformation method also includes extracting the characteristic value of each discharge pulse based on the equivalent time width and equivalent frequency width of the pulse waveform to obtain the characteristic value of the entire partial discharge pulse and its corresponding time-frequency signal point.

3. A power cable partial discharge classification and identification system, characterized in that: The method for classifying and identifying partial discharge of a power cable according to any one of claims 1 to 2 comprises: The data acquisition module is configured to acquire a discharge pulse waveform; The feature extraction module is configured to extract waveform features of the discharge pulse waveform; wherein the waveform feature extraction includes analyzing and classifying the partial discharge pulse data through an equivalent time-frequency transformation method.

4. A computer-readable storage medium, characterized in that A plurality of instructions are stored therein, and the instructions are suitable for being loaded by a processor of a terminal device and executed by a method for classifying and identifying partial discharge of a power cable according to any one of claims 1-2.

5. A terminal device, characterized in that: The invention comprises a processor and a computer-readable storage medium, wherein the processor is used to implement various instructions; the computer-readable storage medium is used to store multiple instructions, wherein the instructions are suitable for being loaded by the processor and executed by a method for classifying and identifying partial discharge of a power cable according to any one of claims 1-2.

Images