Cable partial discharge data processing method and system

Abstract

The invention relates to the technical field of power cables, in particular to a cable partial discharge data processing method and system, and the method comprises the steps: constructing a partial discharge and signal sample library; partial discharge signals are automatically extracted and identified; and carrying out similarity calculation and self-learning on the extracted signals to update a sample library. The method has the beneficial effect of effectively solving the problems of long time consumption and low efficiency of extraction processing of massive partial discharge samples. Furthermore, partial discharge data and statistical characteristics can be automatically processed, whether the signals are partial discharge or not can be quickly judged, the acquisition efficiency of partial discharge samples is remarkably improved, and powerful support is provided for follow-up research work by constructing a partial discharge and interference sample library. In addition, in combination with a data expansion method based on variable noise and the generative adversarial network, the technology can effectively expand partial discharge samples, and the automation degree of partial discharge processing and the data processing efficiency are further improved.

Description

Technical Field

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

[0002] As an important transmission medium in modern power systems, cables play a vital role in power transmission, distribution and use. Their operating status directly affects the stability and safety of power systems. However, during the operation of cables, due to long-term high voltage and high current operation, as well as possible external force damage, equipment defects, water tree intrusion and other factors, cables are prone to various faults. Therefore, real-time monitoring and accurate evaluation of cable operation status are crucial. In order to achieve real-time monitoring and evaluation of cable operation status, researchers have developed a variety of cable status monitoring technologies. Among them, partial discharge (PD) monitoring is an important technical means. By monitoring the PD signal of the cable, potential cable faults can be discovered in time and corresponding treatment measures can be taken to ensure the safe operation of the power system. However, the monitoring and identification of PD signals face many challenges. On the one hand, PD signals are mixed with complex noise and interference signals in industrial sites, making signal extraction and analysis difficult. On the other hand, the similarity between some PD types is very high, which further increases the difficulty of identification. With the advancement of smart grids, traditional data processing technologies are difficult to meet the requirements of power equipment monitoring of massive PD data, and new technologies are needed to improve the efficiency of PD processing. Summary of the invention

[0003] The purpose of this section is to summarize some aspects of embodiments of the present invention and briefly introduce some preferred embodiments. Some simplifications or omissions may be made in this section and the specification abstract and the invention title of this application to avoid blurring the purpose of this section, the specification abstract and the invention title, and such simplifications or omissions cannot be used to limit the scope of the present invention.

[0004] In view of the above existing problems, the present invention is proposed.

[0005] Therefore, the present invention provides a cable partial discharge data processing method and system, which can solve the problems mentioned in the background technology.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0007] In a first aspect, the present invention provides a cable partial discharge data processing method, including constructing a partial discharge and signal sample library;

[0008] Automatically extract and identify partial discharge signals;

[0009] The extracted signals are subjected to similarity calculation and self-learning to update the sample library.

[0010] As a preferred solution of the cable partial discharge data processing method of the present invention, wherein: a partial discharge and signal sample library is constructed, including

[0011] Get target data;

[0012] Construct a partial discharge sample library and an interference signal sample library based on the collected target data;

[0013] Perform feature extraction and normalization on the data in the sample library.

[0014] As a preferred solution of the cable partial discharge data processing method of the present invention, the partial discharge signal is automatically extracted and identified, including

[0015] The automatic identification module of K-Means algorithm is used to classify the partial discharge signals;

[0016] Use the data in the sample library to verify the results of the automatic identification module;

[0017] For signals that the automatic recognition module cannot determine, an expert system is used to assist in analysis and recognition.

[0018] As a preferred solution of the cable partial discharge data processing method of the present invention, the automatic identification module of the K-Means algorithm is used to classify the partial discharge signal, including

[0019] The phase of the partial discharge signal is determined based on the center point distribution and angle relationship of the K-Means clustering results;

[0020] Determine the type of partial discharge signal based on the partial discharge signal characteristics under different phase conditions.

[0021] As a preferred solution of the cable partial discharge data processing method of the present invention, wherein: similarity calculation is performed on the extracted signal, including

[0022] Calculate and normalize the feature vector of the extracted signal;

[0023] The similarity between the extracted signal and the signal in the sample library is measured by calculating the bi-norm between the normalized feature vectors.

[0024] As a preferred solution of the cable partial discharge data processing method of the present invention, wherein: self-learning updates the sample library, including

[0025] Adding the newly identified partial discharge signal to the partial discharge sample library;

[0026] Adding the newly identified interference signal to the interference signal sample library;

[0027] Merge or delete signals with high similarity in the sample library.

[0028] As a preferred solution of the cable partial discharge data processing method of the present invention, the partial discharge signal characteristics include but are not limited to peak voltage, average voltage and root mean square voltage.

[0029] In a second aspect, the present invention provides a cable partial discharge data processing method, comprising: a construction module for constructing a partial discharge and signal sample library;

[0030] An extraction and identification module is used to automatically extract and identify partial discharge signals;

[0031] The calculation and updating module is used to perform similarity calculation on the extracted signals and to update the sample library by self-learning.

[0032] In a third aspect, the present invention provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the above method when executing the computer program.

[0033] In a fourth aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, which implements the steps of the above method when the computer program is executed by a processor.

[0034] Compared with the prior art, the beneficial effects of the present invention are as follows: it effectively solves the problem of time-consuming and low efficiency in extracting and processing massive partial discharge samples. Furthermore, the present invention can not only automatically process partial discharge data, statistically characterize and quickly determine whether a signal is a partial discharge, thereby significantly improving the efficiency of obtaining partial discharge samples, but also provide strong support for subsequent research work by constructing a sample library of partial discharge and interference. In addition, combined with the data expansion method based on variable noise and adversarial generative networks, this technology can also effectively expand partial discharge samples, further improving the degree of automation and data processing efficiency of partial discharge processing. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work. Among them:

[0036] Figure 1 This is a flowchart of the cable partial discharge data processing method.

[0037] Figure 2 A schematic diagram of the internal structure of a computer device. DETAILED DESCRIPTION

[0038] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the accompanying drawings.

[0039] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0040] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive with other embodiments.

[0041] Example 1

[0042] Reference Figure 1 , is the first embodiment of the present invention, which provides a cable partial discharge data processing method, which includes:

[0043] S1. Build a partial discharge and signal sample library.

[0044] Further, obtaining target data, wherein the target data includes industrial partial discharge data and laboratory partial discharge data;

[0045] Construct a partial discharge sample library and an interference signal sample library based on the collected target data;

[0046] Perform feature extraction and normalization on the data in the sample library.

[0047] It should be noted that feature extraction includes but is not limited to peak voltage, average voltage and root mean square voltage. These features can reflect the characteristics of partial discharge signals. The extracted features are further normalized to eliminate the influence of different dimensions and magnitudes, so that the feature vectors can be compared under the same standard.

[0048] S2. Automatically extract and identify partial discharge signals.

[0049] Furthermore, the partial discharge signal is automatically extracted and identified, including

[0050] The automatic identification module of K-Means algorithm is used to classify the partial discharge signals;

[0051] Use the data in the sample library to verify the results of the automatic identification module;

[0052] For signals that the automatic recognition module cannot determine, an expert system is used to assist in analysis and recognition.

[0053] It should be noted that the automatic identification module of the K-Means algorithm can process a large amount of partial discharge data and automatically identify partial discharge signals and interference signals; the results of the automatic identification module are verified by using the data in the sample library to ensure the accuracy of identification; for signals that the automatic identification module cannot determine, an expert system is used to assist in analysis and identification, and these signals are classified manually or semi-automatically.

[0054] It should be further explained that the PD single pulse analysis is used to view the information of a single PD pulse to assist experts in determining whether the signal is PD or interference. The PD single pulse analysis includes the following functions: visualization of time domain signals, selection and visualization of single pulse signals, visualization of pulse signal phase spectrum, scatter plots and bar graphs of custom pulse signal characteristics, setting pulse signal screening thresholds, screening pulse signals according to pulse signal characteristics, and automatic identification of PD signals and interference signals.

[0055] Partial discharge multi-pulse analysis is used to view the information of multiple pulse signals of the same source and automatically determine whether the pulse signal is a partial discharge or interference. Partial discharge multi-pulse analysis includes the following functions: visualization of time domain signals, visualization of pulse signal phase spectrum, scatter plots and bar graphs of custom pulse signal characteristics, and automatic identification of partial discharge signals and interference signals.

[0056] It should be further explained that this method also provides a specific sample library expansion method, namely, partial discharge sample expansion based on variable noise and adversarial generation network:

[0057] The variable noise method requires setting the parameters of the normal distribution of noise first, and then obtaining the distribution of noise. After that, the noise is superimposed on the original sample to obtain the expanded sample data. In order to verify the effectiveness of variable noise expansion, linear regression, support vector machine, and random forest methods can be used to verify the expansion results.

[0058] The probability density function formula of normally distributed noise is:

[0059]

[0060] Where x represents the value of the noise, μ Represents the mean of the normal distribution of noise, and σ represents the standard deviation of the normal distribution of noise. In the PD sample expansion based on variable noise, the mean of the normal distribution of noise is set to 0, and the expanded data under different noise levels are obtained by changing the standard deviation of the normal distribution of noise. The smaller the standard deviation, the more concentrated the noise distribution is, and the larger the standard deviation, the more dispersed the noise distribution is.

[0061] The Generative Adversarial Network is a generative model that uses two neural network models to approximate the real distribution of data with the distribution of generated data. The two neural network models in the Generative Adversarial Network are the generator and the discriminator. The input of the generator is a random variable, and the output is a generated sample. The discriminator determines whether the sample is a real sample or a generated sample. The role of the generator is to make the generated sample as close to the real sample as possible, so that the discriminator cannot distinguish between the real sample and the generated sample. The role of the discriminator is to distinguish the real sample from the generated sample as much as possible. When training the network, the generator and the discriminator are trained alternately.

[0062] The overall optimization goal of the adversarial generation network can be expressed as:

[0063] Gmin Dmax V(G,D)

[0064] Among them, G and D represent the generator and discriminator respectively, and V(G,D) represents the distribution difference between the samples generated by the generator and the distribution difference of the real samples. The optimization goal of the discriminator is to maximize the distribution difference to distinguish the real samples from the generated samples. The optimization goal of the generator is to minimize the distribution difference to make the real samples and the generated samples as close as possible. The model parameters of the generator and the discriminator are updated in each iteration by constructing the loss function.

[0065] The training steps of the adversarial generative network are as follows, which mainly include two steps, namely, steps 1-3 to update the parameters of the discriminator and steps 4-6 to update the parameters of the generator:

[0066] ① Randomly select noise signals from the noise distribution;

[0067] ② Send the noise signal into the generator to obtain the generated data;

[0068] ③ Send the generated data and real data to the discriminator, fix the parameters of the generator, and use the gradient descent algorithm to update the parameters of the discriminator. The update goal is to allow the discriminator to distinguish between the data generated by the generator and the real data;

[0069] ④ The machine selects the noise signal from the noise distribution;

[0070] ⑤ Send the noise signal into the generator to obtain the generated data;

[0071] ⑥ Send the generated data and the real data to the discriminator, fix the parameters of the discriminator, and use the gradient descent algorithm to update the parameters of the generator. The update goal is to make the discriminator unable to distinguish the data generated by the generator from the real data;

[0072] ⑦If the convergence condition is reached, stop, otherwise return to step ①.

[0073] Furthermore, the automatic identification module of the K-Means algorithm is used to classify the partial discharge signals, including determining the phase of the partial discharge signal based on the center point distribution and angle relationship of the K-Means clustering results;

[0074] Determine the type of partial discharge signal based on the partial discharge signal characteristics under different phase conditions.

[0075] It should be noted that the K-Means clustering algorithm overcomes the shortcoming that phase information is difficult to directly obtain in partial discharge detection through signal extraction, coordinate transformation, K-Means clustering, center point translation, and category judgment, and can effectively judge partial discharge signals.

[0076] After obtaining the number of clusters K and the spectrum after the center point is shifted, the recognition rules of the recognition module for the recognition of partial discharge and interference signals are as follows:

[0077] Rule 1: When K = 2, the positive and negative center points are distributed in two opposite areas, and the angle relative to the center point is 180 (1 ± 10%) degrees. The signal that meets the above conditions is a partial discharge signal under single-phase voltage.

[0078] Rule 2: When K=4, the positive and negative center points are distributed in two opposite areas, the angle between the relative center point is 180 (1±10%) degrees, and the angle between adjacent center points is 60 (1±10%) degrees or 120 (1±10%) degrees. The signal that meets the above conditions is a partial discharge signal under biphase voltage.

[0079] Rule 3: When K=6, the positive and negative center points are distributed in two opposite areas, the angle between the relative center point is 180 (1±10%) degrees, and the angle between adjacent center points is 60 (1±10%) degrees. The signal that meets the above conditions is a partial discharge signal under three-phase voltage.

[0080] Rule 4: When rules 1, 2, and 3 are not met, the signal is an interference signal.

[0081] Furthermore, the partial discharge signal characteristics include but are not limited to peak voltage, average voltage and root mean square voltage.

[0082] S3. Calculate the similarity of the extracted signals and update the sample library through self-learning.

[0083] Furthermore, similarity calculation is performed on the extracted signals, including

[0084] Calculate and normalize the feature vector of the extracted signal;

[0085] The similarity between the extracted signal and the signal in the sample library is measured by calculating the bi-norm between the normalized feature vectors.

[0086] It should be noted that the calculation formula of pulse similarity is:

[0087] ||xy||

[0088] Among them, x represents the normalized feature vector of the signal to be judged, and y represents the normalized feature vector of the signal in the sample library. The features in the feature vector include information such as peak voltage, average voltage, and root mean square of voltage. The method of measuring the similarity of pulse signals by calculating the bi-norm between the normalized signal feature vectors can determine whether the signal is a known type of partial discharge signal.

[0089] Preferably, by calculating the similarity of the pulses, it is helpful to obtain the similarity between the pulse to be classified and the samples in the partial discharge library and the interference library. This method optimizes the occupation of the system memory space and improves the efficiency of partial discharge preprocessing.

[0090] Furthermore, the self-learning updates the sample library, including

[0091] Adding the newly identified partial discharge signal to the partial discharge sample library;

[0092] Adding the newly identified interference signal to the interference signal sample library;

[0093] Merge or delete signals with high similarity in the sample library.

[0094] It should be noted that the sample library includes the following functions: statistics of the total number of samples, visualization of the sample signal phase spectrum, scatter plots and bar graphs of custom sample signal characteristics, and deletion of specific sample signals. When it is determined that the signal belongs to a partial discharge signal, the signal is compared with the samples in the partial discharge sample library. If there is a similar partial discharge signal in the partial discharge sample library, the sample is not added to the partial discharge library, otherwise the sample is added to the partial discharge sample library. The rules for whether the interference signal is added to the interference sample library are similar to the above-mentioned rules for adding partial discharges. By screening partial discharge and interference samples, the number of samples in the partial discharge sample library and the interference sample library can be effectively reduced, the occupancy of the system memory space can be optimized, and the efficiency of partial discharge preprocessing can be improved.

[0095] In summary, the beneficial effect of a cable partial discharge data processing method is to effectively solve the problem of time-consuming and inefficient extraction and processing of massive partial discharge samples. Furthermore, the present invention can not only automatically process partial discharge data, statistical features and quickly determine whether the signal is partial discharge, thereby significantly improving the efficiency of obtaining partial discharge samples, but also provide strong support for subsequent research work by constructing a sample library of partial discharge and interference. In addition, combined with the data expansion method based on variable noise and adversarial generative network, this technology can also effectively expand partial discharge samples, further improving the degree of automation and data processing efficiency of partial discharge processing.

[0096] Example 2

[0097] This embodiment provides a cable partial discharge data processing system, which includes a construction module for constructing a partial discharge and signal sample library;

[0098] An extraction and identification module is used to automatically extract and identify partial discharge signals;

[0099] The calculation and updating module is used to perform similarity calculation on the extracted signals and to update the sample library by self-learning.

[0100] The above-mentioned unit modules may be embedded in or independent of a processor in a computer device in the form of hardware, or may be stored in a memory in a computer device in the form of software, so that the processor can call and execute operations corresponding to the above-mentioned modules.

[0101] Example 3

[0102] This embodiment provides a computer device, which may be a terminal, and its internal structure diagram may be as follows: Figure 2 As shown. The computer device includes a processor, a memory, a communication interface, a display screen and an input device connected through a system bus. Among them, the processor of the computer device is used 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 the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be achieved through WIFI, an operator network, NFC (near field communication) or other technologies. When the computer program is executed by the processor, a cable partial discharge data processing method is implemented. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covering the display screen, or a key, trackball or touchpad set on the computer device housing, or an external keyboard, touchpad or mouse, etc.

[0103] This embodiment also provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, it implements: building a partial discharge and signal sample library; automatically extracting and identifying partial discharge signals; performing similarity calculation on the extracted signals and self-learning to update the sample library.

[0104] It should be noted that 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 preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A cable partial discharge data processing method, characterized in that: include, Build a partial discharge and signal sample library; Automatically extract and identify partial discharge signals; The extracted signals are subjected to similarity calculation and self-learning to update the sample library.

2. The cable partial discharge data processing method according to claim 1, characterized in that: The construction of the partial discharge and signal sample library includes Get target data; Construct a partial discharge sample library and an interference signal sample library based on the collected target data; Perform feature extraction and normalization on the data in the sample library.

3. The cable partial discharge data processing method according to claim 2, characterized in that: The automatic extraction and identification of partial discharge signals includes The automatic identification module of K-Means algorithm is used to classify the partial discharge signals; Use the data in the sample library to verify the results of the automatic identification module; For signals that the automatic recognition module cannot determine, an expert system is used to assist in analysis and recognition.

4. The cable partial discharge data processing method according to claim 3, characterized in that: The automatic identification module using the K-Means algorithm classifies the partial discharge signal, including The phase of the partial discharge signal is determined based on the center point distribution and angle relationship of the K-Means clustering results; Determine the type of partial discharge signal based on the partial discharge signal characteristics under different phase conditions.

5. The cable partial discharge data processing method according to claim 4, characterized in that: The similarity calculation of the extracted signals includes: Calculate and normalize the feature vector of the extracted signal; The similarity between the extracted signal and the signal in the sample library is measured by calculating the bi-norm between the normalized feature vectors.

6. The cable partial discharge data processing method according to claim 5, characterized in that: The self-learning update sample library includes Adding the newly identified partial discharge signal to the partial discharge sample library; Adding the newly identified interference signal to the interference signal sample library; Merge or delete signals with high similarity in the sample library.

7. The cable partial discharge data processing method according to any one of claims 4 to 6, characterized in that: The partial discharge signal characteristics include but are not limited to peak voltage, average voltage and root mean square voltage.

8. A cable partial discharge data processing system, characterized in that: include: A construction module for constructing a partial discharge and signal sample library; An extraction and identification module is used to automatically extract and identify partial discharge signals; The calculation and updating module is used to perform similarity calculation on the extracted signals and to update the sample library by self-learning.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images