GIS particle defect evaluation method and equipment fusing vibration and partial discharge

Abstract

The invention relates to a GIS particle defect evaluation method and equipment fusing vibration and partial discharge, and the method comprises the steps: obtaining an ultrasonic signal and an ultrahigh frequency signal caused by the partial discharge of a GIS, generating a phase distribution diagram of the partial discharge, comparing the phase distribution diagram with a reference map, and judging whether a particle defect exists or not; for particle defects, the particle size is calculated based on the average discharge capacity and the maximum discharge capacity of the ultrahigh-frequency signals; and acquiring a vibration signal, combining the calculated particle size, outputting the predicted particle horizontal diffusion area and the maximum jumping height, and calculating a danger degree index of the particle defect so as to evaluate the danger degree of the GIS particle defect. Compared with the prior art, the method has the advantages that the sensitivity, the accuracy and the real-time performance of detection are improved, and more comprehensive evaluation is provided for the health state of the GIS equipment.

Description

Technical Field

[0001] The invention relates to the technical field of gas insulated switchgear, and in particular to a GIS particle defect assessment method and device integrating vibration and partial discharge. Background Art

[0002] Gas Insulated Switchgear (GIS) plays a vital role in modern power systems. Compared with traditional air-insulated switchgear, GIS has the advantages of small footprint, low maintenance requirements, and high operational reliability, and is widely used in high-voltage and ultra-high-voltage power systems. However, insulation failure inside GIS is still one of the main problems affecting its operational stability, among which partial discharge (PD) phenomenon is considered to be an important factor leading to the degradation of insulation performance. PD is usually caused by various defects, including manufacturing defects, assembly defects, and particle defects formed during operation.

[0003] Particle defects are a common problem inside GIS, usually formed by tiny particles of conductive or semi-conductive materials suspended in the insulating gas inside the GIS. Under the action of strong electric fields, these particles may be accelerated and hit the surface of the insulating material during their movement, causing partial discharge. Over time, this partial discharge will gradually degrade the insulating material, resulting in a gradual decline in the insulation performance of the GIS, and may even cause serious electrical failures. Therefore, early detection and evaluation of particle defects are crucial to ensure the safe operation of the GIS.

[0004] Traditional particle defect detection methods mainly rely on electrical signal analysis, such as the use of ultrasonic and ultra-high frequency (UHF) signals to detect partial discharge. However, relying solely on these signals may not fully reveal all the information about particle defects. In fact, the movement and collision of particles inside GIS are often accompanied by certain mechanical vibrations. These vibration signals can reflect the movement characteristics of particles and the development status of defects. If the vibration signal can be combined with the electrical signal for analysis, it will help to more comprehensively evaluate the existence of particle defects and their potential harmfulness.

[0005] Existing studies have shown that vibration signals have important applications in mechanical fault diagnosis, but their application in GIS particle defect detection is still in its infancy. Therefore, it is necessary to propose a new evaluation method that combines vibration signals with partial discharge signals to comprehensively analyze and evaluate the characteristics of particle defects inside GIS. This method can not only improve the accuracy of defect detection, but also predict the development trend of defects, providing strong support for the operation and maintenance of GIS.

[0006] How to achieve GIS particle defect assessment by integrating vibration signals and partial discharge signals has become a technical problem that needs to be solved. Summary of the invention

[0007] The purpose of the present invention is to overcome the defects of the above-mentioned prior art and to provide a GIS particle defect evaluation method and equipment integrating vibration and partial discharge.

[0008] The purpose of the present invention can be achieved by the following technical solutions:

[0009] According to one aspect of the present invention, a GIS particle defect assessment method integrating vibration and partial discharge is provided, the method comprising the following steps:

[0010] Step S1, obtaining ultrasonic signals and UHF signals caused by GIS partial discharge, generating a phase distribution diagram of the partial discharge, and comparing it with a reference spectrum to determine whether there are particle defects;

[0011] Step S2, if it is determined in step S1 that there is a particle defect, the particle size d is calculated based on the average discharge amount and the maximum discharge amount of the UHF signal;

[0012] Step S3, obtaining the vibration signal, combining it with the particle size calculated in step S2, and outputting the predicted particle horizontal diffusion area S and maximum jumping height h;

[0013] Step S4, comprehensively considering the particle horizontal diffusion area S, the maximum jumping height h and the particle size d, calculates the danger level index of the particle defect to evaluate the danger level of the GIS particle defect.

[0014] Preferably, the process of determining whether there are particle defects in step S1 includes:

[0015] Step S1-1, converting the phase distribution map into a two-dimensional image form suitable for convolutional neural network input;

[0016] Step S1-2, the convolutional neural network automatically extracts and learns features through multi-layer convolution operations. At the same time, the convolution layer uses filters to perform local perception on the input image and combines features layer by layer in the convolution layer to build a high-level understanding of the graph;

[0017] Step S1-3, the convolutional neural network outputs the judgment result through the fully connected layer to identify whether there are particle defects in the phase distribution map.

[0018] Preferably, the step S2 further comprises determining the particle size according to the particle diameter, specifically, sorting the particle diameters from large to small, and classifying the particles into three categories according to a preset threshold, namely large particles, medium particles and small particles.

[0019] Preferably, the calculated particle size d is specifically:

[0020] d=k1p max +k2p mean

[0021] Among them, k1 and k2 are proportional coefficients, p max is the maximum discharge capacity, p mean is the average discharge capacity.

[0022] Preferably, the step S3 is specifically:

[0023] Consider the vibration amplitude V m , vibration frequency f and particle size d, the maximum jumping height h, particle horizontal diffusion distance r and particle horizontal diffusion area S are calculated according to the following formula:

[0024]

[0025] S=πr 2

[0026] Among them, k1 and k2 are proportional coefficients, and α, β, p and q are constants related to the measured data.

[0027] Preferably, the calculation of the risk index of particle defects in step S4 is specifically as follows:

[0028]

[0029] Among them, R is the danger index of particle defects; S is the horizontal diffusion area of ​​the particles; h is the maximum jumping height of the particles, which reflects the activity and energy state of the particles under vibration conditions; d is the particle size; k is the proportional constant; α is the influence index of particle size on the danger level; β is an index reflecting the comprehensive influence of S and h on the danger level.

[0030] Preferably, according to the risk level index of the particle defects, risk thresholds of different levels are set, and the risk levels are divided into five levels.

[0031] More preferably, the risk levels are divided into five levels:

[0032] Low risk: The risk index of particle defects is R<10;

[0033] Slight risk: The danger index of particle defects is 10≤R<50;

[0034] Medium risk: The danger index of particle defects is 50≤R<100;

[0035] High risk: The danger index of particle defects is 100≤R<200;

[0036] Very high risk: The danger level index of particle defects is R ≥ 200.

[0037] More preferably, when the danger level is low risk, it indicates that the GIS equipment is in a safe state and only requires routine monitoring; when the danger level is slight risk or medium risk, it indicates that the GIS equipment needs increased attention and should be inspected at an appropriate time; when the danger level is high risk or extremely high risk, it indicates that the GIS equipment is already in a dangerous state and immediate action is needed to prevent serious failures.

[0038] According to another aspect of the present invention, an electronic device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and the method described above is implemented when the processor executes the program.

[0039] Compared with the prior art, the present invention has the following beneficial effects:

[0040] 1) The present invention integrates vibration signals and partial discharge signals. First, a phase distribution diagram of partial discharge is generated based on the ultrasonic signal and UHF signal caused by GIS partial discharge, and compared with the reference spectrum to determine whether there are particle defects. Then, the particle size is calculated based on the average discharge amount and maximum discharge amount of the partial discharge signal, and the particle size is determined. The particle diffusion area and maximum jump height are predicted in combination with the vibration signal. Finally, the danger level of GIS particle defects is quantitatively calculated and divided into five levels according to the danger level. Compared with detection and evaluation based on partial discharge signals alone, the sensitivity and accuracy of detection are improved, a more comprehensive evaluation of the health status of GIS equipment is provided, and potential failure risks are prevented.

[0041] 2) The present invention comprehensively analyzes vibration signals and local discharge signals. The vibration signal, as a key auxiliary parameter, can further reveal the diffusion behavior of particles under various vibration conditions and more comprehensively and accurately predict the evolution trend of particles of different sizes, thereby providing high-sensitivity particle state detection and improving evaluation accuracy and real-time performance.

[0042] 3) The present invention introduces a convolutional neural network to automatically extract and learn features, build a high-level understanding of the spectrum, and identify particle defects; the method for calculating the particle size and the danger index of the particle defects is simple, which ensures the accuracy of the assessment and the real-time nature of the detection, and can timely predict the development trend of faults and reduce false alarms and missed alarms. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 Schematic diagram of the process of the GIS particle defect evaluation method of the present invention;

[0044] Figure 2 is a typical PRPD spectrum of a suspended defect in an embodiment of the present invention;

[0045] Figure 3 is a typical PRPD spectrum of a spike defect in an embodiment of the present invention;

[0046] Figure 4 is a typical PRPD spectrum of surface defects in an embodiment of the present invention;

[0047] Figure 5 It is a typical PRPD spectrum of particle defects in the embodiment of the present invention. DETAILED DESCRIPTION

[0048] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0049] The present invention proposes a GIS particle defect assessment method taking vibration into consideration, aiming to simultaneously consider the ultrasonic signals and UHF signals caused by partial discharge and the vibration signals caused by mechanical defects. Through multi-dimensional data analysis, effective identification and judgment of particle defects in GIS can be achieved, the accuracy of particle defect detection can be improved, and more comprehensive information support can be provided for fault diagnosis of GIS equipment, which is of great significance for improving the operational safety of GIS equipment.

[0050] Example 1

[0051] The present embodiment relates to a GIS particle defect assessment method that integrates vibration and partial discharge. By comprehensively considering the phase distribution diagram (PRPD spectrum) obtained by ultrasonic signals, the PRPD spectrum, average discharge amount and maximum discharge amount obtained by ultra-high frequency signals, and the vibration amplitude and frequency obtained by vibration signals, and comparing them with the existing database, the particle defect type judgment, particle size judgment, development feature judgment and danger level assessment are carried out in sequence, and finally the GIS particle defect assessment considering vibration is realized.

[0052] This method mainly includes three key information acquisition processes:

[0053] 1) Obtaining a PRPD spectrum using an ultrasonic signal detected by an ultrasonic-vibration fusion sensor, wherein the ultrasonic signal includes ultrasonic amplitude information and corresponding AC voltage phase for drawing the PRPD spectrum, and the data volume must meet the drawing requirements;

[0054] 2) Using the UHF signal detected by the UHF sensor to obtain the PRPD spectrum, average discharge and maximum discharge, where the UHF signal includes the UHF signal amplitude and the corresponding AC voltage phase used to draw the PRPD spectrum, as well as the average discharge and maximum discharge obtained by statistics, and the data volume must meet the requirements of drawing and statistical analysis;

[0055] 3) Obtain the vibration amplitude and frequency from the vibration signal detected by the ultrasonic-vibration fusion sensor, where the vibration signal includes the time domain waveform and the vibration amplitude and frequency obtained from the time domain waveform. The data volume must meet the requirements of drawing and statistical analysis.

[0056] The ultrasonic-vibration fusion sensor is a detection device installed on the surface of GIS, which can simultaneously detect the vibration signals caused by mechanical defects and the ultrasonic signals caused by partial discharge.

[0057] The UHF sensor is a detection device installed outside or inside the GIS, which can sense the UHF signal generated by partial discharge.

[0058] Based on the three key information obtained above, the following data preparation is required before using this method:

[0059] Step S101, using an ultrasound-vibration fusion sensor to obtain an ultrasonic signal, and then drawing a PRPD spectrum based on the ultrasonic signal;

[0060] Step S102, using a UHF sensor to obtain a UHF signal, and then drawing PRPD spectra in sequence, and obtaining an average discharge amount and a maximum discharge amount;

[0061] Step S103, using the ultrasonic-vibration fusion sensor to obtain a vibration signal, and then obtain the corresponding vibration amplitude V m , statistical data of vibration frequency f;

[0062] Step S104, preparing a database that can reflect typical characteristics of different defects.

[0063] A database refers to a highly structured data set that specifically stores and manages multidimensional information related to typical particle defects in GIS. The database not only covers the physical and electrical characteristics of defects, such as particle size, morphology, material composition, etc., but also includes typical signal characteristics caused by these defects, including but not limited to the spectrum and amplitude of ultrasonic signals, ultra-high frequency signals and vibration signals. The database aims to provide a standardized reference system to support the accurate matching and comparison of real-time data with known defect patterns. In practical applications, by comparing the collected PRPD spectra with the standard spectra in the database, the system can quickly identify potential defect types and determine whether there are particle defects. In addition, the database stores data on defect development history and trend analysis, so that the system can not only identify the current defect status, but also predict the possible evolution path of the defects, providing in-depth analysis basis for the operation and maintenance of equipment. To ensure effectiveness and accuracy, the database has the feature of regular updates.

[0064] This method will make a comprehensive judgment on the insulation problems caused by particle defects based on the above-mentioned information and the integrated database, thereby achieving an accurate assessment of the degree of danger.

[0065] After completing the preliminary preparations, the assessment work is then carried out according to the following steps. The method includes the following steps:

[0066] Step S1, processing the ultrasonic and UHF signals caused by partial discharge (PD), generating a partial discharge phase distribution (PRPD), and then comparing it with the reference spectrum in the database to determine whether there are particle defects. This process uses a convolutional neural network (CNN) to identify particle defects in the PRPD map.

[0067] First, the PRPD data is converted into a two-dimensional image suitable for CNN input. Different defect types in the PRPD image show significant differences in morphological characteristics, including the change pattern of discharge intensity with phase, amplitude distribution and signal frequency components. For suspended defects, they appear as "bars" in the UHF discharge spectrum. The discharge occurs in the rising stage of the positive semi-axis and the falling stage of the negative semi-axis. In the development stage of local discharge, with the increase of the external applied voltage, the discharge phase center of the positive and negative half-cycles moves toward the zero point. In the UHF discharge spectrum, the discharge frequency at the zero point is more and the color is redder. The discharge of spike defects is mainly divided into two stages. When the voltage is low, only a small amount of discharge pulses appear in the negative semi-axis of the power frequency, and there is basically no discharge in the positive semi-axis. When the voltage continues to increase, the positive semi-axis of the power frequency begins to discharge, and the discharge amplitude is significantly greater than the negative semi-cycle. The peak values ​​of the discharge pulses in the positive and negative semi-cycles are both near the positive and negative peak values ​​of the power frequency phase. With the increase of the external applied voltage, the discharge amplitude gradually increases. For surface defects, in the initial stage, the discharge occurs in the negative half cycle of the power frequency. As the external voltage increases, the discharge amplitude increases, and the phase center of the discharge pulse shifts toward the zero point of the power frequency voltage. When the external voltage reaches a certain level, a discharge pulse appears in the positive half axis of the power frequency. For particle defects, the discharge amplitude is widely distributed, the polarity effect is not obvious, and the discharge signal is distributed throughout the entire power frequency cycle.

[0068] Then, CNN automatically extracts and learns these features through multi-layer convolution operations. At the same time, the convolution layer uses filters to perform local perception on the input image, capturing specific patterns in the image, such as edges, corners, and textures, and combining these features layer by layer in subsequent convolution layers to build a high-level understanding of the graph.

[0069] Finally, the network outputs the judgment result through the fully connected layer to identify whether there are particle defects in the graph.

[0070] Step S2: When it is determined in step S1 that there are particle defects, the particle size is further analyzed, and the particle size is determined by the average discharge amount and the maximum discharge amount of the UHF signal.

[0071] First, a large amount of local discharge signal data is collected to calculate the average discharge amount p of each particle discharge event. mean and the maximum discharge capacity p max , and then calculate the particle size d, specifically:

[0072] d=k1p max +k2p mean

[0073] Among them, k1 and k2 are proportional coefficients, p max is the maximum discharge capacity, p mean is the average discharge capacity.

[0074] Then, the classification algorithm is applied to sort and classify the data, sorting the particle sizes d from large to small, and classifying the particles into three categories according to the preset thresholds:

[0075] The first category is particles with a diameter of more than 1 cm. Their discharge characteristics are usually manifested as higher average discharge and maximum discharge. Due to their larger size, the energy release during the discharge process of these particles is more significant, resulting in stronger signal amplitude and longer duration in the ultra-high frequency (UHF) signal.

[0076] The second category is particles with a diameter between 100 microns and 1 cm. Their discharge characteristics are between large particles and small particles. The average discharge amount and maximum discharge amount are significantly lower than those of large particles but higher than those of small particles.

[0077] The third category is particles with a particle size of less than 100 microns. The energy released during the discharge process is the smallest, and its signal characteristics are manifested as lower average discharge and maximum discharge.

[0078] By analyzing and classifying these data, the size distribution of the particles can be effectively determined, and the output results are "large particles" (particle size greater than 1cm), "medium particles" (particle size between 100μm-1cm) and "small particles" (particle size less than 100μm).

[0079] Step S3, based on the judgment result of step S2, introduce the vibration signal for comprehensive analysis to more accurately predict the development characteristics of the particles. By combining the UHF signal with the vibration signal, the evolution trend of particles of different sizes can be predicted more comprehensively and accurately. As a key auxiliary parameter, the vibration signal can further reveal the diffusion behavior of particles under various vibration conditions. The vibration amplitude is closely related to the diffusion behavior of particles, and there is a significant correlation: the higher the vibration amplitude, the more obvious the horizontal diffusion in the medium.

[0080] For large particles, due to their large mass, their inertial resistance increases at high vibration amplitudes, but their diffusion still shows a certain growth trend. However, since large particles may be more inclined to settle or aggregate under vibration rather than further diffuse, the diffusion rate gradually slows down or even stagnates as the vibration amplitude increases.

[0081] The diffusion behavior of medium particles is most significant at high vibration amplitudes, because their size and mass are moderate, and they can fully diffuse under the drive of the vibration signal, which is usually manifested as a linear relationship between the vibration amplitude and the diffusion level: the higher the vibration amplitude, the wider the diffusion range of medium particles. This trend occurs because medium particles are neither difficult to diffuse like large particles due to inertia, nor easy to agglomerate like small particles due to high surface energy under the action of vibration energy.

[0082] In contrast, although small particles are light and easily affected by vibration, their diffusion characteristics are not significant under high vibration amplitudes. Small particles have high surface energy and are easily attracted and aggregated to form clusters. This aggregation effect may be enhanced under high vibration conditions. Therefore, the diffusion of small particles presents the other end of the "U"-shaped curve within a certain vibration amplitude range, that is, as the vibration amplitude increases, its diffusion trend weakens.

[0083] By comprehensively analyzing the vibration signal and the partial discharge signal, we can more comprehensively understand the dynamic behavior of the particles, thereby improving the accuracy of particle size judgment and providing strong support for the prediction of particle diffusion characteristics. The final output result is the predicted particle horizontal diffusion area S and maximum jump height h. The calculation process of the particle horizontal diffusion area S and maximum jump height h is as follows:

[0084] After the particle size d is obtained according to step S2, the vibration amplitude V is also considered. m , vibration frequency f and particle size d, the maximum jumping height h, particle horizontal diffusion distance r and particle horizontal diffusion area S are calculated according to the following formula:

[0085]

[0086] S=πr 2

[0087] Among them, k1 and k2 are proportional coefficients, and α, β, p and q are constants related to the measured data.

[0088] Step S4, in order to quantitatively evaluate the degree of danger of particle defects, based on the development characteristics of step S3, a calculation formula is designed that comprehensively considers the horizontal diffusion area S of the particles, the maximum jump height h, and the particle size d. This formula reflects the impact of the three variables on the degree of danger, and the results reflect the potential threat of particles to the safety of electrical equipment.

[0089]

[0090] In the above formula, R is the danger index of particle defects; S is the horizontal diffusion area of ​​the particles. Generally, the larger the area, the wider the activity range of the particles in the medium and the greater the potential harm; h is the maximum jumping height of the particles, which reflects the activity and energy state of the particles under vibration conditions; d is the particle size. The smaller the particle size, the easier it is for the particles to enter a narrow space or be captured by the electric field, resulting in partial discharge and insulation failure; k is the proportional constant; α is the influence index of particle size on the danger level, which ranges from 0.5 to 1; β is an index reflecting the combined influence of S and h on the danger level, which ranges from 1 to 2.

[0091] The larger the horizontal diffusion area S, the higher the maximum jump height h, indicating that the diffusion range of the particles is farther. For particles on the inner wall surface of GIS, it is possible to diffuse into the cavity and cause insulation breakdown or jump to the surface of the insulator to cause surface flashover, causing a significant reduction in insulation. The smaller the particle size d, the easier it is to diffuse into a small space or an area with concentrated electric fields, thus posing a potential threat.

[0092] Based on the above judgment, the risk index R of particle defects is comprehensively obtained, and the risk level is divided into 5 levels: low risk, slight risk, medium risk, high risk and extremely high risk:

[0093] (1) Low risk (R<10): The risk level of particle defects is very low and has little impact on the operation of the equipment. Recommended measures: No additional measures are required, only regular monitoring is required.

[0094] (2) Minor risk (10≤R<50): There is a minor risk of particle defects, which may have a minor impact on the equipment, but is unlikely to cause serious failure in the short term. Recommended actions: Continue to monitor the equipment condition and perform inspections during routine maintenance.

[0095] (3) Medium risk (50 ≤ R < 100): The particle defect is of moderate danger and may cause the performance of the equipment to deteriorate. Long-term operation may lead to more serious problems. Recommended measures: Strengthen monitoring and consider further inspection or treatment during the next maintenance cycle.

[0096] (4) High risk (100 ≤ R < 200): The particle defect is highly dangerous and poses a greater threat to the operational safety of the equipment, which may cause equipment failure in the short term. Recommended measures: Arrange maintenance or overhaul as soon as possible, and perform preventive shutdown if necessary.

[0097] (5) Extremely high risk (R ≥ 200): The risk of particle defects is extremely high. The equipment may fail seriously at any time, posing a major safety hazard. Recommended measures: Immediately shut down the equipment and conduct a comprehensive overhaul to eliminate potential safety risks.

[0098] This interval division sets different risk thresholds so that the danger index R can intuitively guide maintenance and operation decisions. The low risk interval indicates that the equipment is in a safe state and only requires routine monitoring; the slight and medium risk intervals indicate that more attention needs to be paid and inspections should be carried out at the appropriate time; the high risk and extremely high risk intervals indicate that the equipment is already in a dangerous state and immediate action is required to prevent serious failures.

[0099] Example 2

[0100] The electronic device of the present invention includes a central processing unit (CPU), which can perform various appropriate actions and processes according to computer program instructions stored in a read-only memory (ROM) or loaded from a storage unit into a random access memory (RAM). In the RAM, various programs and data required for device operation can also be stored. The CPU, ROM and RAM are connected to each other via a bus. An input / output (I / O) interface is also connected to the bus.

[0101] Multiple components in the device are connected to the I / O interface, including: input units, such as keyboards, mice, etc.; output units, such as various types of displays, speakers, etc.; storage units, such as disks, optical disks, etc.; and communication units, such as network cards, modems, wireless communication transceivers, etc. The communication unit allows the device to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunication networks.

[0102] The processing unit performs the various methods and processes described above, such as methods S1 to S4. For example, in some embodiments, methods S1 to S4 may be implemented as a computer software program, which is tangibly contained in a machine-readable medium, such as a storage unit. In some embodiments, part or all of the computer program may be loaded and / or installed on the device via a ROM and / or a communication unit. When the computer program is loaded into the RAM and executed by the CPU, one or more steps of methods S1 to S4 described above may be performed. Alternatively, in other embodiments, the CPU may be configured to execute methods S1 to S4 in any other appropriate manner (e.g., by means of firmware).

[0103] The functions described above herein may be performed at least in part by one or more hardware logic components. For example, without limitation, exemplary types of hardware logic components that may be used include: field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), and the like.

[0104] The program code for implementing the method of the present invention can be written in any combination of one or more programming languages. These program codes can be provided to a processor or controller of a general-purpose computer, a special-purpose computer or other programmable data processing device, so that the program code, when executed by the processor or controller, enables the functions / operations specified in the flow chart and / or block diagram to be implemented. The program code can be executed entirely on the machine, partially on the machine, partially on the machine as a stand-alone software package and partially on a remote machine, or entirely on a remote machine or server.

[0105] In the context of the present invention, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0106] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed by the present invention, and these modifications or replacements should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. A GIS particle defect assessment method integrating vibration and partial discharge, characterized in that: The method comprises the following steps: Step S1, obtaining ultrasonic signals and UHF signals caused by GIS partial discharge, generating a phase distribution diagram of the partial discharge, and comparing it with a reference spectrum to determine whether there are particle defects; Step S2, if it is determined in step S1 that there is a particle defect, the particle size d is calculated based on the average discharge amount and the maximum discharge amount of the UHF signal; Step S3, obtaining the vibration signal, combining it with the particle size calculated in step S2, and outputting the predicted particle horizontal diffusion area S and maximum jumping height h; Step S4, comprehensively considering the particle horizontal diffusion area S, the maximum jumping height h and the particle size d, calculates the danger level index of the particle defect to evaluate the danger level of the GIS particle defect.

2. A GIS particle defect assessment method integrating vibration and partial discharge according to claim 1, characterized in that: The process of determining whether there is a particle defect in step S1 includes: Step S1-1, converting the phase distribution map into a two-dimensional image form suitable for convolutional neural network input; Step S1-2, the convolutional neural network automatically extracts and learns features through multi-layer convolution operations. At the same time, the convolution layer uses filters to perform local perception on the input image and combines features layer by layer in the convolution layer to build a high-level understanding of the graph; Step S1-3, the convolutional neural network outputs the judgment result through the fully connected layer to identify whether there are particle defects in the phase distribution map.

3. The GIS particle defect assessment method integrating vibration and partial discharge according to claim 1 is characterized in that: The step S2 also includes determining the particle size according to the particle size, specifically, sorting the particle sizes from large to small, and classifying the particles into three categories according to a preset threshold, namely large particles, medium particles and small particles.

4. The GIS particle defect assessment method integrating vibration and partial discharge according to claim 1 is characterized in that: The calculated particle size d is specifically: d=k1p max +k2p mean Among them, k1 and k2 are proportional coefficients, p max is the maximum discharge capacity, p mean is the average discharge capacity.

5. The GIS particle defect assessment method integrating vibration and partial discharge according to claim 1 is characterized in that: The step S3 is specifically as follows: Consider the vibration amplitude V m , vibration frequency f and particle size d, the maximum jumping height h, particle horizontal diffusion distance r and particle horizontal diffusion area S are calculated according to the following formula: S＝πr 2 Among them, k1 and k2 are proportional coefficients, and α, β, p and q are constants related to the measured data.

6. The GIS particle defect assessment method integrating vibration and partial discharge according to claim 1 is characterized in that: The calculation of the risk index of particle defects in step S4 is specifically as follows: Among them, R is the danger index of particle defects; S is the horizontal diffusion area of ​​the particles; h is the maximum jumping height of the particles, which reflects the activity and energy state of the particles under vibration conditions; d is the particle size; k is the proportional constant; α is the influence index of particle size on the danger level; β is an index reflecting the comprehensive influence of S and h on the danger level.

7. The GIS particle defect assessment method integrating vibration and partial discharge according to claim 1 is characterized in that: According to the risk level index of the particle defects, different levels of risk thresholds are set, and the risk levels are divided into five levels.

8. The GIS particle defect assessment method integrating vibration and partial discharge according to claim 7 is characterized in that: The five levels of danger are as follows: Low risk: The risk index of particle defects is R<10; Slight risk: The danger index of particle defects is 10≤R<50; Medium risk: The danger index of particle defects is 50≤R<100; High risk: The danger index of particle defects is 100≤R<200; Very high risk: The danger level index of particle defects is R ≥ 200.

9. The GIS particle defect assessment method integrating vibration and partial discharge according to claim 8 is characterized in that: When the risk level is low risk, it indicates that the GIS equipment is in a safe state and only requires routine monitoring; when the risk level is slight risk or medium risk, it indicates that the GIS equipment needs to be paid more attention and inspected at an appropriate time; When the risk level is high risk or extremely high risk, it indicates that the GIS equipment is already in a dangerous state and immediate action is required to prevent serious failures.

10. An electronic device comprising a memory and a processor, wherein a computer program is stored in the memory, wherein: When the processor executes the program, the method according to any one of claims 1 to 9 is implemented.

Images