GIS partial discharge automatic positioning and processing method

Abstract

The invention relates to the technical field of power equipment detection, in particular to a GIS partial discharge automatic positioning and processing method. The method comprises the following steps: S1, building a GIS simulation model; s2, extracting an original signal for effective judgment; s3, calculating a discharge position according to the effective signal; s4, grading the signals, and determining a final discharge position; s5, performing dynamic discharge position updating operation; and S6, determining an early warning level, and taking corresponding measures. The method has the advantages that the electromagnetic wave velocity is corrected by introducing temperature and pressure, so that the positioning accuracy can be effectively improved; the loss of the AI model can be effectively reduced through signal grading processing, and the cost is reduced; according to the development trend of the partial discharge source, the influence area of the discharge source can be monitored in a targeted manner, so that on one hand, the monitoring strength is improved, and on the other hand, resource loss in irrelevant areas is reduced; the early warning stage treatment further reduces the cost on the basis of ensuring effective treatment of discharge of the discharge source.

Description

Technical Field

[0001] This invention relates to the field of power equipment testing technology, and in particular to a method for automatic location and processing of partial discharge in GIS. Background Technology

[0002] Gas-insulated switchgear (GIS) is widely used in power systems due to its advantages such as good insulation, high reliability, and small footprint. However, due to factors such as improper installation, switching operation, and external environment, partial discharge defects can easily occur inside GIS. Long-term accumulation may lead to insulation failure and equipment malfunction. Therefore, partial discharge detection is crucial to ensuring the safe operation of GIS.

[0003] Current mainstream methods for GIS partial discharge localization primarily rely on time-of-flight (TOF) positioning. However, TOF positioning requires capturing discharge signals, and GIS equipment operates in complex environments, making it highly susceptible to external signals such as transformer sprinkler discharge, environmental electromagnetic radiation, and interference from other electrical equipment. Some researchers have incorporated AI into localization methods, achieving promising results. However, AI prediction consumes significant computing power and is costly. Summary of the Invention

[0004] The main objective of this invention is to provide an automatic location and processing method for partial discharge in GIS, thereby solving the problems mentioned in the background art.

[0005] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is: an automatic location and processing method for partial discharge in GIS, comprising the following steps: S1. Build a GIS simulation model; S2. Extract the original signal for effective judgment; S3. Calculate the discharge location based on the valid signal; S4. Classify the signal and determine the final discharge location; S5. Perform dynamic discharge position update operation; S6. Determine the warning level and take corresponding measures.

[0006] Furthermore, the detailed process of step S1 is as follows: S101. Perform geometric modeling in SolidWorks to construct a 3D model; the model includes: GIS core detection-related structures, including the outer surface of pipes, insulating basins, and key functional components; key functional components include: circuit breakers, busbars, joints, and disconnect switches; Key annotations are as follows: Mark the sensor deployment location. On the edge of the unshielded insulating basin, the observation window, and other detectable locations, mark "Sensor Installation Point" and record the unique number and three-dimensional coordinates of the installation point; Label the attributes of the components, including the type and material of each component; The fixed dimensional parameters are marked, including the length of each component, the pipe turning angle, the spacing of sensor mounting holes, and the length of the pipe path between adjacent mounting points; S102. Add supplementary information; Supplementary information includes: simulation scenario, electromagnetic wave propagation speed, and component attenuation coefficient; The simulation scenarios include temperature scenarios, pressure scenarios, and partial discharge source location scenarios; the partial discharge source location scenarios cover all key areas of the GIS, including the inside of insulators, the center of busbars, the inside of circuit breaker compartments, pipe bends, and joint connections. The electromagnetic wave propagation speed library corrects the wave speed based on temperature and pressure: (1); in, The corrected electromagnetic wave speed. The electromagnetic wave speed under standard operating conditions. , These are the current air pressure and temperature, respectively. , These are the air pressure and temperature corresponding to standard operating conditions; The expression for the component attenuation coefficient is: (2); in, For the first The attenuation coefficient of each component, The basic attenuation coefficient at standard temperature, For the first Temperature sensitivity coefficient of each component For the first The aging coefficient of each component; the mathematical expression for the aging coefficient is as follows: (3); in, This is the aging factor over time. For the service life of the equipment, The load aging factor, For the first Section load operation load level, For the first The service life of the load section For aging coefficient under special working conditions, For special operating conditions; For components that operate under high humidity conditions for extended periods, a humidity aging factor is introduced into the component attenuation coefficient for correction. The mathematical expression is as follows: (4); in, Humidity aging coefficient, For the first Average annual relative humidity; For metal components, a vibration aging coefficient is introduced into the component attenuation coefficient for correction, and the mathematical expression is as follows: (5); in, The vibration aging coefficient, For the first Annual average equipment vibration acceleration; If a metal component operates in a high humidity environment for a long time, both the humidity aging coefficient and the vibration aging coefficient should be included in the component attenuation coefficient. Amplitude of electromagnetic wave after attenuation The expression is: (6); in, The amplitude of the electromagnetic wave before attenuation. To adjust the coefficient, The number of components corresponding to the component attenuation coefficient. This refers to the attenuation of electromagnetic waves caused by multiple components along a path. The expression is: (7); in, This represents the total attenuation of the components. S103. Construct a real fault information database, including historical location data, partial discharge signal characteristics, equipment structural parameters, and maintenance measurement data.

[0007] Furthermore, the detailed process of step S2 is as follows: S201. Perform multimodal data acquisition; multimodal data includes: electromagnetic wave signals, ultrasonic signals, vibration signals, local temperature data, and gas concentration data. S202, Perform hierarchical signal preprocessing; For electromagnetic wave signals, interference suppression is performed using the following expression: (8); in, The signal after interference suppression. Electromagnetic wave signals acquired by a UHF sensor. For reference, the environmental interference signals collected by the sensor, The adaptive filter weights at the current moment; The adaptive filter weights are updated iteratively, as shown in the following expression: (9); in, For the adaptive filter weights in the next time step, This is the step size coefficient; For weak signals, an amplification operation is performed, expressed as follows: (10); in, This is the magnification factor. The average amplitude of the signal. This represents the full-scale amplitude of the signal. The enhanced signal can be obtained by multiplying the signal by the corresponding amplification factor. The amplified signal undergoes feature reconstruction, as shown in the following expression: (11); in, For the reconstructed signal, The signal is amplified. The reconstruction function represents the signal phase. With pulse rise time The inherent relationship between them; Multipath correction is performed on electromagnetic wave signals, specifically as follows: Based on the 3D topology of GIS pipelines, the time difference between the direct propagation path and the reflection path of electromagnetic waves is calculated, as shown in the following expression: (12); in, For multipath time difference, The length of the reflection path. This is the direct path length; Based on the multipath time difference correction signal, the expression is as follows: (13); in, This is the signal after multipath time difference correction; S203. Find the sensor A with the earliest acquisition time, and record its location as point A. Record the frequency of all sub-signals in A. S204. Perform initial screening on the sub-signals in the remaining sensors, and remove sub-signals whose frequencies are not in A. S205. Determine the effective signal based on the attenuation rate: Calculate the attenuation rate on the line between adjacent sensors based on the amplitude of the sub-signal, and combine it with the attenuation rate in the simulation model to obtain the following judgment formula: (14); in, , Adjacent sensors and The actual attenuation rate and the model attenuation rate of the line between them A threshold is used to determine the attenuation level; When the attenuation rate of a sub-signal satisfies the judgment formula, the sub-signal is determined to be a valid sub-signal. The valid signal can be obtained by superimposing all valid sub-signals. If no valid sub-signal is found, it indicates that no fault has occurred; continue monitoring. S206. Record the power amplitude of the valid signal at point A. And calculate the noise intensity.

[0008] Furthermore, noise intensity The expression is: (15); in, Noise power, This represents the effective signal power amplitude.

[0009] Furthermore, the detailed process of step S3 is as follows: S301. Use the TDOA time difference method to calculate the rough estimate of the discharge location; S302. Use the correction model to correct the coarsely estimated discharge location, obtaining the corrected estimated discharge location; the distance between the corrected estimated discharge location and point A is... : (16); in, To roughly estimate the distance between the discharge location and point A.

[0010] Furthermore, in step S4, the criteria for signal grading are as follows: The standard for a Level 1 signal is: the signal meets the power amplitude requirement. And noise intensity ; The standard for a level 2 signal is: the signal meets the power amplitude requirement. Noise intensity One of the conditions, and the signal does not meet the classification criteria for a level three signal; The standard for a Level 3 signal is: the signal meets the power amplitude requirement. or noise intensity One of the conditions.

[0011] Furthermore, the detailed process of step S5 is as follows: S501. Use a predictive model to predict the development trend of the partial discharge source. The output of the predictive model is the amplitude growth rate of the partial discharge source within a future time period T. Frequency growth rate And extended prediction directions; S502. Determine the extended prediction area for the future time period T based on the attenuated electromagnetic wave velocity. S503, Setting a safe distance And it will turn off sensors that meet safety conditions, while increasing the sampling frequency of sensors that do not meet safety conditions; The safety conditions are as follows: (17); in, For point Minimum distance from the extended prediction region.

[0012] Furthermore, in step S4, the processing procedure for signals of different levels is as follows: For Level 1 signals, the coarsely estimated discharge location is directly used as the final discharge location, and then the positioning process ends. For secondary signals, the corrected estimated discharge position is used as the final discharge position, and then the positioning process ends. For a level 3 signal, the corrected estimated discharge position is taken as the final discharge position, and then the process proceeds to step S5.

[0013] Furthermore, the warning levels are divided into: primary warning, intermediate warning, and advanced warning; A primary warning must meet the following requirements: and ; Intermediate warning level must meet the following requirements: or And it does not meet the criteria for a high-level early warning; Advanced warnings must meet the following requirements: and .

[0014] Furthermore, the measures corresponding to the warning levels are as follows: For initial warnings, the measures are: continued monitoring; For intermediate warnings, the measures are as follows: live inspection, arrange maintenance personnel to conduct on-site inspections, and use infrared thermal imaging and ultrasonic-assisted detection to check whether there are any abnormalities around the partial emission source; Planned maintenance will be carried out, and the defect will be included in the next round of power outage maintenance plans, with targeted maintenance solutions developed. For advanced warnings, the measures are as follows: special emergency repairs are carried out, emergency maintenance procedures are initiated, the part where the partial discharge source is located is disassembled, and the fault is resolved, such as removing metal particles, replacing deteriorated insulation parts, and repairing poorly contacting conductive parts; then a withstand voltage test is carried out, and after maintenance, a partial discharge withstand voltage test of the GIS is carried out to verify that the defect has been eliminated before it can be put back into operation; The process involves reviewing and analyzing defects, recording their types, development processes, and corrective measures, updating the database, and optimizing subsequent correction and prediction algorithms.

[0015] Beneficial effects: (1) Introducing temperature and pressure to correct electromagnetic wave speed can effectively improve positioning accuracy; (2) Introduce an aging coefficient that includes three dimensions: time, load, and failure, to determine the aging degree of components, making the aging judgment of components more accurate; (3) The aging coefficient also takes into account two scenarios: high humidity and mechanical wear, further increasing the applicability of the aging coefficient; (4) Signal hierarchical processing can effectively reduce the loss of AI models and reduce costs; (5) The development trend of local discharge sources can be monitored in a targeted manner in the areas affected by the discharge sources, which can improve the monitoring efforts on the one hand and reduce resource consumption in irrelevant areas on the other. Attached Figure Description

[0016] The present invention will be further described below with reference to the accompanying drawings and embodiments: Figure 1 This is a flowchart of the steps of the present invention. Detailed Implementation

[0017] Example 1 like Figure 1 As shown, an automatic location and processing method for partial discharge in GIS includes the following steps: S1. Build a GIS simulation model. The detailed process is as follows: S101. Perform geometric modeling in SolidWorks, specifically by combining GIS drawings, field measurement data, and equipment parameters to construct a 3D model. To reduce model complexity, the model only constructs the core GIS detection-related structures, including the outer surface of the pipe, the insulating basin, and key functional components. Key functional components include: circuit breakers, busbars, joints, and disconnect switches. The modeling accuracy is as follows: geometric accuracy not exceeding 0.1mm, only the outer surface of the pipe is modeled, the interior is set as a hollow structure, and the pipe thickness is 0.001m; Key annotations are as follows: Mark the sensor deployment locations. On the edge of the unshielded insulating basin, the observation window, and other detectable locations, label the location as "Sensor Installation Point" and record its unique number and three-dimensional coordinates. For example, the nth installation point would be denoted as... ; Label the attributes of the components, including the type and material of each component; The fixed dimensional parameters are marked, including the length of each component, the pipe turning angle, the spacing of sensor mounting holes, and the length of the pipe path between adjacent mounting points; Finally, output the model file and parts list in STEP format; the parts list includes: part number, type, material, size parameters, and sensor mounting point coordinates, and is generally in Excel format; S102. Add supplementary information; Supplementary information includes: simulation scenario, electromagnetic wave propagation speed, and component attenuation coefficient; The simulation scenarios include temperature scenarios, pressure scenarios, and partial discharge source location scenarios. The temperature scenario uses 25 degrees Celsius as the standard operating temperature, the pressure scenario uses 0.6 MPa as the standard operating pressure, and the partial discharge source location scenario covers all key areas of the GIS, including the inside of insulators, the center of busbars, the inside of circuit breakers, pipe bends, and joint connections. The electromagnetic wave propagation speed library corrects the wave speed based on temperature and pressure: (1); in, The corrected electromagnetic wave speed. The electromagnetic wave speed under standard operating conditions. , These are the current air pressure and temperature, respectively. , These are the air pressure and temperature corresponding to standard operating conditions, typically taken as 0.6 MPa and 25 degrees Celsius. The component attenuation coefficient includes a time-related aging coefficient, which allows for dynamic correction over time. The expression for the component attenuation coefficient is: (2); in, For the first The attenuation coefficient of each component, The basic attenuation coefficient at standard temperature, For the first The temperature sensitivity coefficient of each component is generally determined experimentally. For the first The aging factor of each component; the aging factor encompasses the impact of time, load, and failure dimensions on the aging of the component throughout its life cycle, and its mathematical expression is as follows: (3); in, This is the aging factor over time. For the service life of the equipment, The load aging factor, For the first The load level for section operation is determined as a percentage of the rated load; for example, if the percentage is 90%, the value is 0.9. For the first The service life of the load section For aging coefficient under special working conditions, This refers to the number of times under special operating conditions, which include: lightning strikes, short circuit faults, and operation under over-temperature and over-pressure conditions. For components that operate under high humidity conditions for extended periods, the corrosion caused by the high humidity environment must also be considered. Therefore, a humidity aging factor is introduced into the component attenuation coefficient for correction, and the mathematical expression is as follows: (4); in, Humidity aging coefficient, For the first Average annual relative humidity; For metal components, mechanical wear must also be considered. A vibration aging coefficient is introduced into the component attenuation coefficient for correction, and the mathematical expression is as follows: (5); in, The vibration aging coefficient, For the first Annual average equipment vibration acceleration; If a metal component operates in a high humidity environment for a long time, then both the humidity aging coefficient and the vibration aging coefficient can be included in the component attenuation coefficient. The amplitude of the electromagnetic wave after attenuation The expression is: (6); in, The amplitude of the electromagnetic wave before attenuation. To adjust the coefficient, this value is compared with... It depends on the corresponding parameter type, such as voltage and other parameters unrelated to power. Take 10 parameters related to power. Take 20, This refers to the number of components corresponding to the component attenuation coefficient. For example, insulators have component attenuation coefficients measured in dB per component. The number of insulators, such as in continuous pipelines, is expressed in dB / m as the component attenuation coefficient. The length of the continuous pipe, in meters. This refers to the attenuation of electromagnetic waves caused by multiple components along a path. The expression is: (7); in, This represents the total attenuation of the components. To avoid erroneous corrections caused by instantaneous fluctuations, a multi-dimensional correction trigger mechanism is set up, specifically as follows: The triggering conditions are determined as follows: temperature fluctuation is not less than 5 degrees Celsius, pressure fluctuation is not less than 0.1 MPa, equipment operating years are not less than 3 years, and the frequency increase rate of partial discharge is not less than 15% / h; When any one of the above four triggering conditions is met, the electromagnetic wave propagation speed and attenuation coefficient are automatically corrected, and the triggering period is 5 minutes. S103. Construct a real fault information database, including historical location data, partial discharge signal characteristics, equipment structural parameters, and maintenance measurement data.

[0018] S2. Perform multimodal data acquisition and hierarchical signal preprocessing operations. The detailed process is as follows: S201. Perform multimodal data acquisition; the multimodal data consists of five dimensions: electrical, acoustic, vibrational, thermal, and chemical. Use a UHF sensor to collect electromagnetic wave signals in the 300-1500MHz frequency band; Ultrasonic signals in the 40-80kHz frequency band are acquired using an ultrasonic sensor. Vibration signals in the 10-1000Hz frequency band are collected using vibration sensors; Local temperature data is collected using an infrared sensor. Gas concentration data is collected using a gas sensor; S202, Perform hierarchical signal preprocessing; For electromagnetic wave signals, interference suppression is performed, specifically by using a reference sensor to collect environmental interference signals and filtering them out to obtain the interference-suppressed signal, as shown in the following expression: (8); in, The signal after interference suppression. Electromagnetic wave signals acquired by a UHF sensor. For reference, the environmental interference signals collected by the sensor, The adaptive filter weights at the current moment should ideally have an initial value of 0.01. The adaptive filter weights are updated iteratively, as shown in the following expression: (9); in, For the adaptive filter weights in the next time step, This is the step size coefficient, typically taken as 0.002; For weak signals, an amplification operation is performed, specifically by amplifying the signal's amplitude, as shown in the following expression: (10); in, This is the magnification factor. The average amplitude of the signal. This is the full-scale amplitude of the signal, a fixed parameter of the sensor; The enhanced signal can be obtained by multiplying the signal by the corresponding amplification factor. Since signal amplification also amplifies noise and distortion, a feature reconstruction operation is performed on the amplified signal, as shown in the following expression: (11); in, For the reconstructed signal, The signal is amplified. The reconstruction function represents the signal phase. With pulse rise time There is an inherent correlation between them. Since the rise time difference between the partial discharge signal and the interference signal is large, the phase information of the partial discharge signal can be reconstructed more accurately using the reconstruction function. Multipath correction is performed on electromagnetic wave signals, specifically as follows: Based on the 3D topology of GIS pipelines, the time difference between the direct propagation path and the reflection path of electromagnetic waves is calculated, as shown in the following expression: (12); in, This is the multipath time difference, which is the time difference between the arrival of the reflected path signal and the direct path signal. The length of the reflection path. This is the direct path length; Based on the multipath time difference correction signal, the expression is as follows: (13); in, The signal after multipath time difference correction has an amplitude that is given by formula (3). ; S203. Find the sensor A with the earliest acquisition time, and record its location as point A. Record the frequency of all sub-signals in A. S204. Perform initial screening on the sub-signals in the remaining sensors, and remove sub-signals whose frequencies are not in A. S205. Determine the effective signal based on the attenuation rate: Calculate the attenuation rate on the line between adjacent sensors based on the amplitude of the sub-signal, and combine it with the attenuation rate in the simulation model to obtain the following judgment formula: (14); in, , Adjacent sensors and The actual attenuation rate and the model attenuation rate of the line between them A threshold is used to determine the attenuation level; When the attenuation rate of a sub-signal satisfies the judgment formula, the sub-signal is determined to be a valid sub-signal. The valid signal can be obtained by superimposing all valid sub-signals. If no valid sub-signal is found, it indicates that no fault has occurred; continue monitoring. S206. Record the power amplitude of the valid signal at point A. And calculate the noise intensity. The expression is: (15); in, Noise power.

[0019] S3. Calculate the discharge location based on the valid signal. The detailed process is as follows: S301. Use the commonly used TDOA time difference method to calculate the rough estimate of the discharge location; S302. Use the correction model to correct the coarsely estimated discharge location to obtain the corrected estimated discharge location; The correction model is built using a backpropagation (BP) neural network. The inputs are the time difference calculation error, signal amplitude fluctuation, propagation path attenuation deviation, and environmental noise intensity. The output is the corrected distance. Then the distance between the corrected estimated discharge location and point A is: : (16); in, To roughly estimate the distance between the discharge location and point A.

[0020] S4. Classify the signal, determine the final discharge location, and establish the classification criteria as follows: Level 1 signal meets power amplitude And noise intensity At this time, there is less interference and the positioning is easier; The secondary signal satisfies the power amplitude Noise intensity If one of the conditions is met and the signal does not meet the classification criteria for a Level 3 signal, then the interference is moderate and the positioning difficulty is moderate. Level 3 signals meet power amplitude requirements. or noise intensity One of the conditions is that there is significant interference, making positioning more difficult.

[0021] For Level 1 signals, the coarsely estimated discharge location is directly used as the final discharge location, and then the positioning process ends. For secondary signals, the corrected estimated discharge position is used as the final discharge position, and then the positioning process ends. For a level 3 signal, the corrected estimated discharge position is taken as the final discharge position, and then the process proceeds to step S5.

[0022] S5. Perform dynamic discharge position update operation. The detailed process is as follows: S501. A predictive model is used to predict the development trend of the partial discharge source. The predictive model is built using an LSTM neural network. The input of the predictive model is the amplitude, pulse frequency, and energy of the partial discharge signal, and the output is the amplitude growth rate of the partial discharge source within the future time period T. Frequency growth rate And extended prediction directions; S502. Determine the extended prediction area for the future time period T based on the attenuated electromagnetic wave velocity. S503, Setting a safe distance And it will turn off sensors that meet safety conditions, while increasing the sampling frequency of sensors that do not meet safety conditions; The safety conditions are as follows: (17); in, For point Minimum distance from the extended prediction region.

[0023] S6, according to and Determine the warning level and take corresponding measures, specifically: The warning levels are divided into: primary warning, intermediate warning and advanced warning; A primary warning must meet the following requirements: and ; Intermediate warning level must meet the following requirements: or And it does not meet the criteria for a high-level early warning; Advanced warnings must meet the following requirements: and ; For initial warnings, the measures are: continued monitoring; For intermediate warnings, the measures are as follows: live inspection, arrange maintenance personnel to conduct on-site inspections, and use infrared thermal imaging and ultrasonic-assisted detection to check whether there are any abnormalities around the partial emission source; Planned maintenance will be carried out, and the defect will be included in the next round of power outage maintenance plans, with targeted maintenance solutions developed. For advanced warnings, the measures are as follows: special emergency repairs are carried out, emergency maintenance procedures are initiated, the part where the partial discharge source is located is disassembled, and the fault is resolved, such as removing metal particles, replacing deteriorated insulation parts, and repairing poorly contacting conductive parts; then a withstand voltage test is carried out, and after maintenance, a partial discharge withstand voltage test of the GIS is carried out to verify that the defect has been eliminated before it can be put back into operation; The process involves reviewing and analyzing defects, recording their types, development processes, and corrective measures, updating the database, and optimizing subsequent correction and prediction algorithms.

[0024] The above embodiments are merely preferred technical solutions of the present invention and should not be considered as limitations on the present invention. The scope of protection of the present invention should be limited to the technical solutions described in the claims, including equivalent substitutions of the technical features described in the claims. That is, equivalent substitutions and improvements within this scope are also within the scope of protection of the present invention.

Claims

1. A method for automatic location and processing of partial discharge in GIS, characterized in that, Includes the following steps: S1. Build a GIS simulation model; S2. Extract the original signal for effective judgment; S3. Calculate the discharge location based on the valid signal; S4. Classify the signal and determine the final discharge location; S5. Perform dynamic discharge position update operation; S6. Determine the warning level and take corresponding measures.

2. The method for automatic location and processing of partial discharge in GIS according to claim 1, characterized in that, The detailed process of step S1 is as follows: S101. Perform geometric modeling in SolidWorks to construct a 3D model; the model includes: GIS core detection-related structures, including the outer surface of pipes, insulating basins, and key functional components; key functional components include: circuit breakers, busbars, joints, and disconnect switches; Key annotations are as follows: Mark the sensor deployment location. On the edge of the unshielded insulating basin, the observation window, and other detectable locations, mark "Sensor Installation Point" and record the unique number and three-dimensional coordinates of the installation point; Label the attributes of the components, including the type and material of each component; The fixed dimensional parameters are marked, including the length of each component, the pipe turning angle, the spacing of sensor mounting holes, and the length of the pipe path between adjacent mounting points; S102. Add supplementary information; Supplementary information includes: simulation scenario, electromagnetic wave propagation speed, and component attenuation coefficient; The simulation scenarios include temperature scenarios, pressure scenarios, and partial discharge source location scenarios; the partial discharge source location scenarios cover all key areas of the GIS, including the inside of insulators, the center of busbars, the inside of circuit breaker compartments, pipe bends, and joint connections. The electromagnetic wave propagation speed library corrects the wave speed based on temperature and pressure: （1）； in, The corrected electromagnetic wave speed. The electromagnetic wave speed under standard operating conditions. , These are the current air pressure and temperature, respectively. , These are the air pressure and temperature corresponding to standard operating conditions; The expression for the component attenuation coefficient is: （2）； in, For the first The attenuation coefficient of each component, The basic attenuation coefficient at standard temperature, For the first Temperature sensitivity coefficient of each component For the first The aging coefficient of each component; the mathematical expression for the aging coefficient is as follows: （3）； in, This is the aging factor over time. For the service life of the equipment, The load aging factor, For the first Section load operation load level, For the first The service life of the load section For aging coefficient under special working conditions, For special operating conditions; For components that operate under high humidity conditions for extended periods, a humidity aging factor is introduced into the component attenuation coefficient for correction. The mathematical expression is as follows: （4）； in, Humidity aging coefficient, For the first Average annual relative humidity; For metal components, a vibration aging coefficient is introduced into the component attenuation coefficient for correction, and the mathematical expression is as follows: （5）； in, The vibration aging coefficient, For the first Annual average equipment vibration acceleration; If a metal component operates in a high humidity environment for a long time, both the humidity aging coefficient and the vibration aging coefficient should be included in the component attenuation coefficient. Amplitude of electromagnetic wave after attenuation The expression is: （6）； in, The amplitude of the electromagnetic wave before attenuation. To adjust the coefficient, The number of components corresponding to the component attenuation coefficient. This refers to the attenuation of electromagnetic waves caused by multiple components along a path. The expression is: （7）； in, This represents the total attenuation of the components. S103. Construct a real fault information database, including historical location data, partial discharge signal characteristics, equipment structural parameters, and maintenance measurement data.

3. The method for automatic localization and processing of partial discharge in GIS according to claim 2, characterized in that, The detailed process of step S2 is as follows: S201. Perform multimodal data acquisition; multimodal data includes: electromagnetic wave signals, ultrasonic signals, vibration signals, local temperature data, and gas concentration data. S202, Perform hierarchical signal preprocessing; For electromagnetic wave signals, interference suppression is performed using the following expression: （8）； in, The signal after interference suppression. Electromagnetic wave signals acquired by a UHF sensor. For reference, the environmental interference signals collected by the sensor, The adaptive filter weights at the current moment; The adaptive filter weights are updated iteratively, as shown in the following expression: （9）； in, For the adaptive filter weights in the next time step, This is the step size coefficient; For weak signals, an amplification operation is performed, expressed as follows: （10）； in, This is the magnification factor. The average amplitude of the signal. This represents the full-scale amplitude of the signal. The enhanced signal can be obtained by multiplying the signal by the corresponding amplification factor. The amplified signal undergoes feature reconstruction, as shown in the following expression: （11）； in, For the reconstructed signal, The signal is amplified. The reconstruction function represents the signal phase. With pulse rise time The inherent relationship between them; Multipath correction is performed on electromagnetic wave signals, specifically as follows: Based on the 3D topology of GIS pipelines, the time difference between the direct propagation path and the reflection path of electromagnetic waves is calculated, as shown in the following expression: （12）； in, For multipath time difference, The length of the reflection path. This is the direct path length; Based on the multipath time difference correction signal, the expression is as follows: （13）； in, This is the signal after multipath time difference correction; S203. Find the sensor A with the earliest acquisition time, and record its location as point A. Record the frequency of all sub-signals in A. S204. Perform initial screening on the sub-signals in the remaining sensors, and remove sub-signals whose frequencies are not in A. S205. Determine the effective signal based on the attenuation rate: Calculate the attenuation rate on the line between adjacent sensors based on the amplitude of the sub-signal, and combine it with the attenuation rate in the simulation model to obtain the following judgment formula: （14）； in, , Adjacent sensors and The actual attenuation rate and the model attenuation rate of the line between them A threshold is used to determine the attenuation level; When the attenuation rate of a sub-signal satisfies the judgment formula, the sub-signal is determined to be a valid sub-signal. The valid signal can be obtained by superimposing all valid sub-signals. If no valid sub-signal is found, it indicates that no fault has occurred; continue monitoring. S206. Record the power amplitude of the valid signal at point A. And calculate the noise intensity.

4. The method for automatic location and processing of partial discharge in GIS according to claim 3, characterized in that, Noise intensity The expression is: （15）； in, Noise power, This represents the effective signal power amplitude.

5. The method for automatic location and processing of partial discharge in GIS according to claim 1, characterized in that, The detailed process of step S3 is as follows: S301. Use the TDOA time difference method to calculate the rough estimate of the discharge location; S302. Use the correction model to correct the coarsely estimated discharge location, obtaining the corrected estimated discharge location; the distance between the corrected estimated discharge location and point A is... : （16）； in, To roughly estimate the distance between the discharge location and point A.

6. The method for automatic location and processing of partial discharge in GIS according to claim 4, characterized in that, In step S4, the criteria for signal grading are as follows: The standard for a Level 1 signal is: the signal meets the power amplitude requirement. And noise intensity ; The standard for a level 2 signal is: the signal meets the power amplitude requirement. Noise intensity One of the conditions, and the signal does not meet the classification criteria for a level three signal; The standard for a Level 3 signal is: the signal meets the power amplitude requirement. or noise intensity One of the conditions.

7. The method for automatic location and processing of partial discharge in GIS according to claim 1, characterized in that, The detailed process of step S5 is as follows: S501. Use a predictive model to predict the development trend of the partial discharge source. The output of the predictive model is the amplitude growth rate of the partial discharge source within a future time period T. Frequency growth rate And extended prediction directions; S502. Determine the extended prediction area for the future time period T based on the attenuated electromagnetic wave velocity. S503, Setting a safe distance And it will turn off sensors that meet safety conditions, while increasing the sampling frequency of sensors that do not meet safety conditions; The safety conditions are as follows: （17）； in, For point Minimum distance from the extended prediction region.

8. A method for automatic location and processing of partial discharge in GIS according to claim 6 or 7, characterized in that, In step S4, the processing procedure for signals of different levels is as follows: For Level 1 signals, the coarsely estimated discharge location is directly used as the final discharge location, and then the positioning process ends. For secondary signals, the corrected estimated discharge position is used as the final discharge position, and then the positioning process ends. For a level 3 signal, the corrected estimated discharge position is taken as the final discharge position, and then the process proceeds to step S5.

9. The method for automatic location and processing of partial discharge in GIS according to claim 7, characterized in that, The warning levels are divided into: primary warning, intermediate warning and advanced warning; A primary warning must meet the following requirements: and ; Intermediate warning level must meet the following requirements: or And it does not meet the criteria for a high-level early warning; Advanced warnings must meet the following requirements: and .

10. The method for automatic location and processing of partial discharge in GIS according to claim 9, characterized in that, The measures corresponding to the warning level are as follows: For initial warnings, the measures are: continued monitoring; For intermediate warnings, the measures are as follows: live inspection, arrange maintenance personnel to conduct on-site inspections, and use infrared thermal imaging and ultrasonic-assisted detection to check whether there are any abnormalities around the partial emission source; Planned maintenance will be carried out, and the defect will be included in the next round of power outage maintenance plans, with targeted maintenance solutions developed. For advanced warnings, the measures are as follows: special emergency repairs, initiating emergency maintenance procedures, disassembling the part where the partial discharge source is located, and resolving the fault, such as removing metal particles, replacing deteriorated insulation parts, and repairing poorly contacting conductive parts; Afterwards, a withstand voltage test is conducted, and after maintenance, a partial discharge withstand voltage test is performed on the GIS to verify that the defect has been eliminated before it can be put back into operation. The process involves reviewing and analyzing defects, recording their types, development processes, and corrective measures, updating the database, and optimizing subsequent correction and prediction algorithms.

Images