Transformer partial discharge multi-sensing positioning and energy analysis device and method

Abstract

The invention discloses a transformer partial discharge multi-sensor positioning and energy analysis device and method, belongs to the technical field of power equipment state monitoring, and solves the problems that in the prior art, positioning coordinates of partial discharge points are large in deviation, energy cannot be quantized, and discharge defects cannot be reliably recognized. The ultrahigh-frequency sensor, the ultrasonic sensor and the pulse current sensor are adopted for cooperative work, real discharge, electromagnetic interference, mechanical vibration and other noise are effectively distinguished through multi-signal cross validation, and false alarm and missing alarm are remarkably reduced. The positioning unit adapts to the signal quality change through a dynamic weight strategy, and the positioning precision of the partial discharge point is improved. Meanwhile, by combining the charge transfer amount of the pulse current and the ultrasonic signal, the energy evaluation unit can realize quantitative calculation of the discharge energy, the severity judgment is improved from qualitative judgment to quantitative judgment, and various discharge defects are comprehensively covered.

Description

Technical Field

[0001] This invention relates to the field of power equipment condition monitoring technology, specifically to a transformer partial discharge location and energy analysis device and method based on multi-sensor technology. Background Technology

[0002] As a key piece of equipment in the power grid, the internal insulation health of transformers is of paramount importance. Partial discharge is the most prominent early manifestation of insulation defects, and its effective monitoring is crucial for preventing catastrophic failures.

[0003] Existing detection technologies mainly rely on ultra-high frequency methods, ultrasonic methods, or a combination of both, all of which have inherent limitations: Insufficient Positioning Reliability: While the UHF method offers sensitive response, the signal undergoes multiple reflections within the transformer tank, resulting in complex wave impedance and susceptibility to spatial electromagnetic interference, leading to ambiguous positioning or even false alarms. Although the ultrasonic method boasts good directivity, the propagation path of sound waves between windings and the core is complex, involving attenuation, refraction, and mode conversion, making propagation time extraction difficult and resulting in significant positioning errors when used alone. When combining both methods, current techniques typically assign weights to the velocities and time differences of a set of UHF and ultrasonic signals to obtain equivalent velocities and equivalent time differences. The distance to the discharge point is then calculated based on these equivalent velocities and time differences, and the location coordinates of the discharge point are determined by solving for multiple distances. However, while fixed weights offer high computational efficiency and stability under ideal signal conditions, they cannot adapt to dynamically changing environments. For example, when the ultrasonic signal attenuates due to mechanical vibration interference or path obstruction within the transformer, or when the UHF signal is affected by external electromagnetic pulse interference, the fixed weights cannot mitigate the impact of unreliable signals, leading to deviations or even complete errors in the positioning coordinates.

[0004] The lack of diagnostic dimensions makes quantitative assessment impossible: Current technologies generally stop at detecting and roughly locating discharges, failing to obtain crucial discharge charge transfer amounts, thus making it difficult to calculate the arc energy released in each discharge. This results in a lack of quantitative basis for operation and maintenance decisions, making it impossible to distinguish between harmless and dangerous discharges, leading to a one-size-fits-all conservative strategy and causing unnecessary power outages or delays in handling.

[0005] Limited defect coverage and interference resistance: Single or dual-sensor technology struggles to comprehensively capture all types of discharge defects. For example, weak discharges may be masked by background noise; discharges at certain locations may only be sensitive to specific sensors. The lack of cross-validation mechanisms between multiple signals results in high false alarm and false negative rates when facing external interference or internal noise. Summary of the Invention

[0006] To address the aforementioned problems in the prior art, this invention provides a transformer partial discharge multi-sensor localization and energy analysis device and method, which solves the problems of large positioning coordinate deviation of partial discharge points, inability to quantify energy, and inability to reliably identify discharge defects in the prior art.

[0007] To achieve the above objectives, the technical solution adopted by the present invention is as follows: On the one hand, a transformer partial discharge multi-sensor localization and energy analysis device is provided, comprising: The sensor assembly includes an ultra-high frequency sensor module, an ultrasonic sensor module, and a high-frequency current sensor module, which are used to acquire ultra-high frequency signals, ultrasonic signals, and pulse current signals at the discharge point in the transformer, respectively. The data transmission module, connected to the sensor assembly, is used to synchronously acquire three types of signals and transmit them to the data analysis and processing unit. The data processing unit includes: The multi-signal cross-validation unit is used to evaluate the similarity and compare the pulse timing consistency of the three types of signals of the sensor component. If the pulse start time difference of the three types of sensor signals is within the set time window and the similarity is greater than the preset threshold, the discharge point is regarded as a real discharge event; otherwise, the discharge point is blocked. The positioning unit is used to dynamically assign weights to UHF signals and ultrasonic signals, and obtain the positioning coordinates of the discharge point based on the dynamic weights. The energy assessment unit is used to obtain the charge transfer amount and electric field intensity characterization amount through pulse current signal and ultrasonic signal respectively, and to obtain the arc energy value by combining the discharge gap distance in the positioning coordinates. The discharge defect type of the discharge point is matched by the arc energy value.

[0008] Furthermore, it also includes an assessment and decision-making unit, which is used to determine whether the operation of the transformer is affected by the location coordinates of the discharge point, the arc energy value and the discharge defect type of the actual discharge event, and to issue an alarm when the judgment result is yes.

[0009] Furthermore, it also includes a user interface and visualization unit connected to the assessment and decision-making unit and the data processing unit, which are used to display the actual discharge event judgment results, location coordinates, arc energy value and discharge defect type of the discharge point, and output historical detection data or detection reports.

[0010] On the other hand, a method for providing a transformer partial discharge multi-sensor localization and energy analysis device includes the following steps: S1. Acquire the ultra-high frequency signal, ultrasonic signal and pulse current signal of the discharge point in the transformer, and transmit them to the data processing unit through the data transmission module; S2. Calculate the similarity of the three types of signals. If the pulse start time difference of the three types of signals is within the set time window and the similarity is greater than the preset threshold, then the discharge point is regarded as a real discharge event; otherwise, the discharge point is blocked. S3. Dynamically assign weights to UHF and ultrasonic signals, and obtain the location coordinates of the discharge point based on the dynamic weights; S4. The charge transfer amount and electric field intensity characterization amount are obtained by pulse current signal and ultrasonic signal respectively, and the arc energy value is obtained by combining the discharge gap distance in the positioning coordinates. The discharge defect type of the discharge point is matched by the arc energy value.

[0011] Furthermore, step S3 further includes: S31. Pair multiple UHF sensors and multiple ultrasonic sensors in pairs, and calculate the signal-to-noise ratio of the UHF signal for each sensor pair. The signal-to-noise ratio of ultrasonic signals ; S32, align each sensor pair and Each with a preset effective threshold If, in comparison, only one of the UHF signal and the ultrasonic signal has a signal-to-noise ratio greater than [value missing], [the following is a possible interpretation:] Then assign a value greater than The weight of one signal is 1, and the weight of the other signal is 0; if and All greater than Then, the weights of the UHF signal and the ultrasonic signal are calculated separately. S33. Calculate the equivalent distance from each sensor to the discharge point by weighting the two signals for each sensor, and use at least three equivalent distances to solve for the location coordinates of the discharge point.

[0012] Furthermore, if and All greater than The method for calculating the weights of the UHF signal and the ultrasonic signal respectively is as follows: S321, based on Each sensor is assigned an initial weight for the two signals, and the initial location coordinates of the discharge point are obtained based on the initial weights. ; S322. Treating both the ultra-high frequency signal and the ultrasonic signal as unique signals respectively, obtain the ultra-high frequency method positioning coordinates of the discharge point. Ultrasonic method for locating coordinates ; S323, Calculation and respectively with European distance and and based on Calculate the final weights of the UHF signals separately and the final weight of the ultrasonic signal ; in, This is the feedback gain coefficient; and These are the initial weights for the UHF signal and the ultrasonic signal, respectively.

[0013] Furthermore, the initial weights of the UHF and ultrasonic signals are calculated using the following expressions: in, It is a confidence adjustment factor greater than or equal to 1.

[0014] Furthermore, step S4 further includes: S41. Obtain the charge transfer amount using pulsed current signal and ultrasonic signal respectively. and electric field strength characterization quantity Based on the positioning coordinates of the discharge point output by the positioning unit, the discharge gap distance is calculated. ; S42, the amount of charge transferred Electric field strength characterization quantity and discharge gap distance The electric arc energy value is obtained by inputting it into the pre-trained energy fitting model. ; S43, change the arc energy value The discharge defect type of the discharge point is obtained by matching it with the preset energy threshold range of the discharge defect type.

[0015] Furthermore, training methods for energy-fitting models include: Multiple typical discharge defect types were simulated on the experimental platform, and multiple sets of training data were collected simultaneously. Each set of training data included charge transfer. Electric field strength characterization quantity Discharge gap distance and the actual arc energy value calibrated by a standard energy measuring device ,by As input variables, The energy fitting model is trained using the target output.

[0016] Furthermore, it also includes: Establish a historical database to store the location coordinates, arc energy values, and corresponding occurrence times of each discharge event; generate and output a discharge development trend report based on historical data, including the migration trajectory of the discharge point location and the curve of discharge energy change over time; trigger an early warning signal when the arc energy value at the same location shows a continuous upward trend within a preset period, or when the discharge point migrates to the preset insulation risk area of ​​the transformer.

[0017] This invention discloses a transformer partial discharge multi-sensor localization and energy analysis device and method, the beneficial effects of which are: 1. This invention utilizes the collaborative operation of three types of sensors—ultra-high frequency (UHF), ultrasonic, and pulsed current—along with a multi-signal cross-validation unit for similarity assessment and pulse timing consistency comparison. This effectively distinguishes genuine discharges from external electromagnetic interference or internal mechanical vibrations and other noise, solving the problems of high false alarm and false negative rates associated with single or dual-sensor technologies. Furthermore, by employing a strategy of dynamically assigning weights to UHF and ultrasonic signals through the positioning unit, rather than the fixed weights used in existing technologies, it adapts to changes in signal quality and automatically reduces the influence of unreliable signals, thereby significantly reducing the positioning coordinate deviation of partial discharge points. Moreover, by introducing pulsed current signals to obtain charge transfer amounts and combining them with ultrasonic signals and positioning information, the energy assessment unit achieves quantitative calculation of the discharge arc energy, elevating the severity of the discharge from qualitative judgment to quantitative assessment, comprehensively covering various discharge defects.

[0018] 2. This invention innovatively uses spatial geometric consistency as a weight allocation criterion independent of signal quality by introducing a feedback verification mechanism based on the consistency of positioning results. It constructs a decision-making process of "initial estimation - hypothesis testing - feedback correction." This mechanism, with its adjustable feedback gain coefficient, maintains the stability of weight allocation while ensuring system sensitivity. It not only effectively solves the positioning deviation problem caused by abnormal signal propagation under complex operating conditions but also realizes a paradigm shift in diagnosis from qualitative judgment to quantitative assessment. Combined with multi-sensor fusion design and long-term trend analysis capabilities, the system comprehensively covers various discharge defects while possessing excellent anti-interference and early warning capabilities. It provides a complete technical solution for transformer insulation condition assessment and predictive maintenance, significantly improving the accuracy of power grid operation and maintenance and asset security.

[0019] 3. This invention dynamically adjusts weights by comparing the deviation between the positioning results from a single sensor method and the initial fused positioning results. If the positioning result of a certain method (such as the UHF method) deviates significantly from the initial fused result, it indicates that the method may be unreliable in the current environment, and its weight will be reduced accordingly. The feedback verification mechanism makes the weight allocation more objective and adaptive to actual working conditions, thereby further improving the accuracy and fault tolerance of the positioning algorithm.

[0020] 4. The present invention is based on a nonlinear weight allocation method for signal-to-noise ratio exceeding the threshold. By adjusting the confidence level adjustment factor k, the preference for high signal-to-noise ratio signals can be flexibly adjusted (when k>1, the higher the signal-to-noise ratio, the greater the initial weight). This makes the allocation of initial weights not only dependent on whether the signal is effective, but also more precisely reflects the relative quality of the signal, providing a more reasonable starting point for subsequent feedback adjustments.

[0021] 5. This invention simulates various typical defects on an experimental platform and calibrates the actual energy values ​​using standard devices, so that the trained model has a solid experimental foundation and wide applicability, can accurately reflect the energy characteristics of different discharge defects, and ensures the accuracy and practicality of the energy assessment model. Attached Figure Description

[0022] Figure 1 A schematic diagram of the composition of a transformer partial discharge multi-sensor localization and energy analysis device; Figure 2 A flowchart illustrating the analysis method for a transformer partial discharge multi-sensor localization and energy analysis device. Detailed Implementation

[0023] The specific embodiments of the present invention are described below to enable those skilled in the art to understand the present invention. However, it should be understood that the present invention is not limited to the scope of the specific embodiments. For those skilled in the art, various changes are obvious as long as they are within the spirit and scope of the present invention as defined and determined by the appended claims. All inventions utilizing the concept of the present invention are protected.

[0024] refer to Figure 1 This embodiment provides a transformer partial discharge multi-sensor positioning and energy analysis device, including sensor components, data transmission module and data processing unit.

[0025] The sensor assembly, used for synchronously acquiring multi-physics field signals generated by partial discharge inside the transformer, includes an ultra-high frequency (UHF) sensor module, an ultrasonic sensor module, and a high-frequency current sensor module. The UHF sensor module is used to acquire UHF signals from the discharge points in the transformer; the ultrasonic sensor module is used to acquire ultrasonic signals from the discharge points in the transformer; and the high-frequency current sensor module is used to acquire pulse current signals in the transformer's grounding circuit.

[0026] In this embodiment, the sensor assembly is installed on the transformer tank wall and the grounding lead. The UHF sensor module and the ultrasonic sensor module include multiple UHF sensors and multiple ultrasonic sensors. Specifically, the UHF sensor module is installed on the detection flange pre-set in the transformer tank, and the UHF sensor adopts a broadband planar equiangular spiral antenna structure; the ultrasonic sensor module is uniformly installed on the surface of the transformer tank wall using a magnetic fixation method, and a special coupling agent is applied to the bottom surface of the ultrasonic sensor; the high-frequency current sensor module is installed at the transformer grounding lead, and the high-frequency current sensor adopts a split Rogowski coil structure, and its magnetic core material is selected from high permeability nanocrystalline alloy.

[0027] The data transmission module, connected to the sensor components, is used to synchronously acquire three types of signals and transmit them to the data analysis and processing unit.

[0028] In this embodiment, the data transmission module adopts a synchronous clock circuit based on a temperature-controlled crystal oscillator. The high-stability reference clock generated by this circuit is distributed to each acquisition channel through equal-length transmission lines. A multi-channel synchronous acquisition chipset is adopted, and the clock synchronization between each channel is achieved through differential signal lines. The signal conditioning circuit includes a programmable gain amplifier and an anti-aliasing filter. The acquisition module shell adopts an electromagnetic shielding design, and the internal circuit board is treated with tri-proof measures.

[0029] The data processing unit adopts a modular industrial server architecture, equipped with a multi-core processor and high-speed solid-state drive, and supports a parallel computing framework. The system software runs on a real-time operating system and has task scheduling management and dynamic resource allocation functions. The data processing unit specifically includes a multi-signal cross-validation unit, a positioning unit, and an energy assessment unit.

[0030] The multi-signal cross-validation unit is used to evaluate the similarity of the three types of signals of the sensor component and compare the consistency of the pulse timing. If the pulse start time difference of the three types of sensor signals is within the set time window and the similarity is greater than the preset threshold, the discharge point is regarded as a real discharge event; otherwise, the discharge point is blocked. The positioning unit is used to dynamically assign weights to UHF and ultrasonic signals and obtain the positioning coordinates of the discharge point based on the dynamic weights. The energy assessment unit is used to obtain the charge transfer amount and electric field intensity characterization amount through pulse current signal and ultrasonic signal respectively, and combined with the discharge gap distance in the positioning coordinates, to obtain the arc energy value, and to match the discharge defect type of the discharge point through the arc energy value.

[0031] As a further embodiment, the transformer partial discharge multi-sensor positioning and energy analysis device also includes an evaluation and decision unit. The evaluation and decision unit is used to determine whether the operation of the transformer is affected by the positioning coordinates of the discharge point where the actual discharge event occurs, the arc energy value, and the discharge defect type, and to issue an alarm when the determination result is yes.

[0032] As a further embodiment, the transformer partial discharge multi-sensor positioning and energy analysis device also includes a user interface and a visualization unit connected to the evaluation and decision-making unit and the data processing unit. The user interface and visualization unit are used to display the actual discharge event judgment results, positioning coordinates, arc energy value and discharge defect type of the discharge point, and output historical detection data or detection reports.

[0033] refer to Figure 2 This embodiment also provides a method for transformer partial discharge multi-sensor localization and energy analysis device, including the following steps: S1. Acquire the ultra-high frequency signal, ultrasonic signal and pulse current signal of the discharge point in the transformer, and transmit them to the data processing unit through the data transmission module.

[0034] S2. Calculate the similarity of the three types of signals. If the pulse start time difference of the three types of signals is within a set time window and the similarity is greater than a preset threshold, then the discharge at the discharge point is taken as a real discharge event; otherwise, the discharge point is masked. In this embodiment, the multi-signal cross-validation unit uses a dynamic time warping algorithm to calculate the similarity of the three types of signals of the sensor component and improves the measurement accuracy of the pulse start time difference through an interpolation algorithm.

[0035] S3. Dynamically assign weights to UHF and ultrasonic signals, and obtain the location coordinates of the discharge point based on the dynamic weights.

[0036] S4. The charge transfer amount and electric field intensity characterization amount are obtained by pulse current signal and ultrasonic signal respectively, and the arc energy value is obtained by combining the discharge gap distance in the positioning coordinates. The discharge defect type of the discharge point is matched by the arc energy value.

[0037] As a further aspect of this embodiment, the method further includes: establishing a historical database to store the location coordinates, arc energy values, and corresponding occurrence times of each discharge event; generating and outputting a discharge development trend report based on the historical data, the report including the migration trajectory of the discharge point location and the curve of discharge energy change over time; triggering an early warning signal when the arc energy value at the same location shows a continuous upward trend within a preset period, or when the discharge point migrates towards the transformer's preset insulation risk area. The preset insulation risk area is a weak insulation area, such as the end of the transformer winding.

[0038] Specifically, step S3 further includes: S31. Pair multiple UHF sensors in the UHF sensor module and multiple ultrasonic sensors in the ultrasonic sensor module in pairs, and calculate the signal-to-noise ratio of the UHF signal for each sensor pair. The signal-to-noise ratio of ultrasonic signals .

[0039] The initial weights of the UHF and ultrasonic signals are calculated using the following expressions: in, and These are the initial weights for the UHF signal and the ultrasonic signal, respectively. It is a confidence adjustment factor greater than or equal to 1.

[0040] S32, align each sensor pair and Each with a preset effective threshold If, in comparison, only one of the UHF signal and the ultrasonic signal has a signal-to-noise ratio greater than [value missing], [the following is a possible interpretation:] Then assign a value greater than The weight of one signal is 1, and the weight of the other signal is 0; if and All greater than Then, the weights of the UHF signal and the ultrasonic signal are calculated separately. The method for calculating the weights of the UHF signal and the ultrasonic signal is as follows: S321, based on Each sensor is assigned an initial weight for the two signals, and the initial location coordinates of the discharge point are obtained based on the initial weights. ; S322. Treating both the ultra-high frequency signal and the ultrasonic signal as unique signals respectively, obtain the ultra-high frequency method positioning coordinates of the discharge point. Ultrasonic method for locating coordinates ; S323, Calculation and respectively with European distance and and based on Calculate the final weights of the UHF signals separately and the final weight of the ultrasonic signal ; in, This is the feedback gain coefficient.

[0041] The "dynamic weighting" logic is adopted—first, the initial weights are calculated based on signal quality (the larger the signal-to-noise ratio increment, the more reliable the result); then, the results of the fusion positioning are compared with the positioning results of the UHF method and the ultrasonic method alone. If the deviation is large, the weights are adjusted until the positioning results are stable and consistent, and finally the dynamically optimized weights are obtained, avoiding the limitations of fixed weights.

[0042] S33. Calculate the equivalent distance from each sensor to the discharge point by weighting the two signals for each sensor, and use at least three equivalent distances to solve for the location coordinates of the discharge point.

[0043] For ease of understanding, the expression for calculating the equivalent distance is as follows: in, and These are the propagation speeds of ultra-high frequency signals and ultrasonic signals, respectively. and These represent the propagation time differences between ultra-high frequency signals and ultrasonic signals, respectively. Utilizing spatial geometry principles: each sensor is the "center of a sphere," the equivalent distance is the "radius of a sphere," and the intersection of multiple spherical surfaces in space marks the location of the discharge point. The sensor coordinates and distances are transformed into mathematical equations, which are then solved to obtain the three-dimensional coordinates of the discharge point, thus completing the localization.

[0044] Specifically, step S4 further includes: S41. Obtain the charge transfer amount using pulsed current signal and ultrasonic signal respectively. and electric field strength characterization quantity Based on the positioning coordinates of the discharge point output by the positioning unit, the discharge gap distance is calculated. ; S42, the amount of charge transferred Electric field strength characterization quantity and discharge gap distance The electric arc energy value is obtained by inputting it into the pre-trained energy fitting model. ; S43, change the arc energy value The discharge defect type of the discharge point is obtained by matching it with the preset energy threshold range of the discharge defect type.

[0045] Training methods for energy-fitting models include: Multiple typical discharge defect types were simulated on the experimental platform, and multiple sets of training data were collected simultaneously. Each set of training data included charge transfer. Electric field strength characterization quantity Discharge gap distance and the actual arc energy value calibrated by a standard energy measuring device ,by As input variables, The energy fitting model is trained using the target output. In this embodiment, the energy fitting model can be trained using a three-layer feedforward neural network with an attention mechanism, employing the AdamW optimizer and an early stopping strategy.

[0046] In summary, the beneficial effects of this solution are as follows: By introducing a feedback verification mechanism based on the consistency of positioning results, the system innovatively uses spatial geometric consistency as a weight allocation criterion independent of signal quality, constructing a decision-making process of "initial estimation - hypothesis testing - feedback correction." This mechanism, with the help of an adjustable feedback gain coefficient, maintains the stability of weight allocation while ensuring system sensitivity. It not only effectively solves the positioning deviation problem caused by abnormal signal propagation under complex operating conditions, but also realizes a paradigm shift in diagnosis from qualitative judgment to quantitative assessment. Combined with multi-sensor fusion design and long-term trend analysis function, the system comprehensively covers various discharge defects and has excellent anti-interference capabilities and early warning capabilities. It provides a complete technical solution for transformer insulation condition assessment and predictive maintenance, significantly improving the accuracy of power grid operation and maintenance and asset security capabilities.

[0047] Although specific embodiments of the invention have been described in detail with reference to the accompanying drawings, this should not be construed as limiting the scope of protection of this invention. Various modifications and variations that can be made by those skilled in the art without inventive effort within the scope described in the claims still fall within the scope of protection of this invention.

Claims

1. A transformer partial discharge multi-sensor positioning and energy analysis device, characterized in that, include: The sensor assembly includes an ultra-high frequency sensor module, an ultrasonic sensor module, and a high-frequency current sensor module, which are used to acquire ultra-high frequency signals, ultrasonic signals, and pulse current signals at the discharge point in the transformer, respectively. The data transmission module is connected to the sensor assembly and is used to synchronously acquire signals from the three types of sensors and transmit them to the data analysis and processing unit. The data processing unit includes: The multi-signal cross-validation unit is used to evaluate the similarity and compare the pulse timing consistency of the three types of signals of the sensor component. If the pulse start time difference of the three types of signals is within the set time window and the similarity is greater than the preset threshold, the discharge point is regarded as a real discharge event; otherwise, the discharge point is blocked. The positioning unit is used to dynamically assign weights to UHF signals and ultrasonic signals, and obtain the positioning coordinates of the discharge point based on the dynamic weights. The energy assessment unit is used to obtain the charge transfer amount and electric field intensity characterization amount through pulse current signal and ultrasonic signal respectively, and to obtain the arc energy value by combining the discharge gap distance in the positioning coordinates. The discharge defect type of the discharge point is matched by the arc energy value.

2. The method of the transformer partial discharge multi-sensor localization and energy analysis device according to claim 1, characterized in that, It also includes an assessment and decision-making unit, which is used to determine whether the operation of the transformer is affected by the location coordinates of the discharge point, the arc energy value and the discharge defect type of the actual discharge event, and to issue an alarm when the judgment result is yes.

3. The method of the transformer partial discharge multi-sensor localization and energy analysis device according to claim 2, characterized in that, It also includes a user interface and visualization unit connected to the assessment and decision-making unit and the data processing unit, which are used to display the actual discharge event judgment results, location coordinates, arc energy value and discharge defect type of the discharge point, and output historical detection data or detection reports.

4. The method of the transformer partial discharge multi-sensor localization and energy analysis device according to any one of claims 1 to 3, characterized in that, Including the following steps: S1. Acquire the ultra-high frequency signal, ultrasonic signal and pulse current signal of the discharge point in the transformer, and transmit them to the data processing unit through the data transmission module; S2. Calculate the similarity of the three types of signals. If the pulse start time difference of the three types of signals is within the set time window and the similarity is greater than the preset threshold, then the discharge point is regarded as a real discharge event; otherwise, the discharge point is blocked. S3. Dynamically assign weights to UHF and ultrasonic signals, and obtain the location coordinates of the discharge point based on the dynamic weights; S4. The charge transfer amount and electric field intensity characterization amount are obtained by pulse current signal and ultrasonic signal respectively, and the arc energy value is obtained by combining the discharge gap distance in the positioning coordinates. The discharge defect type of the discharge point is matched by the arc energy value.

5. The method of the transformer partial discharge multi-sensor localization and energy analysis device according to claim 4, characterized in that, Step S3 further includes: S31. Pair multiple ultrasonic sensors in the multiple UHF sensor and ultrasonic sensor modules into pairs, and calculate the signal-to-noise ratio of the UHF signal for each sensor pair. The signal-to-noise ratio of ultrasonic signals ; S32, align each sensor pair and Each with a preset effective threshold If, in comparison, only one of the UHF signal and the ultrasonic signal has a signal-to-noise ratio greater than [value missing], [the following is a possible interpretation:] Then assign a value greater than The weight of one signal is 1, and the weight of the other signal is 0; if All greater than Then, the weights of the UHF signal and the ultrasonic signal are calculated separately. S33. Calculate the equivalent distance from each sensor to the discharge point by weighting the two signals for each sensor, and use at least three equivalent distances to solve for the location coordinates of the discharge point.

6. The method of the transformer partial discharge multi-sensor localization and energy analysis device according to claim 5, characterized in that, like and All greater than The method for calculating the weights of the UHF signal and the ultrasonic signal respectively is as follows: S321, based on Each sensor is assigned an initial weight for the two signals, and the initial location coordinates of the discharge point are obtained based on the initial weights. ; S322. Treating both the ultra-high frequency signal and the ultrasonic signal as unique signals respectively, obtain the ultra-high frequency method positioning coordinates of the discharge point. Ultrasonic method for locating coordinates ; S323, Calculation and respectively with European distance and and based on Calculate the final weights of the UHF signals separately and the final weight of the ultrasonic signal ; in, This is the feedback gain coefficient; and These are the initial weights for the UHF signal and the ultrasonic signal, respectively.

7. The method of the transformer partial discharge multi-sensor localization and energy analysis device according to claim 6, characterized in that, The initial weights of the UHF and ultrasonic signals are calculated using the following expressions: in, This is the confidence level adjustment factor.

8. The method of the transformer partial discharge multi-sensor localization and energy analysis device according to claim 4, characterized in that, Step S4 further includes: S41. Obtain the charge transfer amount using pulsed current signal and ultrasonic signal respectively. and electric field strength characterization quantity Based on the positioning coordinates of the discharge point output by the positioning unit, the discharge gap distance is calculated. ; S42, the amount of charge transferred Electric field strength characterization quantity and discharge gap distance The electric arc energy value is obtained by inputting it into the pre-trained energy fitting model. ; S43, change the arc energy value The discharge defect type of the discharge point is obtained by matching it with the preset energy threshold range of the discharge defect type.

9. The method of the transformer partial discharge multi-sensor localization and energy analysis device according to claim 8, characterized in that, Training methods for energy-fitting models include: Multiple typical discharge defect types were simulated on the experimental platform, and multiple sets of training data were collected simultaneously. Each set of training data included charge transfer. Electric field strength characterization quantity Discharge gap distance and the actual arc energy value calibrated by a standard energy measuring device ,by As input variables, The energy fitting model is trained using the target output.

10. The method of the transformer partial discharge multi-sensor localization and energy analysis device according to claim 4, characterized in that, Also includes: Establish a historical database to store the location coordinates, arc energy values, and corresponding occurrence times of each discharge event; A discharge development trend report is generated and output based on historical data. The report includes the migration trajectory of the discharge point location and the curve of the discharge energy change over time. When the arc energy value at the same location is detected to show a continuous upward trend within a preset period, or when the discharge point migrates to the preset insulation risk area of ​​the transformer, an early warning signal is triggered.

Images