A method, system, device and medium for accurately extracting the apparent discharge quantity of a high-voltage bushing

Abstract

The application discloses a kind of high voltage bushing apparent discharge quantity accurate extraction method, system, equipment and medium, comprising: constructing typical high voltage bushing attenuation transfer function library;The partial discharge signal collected to on-line monitoring is carried out multi-band decomposition, and the amplitude characteristic quantity of each band signal component is extracted;According to the preliminary diagnosis result of defect type, corresponding attenuation transfer function is selected from typical high voltage bushing attenuation transfer function library, and according to the amplitude of attenuation transfer function at each band center frequency, the weight of band is calculated;The original amplitude estimate value of each band is calculated separately, and each original amplitude estimate value is weighted and summed using band weight, to obtain the original pulse amplitude after inversion;Based on the sensor itself calibration coefficient and original pulse amplitude, the internal apparent discharge quantity of bushing is converted, to obtain the internal apparent discharge quantity estimate value of bushing.The present application has controllable amount of calculation, is easy to integrate into existing on-line monitoring system, and has strong engineering practicability.

Description

Technical Field

[0001] This invention relates to the field of online monitoring and fault diagnosis technology for high-voltage electrical equipment, and in particular to a method, system, equipment, and medium for accurately extracting the apparent discharge quantity of high-voltage bushings. Background Technology

[0002] High-voltage bushings are key insulating components in power systems, and their operational reliability is directly related to the safety of main equipment and even the power grid. Partial discharge is an important sign of bushing insulation deterioration. Therefore, online monitoring and accurate quantitative assessment of bushing partial discharge are of great significance for achieving condition-based maintenance and preventing major accidents.

[0003] Currently, online partial discharge monitoring technology based on high-frequency current sensors or ultra-high-frequency sensors has been widely used. However, in the specific structure of high-voltage bushings, due to the complex internal insulating medium and metal structure, the nanosecond-level original pulse signal generated by partial discharge undergoes multipath reflection, dispersion effect, and dielectric loss during its propagation from the discharge point to the external sensor. This results in severe attenuation, distortion, and tailing of the signal waveform, causing a significant difference between the signal characteristics finally acquired by the sensor and the original pulse. Simultaneously, the sensor's frequency response, coupling method, and the limited bandwidth of subsequent monitoring equipment together constitute a bandpass or bandlimited system, typically only capturing a portion of the frequency components in the partial discharge signal that match the system's passband. A large amount of high-frequency or low-frequency energy is filtered out or coupled with extremely low efficiency, further exacerbating the nonlinearity and complexity of the mapping relationship between the measured signal amplitude and the original discharge quantity. In traditional calibration methods, a standard discharge power supply is usually injected directly into the sensor port under laboratory conditions to obtain a single calibration coefficient. This completely ignores the attenuation and distortion caused by the critical link of the discharge pulse propagating in the actual structure inside the high-voltage bushing. Therefore, the apparent discharge quantity obtained by directly multiplying the signal amplitude measured by online monitoring by this calibration coefficient has a significant error compared with the true value of the discharge point. Summary of the Invention

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

[0005] Therefore, this invention provides a method, system, device, and medium for accurately extracting the apparent discharge quantity of high-voltage bushings, solving the problem that existing online monitoring methods for partial discharge of high-voltage bushings fail to accurately reflect the original discharge intensity of the discharge point due to neglecting the attenuation and distortion of the discharge signal in the propagation path inside the bushing and the frequency response limitations of the monitoring system.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution: In a first aspect, the present invention provides a method for accurately extracting the apparent discharge quantity of a high-voltage bushing, comprising: Standard pulse signals are injected into multiple preset positions inside the sample bushing, and the attenuation transfer function at each preset position is calculated to construct a typical high-voltage bushing attenuation transfer function library. The partial discharge signal acquired by online monitoring is decomposed into multiple frequency bands, and the amplitude characteristics of each frequency band signal component are extracted. Based on the preliminary diagnosis results of the defect type, the corresponding attenuation transfer function is selected from the typical high-voltage bushing attenuation transfer function library, and the frequency band weight is calculated based on the amplitude of the attenuation transfer function at the center frequency of each frequency band. Based on the amplitude characteristics of each frequency band signal component, calculate the original amplitude estimate of each frequency band individually, and use the frequency band weight to perform a weighted sum of each original amplitude estimate to obtain the original pulse amplitude after inversion; The apparent discharge quantity inside the bushing is calculated based on the sensor's own calibration coefficient and the original pulse amplitude to obtain an estimated value of the apparent discharge quantity inside the bushing.

[0007] As a preferred embodiment of the method for accurately extracting the apparent discharge quantity of a high-voltage bushing according to the present invention, the construction of a typical high-voltage bushing attenuation transfer function library includes: Select a sample bushing of the same model and specification as the high-voltage bushing to be monitored. On the calibration experimental platform, connect the standard pulse generator to the preset simulated discharge point inside the sample bushing through a coupling element. Install the monitoring sensor at the predetermined installation position outside the bushing. Record the time-domain response signal measured by the sensor when a standard pulse generator injects a known amount of charge; The frequency spectrum is obtained by performing a Fourier transform on the time-domain response signal. Based on the frequency spectrum, the known amount of injected charge, and the injection coefficient of the coupling element, the attenuation transfer function at multiple preset positions is calculated. The attenuation transfer functions of the multiple preset positions are stored in a typical high-voltage bushing attenuation transfer function library.

[0008] As a preferred embodiment of the method for accurately extracting apparent discharge quantity of high-voltage bushings according to the present invention, the step of performing multi-band decomposition on the partial discharge signal acquired by online monitoring and extracting the amplitude characteristic quantity of each frequency band signal component includes: During online monitoring, the partial discharge signal output by the sensor is continuously collected; In response to the detection of a partial discharge pulse event, a time-domain waveform containing the partial discharge pulse event is captured, and the time-domain waveform is decomposed into multiple non-overlapping or partially overlapping frequency band signal components. The amplitude characteristics of each frequency band signal component are extracted, including peak value, root mean square value, and energy.

[0009] As a preferred embodiment of the method for accurately extracting the apparent discharge quantity of a high-voltage bushing according to the present invention, the calculation of the frequency band weight includes: Based on the preliminary diagnosis of the defect type, select the attenuation transfer function corresponding to the current suspected discharge point from the typical high-voltage bushing attenuation transfer function library; The weight value corresponding to each frequency band is determined based on the amplitude of the attenuation transfer function at the center frequency of each frequency band. The weight values ​​of each frequency band are normalized so that the sum of all weights is 1.

[0010] As a preferred embodiment of the method for accurately extracting the apparent discharge quantity of a high-voltage bushing according to the present invention, the method for obtaining the inverted original pulse amplitude includes: Divide the amplitude characteristic of the frequency band signal component by the amplitude of the attenuation transfer function at the center frequency of the frequency band to obtain the original amplitude estimate of the frequency band inversion. Using the frequency band weights as weighting coefficients, the original amplitude estimates retrieved from each frequency band are weighted and summed to obtain the retrieved original pulse amplitude.

[0011] As a preferred embodiment of the method for accurately extracting the apparent discharge quantity of a high-voltage bushing according to the present invention, the method for obtaining the estimated value of the apparent discharge quantity inside the bushing includes: The calibration coefficients of the sensor itself are obtained by calibrating a standard pulse generator directly connected to the sensor input terminal. These calibration coefficients are used to characterize the response characteristics of the sensor and subsequent monitoring circuits to the input signal. Divide the original pulse amplitude after inversion by the calibration coefficient to obtain the estimated value of apparent discharge inside the bushing.

[0012] As a preferred embodiment of the method for accurately extracting the apparent discharge quantity of a high-voltage bushing according to the present invention, the method further includes: assessing the confidence level of the estimated apparent discharge quantity inside the bushing. Based on the original amplitude estimates retrieved individually for each frequency band, the weighted standard deviation is calculated using the frequency band weights; The weighted standard deviation is used as an indicator to measure the consistency of the inversion results for each frequency band.

[0013] Secondly, the present invention provides a system for accurately extracting the apparent discharge quantity of a high-voltage bushing, comprising: The transfer function library construction module is used to inject standard pulse signals at multiple preset positions inside the sample bushing and calculate the attenuation transfer function at each preset position to construct a typical high-voltage bushing attenuation transfer function library. The feature extraction module is used to perform multi-band decomposition on the partial discharge signal acquired by online monitoring and extract the amplitude features of each frequency band signal component; The frequency band weight calculation module is used to select the corresponding attenuation transfer function from the typical high-voltage bushing attenuation transfer function library according to the preliminary diagnosis results of the defect type, and calculate the frequency band weight based on the amplitude of the attenuation transfer function at the center frequency of each frequency band. The inversion calculation module is used to calculate the original amplitude estimate of each frequency band individually based on the amplitude characteristics of the signal components of each frequency band, and to perform a weighted summation of each original amplitude estimate using the frequency band weights to obtain the original pulse amplitude after inversion. The conversion module is used to convert the apparent discharge quantity inside the bushing based on the sensor's own calibration coefficient and the original pulse amplitude, so as to obtain an estimated value of the apparent discharge quantity inside the bushing.

[0014] Thirdly, the present invention provides an electronic device, including a memory and a processor; the memory is used to store computer-executable instructions, and the processor executes the computer-executable instructions to implement the steps of a method for accurately extracting the apparent discharge quantity of a high-voltage bushing.

[0015] Fourthly, the present invention provides a computer-readable storage medium storing computer-executable instructions that, when executed by a processor, implement the steps of a method for accurately extracting the apparent discharge quantity of a high-voltage bushing.

[0016] Compared with existing technologies, the advantages of this invention are as follows: This invention incorporates the propagation attenuation characteristics of discharge pulses within the complex internal structure of high-voltage bushings into the correction system for apparent discharge quantities. Through multi-band decomposition and weighted fusion, it effectively compensates for amplitude losses caused by propagation path and monitoring system bandwidth limitations, significantly improving the accuracy of quantitative results and more realistically reflecting the actual discharge intensity of insulation defects. The weighted fusion mechanism provided by this invention can dynamically allocate weights according to the attenuation degree of each frequency band signal, automatically weakening the influence of frequency bands severely attenuated and subject to noise interference, enhancing anti-interference capability and robustness in complex electromagnetic environments. Furthermore, based on experimental calibration of the equipment's own transmission characteristics, this invention is theoretically applicable to various high-voltage equipment with complex internal structures, unrestricted by sensor type or monitoring frequency band, and possesses good universality. Finally, a typical bushing attenuation transfer function library can be pre-established in the laboratory or before shipment and used as equipment fingerprint data. During online monitoring, only signal processing and calculation inversion are required, with controllable computational load, easy integration into existing online monitoring systems, and strong engineering practicality. Attached Figure Description

[0017] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a schematic diagram of the overall process logic of a method for accurately extracting the apparent discharge quantity of a high-voltage bushing according to an embodiment of the present invention.

[0019] Figure 2 This is a schematic diagram of a typical bushing attenuation transfer function calibration experimental platform for a high-voltage bushing apparent discharge quantity accurate extraction method provided in an embodiment of the present invention.

[0020] Figure 3 This is a schematic diagram illustrating the principle of multi-band weighted fusion inversion for a method for accurately extracting apparent discharge of high-voltage bushings according to an embodiment of the present invention. Detailed Implementation

[0021] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0022] Example 1, referring to Figure 1 As one embodiment of the present invention, a method for accurately extracting the apparent discharge quantity of a high-voltage bushing is provided, such as... Figure 1 The specific steps shown are as follows: S100: Inject standard pulse signals at multiple preset positions inside the sample bushing and calculate the attenuation transfer function at each preset position to construct a typical high-voltage bushing attenuation transfer function library; S200: Performs multi-band decomposition on the partial discharge signal acquired by online monitoring and extracts the amplitude characteristics of each frequency band signal component; S300: Based on the preliminary diagnosis results of the defect type, select the corresponding attenuation transfer function from the typical high-voltage bushing attenuation transfer function library, and calculate the frequency band weight based on the amplitude of the attenuation transfer function at the center frequency of each frequency band. S400: Calculate the original amplitude estimate of each frequency band individually based on the amplitude characteristics of each frequency band signal component, and use the frequency band weight to perform a weighted sum of each original amplitude estimate to obtain the original pulse amplitude after inversion; S500: Based on the sensor's own calibration coefficient and the original pulse amplitude, the apparent discharge quantity inside the bushing is converted to obtain an estimated value of the apparent discharge quantity inside the bushing.

[0023] It should be noted that this invention incorporates the propagation attenuation characteristics of discharge pulses within the complex internal structure of high-voltage bushings into the correction system for apparent discharge quantities. Through multi-band decomposition and weighted fusion, it effectively compensates for amplitude losses caused by propagation path and monitoring system bandwidth limitations, significantly improving the accuracy of quantitative results and more realistically reflecting the actual discharge intensity of insulation defects. The weighted fusion mechanism provided by this invention can dynamically allocate weights according to the attenuation degree of each frequency band signal, automatically weakening the influence of frequency bands severely attenuated and subject to noise interference, enhancing anti-interference capability and robustness in complex electromagnetic environments. Furthermore, based on experimental calibration of the equipment's own transmission characteristics, this invention is theoretically applicable to various high-voltage equipment with complex internal structures, unrestricted by sensor type or monitoring frequency band, and possesses good universality. Finally, a typical bushing attenuation transfer function library can be pre-established in the laboratory or before shipment and used as equipment fingerprint data. During online monitoring, only signal processing and calculation inversion are required, with controllable computational load, easy integration into existing online monitoring systems, and strong engineering practicality.

[0024] Example 2, refer to Figure 2 and Figure 3 Based on the previous embodiment, this embodiment provides a specific implementation method for accurately extracting the apparent discharge quantity of high-voltage bushings.

[0025] In this embodiment of the invention, step S100 involves injecting standard pulse signals at multiple preset positions inside the sample bushing and calculating the attenuation transfer function at each preset position to construct a typical high-voltage bushing attenuation transfer function library. This includes the following sub-steps A1-A4: In A1: Select a sample bushing of the same model and specification as the high-voltage bushing to be monitored. On the calibration experimental platform, connect the standard pulse generator to the preset simulated discharge point inside the sample bushing through a coupling element. Install the monitoring sensor at the predetermined installation position outside the bushing. Specifically, a sample bushing of the same model and specification as the high-voltage bushing to be monitored is selected, and calibration experiments are conducted in the laboratory. For example... Figure 2 As shown, on the calibration experimental platform, a device capable of precisely controlling the injected charge amount is used. The output of a standard pulse generator (unit: pC) is connected to a simulated discharge point pre-set at a location where typical defects may occur inside the bushing via a small coupling capacitor; a sensor for online monitoring is installed at a predetermined installation location outside the bushing.

[0026] It should be noted that the preset positions in this embodiment refer to multiple simulated discharge points pre-set according to the areas where typical defects inside the high-voltage bushing may occur. Specifically, these include positions near the high-voltage conductor, the end of the capacitor screen, and near the grounding screen. These positions cover key areas where the electric field is concentrated and the insulation is weak inside the bushing, representing the most likely source locations of partial discharge in actual operation. This ensures that the attenuation transfer function library can fully reflect the differences in the propagation path from different discharge points to the sensor end.

[0027] In A2: Record the time-domain response signal measured by the sensor when a standard pulse generator injects a known amount of charge. ; In A3: Time-domain response signal The spectrum is obtained by performing a Fourier transform. The attenuation transfer function at multiple preset locations is calculated based on the spectrum, the known amount of injected charge, and the injection coefficient of the coupling element. Specifically, the attenuation transfer function at multiple preset locations is calculated based on the spectrum, the known amount of injected charge, and the injection coefficient of the coupling element. Considering that the injected signal is a charge pulse, while the sensor output is typically a voltage signal, this transfer function essentially characterizes the frequency domain response of a pulse generated per unit charge after propagation. The formula is expressed as: in, This represents the injection coefficient (unit: V / pC or mV / pC) that takes into account the effect of coupling capacitance. It can be accurately calculated or obtained through secondary standard calibration when the injection loop parameters are known.

[0028] In A4: Store the attenuation transfer functions for multiple preset positions into the typical high-voltage bushing attenuation transfer function library.

[0029] Specifically, the above operations are repeated for multiple internal locations to establish a typical high-voltage bushing attenuation transfer function library. ,in, This represents the preset number of typical locations.

[0030] In an optional implementation, the typical high-voltage bushing attenuation transfer function library can also be constructed using the frequency domain sweep measurement method. That is, a vector network analyzer is used instead of a standard pulse generator. The sweep signal source is injected into the simulated discharge point inside the bushing through a coupling element. At the same time, the receiver of the network analyzer measures the response signal at the external sensor. The attenuation transfer function at each preset position is generated by directly acquiring the amplitude and phase frequency response data at different frequency points.

[0031] In an optional implementation, the typical high-voltage bushing attenuation transfer function library can also be constructed using numerical simulation calculation method. That is, a three-dimensional full-wave electromagnetic field simulation model is established based on the actual geometric dimensions, material dielectric parameters and metal structure of the bushing. In the model, excitation sources are set at each preset discharge point and probes are set at the sensor installation position. The frequency domain transmission characteristics from each excitation point to the probe end are obtained through simulation calculation, thereby constructing the transfer function library.

[0032] It should be noted that step S100 above, by conducting standard pulse injection experiments on multiple typical locations inside the same model of sample sleeve in a laboratory environment, can pre-obtain the complete amplitude-frequency attenuation characteristics of the discharge signal as it propagates from different discharge points to the external sensor, forming a reusable device fingerprint database; at the same time, once this database is established, it can cover the monitoring needs of all sleeves of the same model without the need for repeated calibration of each device.

[0033] In this embodiment of the invention, step S200 performs multi-band decomposition on the partial discharge signal acquired by online monitoring and extracts the amplitude characteristics of each frequency band signal component, which includes the following sub-steps B1-B3: In B1: During online monitoring, the partial discharge signal output by the sensor is continuously acquired; In B2: In response to the detection of a partial discharge pulse event, the time-domain waveform containing the partial discharge pulse event is captured, and the time-domain waveform is decomposed into multiple non-overlapping or partially overlapping frequency band signal components. Specifically, a real-time trigger recognition mechanism is used to analyze the waveform. When the signal amplitude exceeds the detection threshold dynamically set according to the background noise or the waveform meets the typical time-domain morphology criteria for partial discharge pulses, a partial discharge pulse event is detected. The detection threshold is usually set to a voltage value slightly higher than the background noise level, such as 20 millivolts. The typical time-domain morphology criteria include a nanosecond-level rising edge, a rising edge duration of no more than 10 nanoseconds, a pulse exhibiting a double exponential or oscillating decay pattern with a rapid rise followed by a slow decay, a pulse width in the range of tens of nanoseconds to several microseconds, and a specific phase distribution pattern within the power frequency cycle. After any trigger condition is met, the system immediately extracts a time window signal containing the complete pulse waveform centered on the trigger time, which serves as the input data for subsequent multi-band decomposition and feature extraction.

[0034] Specifically, after detecting a partial discharge pulse event, the time-domain waveform containing the partial discharge pulse event is extracted. .right Perform multi-resolution analysis or bandpass filter group processing to decompose it into Within a number of non-overlapping or partially overlapping frequency bands. Let the first... The frequency band range is The corresponding frequency band signal components are Its spectrum is .

[0035] In B3: Extract the amplitude characteristics of each frequency band signal component, where the amplitude characteristics include peak value, root mean square value, and energy; Specifically, peak value: ; Specifically, the root mean square (RMS) value: ; Specifically, energy: ; Where T represents the pulse time window length. Preferably, the peak value is used as the characteristic quantity because it is closest to the conventional definition of apparent discharge.

[0036] It should be noted that the above step S200 can decompose the wideband pulse signal into several narrowband components, thereby preserving the amplitude distribution information of the signal in different frequency ranges, providing a fine-grained data basis for subsequent differential correction, avoiding the information loss problem caused by traditional methods that only use the overall peak or single frequency band characteristics, and enabling attenuation correction to be processed separately for different frequency components.

[0037] In this embodiment of the invention, step S300 selects the corresponding attenuation transfer function from the typical high-voltage bushing attenuation transfer function library based on the preliminary diagnosis result of the defect type, and calculates the frequency band weight based on the amplitude of the attenuation transfer function at the center frequency of each frequency band. This includes the following sub-steps C1-C3: In C1: Based on the preliminary diagnosis of the defect type, select the attenuation transfer function corresponding to the current suspected discharge point location from the typical high-voltage bushing attenuation transfer function library; In C2: The weight value corresponding to each frequency band is determined based on the amplitude of the attenuation transfer function at the center frequency of each frequency band; In C3: The weight values ​​of each frequency band are normalized so that the sum of all weights is 1; Specifically, the attenuation transfer function corresponding to the current suspected discharge point location i is selected from the typical high-voltage bushing attenuation transfer function library. For the first Each frequency band, calculate its center frequency. The magnitude of the transfer function at that point .

[0038] Specifically, to characterize the influence of different frequency bands on the propagation of partial discharge pulse signals, the first... Band weight of each frequency band This weight reflects the degree of attenuation of the signal in this frequency band during propagation. The less attenuated a frequency band (i.e., the larger the transfer function amplitude), the more original information its signal components retain, and thus it should be assigned a higher weight or confidence level in the inversion calculation. Therefore, this weight... It can be defined and expressed as follows: in, The power exponent is used to adjust the sensitivity of the weights to the magnitude of the transfer function; it is usually taken as... or .when At this time, more emphasis is placed on frequency bands with smaller attenuation. The weights satisfy the normalization condition: .

[0039] It should be noted that step S300 above enables the frequency band with lighter attenuation and higher signal-to-noise ratio to dominate in subsequent fusion, while the frequency band with severe attenuation is automatically suppressed, thereby realizing an adaptive weighting strategy based on signal quality and effectively improving the stability of the correction results in complex field environments.

[0040] In this embodiment of the invention, step S400 calculates the original amplitude estimate of each frequency band individually based on the amplitude characteristic of each frequency band signal component, and performs a weighted summation of each original amplitude estimate using frequency band weights to obtain the inverted original pulse amplitude. This includes the following sub-steps D1 and D2: In D1: Divide the amplitude characteristic of the frequency band signal component by the amplitude of the attenuation transfer function at the center frequency of the frequency band to obtain the original amplitude estimate of the frequency band inversion alone; In D2: Using the frequency band weight as the weighting coefficient, the original amplitude estimates retrieved from each frequency band individually are weighted and summed to obtain the original pulse amplitude after inversion; Specifically, for the signal characteristics observed in each frequency band It can be regarded as a characteristic quantity of the original discharge pulse in this frequency band. The result after attenuation and noise addition. Theoretically, there is an approximate relationship: Assuming the energy or amplitude distribution of the original pulse across all frequency bands is flat (for a pulse with a nanosecond rising edge, its spectrum can be approximated as flat over a wide range), then it can be considered that... For all It is a constant .

[0041] Specifically, such as Figure 3 The diagram shown illustrates the principle of multi-band weighted fusion inversion, utilizing calculated frequency band weights. The original partial discharge pulse amplitude estimates retrieved from each frequency band are weighted and fused to obtain the final retrieved original pulse amplitude. The formula is expressed as: in, It is the first The original amplitude estimate obtained by inverting the frequency band separately.

[0042] In an alternative implementation, the original pulse amplitude can also be obtained using the overall waveform inversion method. That is, instead of performing multi-band decomposition on the signal, the complete time-domain waveform acquired by the sensor is directly taken as a whole, and the corresponding attenuation transfer function selected from the transfer function library is used to perform time-domain deconvolution or frequency-domain inverse filtering on the entire waveform. The original pulse amplitude is recovered by compensating for the propagation attenuation and system response of the entire frequency band at once.

[0043] In an optional implementation, the original pulse amplitude can also be obtained using an inversion method based on a machine learning regression model. That is, during the construction of the transfer function library, multi-band feature quantities and their corresponding real original amplitudes at different preset positions and different injected charge amounts are collected simultaneously to form a training sample set. A neural network or support vector regression model is then trained to establish a nonlinear mapping relationship between the frequency band feature quantities and the original pulse amplitude. In actual monitoring, the extracted multi-band amplitude features can be directly input into the trained model to output the inverted original pulse amplitude.

[0044] It should be noted that step S400 above can effectively fuse the independent estimation information provided by multiple frequency bands, which not only compensates for the differential attenuation loss of each frequency component during propagation, but also suppresses the influence of single-band random noise through the redundant information of multiple frequency bands, and finally obtains the original pulse amplitude inversion result that is closer to the true intensity of the discharge point.

[0045] In this embodiment of the invention, step S500 calculates the apparent discharge quantity inside the bushing based on the sensor's own calibration coefficient and the original pulse amplitude to obtain an estimated value of the apparent discharge quantity inside the bushing, which includes the following sub-steps E1-E4: In E1: Obtain the calibration coefficients of the sensor itself. The calibration coefficients are obtained by calibrating the sensor by directly connecting a standard pulse generator to the sensor input terminal. They are used to characterize the response characteristics of the sensor and subsequent monitoring circuits to the input signal. It should be noted that, due to the inversion obtained It is the equivalent, unattenuated voltage signal characteristic of the original discharge pulse at the sensor end, in order to obtain the apparent discharge quantity. (pC) requires the use of the system's own calibration coefficients. (Unit: mV / pC). This coefficient is obtained through conventional calibration by directly connecting a standard pulse generator to the sensor input and reflects the response of the sensor and subsequent monitoring circuitry.

[0046] In E2: Divide the original pulse amplitude after inversion by the calibration coefficient to obtain the estimated value of apparent discharge inside the bushing; Specifically, the apparent discharge quantity inside the bushing is calculated using the sensor's own calibration coefficient, expressed by the following formula: From this, we obtain This is the estimated apparent discharge quantity, which is closer to the actual discharge point inside the bushing after propagation attenuation correction.

[0047] In E3: Based on the original amplitude estimates retrieved individually for each frequency band, the weighted standard deviation is calculated using frequency band weights. The formula is expressed as: In E4: the weighted standard deviation is used as an indicator to measure the consistency of the inversion results for each frequency band; Specifically, weighted standard deviation The smaller the value, the better the consistency of the inversion results across different frequency bands, and the better the final result. The higher the confidence level, the better.

[0048] In an optional implementation, the evaluation of the apparent discharge extraction results can also be carried out using the relative deviation analysis method, which is to divide the difference between the maximum and minimum values ​​of the original amplitude estimates retrieved separately for each frequency band by the final result after weighted fusion to obtain the relative deviation index. The smaller the index, the smaller the difference between the retrieval results of each frequency band and the higher the reliability of the fusion result.

[0049] In an optional implementation, the evaluation of the apparent discharge quantity extraction results can also adopt a comprehensive confidence assessment method based on signal-to-noise ratio weighting. That is, while calculating the inversion results of each frequency band, the signal-to-noise ratio of each frequency band signal component is estimated separately. The product of the signal-to-noise ratio of each frequency band and the transfer function amplitude is used as the comprehensive confidence factor. After performing secondary weighting on the inversion results of each frequency band, the equivalent consistency coefficient is calculated. This coefficient can simultaneously reflect the comprehensive impact of propagation attenuation and field noise on the confidence of information in each frequency band.

[0050] It should be noted that step S500 above eliminates the systematic energy loss caused by the propagation path in the final output result, and is compatible with existing monitoring standards and evaluation systems. It provides operators with quantitative indicators that can intuitively judge the degree of insulation degradation, and improves the engineering usability of online monitoring data.

[0051] Example 3 provides an application example of a method for accurately extracting the apparent discharge quantity of a high-voltage bushing, which verifies and illustrates the technical effectiveness of this method.

[0052] This embodiment uses the extraction of apparent discharge quantity inside the high-voltage bushing of a 110kV oil-immersed transformer as an application scenario. First, a typical high-voltage bushing attenuation transfer function library is constructed, and multi-band decomposition and feature extraction are performed on the partial discharge signal acquired through online monitoring. Then, based on the frequency band weight calculation using the attenuation transfer function, the original partial discharge pulse amplitude is retrieved through multi-band weighted fusion. Finally, the apparent discharge quantity inside the bushing is converted and extracted using the sensor's own calibration coefficients. The confidence level of the final extracted apparent discharge quantity result is evaluated to ensure its reliability.

[0053] The specific steps and implementation process are as follows: In the laboratory, a sample of the same type of high-voltage bushing was selected. A standard pulse generator (capable of generating pulses with a rise time of 1ns, a half-peak time of 50ns, and an amplitude corresponding to the charge quantity) was used. The double exponential pulse is connected via a 10pF coupling capacitor to a simulated pinhead defect at the end of the outermost capacitor screen inside the bushing (location). Install an HFCT sensor (same model for online monitoring) with a bandwidth of 1kHz-30MHz on the grounding wire of the bushing end screen. Record the sensor output signal. .right Perform FFT to obtain Known The coupling capacitance and the equivalent impedance of the injection circuit determine Calculate the transfer function: Extract its amplitude-frequency characteristics Assume that its amplitude decreases approximately linearly (dB / decibels) within the frequency band of interest. To simplify the example, define three characteristic frequencies: low frequency... Mid-frequency ,high frequency And assume that the following was measured: This Store in the database.

[0054] Furthermore, the online monitoring system detected a partial discharge pulse at a certain moment. Its waveform peak value If the calibration coefficients of the traditional system are used directly... (Obtained through direct injection calibration at the sensor port), the apparent discharge quantity calculated using the traditional method is: The multi-band decomposition process is as follows: The signal is decomposed into three frequency bands (corresponding to the calibration frequency; due to propagation attenuation, the high-frequency component amplitude is lower) by a set of third-order Butterworth bandpass filters: ① Band 1: [1MHz, 5MHz], center Extracting signals peak .

[0055] ②Band 2: [5MHz, 15MHz], center Extracting signals peak .

[0056] ③ Band 3: [15MHz, 25MHz], center Extracting signals peak .

[0057] Furthermore, assuming that based on the assisted positioning, it is determined that the discharge may occur at a certain location. Nearby. Therefore, it was called. Power of exponent Therefore, the square of the transfer function amplitude at the center frequency of each frequency band is calculated as follows: Summing yields: ; Based on this, the frequency band weights are calculated as follows: Furthermore, the original peak estimates for each frequency band are calculated individually as follows: Based on this, a weighted fusion is performed to obtain the final inverted original peak value: It can be seen that although the high-frequency band inversion value (20.0mV) is large, its weight (0.0085) is extremely small and has little impact on the final result; the low-frequency band inversion value (8.0mV), which has the dominant weight, determines the main result.

[0058] Furthermore, system calibration coefficients are used. The final result of the apparent discharge quantity inside the bushing is calculated as follows: Furthermore, the confidence index for the aforementioned extraction results of apparent discharge is calculated as follows: Further calculations yielded a relative standard deviation of approximately .

[0059] Based on the above comparative analysis: the apparent discharge quantity calculated using the traditional online monitoring method is as follows: The apparent discharge quantity of partial discharge inside the bushing obtained by calculating and extracting using the method of this invention is as follows: Confidence index (Compared to 21.7%). This is mainly due to the uncorrected traditional method overestimating the apparent discharge (25pC vs 14.5pC), with an error rate exceeding 70%. This is because the signal received by the sensor (15mV peak) is the result of severe attenuation of the original pulse, especially the high-frequency components carrying a large amount of energy, which are almost completely attenuated. This invention uses multi-band attenuation correction to invert the equivalent original signal peak value without attenuation, which is approximately 8.72mV, and then calculates a discharge quantity of 14.5pC that is closer to the true value. The weights are mostly concentrated in the low-frequency band (Band 1) with less attenuation, reflecting the method's reliance on high signal-to-noise ratio information. The confidence assessment indicates that there is some uncertainty in the inversion results, mainly due to the difference in the inversion values ​​of the mid-frequency band (Band 2). This embodiment clearly demonstrates the entire process of this invention from data acquisition and processing to final quantitative calculation, and its significant improvement over traditional methods. In practical applications, the frequency band division can be more refined, the transfer function can be obtained through sampling at more locations and with denser frequencies, and more advanced signal processing and fusion algorithms can be used to further improve accuracy and reliability.

[0060] Example 4: This example provides a system for accurately extracting the apparent discharge quantity of a high-voltage bushing, comprising: The transfer function library construction module is used to inject standard pulse signals at multiple preset positions inside the sample bushing and calculate the attenuation transfer function at each preset position to construct a typical high-voltage bushing attenuation transfer function library. The feature extraction module is used to perform multi-band decomposition on the partial discharge signal acquired by online monitoring and extract the amplitude features of each frequency band signal component; The frequency band weight calculation module is used to select the corresponding attenuation transfer function from the typical high-voltage bushing attenuation transfer function library based on the preliminary diagnosis results of the defect type, and calculate the frequency band weight based on the amplitude of the attenuation transfer function at the center frequency of each frequency band. The inversion calculation module is used to calculate the original amplitude estimate of each frequency band based on the amplitude characteristics of each frequency band signal component, and to perform a weighted summation of each original amplitude estimate using frequency band weights to obtain the original pulse amplitude after inversion. The conversion module is used to convert the apparent discharge quantity inside the bushing based on the sensor's own calibration coefficient and the original pulse amplitude, so as to obtain the estimated value of the apparent discharge quantity inside the bushing.

[0061] It should be noted that the technical solution of the high-voltage bushing apparent discharge quantity precise extraction system is based on the same concept as the technical solution of the high-voltage bushing apparent discharge quantity precise extraction method described above. For details not described in detail in the technical solution of the high-voltage bushing apparent discharge quantity precise extraction system in this embodiment, please refer to the description of the technical solution of the high-voltage bushing apparent discharge quantity precise extraction method described above.

[0062] The above-mentioned unit modules can be embedded in the processor of the electronic device in hardware form or independent of it, or they can be stored in the memory of the electronic device in software form, so that the processor can call and execute the corresponding operations of the above modules.

[0063] This embodiment also provides an electronic device, which includes a processor, a memory, a communication interface, a display screen, and an input device connected via a system bus. The processor provides computing and control capabilities. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through WiFi, carrier networks, NFC (Near Field Communication), or other technologies. When the computer program is executed by the processor, it implements a method for accurately extracting the apparent discharge quantity of a high-voltage bushing. The display screen can be a liquid crystal display (LCD) or an e-ink display. The input device can be a touch layer covering the display screen, buttons, a trackball, or a touchpad mounted on the device's casing, or an external keyboard, touchpad, or mouse.

[0064] This embodiment also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method proposed in the above embodiments.

[0065] The storage medium proposed in this embodiment belongs to the same inventive concept as the method proposed in the above embodiments. Technical details not described in detail in this embodiment can be found in the above embodiments, and this embodiment has the same beneficial effects as the above embodiments.

[0066] Based on the above description of the implementation methods, those skilled in the art can clearly understand that the present invention can be implemented using software and necessary general-purpose hardware, and of course, it can also be implemented using hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory, random access memory, flash memory, hard disk, or optical disk, and includes several instructions to cause an electronic device (which may be a personal computer, server, or network device, etc.) to execute the method of the embodiments of the present invention.

[0067] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the present invention.

Claims

1. A method for accurately extracting the apparent discharge quantity of a high-voltage bushing, characterized in that, include: Standard pulse signals are injected into multiple preset positions inside the sample bushing, and the attenuation transfer function at each preset position is calculated to construct a typical high-voltage bushing attenuation transfer function library. The partial discharge signal acquired by online monitoring is decomposed into multiple frequency bands, and the amplitude characteristics of each frequency band signal component are extracted. Based on the preliminary diagnosis results of the defect type, the corresponding attenuation transfer function is selected from the typical high-voltage bushing attenuation transfer function library, and the frequency band weight is calculated based on the amplitude of the attenuation transfer function at the center frequency of each frequency band. Based on the amplitude characteristics of each frequency band signal component, calculate the original amplitude estimate of each frequency band individually, and use the frequency band weight to perform a weighted sum of each original amplitude estimate to obtain the original pulse amplitude after inversion; The apparent discharge quantity inside the bushing is calculated based on the sensor's own calibration coefficient and the original pulse amplitude to obtain an estimated value of the apparent discharge quantity inside the bushing.

2. The method for accurately extracting the apparent discharge quantity of a high-voltage bushing as described in claim 1, characterized in that, The construction of a typical high-voltage bushing attenuation transfer function library includes: Select a sample bushing of the same model and specification as the high-voltage bushing to be monitored. On the calibration experimental platform, connect the standard pulse generator to the preset simulated discharge point inside the sample bushing through a coupling element. Install the monitoring sensor at the predetermined installation position outside the bushing. Record the time-domain response signal measured by the sensor when a standard pulse generator injects a known amount of charge; The frequency spectrum is obtained by performing a Fourier transform on the time-domain response signal. Based on the frequency spectrum, the known amount of injected charge, and the injection coefficient of the coupling element, the attenuation transfer function at multiple preset positions is calculated. The attenuation transfer functions of the multiple preset positions are stored in a typical high-voltage bushing attenuation transfer function library.

3. The method for accurately extracting the apparent discharge quantity of a high-voltage bushing as described in claim 2, characterized in that, The process of multi-band decomposing the partial discharge signal acquired by online monitoring and extracting the amplitude characteristics of each frequency band signal component includes: During online monitoring, the partial discharge signal output by the sensor is continuously collected; In response to the detection of a partial discharge pulse event, a time-domain waveform containing the partial discharge pulse event is captured, and the time-domain waveform is decomposed into multiple non-overlapping or partially overlapping frequency band signal components. The amplitude characteristics of each frequency band signal component are extracted, including peak value, root mean square value, and energy.

4. The method for accurately extracting the apparent discharge quantity of a high-voltage bushing as described in claim 3, characterized in that, The calculation of the frequency band weight includes: Based on the preliminary diagnosis of the defect type, select the attenuation transfer function corresponding to the current suspected discharge point from the typical high-voltage bushing attenuation transfer function library; The weight value corresponding to each frequency band is determined based on the amplitude of the attenuation transfer function at the center frequency of each frequency band. The weight values ​​of each frequency band are normalized so that the sum of all weights is 1.

5. The method for accurately extracting the apparent discharge quantity of a high-voltage bushing as described in claim 4, characterized in that, The obtained inverted original pulse amplitude includes: Divide the amplitude characteristic of the frequency band signal component by the amplitude of the attenuation transfer function at the center frequency of the frequency band to obtain the original amplitude estimate of the frequency band inversion. Using the frequency band weights as weighting coefficients, the original amplitude estimates retrieved from each frequency band are weighted and summed to obtain the retrieved original pulse amplitude.

6. The method for accurately extracting the apparent discharge quantity of a high-voltage bushing as described in claim 5, characterized in that, The estimated value of the apparent discharge inside the bushing includes: The calibration coefficients of the sensor itself are obtained by calibrating a standard pulse generator directly connected to the sensor input terminal. These calibration coefficients are used to characterize the response characteristics of the sensor and subsequent monitoring circuits to the input signal. Divide the original pulse amplitude after inversion by the calibration coefficient to obtain the estimated value of apparent discharge inside the bushing.

7. The method for accurately extracting the apparent discharge quantity of a high-voltage bushing as described in claim 6, characterized in that, The method further includes a confidence assessment of the estimated apparent discharge quantity inside the bushing: Based on the original amplitude estimates retrieved individually for each frequency band, the weighted standard deviation is calculated using the frequency band weights; The weighted standard deviation is used as an indicator to measure the consistency of the inversion results for each frequency band.

8. A system for accurately extracting apparent discharge quantity of high-voltage bushings, realizing the method for accurately extracting apparent discharge quantity of high-voltage bushings as described in any one of claims 1 to 7, characterized in that, include: The transfer function library construction module is used to inject standard pulse signals at multiple preset positions inside the sample bushing and calculate the attenuation transfer function at each preset position to construct a typical high-voltage bushing attenuation transfer function library. The feature extraction module is used to perform multi-band decomposition on the partial discharge signal acquired by online monitoring and extract the amplitude features of each frequency band signal component; The frequency band weight calculation module is used to select the corresponding attenuation transfer function from the typical high-voltage bushing attenuation transfer function library according to the preliminary diagnosis results of the defect type, and calculate the frequency band weight based on the amplitude of the attenuation transfer function at the center frequency of each frequency band. The inversion calculation module is used to calculate the original amplitude estimate of each frequency band individually based on the amplitude characteristics of the signal components of each frequency band, and to perform a weighted summation of each original amplitude estimate using the frequency band weights to obtain the original pulse amplitude after inversion. The conversion module is used to convert the apparent discharge quantity inside the bushing based on the sensor's own calibration coefficient and the original pulse amplitude, so as to obtain an estimated value of the apparent discharge quantity inside the bushing.

9. An electronic device comprising a memory and a processor, characterized in that: The memory is used to store computer-executable instructions, and when the processor executes the computer-executable instructions, it implements the steps of the method for accurately extracting the apparent discharge quantity of a high-voltage bushing as described in any one of claims 1 to 7.

10. A computer-readable storage medium having computer-executable instructions stored thereon, characterized in that: When the computer-executable instructions are executed by the processor, they implement the steps of the method for accurately extracting the apparent discharge quantity of a high-voltage bushing as described in any one of claims 1 to 7.

Images