GIS ultrahigh frequency sensor ultrasonic cooperative detection system based on piezoelectric energization

Abstract

The invention discloses a GIS ultrahigh frequency sensor ultrasonic cooperative detection system based on piezoelectric energization, and relates to the field of electric power detection, and the system comprises an excitation module which is used for transmitting a piezoelectric excitation signal with a specific frequency to a preset detection region in GIS equipment, so as to trigger the potential partial discharge acoustoelectric signal radiation of the region; the sensing module is used for receiving ultrahigh-frequency electromagnetic signals and ultrasonic vibration signals radiated by partial discharge in the GIS equipment; the excitation frequency, the filtering bandwidth and the sampling rate are dynamically optimized according to the detection result, the detection accuracy and efficiency are continuously improved, and reliable and effective guarantee is provided for safe operation and defect early warning of GIS equipment.

Description

Technical Field

[0001] This invention relates to the field of power detection technology, specifically to a piezoelectric-enabled GIS UHF sensor ultrasonic collaborative detection system. Background Technology

[0002] Gas-insulated switchgear is a core piece of equipment in high-voltage power grids. Its insulation performance directly determines the safe operation of the power grid. During operation, it is prone to insulation degradation due to internal partial discharge. If it is not detected in time, it will lead to equipment failure and affect the stability of the power grid.

[0003] However, traditional testing methods often employ either ultra-high frequency or ultrasonic technology. The former is susceptible to electromagnetic interference, while the latter is susceptible to mechanical noise. Both methods suffer from missed or false detections, making it difficult to accurately locate defects.

[0004] To address this, we propose a piezoelectric-enabled UHF sensor-ultrasonic collaborative detection system for GIS. Summary of the Invention

[0005] In view of the above-mentioned shortcomings of the existing technology, the present invention provides an ultrasonic collaborative detection system for GIS UHF sensors based on piezoelectric energy, which can effectively solve the problems of the existing technology.

[0006] To achieve the above objectives, the present invention is implemented through the following technical solutions;

[0007] This invention discloses a piezoelectric-enabled GIS UHF sensor ultrasonic collaborative detection system, comprising:

[0008] The system comprises the following modules: an excitation module, which transmits a piezoelectric excitation signal of a specific frequency to a preset detection area within the GIS equipment to trigger the radiation of acoustic and electrical signals of potential partial discharge in that area; a sensing module, which receives the ultra-high frequency electromagnetic signals and ultrasonic vibration signals radiated by partial discharge within the GIS equipment; a conditioning module, which receives the ultra-high frequency signals and ultrasonic signals output from the sensing module and performs filtering, amplification, and impedance matching processing to output a standardized electrical signal; a collaborative analysis module, which performs time-domain correlation matching and frequency-domain feature fusion calculation on the standardized ultra-high frequency signals and ultrasonic signals output from the conditioning module; a storage module, which classifies and stores the feature fusion calculation results data and the original standardized signals output from the analysis module to establish a data index association table; and a feedback module, which matches the feature fusion calculation results data with a preset GIS partial discharge feature threshold and defect type feature library to output detection status information such as whether the equipment has partial discharge, discharge intensity, and suspected defect type, and feeds it back to the preset receiving end.

[0009] The excitation module is interconnected with the sensing module and the conditioning module via a wireless network. The conditioning module is interconnected with the collaborative analysis module via a wireless network. The collaborative analysis module is interconnected with the storage module and the feedback module via a wireless network.

[0010] Furthermore, during the operation phase of the excitation module, the equivalent radius of the cavity corresponding to the preset detection area of ​​the GIS equipment, the relative permittivity of the insulating medium within the preset detection area, and the sound velocity in the insulating medium are first obtained. Then, the optimal excitation frequency is calculated based on the above parameters. , For correction factor, >0; Preset the sound velocity in the insulating medium within the detection area of ​​the GIS equipment; Preset the equivalent radius of the cavity corresponding to the detection area of ​​the GIS equipment; Preset the relative permittivity of the insulating medium within the detection area for GIS equipment;

[0011] The optimal excitation frequency is denoted as the specific frequency applied in the excitation module.

[0012] Furthermore, the sensing module includes an ultra-high frequency sensor array and an ultrasonic sensor array;

[0013] The ultra-high frequency (UHF) sensor array consists of at least three UHF sensors. Each UHF sensor is evenly distributed circumferentially along the cavity of the preset detection area of ​​the GIS equipment, and the signal acquisition direction of each UHF sensor points to the center of the detection area. An ultrasonic sensor is arranged adjacent to each UHF sensor along the radial position of the GIS equipment, and the vibration acquisition surface of the ultrasonic sensor is in close contact with the outer wall of the cavity of the GIS equipment. During the operation of the sensing module, the same trigger signal is output to the UHF sensor array and the ultrasonic sensor array, so that the two types of sensors start acquiring signals at the same time, so that the acquisition time of the UHF electromagnetic signal radiated by partial discharge is synchronized with that of the ultrasonic vibration signal.

[0014] Furthermore, during the operation of the conditioning module, the filtering bandwidth is adjusted to filter out interference frequency components that are unrelated to the partial discharge signal, thereby completing the filtering operation.

[0015] Based on amplification gain The filtered UHF signal and the ultrasonic signal are amplified separately. Indicates the preset reference gain. Indicates the preset reference voltage. This represents the peak voltage of the signal after filtering.

[0016] The conditioning module is equipped with an adjustable impedance matching unit, which is integrated by a combination of a variable capacitor and a variable inductor. The adjustable impedance matching unit adjusts the equivalent impedance of the impedance matching network in real time through the combination of the variable capacitor and the variable inductor, so that the input impedance of the conditioning module is consistent with the output impedance of the sensor in the sensing module.

[0017] Furthermore, the time-domain correlation matching process in the collaborative analysis module is as follows:

[0018] The standardized UHF signal output from the conditioning module is extracted and denoted as... and standardized ultrasound signals, denoted as , where t is time;

[0019] Calculate the cross-correlation coefficients of two signals under different time delays:

[0020] ;

[0021] In the formula: ∈[-1,1]; These are the start and end times of signal acquisition, respectively.

[0022] Among them, the number of cross-relationships middle Representing the time delay variable, based on the above formula, we obtain... The event delay with the largest value is denoted as , to transmit ultrasound signals Adjusted to This allows the UHF signal and the ultrasonic signal to be aligned and matched in the time domain.

[0023] Furthermore, the frequency domain feature fusion calculation process in the collaborative analysis module is as follows:

[0024] Fourier transforms are performed on the normalized UHF signal and the normalized ultrasound signal after time-domain correlation matching, respectively, to obtain the frequency domain characteristics of the UHF signal, denoted as . and the frequency domain characteristics of ultrasound signals ,in Angular frequency;

[0025] Calculate each frequency point The corresponding signal-to-noise ratio of ultra-high frequency signals and the signal-to-noise ratio of ultrasonic signals, that is, the ratio of the amplitude of the signal to the amplitude of the noise at that frequency point;

[0026] Calculate each frequency point based on the signal-to-noise ratio. Corresponding weighting coefficients:

[0027] , Representing frequency points The corresponding UHF signal-to-noise ratio and frequency point The corresponding ultrasonic signal-to-noise ratio, These are the weighting coefficients for the frequency domain characteristics of UHF signals and the weighting coefficients for the frequency domain characteristics of ultrasonic signals, respectively.

[0028] Finally, the frequency domain features after fusion are calculated. :

[0029] .

[0030] Furthermore, during the operation of the storage module, the feature fusion calculation results data output by the collaborative analysis module and the original standardized signal output by the conditioning module are divided into three levels according to detection time, detection area, and signal type:

[0031] The first level uses the detection timestamp as the classification identifier; the second level uses the preset detection area number of the GIS equipment as the classification identifier under each timestamp; and the third level uses UHF standardized signal, ultrasonic standardized signal, and feature fusion result as the classification identifier under each area identifier.

[0032] Then, a data index association table is established based on the classification results. The fields of the data index association table include: detection timestamp, detection area number, signal type identifier, data storage path, data size, and data generation time.

[0033] The data index association table dynamically adjusts the index priority based on the detection status information output by the feedback module:

[0034] For detection data that is determined to have partial discharge, its index priority is raised to a preset high priority range;

[0035] For detection data that is determined to have no partial discharge, its index priority is set to a preset low priority range.

[0036] Furthermore, the logic for the feedback module to output detection status information during the operation phase is as follows:

[0037] Determine if partial discharge exists and its intensity:

[0038] Extract the mean value of the fused frequency domain features from the feature fusion calculation results output by the collaborative analysis module, and compare the mean value with the preset GIS partial discharge feature threshold.

[0039] If the average value does not exceed the threshold, it is determined that there is no partial discharge;

[0040] If the mean exceeds the threshold, it is determined that there is partial discharge, and the discharge intensity level is expressed as the ratio of the mean to the threshold.

[0041] Identify suspected defect types:

[0042] The feature fusion calculation results are converted into feature vectors. , This represents the values ​​of the fused frequency domain features at n key frequency points, and extracts the standard feature vectors corresponding to each standard defect type from a pre-set defect type feature library. Let m be the defect type number, and then calculate the cosine similarity between the two vectors:

[0043] ;

[0044] in, ∈[0,1], the closer its value is to 1, the higher the feature matching degree. The standard defect type with the highest similarity and greater than the preset similarity threshold is taken as the suspected defect type.

[0045] Finally, the feedback module integrates the results of the judgment and identification of whether partial discharge exists, the level of discharge intensity, and the type of suspected defect into detection status information, and then feeds it back to the preset receiving end.

[0046] Furthermore, during the operation of the feedback module, dynamic adjustment commands are simultaneously output to the excitation module, conditioning module, and sensing module based on the detection status information:

[0047] Send a frequency fine-tuning command to the excitation module:

[0048] When the detection status information indicates the presence of partial discharge and the suspected defect type is identified, the preset feature frequency set corresponding to that suspected defect type is extracted. k is the number of characteristic frequencies, and the following adjustments are performed;

[0049] Will Revised to , Indicates the frequency correction factor. Represents the maximum cosine similarity;

[0050] Send a filter bandwidth adjustment command to the conditioning module:

[0051] The adaptive filter unit is controlled to set the center frequency of the filter passband to... The mean value of the filtering bandwidth is reduced to x times the original bandwidth, where x is the bandwidth compression factor.

[0052] Send a sampling rate adjustment command to the sensing module:

[0053] The sampling rate of the UHF sensor array and the ultrasonic sensor array is increased to y times the original sampling rate, where y is the sampling rate amplification factor, and the sampling duration is extended to a preset duration.

[0054] Compared with the known prior art, the technical solution provided by this invention has the following beneficial effects:

[0055] This invention provides a piezoelectric-enabled UHF sensor ultrasonic collaborative detection system for GIS. During operation, the system calculates the optimal excitation frequency by combining the cavity parameters of the GIS equipment detection area with the characteristics of the insulating medium, accurately triggering the acoustic and electrical signal radiation of potential partial discharge. A multi-sensor array synchronously acquires UHF electromagnetic signals and ultrasonic vibration signals, ensuring temporal consistency and comprehensive directional coverage of signal acquisition. Targeted filtering, amplification, and impedance matching processes remove irrelevant interference, enhance signal strength, and ensure standardization. Time-domain correlation matching achieves precise alignment of acoustic and electrical signals, and frequency-domain signal-to-noise ratio weighted fusion enhances feature extraction accuracy. Data is categorized and stored according to detection time, region, and signal type, with dynamic index priority adjustment for convenient and efficient management and traceability. The system can accurately determine whether partial discharge exists in the equipment, the discharge intensity level, and the type of suspected defects. Furthermore, it can dynamically optimize the excitation frequency, filtering bandwidth, and sampling rate based on the detection results, continuously improving detection accuracy and efficiency, and providing reliable and effective protection for the safe operation and defect early warning of GIS equipment. Attached Figure Description

[0056] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without any creative effort.

[0057] Figure 1 This is a schematic diagram of a piezoelectric-enabled GIS UHF sensor ultrasonic collaborative detection system. Detailed Implementation

[0058] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0059] The present invention will be further described below with reference to embodiments.

[0060] Example:

[0061] This embodiment presents a piezoelectric-enabled GIS UHF sensor-ultrasonic collaborative detection system, such as... Figure 1 As shown, it includes:

[0062] The excitation module is used to transmit a piezoelectric excitation signal of a specific frequency to a preset detection area within the GIS equipment, so as to trigger the radiation of acoustic and electrical signals of potential local discharge in that area.

[0063] During the excitation module operation phase, the equivalent radius of the cavity corresponding to the preset detection area of ​​the GIS equipment, the relative permittivity of the insulating medium within the preset detection area, and the sound velocity in the insulating medium are first obtained. Then, the optimal excitation frequency is calculated based on the above parameters, and the optimal excitation frequency is recorded as the specific frequency applied in the excitation module.

[0064] ;

[0065] In the formula: To achieve the optimal excitation frequency; For correction factor, >0; Preset the sound velocity in the insulating medium within the detection area of ​​the GIS equipment; Preset the equivalent radius of the cavity corresponding to the detection area of ​​the GIS equipment; Preset the relative permittivity of the insulating medium within the detection area for GIS equipment;

[0066] Specifically, when the insulation dielectric loss is greater and the cavity structure deviation is more obvious within the preset detection area of ​​the GIS equipment, such as cavity roundness deviation or excessive inner wall roughness, the deviation will be more pronounced. The larger the value, the smaller the insulation dielectric loss and the smaller the cavity structure deviation in that area. The smaller the value;

[0067] The above formula is based on the core physical parameters of the preset detection area of ​​the GIS equipment. It uses the sound velocity in the insulating medium, the equivalent radius of the cavity, and the relative permittivity as key variables to construct an excitation frequency model that is deeply adapted to the actual structure of the equipment and the characteristics of the medium. At the same time, a correction coefficient is introduced to quantify the impact of insulating medium loss and cavity structure deviation. Through this dynamic calculation method, the problem of using a fixed frequency that cannot be accurately adapted to different equipment operating conditions can be avoided, ensuring that the excitation signal can effectively trigger the acoustic and electrical signal radiation of potential partial discharge in the detection area, providing a clearer detection object for subsequent signal perception.

[0068] The sensing module is used to receive ultra-high frequency electromagnetic signals and ultrasonic vibration signals radiated by partial discharge within the GIS equipment;

[0069] The sensing module includes an ultra-high frequency sensor array and an ultrasonic sensor array;

[0070] The UHF sensor array consists of at least three UHF sensors. Each UHF sensor is evenly distributed around the cavity of the preset detection area of ​​the GIS equipment, and the signal acquisition direction of each UHF sensor is pointing to the center of the detection area. Each UHF sensor is adjacent to an ultrasonic sensor along the radial position of the GIS equipment, and the vibration acquisition surface of the ultrasonic sensor is in close contact with the outer wall of the cavity of the GIS equipment.

[0071] During the operation phase of the sensing module, the same trigger signal is output to the UHF sensor array and the ultrasonic sensor array, so that the two types of sensors start to collect signals at the same time, so that the acquisition time of the UHF electromagnetic signal radiated by partial discharge is synchronized with that of the ultrasonic vibration signal.

[0072] The conditioning module is used to receive the ultra-high frequency signal and the ultrasonic signal output by the sensing module, and perform filtering, amplification and impedance matching processing respectively to output a standardized electrical signal.

[0073] During the conditioning module's operation, the filtering bandwidth is adjusted to filter out interference frequency components unrelated to the partial discharge signal, thus completing the filtering operation.

[0074] Based on amplification gain The filtered UHF signal and the ultrasonic signal are amplified separately. Indicates the preset reference gain. Indicates the preset reference voltage. This represents the peak voltage of the signal after filtering.

[0075] The conditioning module is equipped with an adjustable impedance matching unit, which is integrated by a combination of variable capacitor and variable inductor. The adjustable impedance matching unit adjusts the equivalent impedance of the impedance matching network in real time through the combination of variable capacitor and variable inductor, so that the input impedance of the conditioning module is consistent with the output impedance of the sensor in the sensing module.

[0076] The collaborative analysis module is used to perform time-domain correlation matching and frequency-domain feature fusion calculation on the standardized UHF signal and ultrasound signal output by the conditioning module.

[0077] The temporal correlation matching process in the collaborative analysis module is as follows:

[0078] The standardized UHF signal output from the conditioning module is extracted and denoted as... and standardized ultrasound signals, denoted as , where t is time;

[0079] Calculate the cross-correlation coefficients of two signals under different time delays:

[0080] ;

[0081] In the formula: ∈[-1,1]; These are the start and end times of signal acquisition, respectively.

[0082] Among them, the number of cross-relationships middle Representing the time delay variable, based on the above formula, we obtain... The event delay with the largest value is denoted as , to transmit ultrasound signals Adjusted to This allows the UHF signal and the ultrasonic signal to be aligned and matched in the time domain;

[0083] The above formula calculates the time delays of the two types of signals at different times, using the start and end times of signal acquisition as the integration interval, for the standardized UHF and ultrasound signals output by the conditioning module. Cross-correlation coefficients under ;

[0084] Because the electromagnetic signals generated by partial discharge and the vibration signals are physically related but propagate at different speeds, a time difference occurs. This can be addressed by searching for... Maximum time delay By adjusting the ultrasonic signal to be aligned with the UHF signal in the time domain, the problem of time asynchrony between the two types of sensor acquisition signals can be solved, laying a precise time domain foundation for subsequent frequency domain feature fusion and avoiding fusion analysis errors caused by time misalignment.

[0085] The frequency domain feature fusion calculation process in the collaborative analysis module is as follows:

[0086] Fourier transforms are performed on the normalized UHF signal and the normalized ultrasound signal after time-domain correlation matching, respectively, to obtain the frequency domain characteristics of the UHF signal, denoted as . and the frequency domain characteristics of ultrasound signals ,in Angular frequency;

[0087] Calculate each frequency point The corresponding signal-to-noise ratio of ultra-high frequency signals and the signal-to-noise ratio of ultrasonic signals, that is, the ratio of the amplitude of the signal to the amplitude of the noise at that frequency point;

[0088] Calculate each frequency point based on the signal-to-noise ratio. Corresponding weighting coefficients:

[0089] , Representing frequency points The corresponding UHF signal-to-noise ratio and frequency point The corresponding ultrasonic signal-to-noise ratio, These are the weighting coefficients for the frequency domain characteristics of UHF signals and the weighting coefficients for the frequency domain characteristics of ultrasonic signals, respectively.

[0090] Finally, the frequency domain features after fusion are calculated. :

[0091] ;

[0092] In the above scheme, the weighting coefficient formula is based on each angular frequency. Using the signal-to-noise ratio (SNR) of UHF signals and the SNR of ultrasonic signals as the basis for calculation, the ratio of the SNR of a single signal to the sum of the SNRs of the two types of signals is used as the weight of the frequency domain feature of that signal. This ensures that frequencies with higher SNR and stronger signal reliability are assigned higher weights. The fusion feature formula is then based on this weight, performing a weighted summation of the UHF frequency domain features and ultrasonic frequency domain features after Fourier transform. This design abandons the limitations of traditional fixed-weight fusion. By dynamically allocating the weights of each frequency point, the fusion result prioritizes the retention of high-reliability signal components and suppresses low SNR interference, while simultaneously achieving complementary advantages of the two types of signals in the frequency domain, significantly improving the accuracy and anti-interference capability of the frequency domain features.

[0093] The storage module is used to classify and store the feature fusion calculation results data and the original standardized signals output by the analysis module in order to establish a data index association table;

[0094] During the storage module's operation phase, the feature fusion calculation results data output by the collaborative analysis module and the raw standardized signal output by the conditioning module are divided into three levels according to detection time, detection area, and signal type:

[0095] The first level uses the detection timestamp as the classification identifier; the second level uses the preset detection area number of the GIS equipment as the classification identifier under each timestamp; and the third level uses UHF standardized signal, ultrasonic standardized signal, and feature fusion result as the classification identifier under each area identifier.

[0096] Then, a data index association table is established based on the classification results. The fields of the data index association table include: detection timestamp, detection area number, signal type identifier, data storage path, data size, and data generation time.

[0097] The data index association table dynamically adjusts the index priority based on the detection status information output by the feedback module:

[0098] For detection data that is determined to have partial discharge, its index priority is raised to a preset high priority range;

[0099] For detection data that is determined to have no partial discharge, its index priority is set to a preset low priority range;

[0100] The feedback module is used to match the feature fusion calculation results data with the preset GIS partial discharge feature threshold and defect type feature library, output the detection status information of whether the device has partial discharge, discharge intensity and suspected defect type, and feed it back to the preset receiving end;

[0101] The logic for outputting detection status information during the feedback module's operation is as follows:

[0102] Determine if partial discharge exists and its intensity:

[0103] Extract the mean value of the fused frequency domain features from the feature fusion calculation results output by the collaborative analysis module, and compare the mean value with the preset GIS partial discharge feature threshold.

[0104] If the average value does not exceed the threshold, it is determined that there is no partial discharge;

[0105] If the mean exceeds the threshold, it is determined that there is partial discharge, and the discharge intensity level is expressed as the ratio of the mean to the threshold.

[0106] Identify suspected defect types:

[0107] The feature fusion calculation results are converted into feature vectors. , This represents the values ​​of the fused frequency domain features at n key frequency points, and extracts the standard feature vectors corresponding to each standard defect type from a pre-set defect type feature library. Let m be the defect type number, and then calculate the cosine similarity between the two vectors:

[0108] ;

[0109] in, ∈[0,1], the closer its value is to 1, the higher the feature matching degree. The standard defect type with the highest similarity and greater than the preset similarity threshold is taken as the suspected defect type.

[0110] Finally, the feedback module integrates the judgment and discrimination results of whether partial discharge exists, the discharge intensity level, and the suspected defect type into detection status information, and then feeds it back to the preset receiving end.

[0111] During the operation of the feedback module, dynamic adjustment commands are synchronously output to the excitation module, conditioning module, and sensing module based on the detection status information:

[0112] Send a frequency fine-tuning command to the excitation module:

[0113] When the detection status information indicates the presence of partial discharge and the suspected defect type is identified, the preset feature frequency set corresponding to that suspected defect type is extracted. k is the number of characteristic frequencies, and the following adjustments are performed;

[0114] Will Revised to , Indicates the frequency correction factor. Represents the maximum cosine similarity;

[0115] The above formula is based on the initial optimal excitation frequency, combined with the maximum cosine similarity (the similarity between the fused features and the best-matching standard defect features) obtained by the feedback module, and introduces a frequency correction coefficient. A modified model is constructed, where α increases with the increase of maximum similarity and decreases as it approaches a preset threshold. Ultimately, the initial optimal excitation frequency is modified to a more targeted excitation frequency. This design ensures that: when the defect type matching degree is higher, the fine-tuned excitation frequency can more accurately excite the partial discharge signal of that type of defect. The value constraints ensure that the fine-tuning range is reasonable, avoid over-adjustment that leads to signal distortion, and further improve the signal acquisition quality of subsequent detection.

[0116] Send a filter bandwidth adjustment command to the conditioning module:

[0117] The adaptive filter unit is controlled to set the center frequency of the filter passband to... The mean value of the filtering bandwidth is reduced to x times the original bandwidth, where x is the bandwidth compression factor.

[0118] Send a sampling rate adjustment command to the sensing module:

[0119] The sampling rate of the UHF sensor array and the ultrasonic sensor array is increased to y times the original sampling rate, where y is the sampling rate amplification factor, and the sampling time is extended to a preset time.

[0120] Among them, frequency correction coefficient ∈[0.05,0.2], and The closer the value is to 1, the larger the value becomes. When the maximum cosine similarity is closer to the preset similarity threshold, the smaller the value becomes. x∈ (0,1), when the defect feature frequency is concentrated and the background interference is small, the value of x becomes larger. When the defect feature frequency is dispersed or the background interference is large, the value of x becomes smaller. The sampling rate amplification coefficient y∈[1.2,3.0], when the frequency of the discharge signal corresponding to the suspected defect type is higher and the duration of the signal is shorter, the value of y becomes larger. When the frequency of the discharge signal corresponding to the suspected defect type is lower and the duration of the signal is longer, the value of y becomes smaller.

[0121] The incentive module is interconnected with the sensing module and the conditioning module via a wireless network. The conditioning module is interconnected with the collaborative analysis module via a wireless network. The collaborative analysis module is interconnected with the storage module and the feedback module via a wireless network.

[0122] In this embodiment, the excitation module transmits a piezoelectric excitation signal of a specific frequency to a preset detection area within the GIS equipment to trigger the acoustic-electric signal radiation of potential partial discharge in that area. The sensing module then receives the ultra-high frequency electromagnetic signal and ultrasonic vibration signal radiated by partial discharge within the GIS equipment. The conditioning module then receives the ultra-high frequency signal and ultrasonic signal output from the sensing module and performs filtering, amplification, and impedance matching processing to output a standardized electrical signal. The collaborative analysis module further performs time-domain correlation matching and frequency-domain feature fusion calculation on the standardized ultra-high frequency signal and ultrasonic signal output from the conditioning module. The storage module classifies and stores the feature fusion calculation results data output by the analysis module and the original standardized signal to establish a data index association table. Finally, the feedback module matches the feature fusion calculation results data with the preset GIS partial discharge feature threshold and defect type feature library to output the detection status information of whether the equipment has partial discharge, the discharge intensity, and the suspected defect type, and feeds it back to the preset receiving end.

[0123] In the above embodiments, the system can calculate the optimal excitation frequency to trigger potential discharge signals by combining the equipment cavity parameters and medium characteristics in the GIS equipment inspection scenario. It can also simultaneously collect and optimize relevant electromagnetic and vibration signals, accurately determine whether discharge exists, the intensity of discharge, and the type of suspected defects, classify and store the data, and dynamically adjust the detection parameters to improve subsequent accuracy. This effectively ensures the safe and stable operation of GIS equipment and reduces the risk of failure.

[0124] Regarding the system in the above embodiments, the following is an application example of the system:

[0125] A substation selected this system for partial discharge testing of its 110kV gas-insulated switchgear (GIS). The test area was the circuit breaker cavity. The specific testing procedure is as follows:

[0126] In the initial stage of testing, the system excitation module first obtains the core parameters of the circuit breaker cavity: the equivalent radius of the cavity is 0.3m, the relative permittivity of the insulating medium (SF6 gas) inside the cavity is 1.002, and the sound velocity in the medium is 340m / s. Because previous testing revealed that the cavity's dielectric loss is low, and the roundness deviation and inner wall roughness are within acceptable ranges, a correction factor of 0.8 is used. Based on these parameters, the optimal excitation frequency is calculated to be 50kHz. The excitation module then transmits a 50kHz piezoelectric excitation signal to the testing area, triggering the radiation of acoustic and electrical signals from potential partial discharges within the area.

[0127] Subsequently, the system's sensing module starts operating. This module's UHF sensor array consists of three UHF sensors, evenly distributed circumferentially along the circuit breaker cavity. Each UHF sensor's signal acquisition direction is pointed towards the cavity center. Furthermore, an ultrasonic sensor is placed adjacent to each UHF sensor radially along the GIS equipment. The vibration acquisition surfaces of all ultrasonic sensors are in close contact with the outer wall of the cavity. The sensing module outputs the same trigger signal to both types of sensors. The two types of sensors synchronously acquire signals from 10:00:00 to 10:00:05, ensuring that the UHF electromagnetic signal and ultrasonic vibration signal acquisition times are consistent.

[0128] After acquisition, the signal is transmitted to the conditioning module. The conditioning module first adjusts the filter bandwidth to filter out interference frequency components below 20kHz and above 100kHz that are unrelated to the partial discharge signal; then, the signal is amplified, with a preset reference gain of 20dB and a preset reference voltage of 5V. Calculations show that the peak voltage of the UHF signal after filtering is 2V and the peak voltage of the ultrasonic signal is 1.5V, thus completing the amplification of the two types of signals. At the same time, the adjustable impedance matching unit inside the module adjusts the combination of variable capacitor and variable inductor to keep the input impedance of the conditioning module consistent with the output impedance of the sensor, ultimately outputting standardized UHF and ultrasonic signals.

[0129] Next, the collaborative analysis module processes the standardized signal. In the time-domain correlation matching stage, after extracting the standardized UHF signal and the ultrasound signal, the cross-correlation coefficient under different time delays is calculated. The time delay that maximizes the cross-correlation coefficient is found to be 0.002s. The ultrasound signal is then adjusted to match this time delay, achieving time-domain alignment of the two types of signals. In the frequency-domain feature fusion stage, the time-domain aligned signals are subjected to Fourier transform to obtain their respective frequency-domain features. Then, the signal-to-noise ratio (SNR) at each frequency point is calculated. For example, at a frequency of 30kHz, the SNR of the UHF signal is 25dB, and the SNR of the ultrasound signal is 18dB. The corresponding frequency-domain feature weighting coefficients for the UHF signal and the ultrasound signal at that frequency point are calculated to be 0.58 and 0.42, respectively. The fused frequency-domain feature is obtained by weighted summation. This process is repeated for all frequency points to complete the frequency-domain feature fusion.

[0130] The storage module then categorizes and stores the data according to a three-level classification rule: the first level uses the detection timestamp from 10:00:00 to 10:00:05 as the classification identifier; the second level uses the circuit breaker cavity number "A01" under the same timestamp as the classification identifier; and the third level uses UHF standardized signal, ultrasonic standardized signal, and feature fusion result as classification identifiers under the same area identifier. At the same time, a data index association table is established, which includes fields such as detection timestamp, detection area number "A01", various signal type identifiers, data storage path, data size 20MB, and data generation time 10:00:06.

[0131] Finally, the feedback module performs result judgment and command issuance. For partial discharge and intensity judgment, the extracted fusion frequency domain features have a mean value of 120mV. The preset GIS partial discharge feature threshold is 80mV. Because the mean value exceeds the threshold, partial discharge is determined to exist, with a discharge intensity level of 1.5 (obtained from the ratio of 120mV to 80mV). For defect type discrimination, the feature fusion result is converted into a feature vector, and the cosine similarity is calculated with the feature vectors of each standard defect in the preset defect type feature library (including corona discharge, surface discharge, and air gap discharge). The similarity with the standard feature vector of air gap discharge is 0.92 (exceeding the preset similarity threshold of 0.7), thus the suspected defect type is determined to be air gap discharge. The feedback module sends the detection status information of "partial discharge present, discharge intensity level 1.5, suspected air gap discharge" to the substation monitoring center. Simultaneously, it issues dynamic adjustment commands: a frequency fine-tuning command is sent to the excitation module, adjusting the original optimal excitation frequency of 50kHz to 58.88kHz due to a maximum cosine similarity of 0.92 and a frequency correction coefficient of 0.18; a filter bandwidth adjustment command is sent to the conditioning module, setting the filter passband center frequency to 58.88kHz and reducing the bandwidth to 0.6 times the original bandwidth (due to concentrated defect feature frequencies and low background interference); and a sampling rate adjustment command is sent to the sensing module, increasing the sampling rate to 2.5 times the original sampling rate (due to the high frequency and short duration of the air gap discharge signal) and extending the sampling duration to 10 seconds to support subsequent accurate detection.

[0132] In summary, during operation, the system in the above embodiments calculates the optimal excitation frequency by combining the cavity parameters of the GIS equipment detection area with the characteristics of the insulating medium, accurately triggering the acoustic and electrical signal radiation of potential partial discharge. It simultaneously acquires ultra-high frequency electromagnetic signals and ultrasonic vibration signals through a multi-sensor array, ensuring the temporal consistency and directional comprehensiveness of signal acquisition. Targeted filtering, amplification, and impedance matching processes remove irrelevant interference, enhance signal strength, and ensure standardization. Time-domain correlation matching achieves precise alignment of acoustic and electrical signals, and frequency-domain signal-to-noise ratio weighted fusion enhances feature extraction accuracy. Data is categorized and stored according to detection time, region, and signal type, and index priority is dynamically adjusted for convenient and efficient management and traceability. The system can accurately determine whether partial discharge exists in the equipment, the discharge intensity level, and the type of suspected defects. Furthermore, it can dynamically optimize the excitation frequency, filtering bandwidth, and sampling rate based on the detection results, continuously improving detection accuracy and efficiency, and providing reliable and effective protection for the safe operation and defect early warning of GIS equipment.

[0133] 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 the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions will not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A piezoelectrically energized GIS ultra-high frequency sensor ultrasonic cooperative detection system, characterized in that, The method comprises the following steps: An excitation module is used to transmit a piezoelectric excitation signal of a specific frequency to a preset detection area in a GIS device to trigger the sound and electricity signal radiation of potential partial discharge in the area; A sensing module is used to receive the ultra-high frequency electromagnetic signal and ultrasonic vibration signal radiated by the partial discharge in the GIS device; A conditioning module is used to filter, amplify and impedance match the ultra-high frequency signal and ultrasonic signal output by the sensing module to output standardized electrical signals; A collaborative analysis module is used to perform time-domain correlation matching and frequency-domain feature fusion calculation on the standardized ultra-high frequency signal and ultrasonic signal output by the conditioning module; A storage module is used to store the feature fusion calculation result data and the original standardized signal output by the analysis module to establish a data index association table; A feedback module is used to match the feature fusion calculation result data with the preset GIS partial discharge feature threshold and defect type feature library, output the detection state information of whether the device has partial discharge, discharge intensity and suspected defect type, and feed back to the preset receiving end.

2. The piezoelectrically energized GIS UHF sensor ultrasonic co-detection system of claim 1, wherein, In the running stage of the excitation module, the equivalent radius of the cavity corresponding to the preset detection area of the GIS device, the relative dielectric constant of the insulating medium in the preset detection area and the sound speed in the insulating medium are first obtained, and then the optimal excitation frequency is calculated based on the above parameters, and the optimal excitation frequency is recorded as the specific frequency applied in the excitation module: ； In the formula: is the optimal excitation frequency; is the correction coefficient, > 0; is the sound speed in the insulating medium in the preset detection area of the GIS device; is the equivalent radius of the cavity corresponding to the preset detection area of the GIS device; is the relative dielectric constant of the insulating medium in the preset detection area of the GIS device.

3. The piezoelectrically energized GIS UHF sensor ultrasonic co-detection system of claim 1, wherein, The sensing module comprises an ultra-high frequency sensor array and an ultrasonic sensor array; The ultra-high frequency sensor array is composed of at least three ultra-high frequency sensors, each of which is uniformly distributed along the circumference of the cavity of the preset detection area of the GIS device, and the signal collection direction of each ultra-high frequency sensor is directed to the center position of the detection area, and one ultrasonic sensor is arranged adjacent to each ultra-high frequency sensor along the radial position of the GIS device, and the vibration collection surface of the ultrasonic sensor is closely attached to the outer wall of the cavity of the GIS device; In the running stage of the sensing module, the same trigger signal is output to the ultra-high frequency sensor array and the ultrasonic sensor array, so that the two types of sensors start collecting signals at the same time, so that the collection time of the ultra-high frequency electromagnetic signal and the ultrasonic vibration signal radiated by the partial discharge is synchronized.

4. The piezoelectrically energized GIS UHF sensor ultrasonic co-detection system of claim 1, wherein, In the running stage of the conditioning module, the filter bandwidth is adjusted to filter out the interference frequency components irrelevant to the partial discharge signal to complete the filtering operation; Based on the amplification gain again The filtered ultra-high frequency signal and the ultrasonic signal are amplified respectively, represents a preset reference gain, represents a preset reference voltage, represents a peak voltage of the filtered signal. An adjustable impedance matching unit is arranged in the conditioning module, which is integrated by a variable capacitor and a variable inductor, and the adjustable impedance matching unit adjusts the equivalent impedance of the impedance matching network in real time through the combination of the variable capacitor and the variable inductor, so that the input impedance of the conditioning module is consistent with the output impedance of the sensor in the sensing module.

5. The piezoelectrically energized GIS UHF sensor ultrasonic co-detection system of claim 1, wherein, The time-domain correlation matching process in the collaborative analysis module is as follows: The standardized ultra-high frequency signal outputted by the extraction conditioning module is denoted as , and the standardized ultrasonic signal is denoted as , wherein t is time. The cross-correlation coefficient of the two signals under different time delays is calculated: ； In the formula: ∈ [-1, 1]; respectively the start time and the end time of signal acquisition; Wherein, the cross-correlation coefficient In the middle Indicates the time delay variable, based on the above formula to get the event delay, recorded as The maximum value, recorded as The ultrasonic signal Adjust to Make the ultra-high frequency signal and ultrasonic signal in time domain on the alignment.

6. The piezoelectrically energized GIS UHF sensor ultrasonic co-detection system of claim 1, wherein, The frequency-domain feature fusion calculation process in the collaborative analysis module is as follows: The normalized ultra-high frequency signal and the normalized ultrasonic signal after time domain correlation matching are respectively subjected to Fourier transform to obtain frequency domain features of the ultra-high frequency signal, denoted as , and frequency domain features of the ultrasonic signal , wherein is an angular frequency; Calculate each frequency point The corresponding VHF signal signal-to-noise ratio and the ultrasonic signal signal-to-noise ratio, that is, the ratio of the amplitude of the signal at the frequency point to the noise amplitude; Calculate each frequency point according to the signal-to-noise ratio Corresponding weight coefficient: , respectively represent the frequency points of the corresponding very high frequency signal signal-to-noise ratio, frequency points of the corresponding ultrasonic signal signal-to-noise ratio, respectively are weight coefficients of the very high frequency signal frequency domain features, weight coefficients of the ultrasonic signal frequency domain features; Finally, the fused frequency domain features are calculated : 。 7. The piezoelectrically energized GIS UHF sensor ultrasonic co-detection system of claim 1, wherein, In the running stage of the storage module, the feature fusion calculation result data output by the collaborative analysis module and the original standardized signal output by the conditioning module are classified into three levels according to the detection time, detection area and signal type: The first level takes the detection timestamp as the classification identifier, the second level takes the preset detection area number of the GIS device under each time identifier as the classification identifier, and the third level takes the ultra-high frequency standardized signal, the ultrasonic standardized signal and the feature fusion result under each area identifier as the classification identifier; A data index association table is established based on the classification result, and fields of the data index association table include: detection timestamp, detection area number, signal type identifier, data storage path, data size, and data generation time; The data index association table dynamically adjusts the index priority according to the detection state information output by the feedback module: For the detection data determined to have partial discharge, the index priority is raised to a preset high priority interval; For the detection data determined to have no partial discharge, the index priority is set to a preset low priority interval.

8. The piezoelectrically energized GIS UHF sensor ultrasonic co-detection system of claim 1, wherein, The logic of the feedback module for outputting the detection state information in the running stage is as follows: Determine whether there is partial discharge and the discharge intensity: Extract the mean value of the fusion frequency domain feature in the feature fusion calculation result output by the collaborative analysis module, and compare the mean value with the preset GIS partial discharge feature threshold value; If the mean value does not exceed the threshold value, it is determined that there is no partial discharge; If the mean value exceeds the threshold value, it is determined that there is partial discharge, and the discharge intensity level is represented by the ratio of the mean value to the threshold value; Determine the suspected defect type: Convert the feature fusion calculation result into a feature vector , The value of the fusion frequency domain feature at n key frequency points is represented, and the standard feature vector corresponding to each standard defect type is extracted from a pre-set defect type feature library , m is the defect type number, and the cosine similarity of the two vectors is calculated again: ； wherein, ∈ [0, 1], the value is closer to 1, indicating that the feature matching degree is higher, and the standard defect type with the maximum similarity and greater than the preset similarity threshold is taken as the suspected defect type; The feedback module finally integrates the determination of whether there is partial discharge, the discharge intensity level, the suspected defect type and the determination result into the detection state information, and feeds back to the preset receiving end.

9. The piezoelectrically energized GIS UHF sensor ultrasonic co-detection system of claim 1, wherein, In the running stage of the feedback module, dynamic adjustment instructions are output to the excitation module, the conditioning module and the perception module according to the detection state information: Send frequency fine-tuning instructions to the excitation module: When the detection state information is that there is partial discharge and the suspected defect type has been determined, a preset characteristic frequency set corresponding to the suspected defect type is extracted , k is the number of characteristic frequencies, and the following adjustment is performed; will be described. amended to , denotes a frequency correction coefficient, denotes a maximum cosine similarity; Send filter bandwidth adjustment instructions to the conditioning module: The control adaptive filter unit sets the filter passband center frequency to the mean value of and the filter bandwidth is reduced to x times of the original bandwidth, x being a bandwidth compression factor. Send sampling rate adjustment instructions to the perception module: Control the ultra-high frequency sensor array and the ultrasonic sensor array to increase the sampling rate to y times of the original sampling rate, and the sampling time length is extended to a preset time length, where y is the sampling rate amplification coefficient.

10. The piezoelectrically energized GIS UHF sensor ultrasonic co-detection system of claim 1, wherein, The excitation module is connected with the perception module and the conditioning module through wireless network interaction, the conditioning module is connected with the collaborative analysis module through wireless network interaction, and the collaborative analysis module is connected with the storage module and the feedback module through wireless network interaction.

Images