Integrated high-frequency partial discharge signal primary and secondary fusion ring network box

Abstract

The application relates to a primary and secondary fusion ring network box integrating high-frequency partial discharge signals and belongs to the technical field of ring network boxes. The ring network box comprises an acquisition module, a separation module and a mapping module. The acquisition module is used for acquiring high-frequency electromagnetic signals and synchronous power frequency phases of multiple monitoring nodes in the primary and secondary fusion ring network box. The separation module is used for separating the high-frequency electromagnetic signals based on spatial time difference correlation to obtain a purified partial discharge pulse sequence. The mapping module is used for mapping the purified partial discharge pulse sequence and the synchronous power frequency phases in combination with internal topological structure parameters of the primary and secondary fusion ring network box to calculate an insulation defect state matrix. A driving module is used for generating corresponding secondary linkage control instructions according to the insulation defect state matrix, driving an executing mechanism of the primary and secondary fusion ring network box based on the secondary linkage control instructions to carry out fault isolation or early warning output, and solving the technical problem that the prior art cannot effectively distinguish real insulation abnormalities from interference signals, leading to insufficient insulation defect state judgment accuracy.

Description

Technical Field

[0001] This invention relates to the field of ring main unit technology, and in particular to a primary and secondary fusion ring main unit that integrates high-frequency partial discharge signals. Background Technology

[0002] Existing integrated primary and secondary ring main units typically have status sensors, partial discharge monitoring units, and power distribution automation terminals installed inside the unit. These sensors collect electrical quantities, environmental quantities, and insulation-related signals, which are then analyzed by a local controller or communication terminal. Based on preset thresholds, feature rules, or historical samples, the system determines whether there are insulation abnormalities in the equipment. When the monitoring results exceed the set conditions, an alarm message is sent to the operation and maintenance platform, or the switching equipment is controlled to perform corresponding operating operations.

[0003] Because ring main units simultaneously house switches, transformers, cable joints, and communication control equipment, equipment switching, load changes, and external electromagnetic environment can easily generate transient interference in the monitoring signals. Existing judgment methods primarily rely on signal amplitude or single characteristics for identification, making it difficult to effectively distinguish between genuine insulation anomalies and interference signals. This results in insufficient accuracy in determining the insulation defect status and may lead to false alarms or untimely anomaly handling. Therefore, it is necessary to propose a control method that can improve the accuracy of insulation defect identification in ring main units. Summary of the Invention

[0004] The purpose of this invention is to at least partially solve one of the technical problems existing in the prior art.

[0005] To achieve the above objectives, the present invention provides a primary and secondary fusion ring network box integrating high-frequency partial discharge signals, comprising: The acquisition module is used to acquire the high-frequency electromagnetic signals and synchronous power frequency phase of multiple monitoring nodes in the primary and secondary fusion ring network box; The separation module is used to separate the high-frequency electromagnetic signal based on spatial time difference correlation to obtain a purified partial discharge pulse sequence; The mapping module is used to combine the internal topology parameters of the primary and secondary fusion ring network box to map the purified partial discharge pulse sequence with the synchronous power frequency phase and calculate the insulation defect state matrix. The drive module is used to generate corresponding secondary linkage control commands based on the insulation defect state matrix, and drive the actuator of the primary and secondary fusion ring network box to perform fault isolation or early warning output based on the secondary linkage control commands.

[0006] Furthermore, the high-frequency electromagnetic signals and synchronous power frequency phase of multiple monitoring nodes within the primary and secondary fusion ring network box are acquired, including: The acquisition module is used to synchronously acquire the high-frequency electromagnetic signals at the microsecond level through an array of ultra-high frequency sensors arranged in different air chambers of the primary and secondary fusion ring network box. The extraction module is used to extract the fundamental zero-crossing point of the output voltage of the secondary voltage transformer of the primary and secondary integrated ring network box; The compensation module is used to perform phase compensation on the fundamental zero-crossing point according to the preset high-frequency phase shift parameters of the transformer, and generate the synchronous power frequency phase that is time-aligned with the high-frequency electromagnetic signal.

[0007] Furthermore, the high-frequency electromagnetic signal is separated based on spatial time-difference correlation to obtain a purified partial discharge pulse sequence, including: The module is used to calculate the maximum reflection delay of electromagnetic waves in the cavity based on the maximum diagonal size of the metal cavity of the primary and secondary fusion ring network box, and to construct a multipath reflection suppression window. A shielding module is used to extract the initial arrival timestamp of the first peak in the high-frequency electromagnetic signal and shield subsequent reflected peaks within the multipath reflection suppression window. The calculation module is used to calculate the spatial time difference of arrival based on the initial arrival timestamp of each monitoring node after shielding, and to determine whether the signal source is located inside or outside the enclosure by the spatial time difference of arrival, thereby realizing the spatial time difference correlation separation. The determination module is used to eliminate signals determined to be external interference and to use the direct wave signals determined to be internal sources as the purified partial discharge pulse sequence.

[0008] Furthermore, combining the internal topology parameters of the primary and secondary fusion ring network box, the purified partial discharge pulse sequence is mapped to the synchronous power frequency phase to calculate the insulation defect state matrix, including: Analyze the internal topology parameters of the primary and secondary fusion ring network box, and extract the medium type and corresponding spatial attenuation coefficient between the source location of the purified partial discharge pulse sequence and each monitoring node; The apparent amplitude of the purified partial discharge pulse sequence is inversely compensated using the spatial attenuation coefficient to obtain the true discharge amplitude; The actual discharge amplitude and the occurrence time of the purified partial discharge pulse sequence are projected onto the corresponding phase angle interval of the synchronous power frequency phase to generate a phase-resolved two-dimensional partial discharge spectrum. Extract the discharge characteristic parameters of the phase-resolved partial discharge two-dimensional spectrum and match the defect type; The insulation defect state matrix is ​​calculated by combining the defect type, source location coordinates, and actual discharge amplitude.

[0009] Furthermore, the apparent amplitude of the purified partial discharge pulse sequence is inversely compensated using the spatial attenuation coefficient to obtain the true discharge amplitude, including: The apparent amplitude of the purified partial discharge pulse sequence is decomposed into multiple frequency bands of apparent spectral components by performing a fast Fourier transform. Based on the medium type, extract the number of medium interfaces and the medium wave impedance along the propagation path from the source location of the purified partial discharge pulse sequence to each monitoring node. Based on the number of medium interfaces and the medium wave impedance, calculate the frequency-varying transmission coefficient of electromagnetic waves crossing the interface for each frequency band. The spatial attenuation coefficient is converted into the frequency-varying attenuation factor of the corresponding frequency band. The frequency-varying transmission coefficient and the frequency-varying attenuation factor are used to perform frequency-domain segmented inverse compensation on the apparent spectral component to obtain the compensated spectral component. The time-domain signal is reconstructed by performing an inverse fast Fourier transform on the compensated spectral components, and the peak value of the reconstructed time-domain signal is extracted as the true discharge amplitude.

[0010] Furthermore, the actual discharge amplitude and the occurrence time of the purified partial discharge pulse sequence are projected onto the corresponding phase angle interval of the synchronous power frequency phase to generate a phase-resolved two-dimensional partial discharge spectrum, including: Based on the medium type, the electromagnetic wave relative permittivity of the corresponding medium is extracted. Combined with the physical length of the propagation path of each monitoring node, the cross-medium propagation time delay of the purified partial discharge pulse sequence from the source position to each monitoring node is calculated. The distortion compensation phase angle is calculated by multiplying the cross-medium propagation time delay by the power frequency angular frequency. The phase shift correction of the synchronous power frequency phase corresponding to the occurrence time of the purified partial discharge pulse sequence is performed using the distortion compensation phase angle to obtain the corrected power frequency phase; The actual discharge amplitude and the occurrence time of the purified partial discharge pulse sequence are projected onto the corresponding phase angle interval of the corrected power frequency phase to generate the phase-resolved partial discharge two-dimensional spectrum.

[0011] Furthermore, the discharge characteristic parameters of the phase-resolved partial discharge two-dimensional spectrum are extracted and the defect type is matched, including: The phase-resolved partial discharge two-dimensional spectrum is divided into equal phase angle windows, and the pulse cluster contour gradient and density distribution entropy in each phase angle window are extracted and combined to generate a multidimensional discharge feature parameter tensor. The multidimensional discharge feature parameter tensor is used as the discharge feature parameter. Obtain a pre-constructed topological defect fingerprint database containing electric field distortion features at multi-medium interface points; Calculate the multidimensional spatial distance between the multidimensional discharge feature parameter tensor and each candidate defect type in the topological defect fingerprint database, and output the candidate defect type corresponding to the minimum multidimensional spatial distance to achieve the matching defect type.

[0012] Furthermore, based on the number of dielectric interfaces and the dielectric wave impedance, the frequency-varying transmission coefficient of electromagnetic waves crossing the interfaces in each frequency band is calculated, including: Extract the surface roughness parameters of the medium interface, and combine them with the center frequency of each frequency band to calculate the interface scattering loss factor of electromagnetic waves in the corresponding frequency band. Calculate the fundamental transmission coefficient of the ideal plane based on the dielectric wave impedance of the adjacent medium; Multiply the basic transmission coefficient by the interface scattering loss factor, and perform a multiplication attenuation calculation according to the number of medium interfaces to obtain the frequency-varying transmission coefficient of electromagnetic waves in each frequency band when crossing the interface.

[0013] Furthermore, based on the insulation defect state matrix, corresponding secondary linkage control commands are generated. These commands then drive the actuators of the primary and secondary fusion ring network box to perform fault isolation or early warning output, including: Extract the discharge amplitude growth rate and discharge frequency density of the target defect node in the insulation defect state matrix; Based on the discharge amplitude growth rate and discharge frequency density, the margin time for the target defect node to reach the critical state of insulation breakdown is calculated. When the margin time is less than the preset emergency trip threshold, a hard isolation command is generated as the secondary linkage control command to drive the circuit breaker in the actuator to disconnect the corresponding fault branch and perform the fault isolation. When the margin time is greater than or equal to the emergency trip threshold and less than the alarm threshold, a flexible scheduling instruction is generated as the secondary linkage control instruction, and the early warning output is sent to the main station through the secondary communication module.

[0014] Furthermore, based on the discharge amplitude growth rate and discharge frequency density, the margin time for the target defect node to reach the critical state of insulation breakdown is calculated, including: The dynamic degradation acceleration factor of the target defect node is calculated by nonlinearly exponentially coupling the discharge amplitude growth rate with the discharge frequency density. Extract the current cumulative discharge energy of the target defect node and obtain the preset ultimate breakdown energy threshold of the corresponding insulating medium; Using the current accumulated discharge energy as the initial state, an energy evolution integral equation is constructed using the dynamic degradation acceleration factor. The upper limit of the integration time required to reach the ultimate breakdown energy threshold by solving the energy evolution integral equation is then used as the margin time to reach the critical state of insulation breakdown.

[0015] This invention provides a primary and secondary fusion ring network enclosure for integrating high-frequency partial discharge (PD) signals, comprising: an acquisition module for acquiring high-frequency electromagnetic signals and synchronous power frequency phases of multiple monitoring nodes within the primary and secondary fusion ring network enclosure; a separation module for separating the high-frequency electromagnetic signals based on spatial time difference correlation to obtain a purified PD pulse sequence; a mapping module for mapping the purified PD pulse sequence to the synchronous power frequency phase using internal topology parameters of the primary and secondary fusion ring network enclosure to calculate an insulation defect state matrix; and a driving module for generating corresponding secondary linkage control commands based on the insulation defect state matrix and driving the system based on the secondary linkage control commands. The actuator of the primary and secondary integrated ring main unit performs fault isolation or early warning output, solving the technical problem that existing technologies cannot effectively distinguish between real insulation anomalies and interference signals, resulting in insufficient accuracy in determining the state of insulation defects. It realizes multi-node collaborative acquisition of high-frequency partial discharge signals within the ring main unit, spatial time difference correlation separation, and power frequency phase correlation characterization. It can distinguish between real partial discharge pulses and interference signals in complex electromagnetic environments, and determine the associated location and state characteristics of insulation defects by combining internal topology parameters. Furthermore, it generates fault isolation or early warning commands based on the state of insulation defects, improving the accuracy of insulation anomaly determination, the timeliness of control response, and the operational safety of the ring main unit. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a schematic diagram of the primary and secondary fusion ring network box integrating high-frequency partial discharge signals in one embodiment of the present invention; Figure 2 This is one embodiment of the present invention. Figure 1 A schematic diagram of the module structure; Figure 3 This is one embodiment of the present invention. Figure 1 A schematic diagram of the separation module.

[0018] The objectives, features, and advantages of this invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0019] The embodiments of the present invention are described in detail below. Examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention. The step numbers in the following embodiments are set only for ease of explanation, and there is no limitation on the order between the steps. The execution order of each step in the embodiments can be adaptively adjusted according to the understanding of those skilled in the art.

[0020] The following describes in detail, with reference to the accompanying drawings, an integrated primary and secondary fusion ring network box for high-frequency partial discharge signals according to an embodiment of the present invention.

[0021] like Figures 1-3 According to one embodiment of the present invention, a primary and secondary fusion ring network box integrating high-frequency partial discharge signals includes: Module 1 is used to acquire the high-frequency electromagnetic signals and synchronous power frequency phase of multiple monitoring nodes within the primary and secondary fusion ring network box.

[0022] The acquisition module 11 is used to synchronously acquire the high-frequency electromagnetic signals at the microsecond level through an array of ultra-high frequency sensors arranged in different air chambers of the primary and secondary fusion ring network box. Extraction module 12 is used to extract the fundamental zero-crossing point of the output voltage of the secondary voltage transformer of the primary and secondary integrated ring network box; The compensation module 13 is used to perform phase compensation on the fundamental zero-crossing point according to the preset high-frequency phase shift parameters of the transformer, and generate the synchronous power frequency phase that is time-aligned with the high-frequency electromagnetic signal.

[0023] Specifically, an array of ultra-high frequency (UHF) sensors is first arranged in different chambers of the primary and secondary fusion ring network enclosure. These chambers include switch chambers, busbar chambers, and cable chambers, with each UHF sensor connected to a high-speed acquisition unit. The acquisition unit provides each sensor with the same clock reference and a unified trigger signal. Upon triggering, the pre-trigger buffer and real-time sampling channel are simultaneously activated to continuously record the high-frequency electromagnetic pulse waveforms in each chamber. Each sampled data segment is appended with the sensor number, chamber number, and acquisition timestamp. The timestamp resolution is set to microseconds, and the sampling window covers the entire time period before and after the pulse arrival. For data with amplitudes exceeding the range or acquisition channel saturation, the occurrence time is retained, and the data quality status is marked for validity assessment during subsequent processing.

[0024] Subsequently, the voltage sampling sequence output from the secondary voltage transformer of the integrated primary and secondary ring main unit is read. First, DC bias correction and power frequency bandpass processing are applied to the sampling sequence to reduce the impact of high-frequency interference and random noise on the fundamental frequency determination. The fundamental frequency zero-crossing interval is identified according to the positive and negative changes in voltage values ​​between adjacent sampling points, and the actual zero-crossing time is estimated by the amplitude change trend between adjacent sampling points. When the voltage changes from negative to positive, the rising zero-crossing point is recorded; when it changes from positive to negative, the falling zero-crossing point is recorded. A period consistency check is performed on the zero-crossing points of multiple consecutive power frequency cycles, eliminating false zero-crossing points with excessively short durations or intervals significantly deviating from the power frequency cycle, thus forming a fundamental frequency zero-crossing point sequence arranged in ascending order of time.

[0025] Because the secondary voltage transformer and its signal conditioning circuit generate inherent phase shifts for different frequency components, the high-frequency phase shift parameters of the transformer in the power frequency measurement link are pre-calibrated using a standard signal source and written into the acquisition controller. When processing each fundamental zero-crossing point, the phase shift parameters are converted into the corresponding time offset based on the current power frequency cycle, and then the zero-crossing time is corrected in the opposite direction to the actual phase shift direction of the transformer; the transformer exhibits a delayed zero-crossing mark when lagging and a delayed zero-crossing mark when leading. When the power frequency cycle changes, the time conversion reference is re-determined based on the latest continuous cycle, so that the compensation amount is updated synchronously with the power frequency change. The compensated zero-crossing point is aligned with the UHF sensor array using the same clock timestamp, and a synchronous power frequency phase sequence containing the zero-crossing time, cycle number, and phase state is generated.

[0026] In this embodiment, the above scheme reduces the time deviation between the ultra-high frequency signals of different air chambers and the power frequency reference by unifying clock acquisition, accurately extracting the fundamental zero-crossing point, and compensating for the phase shift of the current transformer, thus providing a reliable phase reference for subsequent partial discharge pulse separation and insulation state mapping.

[0027] Separation module 2 is used to separate the high-frequency electromagnetic signal based on spatial time difference correlation to obtain a purified partial discharge pulse sequence.

[0028] Module 21 is used to calculate the maximum reflection delay of electromagnetic waves in the cavity based on the maximum diagonal size of the metal cavity of the primary and secondary fusion ring network box, and to construct a multipath reflection suppression window. The shielding module 22 is used to extract the initial arrival timestamp of the first peak in the high-frequency electromagnetic signal and shield subsequent reflected peaks within the multipath reflection suppression window. Calculation module 23 is used to calculate the spatial time difference of arrival based on the initial arrival timestamps of each monitoring node after shielding, and to determine whether the signal source is located inside or outside the enclosure through the spatial time difference of arrival, thereby realizing the spatial time difference correlation separation. The determination module 24 is used to remove signals determined to be external interference and to use the direct wave signals determined to be internal sources as the purified partial discharge pulse sequence.

[0029] Specifically, the external dimensions of the primary and secondary fusion ring network box, the boundaries of each gas chamber, and the installation positions of the UHF sensors are read to determine the maximum diagonal dimension of the metal cavity. The actual propagation speed of electromagnetic waves within the cavity is used as the time conversion benchmark. For the longest possible additional propagation path formed after electromagnetic waves are reflected by the metal wall, a conservative round-trip propagation path is used to estimate the upper limit of the reflection delay. Simultaneously, acquisition trigger jitter, sensor response differences, and timestamp quantization errors are taken into account. The continuous time range not less than the upper limit of the delay is discretized to form a multipath reflection suppression window consistent with the sampling time base of each monitoring node. For gas chambers filled with insulating gas or containing insulating partitions, the propagation speed is corrected based on the relative permittivity to ensure that the window boundary covers the possible reflection delay within the actual cavity.

[0030] Background noise is estimated for the high-frequency electromagnetic signals collected by each monitoring node. Based on the noise baseline and a preset trigger threshold, the earliest effective abrupt change segment of the waveform is searched, and the starting edge, peak position, and corresponding timestamp of the first peak are determined within this segment. After the first peak is determined, its initial arrival time is used as the initial arrival timestamp of that node. Within the subsequent multipath reflection suppression window, subsequent peaks appearing on the same channel are marked with a shielding flag and are no longer considered independent arrival events for spatial positioning. If the first peak experiences clipping or local saturation, the arrival time is corrected by combining the starting edge and adjacent unsaturated sampling points, while retaining the signal quality marker.

[0031] For the initial arrival timestamps of each monitoring node, node data belonging to the same discharge pulse event are merged according to their time proximity, and the spatial arrival time difference between nodes is calculated. Combining the sensor installation coordinates, chamber boundaries, and enclosure geometry, it is checked whether the time difference can be simultaneously explained by a direct propagation path from a certain location inside the enclosure. When the temporal sequence and time difference of multiple nodes are both within the allowable propagation distance and measurement error range, the event is determined to be an internal source within the enclosure. When the time difference exhibits timing characteristics of propagation from outside the enclosure, or when no effective source location satisfying the constraints of each node can be found inside the enclosure, the event is determined to be external interference.

[0032] For events identified as external interference, their corresponding waveform segments and event markers at each monitoring node are removed. For events identified as internal sources, only the first direct peak and its preset valid waveform segment are retained and written to the event cache with a unified timestamp. The monitoring node, arrival time, peak amplitude, and polarity information are also recorded. After event-level deduplication and time-sequencing, a purified partial discharge pulse sequence is formed.

[0033] In this embodiment, the above scheme establishes a reflection suppression boundary based on the cavity size, and combines the arrival time difference of the direct wave at multiple nodes to distinguish between the internal discharge power supply and external interference, thereby reducing the impact of reflected wave peaks and external electromagnetic disturbances on partial discharge identification.

[0034] The mapping module 3 is used to combine the internal topology parameters of the primary and secondary fusion ring network box to map the purified partial discharge pulse sequence with the synchronous power frequency phase, and calculate the insulation defect state matrix.

[0035] The internal topology parameters of the primary and secondary fusion ring network box are analyzed, and the medium type and corresponding spatial attenuation coefficient between the source location of the purified partial discharge pulse sequence and each monitoring node are extracted.

[0036] Specifically, the gas chamber connection relationship, partition arrangement, busbar and cable channel parameters of the primary and secondary fusion ring network box are read. Starting from the source position determined by the purified partial discharge pulse sequence, the propagation path is traced to the corresponding monitoring node, and the types of insulating gas, solid insulating components, metal partitions, and cable media passed through along the path are recorded. Then, based on the medium material, signal frequency band, and boundary structure, the spatial attenuation coefficient corresponding to each medium is extracted from a pre-calibrated attenuation parameter table. The coefficient represents the degree of attenuation of the signal amplitude per unit propagation distance; the larger the value, the more obvious the attenuation. It is associated with each path segment and stored. In this embodiment, the above scheme establishes the medium propagation information from the source position to the monitoring node, providing a basis for subsequent partial discharge signal amplitude correction and insulation status judgment.

[0037] The apparent amplitude of the purified partial discharge pulse sequence is inversely compensated using the spatial attenuation coefficient to obtain the true discharge amplitude.

[0038] The apparent amplitude of the purified partial discharge pulse sequence is decomposed into multiple frequency bands of apparent spectral components by performing a fast Fourier transform. Based on the medium type, the number of medium interfaces and the medium impedance along the propagation path from the source location of the purified partial discharge pulse sequence to each monitoring node are extracted.

[0039] Specifically, for the purified partial discharge pulse sequence received by each monitoring node, an effective time segment containing the first direct peak is extracted according to a uniform sampling interval. The sampling baseline is removed first, and then a smoothing window is applied to reduce spectral leakage caused by truncation. Subsequently, a Fast Fourier Transform is performed on the corrected apparent amplitude sequence to decompose the time-domain pulse into multiple consecutively arranged apparent spectral components. Amplitude correction is performed based on the number of sampling points, window function gain, and acquisition channel calibration coefficients to form a spectral record with frequency range, center frequency, and component amplitude.

[0040] Based on this, the medium sequence is read according to the propagation path from the source location to each monitoring node. The type change between the insulating gas and the solid insulating component, metal partition, or cable medium is determined as the medium interface, and the number of interfaces on each path is counted. Then, based on the dielectric properties, magnetic permeability properties, and calibration data of each medium in the corresponding frequency band, the corresponding medium wave impedance is extracted and associated with the interface location one by one; the greater the difference in medium wave impedance, the more obvious the reflection tendency at the interface. In this embodiment, the above scheme obtains the frequency division characterization of the partial discharge pulse and the path medium boundary parameters, providing a basis for subsequent reflection effect correction and insulation defect state calculation.

[0041] Based on the number of medium interfaces and the medium wave impedance, the frequency-varying transmission coefficient of electromagnetic waves in each frequency band when crossing the interface is calculated.

[0042] Extract the surface roughness parameters of the medium interface, and combine them with the center frequency of each frequency band to calculate the interface scattering loss factor of electromagnetic waves in the corresponding frequency band. Calculate the fundamental transmission coefficient of the ideal plane based on the dielectric wave impedance of the adjacent medium; Multiply the basic transmission coefficient by the interface scattering loss factor, and perform a multiplication attenuation calculation according to the number of medium interfaces to obtain the frequency-varying transmission coefficient of electromagnetic waves in each frequency band when crossing the interface.

[0043] Specifically, the surface roughness parameters of each medium interface are read. These parameters can be obtained from interface processing and inspection data, assembly records, or a pre-calibrated material database, and are uniformly characterized using root mean square roughness. For each frequency band, the corresponding electromagnetic wave wavelength is determined based on its center frequency. Then, the relative relationship between surface roughness and this wavelength is compared to generate a dimensionless interface scattering loss factor. The higher the center frequency or the greater the relative surface roughness, the more obvious the interface scattering, and the smaller the loss factor. When the interface is approximately smooth, the loss factor is close to one.

[0044] Subsequently, based on the dielectric impedance of adjacent media in the corresponding frequency band, the base transmission coefficient is determined according to the ideal plane normal incidence condition, and is uniformly expressed using the electromagnetic wave amplitude transmission coefficient, with its value between zero and one. The greater the difference in dielectric impedance, the lower the base transmission coefficient under the ideal plane; when the dielectric impedances are similar, the base transmission coefficient increases accordingly.

[0045] For each frequency band, a cumulative transmission coefficient value is established, initially taking the lossless propagation state. Then, each medium interface along the path from the source position to the monitoring node is read sequentially. The current cumulative value is multiplied by the basic transmission coefficient and interface scattering loss factor of that interface, and this process is repeated according to the order of the interfaces in the propagation path until the attenuation calculation of all interfaces is completed. The result is used as the frequency-varying transmission coefficient of that frequency band. In this embodiment, the above scheme simultaneously considers the influence of medium impedance differences, surface roughness, and the number of interfaces on spectral propagation, improving the accuracy of partial discharge signal cross-medium transmission correction.

[0046] The spatial attenuation coefficient is converted into the frequency-varying attenuation factor of the corresponding frequency band. The frequency-varying transmission coefficient and the frequency-varying attenuation factor are used to perform frequency-domain segmented inverse compensation on the apparent spectral component to obtain the compensated spectral component. The time-domain signal is reconstructed by performing an inverse fast Fourier transform on the compensated spectral components, and the peak value of the reconstructed time-domain signal is extracted as the true discharge amplitude.

[0047] Specifically, the propagation distance from the purified partial discharge pulse source to each monitoring node and the spatial attenuation coefficient of each path segment are read. An attenuation parameter mapping relationship is established based on the center frequency of the corresponding frequency band, resulting in a frequency-varying attenuation factor with a value between zero and one. The longer the propagation distance, the higher the center frequency, or the stronger the medium attenuation, the smaller the factor. Subsequently, the frequency-varying transmission coefficient and the frequency-varying attenuation factor of each frequency band are used together as the amplitude retention degree of that frequency band, and reverse gain compensation is applied to the apparent spectral components. During the compensation process, the original phase information and the conjugate relationship between positive and negative frequency components are maintained, and a minimum retention coefficient and a maximum compensation gain are set to avoid amplifying background noise due to excessively small path parameters.

[0048] After compensation for each frequency band is completed, an inverse fast Fourier transform is performed according to the original sampling frequency and spectral arrangement order to reconstruct a time-domain signal consistent with the original sampling time reference. Baseline correction is performed on the reconstructed signal, searching for the sampling point with the largest absolute amplitude within the effective pulse window, and residual interference is eliminated by combining this with a noise threshold. This largest amplitude is taken as the true discharge amplitude of the corresponding partial discharge event, and its dimensions are consistent with the input apparent amplitude. In this embodiment, the above scheme eliminates propagation loss and interface transmission effects through segmented compensation in the frequency domain and restores the correction result to the time domain, improving the accuracy of partial discharge amplitude assessment.

[0049] The actual discharge amplitude and the occurrence time of the purified partial discharge pulse sequence are projected onto the corresponding phase angle interval of the synchronous power frequency phase to generate a phase-resolved two-dimensional partial discharge spectrum.

[0050] Based on the medium type, the electromagnetic wave relative permittivity of the corresponding medium is extracted. Combined with the physical length of the propagation path of each monitoring node, the cross-medium propagation time delay of the purified partial discharge pulse sequence from the source position to each monitoring node is calculated. The distortion compensation phase angle is calculated by multiplying the cross-medium propagation time delay by the power frequency angular frequency. The phase shift correction of the synchronous power frequency phase corresponding to the occurrence time of the purified partial discharge pulse sequence is performed using the distortion compensation phase angle to obtain the corrected power frequency phase; The actual discharge amplitude and the occurrence time of the purified partial discharge pulse sequence are projected onto the corresponding phase angle interval of the corrected power frequency phase to generate the phase-resolved partial discharge two-dimensional spectrum.

[0051] Specifically, the aforementioned medium path table is read, and for each propagation path segment between the source location and each monitoring node, the relative permittivity of the insulating gas, solid insulating components, and cable medium is extracted according to the corresponding frequency band. Structures such as metal partitions, which are not considered propagation media, are treated as reflection boundaries and are not included in the continuous propagation distance of the electromagnetic wave. Using the free-space electromagnetic wave propagation speed as a benchmark, the propagation speed is corrected according to the relative permittivity of each medium. The propagation time of each path segment is then determined based on its physical length, and accumulated segment by segment along the propagation direction to obtain the cross-medium propagation time delay of the purified partial discharge pulse from the source location to the corresponding monitoring node. When the relative permittivity of the medium is large or the path length increases, the corresponding propagation time delay increases accordingly. For media with frequency dispersion characteristics, the corresponding permittivity data is retrieved according to the center frequency of each frequency band.

[0052] Subsequently, using the power frequency angular frequency as the rate of change of the power frequency phase over time, the cross-medium propagation time delay is converted into a corresponding phase lag, yielding the distortion compensation phase angle. The power frequency angular frequency is measured in radians per second, the propagation time delay in seconds, and the resulting phase angle is expressed in radians. The longer the propagation delay or the higher the power frequency angular frequency, the greater the resulting phase lag. The phase lag is converted to a preset phase angle range for a single power frequency cycle, while retaining a marker spanning the entire cycle to avoid losing timing information after phase angle normalization.

[0053] For each purified partial discharge pulse event, the synchronous power frequency phase is interpolated based on its position between adjacent fundamental zero-crossing points to obtain the original power frequency phase corresponding to the event. Then, based on the propagation direction from the pulse source location to the current monitoring node, the phase lag caused by propagation is subtracted from the original power frequency phase to obtain the corrected power frequency phase equivalent to the time of the source location occurrence. When the propagation path causes the phase to exceed one power frequency cycle, it is wound back according to the cycle boundary while retaining the cycle number to maintain phase continuity and event sequence consistency.

[0054] After phase correction, the corrected power frequency phase is used as the horizontal axis of the two-dimensional spectrum, and the actual discharge amplitude is used as the vertical axis. Each partial discharge event is projected onto the corresponding phase angle interval and amplitude interval. The phase interval is divided according to the preset phase angle resolution, and the amplitude interval is divided according to the amplitude level. For each valid event received, the event count or amplitude weight is accumulated in the corresponding grid, ultimately forming a phase-resolved two-dimensional partial discharge spectrum. For duplicate records of the same discharge pulse formed at different monitoring nodes, they are merged based on the event time correlation results to avoid the same event being counted repeatedly. In this embodiment, the above scheme compensates for the time phase distortion caused by the partial discharge pulse propagating across different media, so that the actual discharge amplitude can accurately correspond to its source-end power frequency phase angle, providing a more stable spectrum basis for insulation defect type identification and condition assessment.

[0055] The discharge characteristic parameters of the phase-resolved partial discharge two-dimensional spectrum are extracted and the defect type is matched.

[0056] The phase-resolved partial discharge two-dimensional spectrum is divided into equal phase angle windows, and the pulse cluster contour gradient and density distribution entropy in each phase angle window are extracted and combined to generate a multidimensional discharge feature parameter tensor. The multidimensional discharge feature parameter tensor is used as the discharge feature parameter. Obtain a pre-constructed topological defect fingerprint database containing electric field distortion features at multi-medium interface points; Calculate the multidimensional spatial distance between the multidimensional discharge feature parameter tensor and each candidate defect type in the topological defect fingerprint database, and output the candidate defect type corresponding to the minimum multidimensional spatial distance to achieve the matching defect type.

[0057] Specifically, the phase-resolved partial discharge (PD) two-dimensional spectrum is divided into multiple phase angle windows of equal width along the power frequency phase angle direction, with each phase angle window using the same amplitude scale and pixel resolution. PD pulses falling into each phase angle window are clustered, forming pulse clusters based on pulse amplitude and phase angle coordinates. The gradient of the outer contour of the pulse cluster is extracted in the phase-amplitude plane. The contour gradient reflects the steepness of the pulse cluster boundary as the phase angle or amplitude changes; a larger gradient indicates concentrated discharge distribution or abrupt boundary changes. Simultaneously, the phase angle window is divided into several statistical units, and the dispersion of pulse occurrence frequency within each unit is calculated to form a density distribution entropy; the more dispersed the density and the higher the randomness, the larger the entropy value. The contour gradient, density distribution entropy, and corresponding amplitude statistics of each phase angle window are stacked in a fixed order, maintaining the sequential relationship between the phase angle windows, to obtain a multidimensional discharge feature parameter tensor, which serves as the discharge feature parameter for subsequent defect matching.

[0058] When establishing a topological defect fingerprint database in advance, electromagnetic field models with different dielectric combinations, geometric shapes, and defect states are constructed for typical locations within the primary and secondary integrated ring network enclosure, such as gas-solid insulation interfaces, bushing connections, shielding structure edges, and cable terminals. Alternatively, phase-resolved partial discharge spectra are collected using calibrated test samples. The contour gradient, density distribution entropy, and amplitude statistics of each sample are extracted under the same phase window division rules, and identifiers such as defect location, defect type, dielectric combination, and applicable frequency band are added to form a topological defect fingerprint consistent with the tensor dimension of the multidimensional discharge characteristic parameters. For samples obtained with different operating voltages or sampling gains, amplitude benchmark unification and feature scale calibration are performed first to avoid fingerprint shifts caused by differences in operating conditions.

[0059] During matching, the input multidimensional discharge feature parameter tensor and each candidate defect fingerprint in the fingerprint database are standardized on the same scale. Then, corresponding weights are set according to the repeatability and defect sensitivity of each feature, and the comprehensive spatial distance between the two is calculated in dimensions such as contour morphology, density entropy value, phase angle distribution, and amplitude statistics. The smaller the distance, the closer the topological response of the input discharge feature is to that candidate defect. After comparing all candidates, the candidate defect type with the smallest multidimensional spatial distance is selected as the matching defect type, and the corresponding distance value and the feature dimensions involved in the matching are saved. In this embodiment, the above scheme uses the phase angle distribution morphology and topological electric field distortion fingerprint for joint comparison, improving the specificity and interpretability of ring network box insulation defect type identification.

[0060] The insulation defect state matrix is ​​calculated by combining the defect type, source location coordinates, and actual discharge amplitude.

[0061] Specifically, the matching defect type is first converted into a numerical identifier according to the coding rules of the topological defect fingerprint database, and the source location coordinates are unified to the three-dimensional equipment coordinate system of the primary and secondary fusion ring network box, with the coordinate unit remaining in meters; the actual discharge amplitude uses the compensated amplitude unit and is normalized. Each partial discharge event is treated as a row, with the defect type code, the three-dimensional coordinates of the source location, and the actual discharge amplitude written sequentially, arranged according to the event occurrence time to form an insulation defect state matrix, while retaining the unit and code correspondence of each field. In this embodiment, the above scheme uniformly represents the defect category, spatial location, and discharge intensity, providing a structured state basis for subsequent fault isolation and early warning decisions.

[0062] The drive module 4 is used to generate corresponding secondary linkage control commands based on the insulation defect state matrix, and drive the actuator of the primary and secondary fusion ring network box to perform fault isolation or early warning output based on the secondary linkage control commands.

[0063] Extract the discharge amplitude growth rate and discharge frequency density of the target defect node in the insulation defect state matrix.

[0064] Specifically, based on the defect type code and source location coordinates of the target defect node, historical records of the same node are filtered from the insulation defect state matrix and sorted by event occurrence time. A preset sliding time window is used to analyze the changing trend of the actual discharge amplitude. The discharge amplitude growth rate is determined by the amplitude difference or fitting slope of adjacent valid records, in amplitudes per second. Simultaneously, the number of valid discharge pulses within the time window is counted and divided by the window duration to obtain the discharge frequency density, in events per second. Significantly outlier records are removed, and results are marked as valid when data is insufficient. In this embodiment, the above scheme can simultaneously reflect the degradation rate and discharge activity level of the target defect.

[0065] Based on the discharge amplitude growth rate and discharge frequency density, the margin time for the target defect node to reach the critical state of insulation breakdown is calculated.

[0066] The dynamic degradation acceleration factor of the target defect node is calculated by nonlinearly exponentially coupling the discharge amplitude growth rate with the discharge frequency density. Extract the current cumulative discharge energy of the target defect node and obtain the preset ultimate breakdown energy threshold of the corresponding insulating medium; Using the current accumulated discharge energy as the initial state, an energy evolution integral equation is constructed using the dynamic degradation acceleration factor. The upper limit of the integration time required to reach the ultimate breakdown energy threshold by solving the energy evolution integral equation is then used as the margin time to reach the critical state of insulation breakdown.

[0067] Specifically, the discharge amplitude growth rate and discharge frequency density of the target defect node within a continuous monitoring time window are read and then dimensionlessly processed using reference growth rates and frequency densities calibrated under normal insulation conditions to avoid direct coupling between the two types of parameters due to different units. Nonlinear exponential coupling is applied to the dimensionless results, so that the dynamic degradation acceleration factor increases rapidly when both indicators rise simultaneously; when either indicator significantly exceeds the reference level, the acceleration factor increases accordingly, and when the discharge activity approaches a steady state, the factor approaches the baseline value. To prevent computational overflow caused by abnormal sampling, a reasonable upper limit is set for the factor, and it is updated in each prediction cycle.

[0068] Subsequently, the current accumulated discharge energy of the target defect node is read from the insulation defect state matrix and the corresponding partial discharge event record. The accumulated energy is accumulated sequentially according to the confirmed valid discharge pulses and expressed in a unified energy unit. Then, based on the matched defect type, corresponding insulation medium, medium thickness, and operating environment parameters, the ultimate breakdown energy threshold is extracted from the pre-established medium tolerance database, and the threshold is converted to a unit and benchmark consistent with the current accumulated energy. When the current accumulated discharge energy has reached the threshold, the critical margin time is directly set to zero.

[0069] When performing energy evolution calculations, the current accumulated discharge energy is used as the integration starting point. The future energy change process is established based on the discharge energy growth trend within the most recent time window, and a dynamic degradation acceleration factor is applied to the energy accumulation rate for each predicted time period. The calculation process uses discrete time steps to advance segment by segment. Within each time step, the predicted energy is updated and the elapsed time is accumulated until the predicted energy first reaches the ultimate breakdown energy threshold. The accumulated time corresponding to this first reaching moment is the upper limit of the integration time. To maintain a safe assessment, the energy growth rate uses a predicted value no lower than the current effective trend, and the larger the dynamic degradation acceleration factor, the shorter the time required to reach the threshold. If the threshold is not reached within the preset prediction range, the non-limited state and the corresponding maximum evaluation time are recorded. In this embodiment, the above scheme integrates discharge activity changes and insulation energy accumulation into the dynamic evolution assessment, obtaining an insulation breakdown critical margin time that can be used for fault early warning.

[0070] When the margin time is less than the preset emergency trip threshold, a hard isolation command is generated as the secondary linkage control command to drive the circuit breaker in the actuator to disconnect the corresponding fault branch and perform the fault isolation. When the margin time is greater than or equal to the emergency trip threshold and less than the alarm threshold, a flexible scheduling instruction is generated as the secondary linkage control instruction, and the early warning output is sent to the main station through the secondary communication module.

[0071] Specifically, emergency tripping thresholds and alarm thresholds are pre-set based on the operating time of the circuit breaker tripping mechanism, the importance of the faulty branch, communication delay, and the safety requirements of the insulation medium, ensuring that the alarm threshold is greater than or equal to the emergency tripping threshold. After reading the margin time of the target defective node, the controller first performs time unit unification and data validity verification; when the margin time is less than the emergency tripping threshold, it is determined that the insulation state is approaching the critical danger zone, and a hard isolation command containing the faulty branch identifier, circuit breaker address, tripping reason, and execution priority is immediately generated.

[0072] The hard isolation command is sent to the tripping drive of the corresponding circuit breaker via the local control circuit, driving the tripping mechanism to disconnect the faulty branch, and collecting tripping feedback through auxiliary contacts or position sensors. If no disconnection state is detected within the preset response time, the fault command is retained and an execution failure message is sent to prevent the faulty branch from being re-energized. After the circuit breaker completes the tripping, the status of the faulty branch is latched until maintenance confirmation or manual reset.

[0073] When the margin time reaches the emergency tripping threshold but is less than the alarm threshold, the controller does not directly disconnect the faulty branch. Instead, it generates a flexible dispatch command, which is sent to the master station via the secondary communication module. The message includes the defect type, source location coordinates, actual discharge amplitude, discharge growth trend, discharge frequency density, and margin time, along with the ring main unit identifier and timestamp, for the master station to arrange load transfer, on-site verification, or maintenance plans. When the margin time reaches the alarm threshold, normal monitoring is maintained. In this embodiment, the above scheme implements hard isolation or flexible early warning based on the time margin of insulation risk, balancing fault handling speed and power supply continuity.

[0074] In this embodiment, the specific implementation of each unit in the above system embodiment is the same as that in the above method embodiment, and will not be repeated here.

Claims

1. A primary and secondary fusion ring network box integrating high-frequency partial discharge signals, characterized in that, include: The acquisition module is used to acquire the high-frequency electromagnetic signals and synchronous power frequency phase of multiple monitoring nodes in the primary and secondary fusion ring network box; The separation module is used to separate the high-frequency electromagnetic signal based on spatial time difference correlation to obtain a purified partial discharge pulse sequence; The mapping module is used to combine the internal topology parameters of the primary and secondary fusion ring network box to map the purified partial discharge pulse sequence with the synchronous power frequency phase and calculate the insulation defect state matrix. The drive module is used to generate corresponding secondary linkage control commands based on the insulation defect state matrix, and drive the actuator of the primary and secondary fusion ring network box to perform fault isolation or early warning output based on the secondary linkage control commands.

2. The integrated primary and secondary fusion ring network box for high-frequency partial discharge signals according to claim 1, characterized in that, Acquire high-frequency electromagnetic signals and synchronous power frequency phases from multiple monitoring nodes within the primary and secondary fusion ring network enclosure, including: The acquisition module is used to synchronously acquire the high-frequency electromagnetic signals at the microsecond level through an array of ultra-high frequency sensors arranged in different air chambers of the primary and secondary fusion ring network box. The extraction module is used to extract the fundamental zero-crossing point of the output voltage of the secondary voltage transformer of the primary and secondary integrated ring network box; The compensation module is used to perform phase compensation on the fundamental zero-crossing point according to the preset high-frequency phase shift parameters of the transformer, and generate the synchronous power frequency phase that is time-aligned with the high-frequency electromagnetic signal.

3. The integrated primary and secondary fusion ring network box for high-frequency partial discharge signals according to claim 1, characterized in that, The high-frequency electromagnetic signal is separated based on spatial time difference correlation to obtain a purified partial discharge pulse sequence, including: The module is used to calculate the maximum reflection delay of electromagnetic waves in the cavity based on the maximum diagonal size of the metal cavity of the primary and secondary fusion ring network box, and to construct a multipath reflection suppression window. A shielding module is used to extract the initial arrival timestamp of the first peak in the high-frequency electromagnetic signal and shield subsequent reflected peaks within the multipath reflection suppression window. The calculation module is used to calculate the spatial time difference of arrival based on the initial arrival timestamp of each monitoring node after shielding, and to determine whether the signal source is located inside or outside the enclosure by the spatial time difference of arrival, thereby realizing the spatial time difference correlation separation. The determination module is used to eliminate signals determined to be external interference and to use the direct wave signals determined to be internal sources as the purified partial discharge pulse sequence.

4. The integrated primary and secondary fusion ring network box for high-frequency partial discharge signals according to claim 1, characterized in that, Based on the internal topology parameters of the primary and secondary fusion ring network box, the purified partial discharge pulse sequence is mapped to the synchronous power frequency phase to calculate the insulation defect state matrix, including: Analyze the internal topology parameters of the primary and secondary fusion ring network box, and extract the medium type and corresponding spatial attenuation coefficient between the source location of the purified partial discharge pulse sequence and each monitoring node; The apparent amplitude of the purified partial discharge pulse sequence is inversely compensated using the spatial attenuation coefficient to obtain the true discharge amplitude; The actual discharge amplitude and the occurrence time of the purified partial discharge pulse sequence are projected onto the corresponding phase angle interval of the synchronous power frequency phase to generate a phase-resolved two-dimensional partial discharge spectrum. Extract the discharge characteristic parameters of the phase-resolved partial discharge two-dimensional spectrum and match the defect type; The insulation defect state matrix is ​​calculated by combining the defect type, source location coordinates, and actual discharge amplitude.

5. The integrated primary and secondary fusion ring network box for high-frequency partial discharge signals according to claim 4, characterized in that, The apparent amplitude of the purified partial discharge pulse sequence is inversely compensated using the spatial attenuation coefficient to obtain the true discharge amplitude, including: The apparent amplitude of the purified partial discharge pulse sequence is decomposed into multiple frequency bands of apparent spectral components by performing a fast Fourier transform. Based on the medium type, extract the number of medium interfaces and the medium wave impedance along the propagation path from the source location of the purified partial discharge pulse sequence to each monitoring node. Based on the number of medium interfaces and the medium wave impedance, calculate the frequency-varying transmission coefficient of electromagnetic waves crossing the interface for each frequency band. The spatial attenuation coefficient is converted into the frequency-varying attenuation factor of the corresponding frequency band. The frequency-varying transmission coefficient and the frequency-varying attenuation factor are used to perform frequency-domain segmented inverse compensation on the apparent spectral component to obtain the compensated spectral component. The time-domain signal is reconstructed by performing an inverse fast Fourier transform on the compensated spectral components, and the peak value of the reconstructed time-domain signal is extracted as the true discharge amplitude.

6. The integrated primary and secondary fusion ring network box for high-frequency partial discharge signals according to claim 5, characterized in that, Projecting the actual discharge amplitude and the occurrence time of the purified partial discharge pulse sequence onto the corresponding phase angle interval of the synchronous power frequency phase generates a phase-resolved two-dimensional partial discharge spectrum, including: Based on the medium type, the electromagnetic wave relative permittivity of the corresponding medium is extracted. Combined with the physical length of the propagation path of each monitoring node, the cross-medium propagation time delay of the purified partial discharge pulse sequence from the source position to each monitoring node is calculated. The distortion compensation phase angle is calculated by multiplying the cross-medium propagation time delay by the power frequency angular frequency. The phase shift correction of the synchronous power frequency phase corresponding to the occurrence time of the purified partial discharge pulse sequence is performed using the distortion compensation phase angle to obtain the corrected power frequency phase; The actual discharge amplitude and the occurrence time of the purified partial discharge pulse sequence are projected onto the corresponding phase angle interval of the corrected power frequency phase to generate the phase-resolved partial discharge two-dimensional spectrum.

7. The integrated primary and secondary fusion ring network box for high-frequency partial discharge signals according to claim 4, characterized in that, Extracting discharge characteristic parameters from the phase-resolved partial discharge two-dimensional spectrum and matching defect types includes: The phase-resolved partial discharge two-dimensional spectrum is divided into equal phase angle windows, and the pulse cluster contour gradient and density distribution entropy in each phase angle window are extracted and combined to generate a multidimensional discharge feature parameter tensor. The multidimensional discharge feature parameter tensor is used as the discharge feature parameter. Obtain a pre-constructed topological defect fingerprint database containing electric field distortion features at multi-medium interface points; Calculate the multidimensional spatial distance between the multidimensional discharge feature parameter tensor and each candidate defect type in the topological defect fingerprint database, and output the candidate defect type corresponding to the minimum multidimensional spatial distance to achieve the matching defect type.

8. The integrated primary and secondary fusion ring network box for high-frequency partial discharge signals according to claim 5, characterized in that, Based on the number of medium interfaces and the dielectric wave impedance, the frequency-varying transmission coefficient of electromagnetic waves crossing the interfaces in each frequency band is calculated, including: Extract the surface roughness parameters of the medium interface, and combine them with the center frequency of each frequency band to calculate the interface scattering loss factor of electromagnetic waves in the corresponding frequency band. Calculate the fundamental transmission coefficient of the ideal plane based on the dielectric wave impedance of the adjacent medium; Multiply the basic transmission coefficient by the interface scattering loss factor, and perform a multiplication attenuation calculation according to the number of medium interfaces to obtain the frequency-varying transmission coefficient of electromagnetic waves in each frequency band when crossing the interface.

9. The integrated primary and secondary fusion ring network box for high-frequency partial discharge signals according to claim 1, characterized in that, Based on the insulation defect state matrix, corresponding secondary linkage control commands are generated. These commands then drive the actuators of the primary and secondary fusion ring network enclosure to perform fault isolation or early warning output, including: Extract the discharge amplitude growth rate and discharge frequency density of the target defect node in the insulation defect state matrix; Based on the discharge amplitude growth rate and discharge frequency density, the margin time for the target defect node to reach the critical state of insulation breakdown is calculated. When the margin time is less than the preset emergency trip threshold, a hard isolation command is generated as the secondary linkage control command to drive the circuit breaker in the actuator to disconnect the corresponding fault branch and perform the fault isolation. When the margin time is greater than or equal to the emergency trip threshold and less than the alarm threshold, a flexible scheduling instruction is generated as the secondary linkage control instruction, and the early warning output is sent to the main station through the secondary communication module.

10. The integrated primary and secondary fusion ring network box for high-frequency partial discharge signals according to claim 9, characterized in that, Based on the discharge amplitude growth rate and discharge frequency density, the margin time for the target defect node to reach the critical state of insulation breakdown is calculated, including: The dynamic degradation acceleration factor of the target defect node is calculated by nonlinearly exponentially coupling the discharge amplitude growth rate with the discharge frequency density. Extract the current cumulative discharge energy of the target defect node and obtain the preset ultimate breakdown energy threshold of the corresponding insulating medium; Using the current accumulated discharge energy as the initial state, an energy evolution integral equation is constructed using the dynamic degradation acceleration factor. The upper limit of the integration time required to reach the ultimate breakdown energy threshold by solving the energy evolution integral equation is then used as the margin time to reach the critical state of insulation breakdown.

Images