A ring main unit partial discharge mode recognition method and system

Abstract

This invention discloses a method and system for partial discharge pattern recognition in a ring main unit, belonging to the field of power equipment condition monitoring technology. The method includes: acquiring test data transmitted between carrier communication nodes to determine the fluctuation data of the bit error rate (BER) over time; acquiring the phase synchronization relationship between the fluctuation data and the power frequency voltage; based on the phase synchronization relationship, aligning and averaging the fluctuation data from multiple power frequency cycles according to the power frequency phase to determine the distribution characteristics of the average BER as a function of the power frequency phase; acquiring the similarity between the distribution characteristics and the standard phase distribution characteristics of multiple preset partial discharge patterns; and determining the current partial discharge pattern of the ring main unit based on the similarity. This invention fully reuses existing carrier communication nodes, requires no additional sensors, and achieves low-cost, high-reliability, non-intrusive partial discharge pattern recognition by using the power frequency phase characteristics of the BER fluctuations and performing pattern matching.

Description

Technical Field

[0001] This application relates to the field of power equipment condition monitoring, and in particular to a method and system for partial discharge pattern recognition of ring main units. Background Technology

[0002] As a critical node device in the power distribution network, the internal insulation condition of the ring main unit directly affects the reliability of power supply. Partial discharge is an important sign of insulation degradation, and the identification of different types of partial discharge is crucial for formulating maintenance strategies.

[0003] Existing partial discharge detection methods, such as pulsed current methods, ultra-high frequency methods, and ultrasonic methods, typically require the installation of dedicated sensors and data acquisition systems on ring main units. However, the limited space and numerous outgoing circuits within ring main units severely restrict the locations available for sensor installation. Adding a large number of sensors significantly increases equipment modification costs and maintenance complexity, limiting the widespread application of online partial discharge monitoring in distribution networks. While some solutions attempt to detect insulation defects using changes in the characteristics of power line carrier communication channels, these methods can only determine the presence or absence of partial discharge, failing to effectively distinguish discharge patterns. Furthermore, communication errors caused by partial discharge are generally considered sporadic and random, making it difficult to establish stable and identifiable characteristic patterns, thus failing to meet the needs of refined condition assessment. Summary of the Invention

[0004] This application provides a method and system for partial discharge pattern recognition in a ring network box to improve the above-mentioned problems.

[0005] To achieve the above objectives, this application adopts the following technical solution: In a first aspect, embodiments of this application propose a partial discharge pattern recognition method for a ring network enclosure, applied to a partial discharge pattern recognition system for a ring network enclosure. The system includes a control terminal and a ring network enclosure, with carrier communication nodes installed within the ring network enclosure. The method is executed by the control terminal and includes: acquiring test data transmitted between the carrier communication nodes; determining fluctuation data of the bit error rate (BER) over time based on the test data; acquiring the phase synchronization relationship between the fluctuation data and the power frequency voltage; based on the phase synchronization relationship, aligning the fluctuation data of multiple power frequency cycles according to the power frequency phase and then superimposing and averaging them to determine the average BER, and determining the distribution characteristics of the average BER as a function of the power frequency phase; acquiring the similarity between the distribution characteristics and the standard phase distribution characteristics of multiple preset partial discharge patterns; and determining the current partial discharge pattern of the ring network enclosure based on the multiple similarities.

[0006] Therefore, by reusing existing carrier communication nodes within the ring network box, non-invasive identification of partial discharge modes is achieved without installing any dedicated sensors, overcoming the shortcomings of traditional methods such as difficult and costly sensor installation. By utilizing the correlation characteristics between bit error rate fluctuations and power frequency voltage phase, communication quality changes are transformed into analyzable discharge characteristics, breaking through the industry's technical prejudice that partial discharge only causes random communication errors, and providing an economical and efficient new approach for power distribution equipment condition monitoring.

[0007] In conjunction with the first aspect, optionally, the carrier communication node includes a first carrier communication node and a second carrier communication node, acquires test data transmitted between the carrier communication nodes, and determines the fluctuation data of the bit error rate over time based on the test data, including: instructing the first carrier communication node to send a test frame of the target bit sequence to the second carrier communication node; comparing the bit sequence acquired by the second carrier communication node with the target bit sequence stored locally bit by bit to determine the number of erroneous bits within a preset time window, and acquiring the bit error rate of the preset time window; and determining the fluctuation data based on the bit error rates of multiple preset time windows.

[0008] Therefore, by actively sending test frames with known target bit sequences and comparing them bit by bit, non-stationary interference from normal business data traffic can be eliminated, and the bit error rate statistics can fully reflect the influence of transient interference such as partial discharge on the channel. By adopting a sliding statistical method with a preset time window, the statistical stability can be improved while ensuring time resolution, and high sensitivity and high reliability detection of weak discharge interference can be achieved.

[0009] In conjunction with the first aspect, optionally, the phase synchronization relationship between the fluctuation data and the power frequency voltage is obtained, including: obtaining the power frequency voltage signal based on the carrier communication node; performing zero-crossing detection on the power frequency voltage signal and determining the pulse sequence at the start of the power frequency cycle; comparing the timestamp of the fluctuation data with the pulse sequence; and determining the correspondence between the bit error rate and the power frequency phase at each moment based on the comparison result.

[0010] Therefore, it can be seen that using the power frequency voltage signal coupled by the carrier communication node itself for zero-crossing detection and phase synchronization does not require additional wiring or external synchronization devices, and achieves the same source synchronization between the measurement signal and the phase reference, avoiding the time drift problem between independent acquisition systems; the bit error rate fluctuation data is directly mapped to the power frequency phase axis, providing an accurate and reliable timing basis for the subsequent extraction of phase-dependent discharge characteristics.

[0011] In conjunction with the first aspect, optionally, based on the phase synchronization relationship, the fluctuation data of multiple power frequency cycles are aligned according to the power frequency phase and then superimposed and averaged to determine the average bit error rate, and the distribution characteristics of the average bit error rate with the power frequency phase are determined, including: dividing each power frequency cycle into M phase intervals, where M is an integer greater than 1; based on the correspondence relationship, averaging the bit error rate values ​​falling into the same phase interval in multiple power frequency cycles to determine the average bit error rate of the corresponding phase interval; constructing a target curve with the power frequency phase as the horizontal axis and the average bit error rate as the vertical axis, and using the target curve as the distribution characteristic.

[0012] Therefore, by aligning the bit error rate fluctuation data of multiple power frequency cycles and then superimposing and averaging them, the random noise and accidental bit errors without phase regularity are effectively suppressed, the partial discharge related feature components synchronized with the power frequency voltage are significantly enhanced, and the signal-to-noise ratio is greatly improved. The discrete average bit error rate is constructed as a target curve as a distribution feature, realizing the transformation from random data to deterministic patterns, and providing highly recognizable input features for subsequent pattern recognition.

[0013] In conjunction with the first aspect, optionally, the similarity between the distribution characteristics and the standard phase distribution characteristics of multiple preset partial discharge modes is obtained. Based on the multiple similarities, the current partial discharge mode of the ring network box is determined, including: cyclically shifting the target curve along the power frequency phase axis with a preset phase step to obtain multiple shift curves, each shift curve corresponding to a shift phase amount; determining the sum of the absolute values ​​of the average bit error rate differences between the standard curve and the shift curve in M ​​phase intervals for each preset partial discharge mode; determining the shift curve with the smallest sum of absolute values ​​as the target shift curve; obtaining the shift phase amount corresponding to the target shift curve as the target shift phase amount; and determining the similarity between the distribution characteristics and the standard phase distribution characteristics of the preset partial discharge mode based on the minimum sum of absolute values ​​and the matching of the target shift phase amount with the preset shift amount range corresponding to the preset partial discharge mode.

[0014] Therefore, by performing a cyclic shift traversal to search for the optimal phase alignment of the target curve, the overall phase shift caused by communication path delay, transformer angle difference, etc., can be automatically compensated, ensuring that the similarity assessment is not affected by the inherent delay of the system. Combining the dual constraints of minimizing the sum of absolute errors and the preset shift range, both the matching degree of the curve amplitude shape and the physical rationality of the phase correction are considered, effectively avoiding misidentification caused by forced misalignment matching, and greatly improving the robustness and accuracy of pattern matching.

[0015] In conjunction with the first aspect, optionally, based on the minimum sum of absolute values ​​and the matching between the target shift phase amount and the preset shift range corresponding to the preset partial discharge mode, the similarity between the distribution characteristics and the standard phase distribution characteristics of the preset partial discharge mode is determined, including: obtaining a preset maximum allowable difference value, determining the difference between the maximum allowable difference value and the minimum sum of absolute values, and using the difference value as a first similarity component, wherein when the difference value is less than zero, the first similarity component is set to zero; when the target shift phase amount falls within the preset shift range, the second similarity component is set to 1, otherwise it is set to 0; and the product of the first similarity component and the second similarity component is used as the similarity.

[0016] Therefore, by decoupling the similarity into a first similarity component and a second similarity component, and fusing them through a product, the similarity can not only finely distinguish between different degrees of amplitude matching, but also rigidly reject physically unacceptable large phase shifts. This calculation method is simple, highly interpretable, and can effectively prevent false high similarity matches caused by accidental similarity due to noise, thus improving the reliability and specificity of the recognition results.

[0017] In conjunction with the first aspect, optionally, after determining the current partial discharge mode of the ring network box based on multiple similarities, the method further includes: comparing the maximum value among the multiple similarities with a preset identification threshold; when the comparison result is that the maximum value is less than the preset identification threshold, outputting a labeling request containing the current distribution features; obtaining the mode label returned in response to the labeling request; and supplementing the current distribution features and the mode label as newly added standard phase distribution features to the standard phase distribution features of multiple preset partial discharge modes.

[0018] Therefore, it can be seen that by actively triggering the annotation request and obtaining expert mode labels when the maximum similarity is below the threshold, the organic synergy between automatic recognition and human experience is achieved.

[0019] In conjunction with the first aspect, optionally, the current distribution features and mode labels are added as new standard phase distribution features to the standard phase distribution features of multiple preset partial discharge modes, including: obtaining the existing standard phase distribution feature set corresponding to the mode label; if the set is empty, adding the current distribution features and mode labels to the set; otherwise, determining the similarity between the current distribution features and each existing standard phase distribution feature in the set; when all similarities are less than a preset addition threshold, adding the current distribution features and mode labels to the set.

[0020] Therefore, it can be seen that before adding new standard features, by judging the similarity of existing features under the same pattern label, only distribution features with sufficient novelty are included in the standard library. This effectively prevents the accumulation of redundant samples caused by fluctuations in operating conditions. While ensuring the simplicity and representativeness of the standard library, it avoids the waste of storage resources and the decline in matching efficiency, and maintains the clarity and stability of the classification boundary under long-term operation.

[0021] The second aspect of this application proposes a partial discharge pattern recognition system for a ring network enclosure, including a control terminal and a ring network enclosure, wherein a carrier communication node is installed inside the ring network enclosure, and the system includes: The first acquisition module is used to acquire test data transmitted between carrier communication nodes and determine the fluctuation data of bit error rate over time based on the test data. The second acquisition module is used to acquire the phase synchronization relationship between the fluctuation data and the power frequency voltage; The determination module is used to determine the average bit error rate by aligning the fluctuation data of multiple power frequency cycles according to the power frequency phase based on the phase synchronization relationship and then superimposing and averaging them, and to determine the distribution characteristics of the average bit error rate as the power frequency phase changes. The third acquisition module is used to acquire the similarity between the distribution characteristics and the standard phase distribution characteristics of multiple preset partial discharge modes, and to determine the current partial discharge mode of the ring network box based on the multiple similarities.

[0022] A third aspect of this invention provides an electronic device, which includes: At least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the method proposed in the first aspect of the present invention.

[0023] A fourth aspect of the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method as described in the first aspect of the present invention. Attached Figure Description

[0024] Figure 1 This is a flowchart illustrating a partial discharge pattern recognition method for a ring network box proposed in an embodiment of this application.

[0025] Figure 2 This is a schematic diagram of the structure of a partial discharge pattern recognition system for a ring network box proposed in an embodiment of this application.

[0026] Figure 3 This is a schematic diagram of the structure of a processing device proposed in this application. Detailed Implementation

[0027] 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. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0028] This invention proposes a partial discharge pattern recognition method for a ring network enclosure, applied to a ring network enclosure partial discharge pattern recognition system. The ring network enclosure partial discharge pattern recognition system includes a control terminal and multiple carrier communication nodes located within the ring network enclosure. The carrier communication nodes exchange data using power lines and are capable of sensing power frequency voltage signals. The control terminal communicates with at least one carrier communication node wirelessly or via wired means to obtain test statistics and issue commands. In this embodiment, the carrier communication nodes employ existing broadband power line carrier (HPLC) chips, supporting the transmission of custom test frames and reporting of received bit sequences. For details, please refer to... Figure 1 The method is executed by the control terminal and includes the following steps: S101: Obtain test data transmitted between carrier communication nodes, and determine the fluctuation data of bit error rate over time based on the test data.

[0029] Understandably, the ring network enclosure is originally equipped with carrier communication nodes for automated data exchange, such as broadband power line carrier modules. These nodes can transmit test data to each other. The test data can be any pre-defined signal sequence whose correctness can be verified at the receiving end. By having the communication nodes send and receive test data within a specified time period, and comparing the actual received data with the expected correct data at the receiving end, the number of erroneous bits within a certain time period can be counted, and the bit error rate can be calculated. With the occurrence of partial discharge, high-frequency electromagnetic interference is superimposed on the power line channel, causing instantaneous deterioration of the communication signal quality and a momentary or short-term increase in the bit error rate. Therefore, by continuously sending and receiving test data and statistically analyzing the bit error rate, a time-varying bit error rate sequence, i.e., fluctuation data, can be obtained. This fluctuation data retains the subtle temporal characteristics of the interference caused by partial discharge.

[0030] Specifically, at least two carrier communication nodes are provided within the ring network enclosure, which are distinguished as a first carrier communication node acting as the sender and a second carrier communication node acting as the receiver. In this application, step S101 may include the following steps: S1011: Instructs the first carrier communication node to send a test frame of the target bit sequence to the second carrier communication node.

[0031] Understandably, the control terminal instructs the first carrier communication node to send a test frame to the second carrier communication node within a specified test period. The payload of this test frame is a predefined target bit sequence, i.e., a binary sequence that is known and completely identical to both the sender and receiver. The target bit sequence can be a pseudo-random sequence with good autocorrelation properties or a simple alternating code pattern. The key is the determinism of the sequence, i.e., the receiver can accurately determine whether each received bit has an error based on the locally stored original sequence.

[0032] S1012: Compare the bit sequence obtained by the second carrier communication node with the target bit sequence stored locally bit by bit to determine the number of erroneous bits within the preset time window and obtain the bit error rate of the preset time window.

[0033] After the second carrier communication node receives the test frame from the first carrier communication node, it sends the demodulated bit sequence, i.e., the actual received sequence, to the control terminal or processes it locally. In this step, the actual received sequence is compared bit by bit with the target bit sequence pre-stored locally. For each bit, if the received bit does not match the target bit, it is determined to be an erroneous bit. By counting the total number of erroneous bits occurring within a preset time window, such as a fixed duration of 1 millisecond or 2 milliseconds, and dividing this number of erroneous bits by the total number of bits transmitted within that time window, i.e., the product of the test frame length and the number of frames within the window, the bit error rate corresponding to the preset time window can be obtained.

[0034] The selection of the preset time window duration balances sensitivity and statistical stability: if the window is too short, such as 0.1 milliseconds, the number of bits per window is too small, resulting in a large variance in the bit error rate (BER) estimate; if the window is too long, such as 10 milliseconds, it may mask some details of short-duration discharge interference. Experiments show that setting the window length between 0.5 and 2 milliseconds, and using a 50% overlapping sliding window method, can effectively smooth random noise without sacrificing time resolution. Furthermore, frequency-division statistics can be performed on the BER of different carrier frequency sub-channels to utilize the differences in attenuation characteristics of partial discharge interference in different frequency bands to enhance anti-interference capabilities.

[0035] S1013: Determine fluctuation data based on the bit error rate of multiple preset time windows.

[0036] Understandably, in step S1012, a bit error rate (BER) value is obtained after each preset time window's statistical calculation. As time progresses and test frames are continuously transmitted, the system continuously calculates and records the BER corresponding to each time window, thus obtaining a time-indexed BER sequence. These BER values ​​are not constant but fluctuate with changes in the electromagnetic environment on the power line, including the occurrence of partial discharge. Therefore, the sequence composed of BER values ​​from multiple consecutive time windows constitutes the fluctuating data. This fluctuating data is essentially a time series, where each data point corresponds to the BER level of the test channel at a specific moment. To obtain a continuous fluctuating curve, the preset time window can be a sliding window, meaning there is partial overlap between adjacent windows. This improves the smoothness of the statistical results without sacrificing time resolution.

[0037] S102: Obtain the phase synchronization relationship between fluctuation data and power frequency voltage.

[0038] Partial discharge is closely related to the waveform of the AC voltage applied to the insulating medium, and it typically occurs or intensifies only in certain phase intervals of the power frequency cycle. Therefore, simple time-domain bit error rate fluctuations cannot directly reflect the discharge mode, and it is necessary to establish a correspondence between fluctuation data and the phase of the power frequency voltage.

[0039] Specifically, the power frequency voltage refers to the 50Hz or 60Hz AC voltage of the power system in which the ring main unit operates. The phase synchronization relationship is obtained by determining the electrical angle position within the power frequency cycle corresponding to each bit error rate data sampling moment. The power frequency voltage signal is synchronously acquired, and its periodic reference point is detected, such as the moment when the voltage crosses zero from negative to positive. This reference is used to generate a phase-time axis from 0° to 360°. Then, the time-stamped fluctuation data is mapped onto this phase axis according to the time information, thereby assigning a power frequency phase angle to each bit error rate value.

[0040] Understandably, since the carrier communication node itself is coupled through the power line and can sense the power frequency voltage waveform, the establishment of this correspondence can be accomplished by the communication node itself or a simple zero-crossing detection circuit, without the need for an additional complex synchronous acquisition system.

[0041] For example, step S102 may include the following steps: S1021: Obtain power frequency voltage signal based on carrier communication node.

[0042] Understandably, when a carrier communication node within a ring network box exchanges data via power lines, its coupling unit is directly connected to the power frequency grid. Therefore, this node naturally possesses the ability to acquire power frequency voltage signals. In this step, it is not necessary to add an additional voltage transformer or a dedicated acquisition circuit. Instead, the signal coupling channel of the carrier communication node itself is used to extract the power frequency voltage waveform superimposed on the power lines or its key characteristic parameters.

[0043] Since the power frequency voltage signal is a background physical quantity that can be synchronously sensed by the carrier communication node during normal operation, this acquisition process requires no hardware modification; only the corresponding sampling or detection function needs to be enabled at the software level. The acquired power frequency voltage signal provides the original time base reference for subsequent zero-crossing detection and periodic synchronization.

[0044] As an alternative implementation, the power frequency voltage signal can also be obtained wirelessly from the secondary side of the voltage transformer (PT) of the distribution line where the ring network box is located, instead of being directly obtained by the carrier communication node. The control terminal then performs zero-crossing detection and phase allocation. This method is suitable for scenarios with existing centralized distribution automation master stations, reducing the processing capacity requirements of individual carrier communication nodes.

[0045] S1022: Perform zero-crossing detection on the power frequency voltage signal and determine the pulse sequence at the start of the power frequency cycle.

[0046] Zero-crossing detection refers to capturing the moment when a power frequency voltage signal crosses the zero level as it transitions from the negative half-cycle to the positive half-cycle (or from the positive half-cycle to the negative half-cycle). Since the power frequency voltage waveform crosses zero twice in each cycle, the one with a fixed polarity change is usually selected as the start marker of each power frequency cycle. By performing real-time zero-crossing detection on the acquired power frequency voltage signal, a set of pulse signals arranged sequentially in time can be generated. The rising or falling edge of each pulse precisely corresponds to the start time of a power frequency cycle (i.e., the 0° phase point). This pulse sequence forms a uniform time scale on the time axis, and the time interval between adjacent pulses is a complete power frequency cycle.

[0047] To improve the accuracy of zero-crossing detection, bandpass filtering can be applied to the power frequency signal before zero-crossing detection to remove high-frequency harmonics and transient interference, preventing zero-crossing jitter caused by voltage waveform distortion. In addition to zero-crossing detection, phase-locked loops (PLLs) or software PLLs can be used to directly extract continuous phase information from the power frequency voltage to obtain a smoother and more interference-resistant phase reference.

[0048] S1023: Compare the timestamps of the fluctuation data with the pulse sequence.

[0049] Understandably, for each bit error rate sampling point, the power frequency cycle interval in which its timestamp lies can be determined through interpolation or lookup, i.e., between which two adjacent zero-crossing pulses it lies in, and the time offset of that sampling point relative to the starting pulse of its cycle can be calculated. This offset directly reflects the relative timing position of the bit error rate value within the power frequency cycle.

[0050] S1024: Based on the comparison results, determine the correspondence between the bit error rate and the power frequency phase at each moment.

[0051] Based on the time offset of each bit error rate sampling point relative to the start of its corresponding power frequency cycle, and combined with the known length of the power frequency cycle (e.g., 20 milliseconds for a 50Hz power grid), this time offset can be proportionally mapped to a phase angle value between 0° and 360°. For example, a bit error rate sampling point with an offset of 5 milliseconds corresponds to a power frequency phase of (5ms / 20ms) × 360° = 90°.

[0052] Through this mapping, the original bit error rate sequence, which only had absolute time labels, was reorganized into a dataset indexed by the power frequency phase. Each power frequency phase angle may correspond to one or more bit error rate values ​​from different periods, thus establishing a correspondence between the bit error rate at each moment and the power frequency phase at that moment.

[0053] S103: Based on the phase synchronization relationship, the fluctuation data of multiple power frequency cycles are aligned according to the power frequency phase and then superimposed and averaged to determine the average bit error rate, and the distribution characteristics of the average bit error rate with the power frequency phase change are determined.

[0054] Within a single power frequency cycle, due to interference from non-discharge factors such as random noise and load switching, the phase distribution of the bit error rate often exhibits a chaotic state, and the characteristic peaks generated by partial discharge may be submerged. However, partial discharge interference has repeatability that is synchronized with the power frequency voltage, meaning that each discharge tends to cause an increase in bit error rate within a similar phase range.

[0055] Therefore, in this step, fluctuation data from multiple consecutive power frequency cycles are collected, meaning each cycle contains a complete segment of bit error rate (BER) information corresponding to the 0°–360° phase. These cycles are then aligned according to the power frequency phase, i.e., the 0° points of all cycles coincide. Next, the BER values ​​for each identical phase interval, such as the area near a certain phase angle, are averaged across different cycles to obtain the average BER for that phase interval. Traversing the entire power frequency cycle results in a curve or graph of the average BER changing with the power frequency phase, i.e., the distribution characteristics.

[0056] Specifically, in this application, step S103 may include the following steps: S1031: Divide each power frequency cycle into M phase intervals, where M is an integer greater than 1.

[0057] The power frequency cycle corresponds to a complete electrical angle range from 0° to 360°. In order to characterize the variation of bit error rate with power frequency phase in a finite discrete form, this continuous angle range needs to be divided into several sub-intervals of equal width, i.e., M phase intervals. The value of M determines the resolution of the phase distribution characteristics: the larger M is, the smaller the angle range covered by each phase interval, and the stronger the ability to capture details of bit error rate changes; conversely, the smaller M is, the wider the interval, the more data points are concentrated in each interval, and the smoother the average result.

[0058] In practice, the value of M can be selected based on the length of the power frequency cycle and the sampling density of the fluctuation data, ensuring that a sufficient number of bit error rate samples are accumulated within each phase interval to guarantee the statistical reliability of the average value. For example, if the power frequency cycle is 20 milliseconds, dividing it into 360 equal phase intervals, each interval corresponding to a phase width of 1°, can achieve a high phase resolution. This step does not limit the specific value of M, only requiring it to be an integer greater than 1, thus ensuring the practical operability of dividing the power frequency cycle.

[0059] S1032: Based on the correspondence, the bit error rate values ​​falling into the same phase interval in multiple power frequency cycles are averaged to determine the average bit error rate of the corresponding phase interval.

[0060] Specifically, for all the bit error rate (BER) data points collected across multiple power frequency cycles, they are first categorized into the corresponding phase intervals defined in step S1031 based on the power frequency phase value corresponding to each data point. For example, all BER values ​​with power frequency phases between 15.5° and 16.5° will be uniformly categorized into the phase interval representing the vicinity of 16°. Subsequently, the arithmetic mean of all BER values ​​accumulated within each phase interval is calculated to obtain the average BER corresponding to that phase interval.

[0061] Because communication errors caused by partial discharge tend to recur within a specific phase interval of the power frequency cycle, while random background noise and accidental errors are scattered across different cycles without a stable phase relationship, averaging the bit error rate values ​​of multiple cycles within the same phase interval preserves or even enhances the phase-consistent discharge-related error components, while random noise components cancel each other out and attenuate in amplitude due to phase incoherence. The resulting average bit error rate essentially removes irregular interference components, highlighting deterministic bit error fluctuation information synchronized with the power frequency voltage.

[0062] S1033: Construct a target curve with the power frequency phase as the horizontal axis and the average bit error rate as the vertical axis, and use the target curve as a distribution feature.

[0063] After calculating the average bit error rate (BER) values ​​for each of the M phase intervals, a complete set of discrete data point pairs is obtained. Each pair consists of the center phase value of a power frequency phase interval and the corresponding average BER value. These discrete points are then connected sequentially according to the power frequency phase from smallest to largest, or arranged directly as a discrete sequence, and plotted as a curve in a two-dimensional coordinate system. The horizontal axis of this curve represents the power frequency phase, and the vertical axis represents the average BER under that phase. This curve is defined as the target curve for this measurement condition, which is also the final distribution characteristic.

[0064] The target curve is similar in form to the traditional partial discharge PRPD (phase-resolved partial discharge) spectrum, but the essential difference lies in the fact that the vertical axis of the traditional PRPD spectrum is usually the discharge amplitude or the number of discharges, relying on high-frequency acquisition by a dedicated sensor; while the vertical axis of this invention is the discharge interference intensity indirectly characterized by the communication bit error rate. This innovation enables the carrier communication module to simultaneously act as a data transceiver and a discharge detection module, realizing the integration of communication and sensing.

[0065] S104: Obtain the similarity between the distribution characteristics and the standard phase distribution characteristics of multiple preset partial discharge modes, and determine the current partial discharge mode of the ring network box based on the multiple similarities.

[0066] Understandably, different types of partial discharge, such as internal air gap discharge, surface discharge, and corona discharge, typically exhibit different distribution patterns at the power frequency phase due to differences in discharge mechanisms and electrode structures. Therefore, based on accumulated experience in the laboratory or field, standard phase distribution characteristics for each typical discharge mode can be established in advance, forming a reference template library.

[0067] This standard phase distribution feature library can be obtained through full-scale discharge experiments on a typical ring network box defect model. For example, for internal air gap discharge, tiny holes can be artificially created in the insulation components and high voltage can be applied to excite discharge. The same carrier communication test system of this invention can be used to collect bit error rate fluctuation data, and then phase synchronization and superposition averaging can be performed to obtain its standard distribution characteristics. At the same time, in actual field operation, after power outage maintenance to confirm the defect type, typical distribution characteristics from historical monitoring data can be added to the standard library after expert annotation, so as to achieve continuous expansion of the library.

[0068] As one implementation method, the distribution characteristics under the current operating conditions are compared one by one with the standard characteristics of each preset pattern in the template library to calculate the similarity. The similarity evaluation can be based on multi-dimensional information such as graphic shape, amplitude distribution, and peak phase, and a quantitative value can be obtained through appropriate distance measurement or correlation algorithm.

[0069] Furthermore, as one implementation, step S104 may include the following steps: S1041: The target curve is cyclically shifted along the power frequency phase axis with a preset phase step to obtain multiple shift curves, each shift curve corresponding to a shift phase amount.

[0070] The horizontal axis of the target curve represents the power frequency phase. However, under actual field conditions, due to inherent time delays in the signal coupling path of carrier communication nodes, voltage transformer phase differences on power lines, and response times of detection circuits, the measured target curve may exhibit an overall phase shift compared to the preset standard curve. This means the entire curve shape may shift to the left or right by a fixed angle on the phase axis. If the two curves are directly compared without correction, even if their actual discharge modes are the same, they may be misjudged as having low similarity due to phase misalignment.

[0071] Therefore, a preset phase step size, such as 1°, can be selected, and the target curve can be gradually shifted left or right along the power frequency phase axis, shifting by one step each time, and traversing all possible offsets within a preset search range, such as -30° to +30°. When the shift causes a portion of the curve to exceed the boundary of 0° or 360°, that portion re-enters from the other end of the phase axis, thus maintaining data integrity within the 360° period. After each shift, a new curve is obtained, called the shift curve, and the shift phase amount corresponding to that shift is recorded. After a complete traversal, a set containing multiple shift curves is obtained, each curve corresponding to a different shift phase amount.

[0072] This cyclic shift search strategy is equivalent to performing a one-dimensional cross-correlation operation in the phase dimension. However, compared to directly calculating the cross-correlation coefficient, the error metric based on the sum of absolute values ​​in subsequent steps is more tolerant of the scaling of the vertical axis between the target curve and the standard curve due to differences in discharge power intensity, communication path attenuation, etc. This is because the sum of absolute values ​​is an L1 norm metric, which is more robust to abnormally large deviations in individual phase intervals and is less likely to completely overturn the similarity evaluation due to a few noise spikes.

[0073] S1042: Determine the sum of the absolute values ​​of the average bit error rate differences between the standard curve and the shift curve of each preset partial discharge mode in M ​​phase intervals, and determine the shift curve with the smallest sum of absolute values ​​as the target shift curve.

[0074] Specifically, for a given preset mode, its standard phase distribution characteristic is also a curve composed of the average bit error rate (BER) values ​​over M phase intervals. For each shift curve, all M phase intervals are traversed. Within each interval, the absolute value of the difference between the average BER of the standard curve and the average BER of the shift curve in that interval is taken. These M absolute values ​​are then summed to obtain a non-negative number representing the overall difference. The smaller this sum, the closer the shapes of the two curves are to that shift amount.

[0075] After calculating the sum of the absolute values ​​for all shift curves, these sums are compared, and the one with the smallest value is selected as the target shift curve for this preset mode. This step essentially finds the optimal phase correction amount required under the assumption that the current discharge belongs to this mode, as well as the minimum amplitude error achievable under this optimal correction.

[0076] S1043: Obtain the shift phase corresponding to the target shift curve as the target shift phase.

[0077] While determining the target shift curve, the associated shift phase quantity is extracted separately and named the target shift phase quantity. This quantity represents the magnitude and direction of the overall translation of the measured curve on the phase axis in order to achieve the best match between the current measured distribution characteristics and the standard characteristics of the preset mode.

[0078] The target shift phase may be positive (shifted to the right, i.e., lag) or negative (shifted to the left, i.e., lead), and its range depends on the search boundary set in step S1041.

[0079] S1044: Based on the minimum sum of absolute values ​​and the matching between the target shift phase amount and the preset shift range corresponding to the preset partial discharge mode, determine the similarity between the distribution characteristics and the standard phase distribution characteristics of the preset partial discharge mode.

[0080] Understandably, the smaller the minimum sum of absolute values, the higher the agreement between the amplitude waveforms of the two curves after phase correction under the current preset mode, and the stronger the evidence supporting the attribution to this mode; conversely, if the minimum difference is still large, then there is an essential difference in the discharge intensity distribution between the two modes, and the similarity should be low.

[0081] On the other hand, the target phase shift must fall within a reasonable offset range specified in advance by the preset mode. This range is determined based on prior knowledge such as the discharge mechanism and system delay characteristics. For example, a certain type of discharge usually only allows a phase deviation within ±10°. If the optimal match requires shifting the curve by 30° to align, it is very likely just a random local shape similarity. Such a match does not conform to the laws of physics, and the similarity should be rejected or weakened.

[0082] Understandably, this step does not limit the specific mathematical expression of similarity, but its core technical approach is to combine amplitude difference and phase compliance to jointly determine the final similarity value. When the target shift phase amount falls within the preset range, the similarity is inversely proportional to the amplitude difference; when it exceeds the range, the similarity is set to the minimum or directly to zero.

[0083] As an example, it can be done in the following way: S201: Obtain the preset maximum allowable difference, determine the difference between the maximum allowable difference and the minimum value of the sum of absolute values, and use the difference as the first similarity component. When the difference is less than zero, the first similarity component is set to zero.

[0084] Understandably, the maximum permissible difference is a non-negative threshold calibrated empirically or experimentally. It represents the maximum cumulative amplitude error that can be tolerated when judging two curves to belong to the same discharge mode, i.e., the minimum sum of absolute values. Physically, it can be understood as the upper limit of error at which the similarity begins to degrade from full marks to zero. Subtracting the minimum sum of absolute values ​​from this maximum permissible difference yields a difference value. The larger this difference value, the smaller the cumulative amplitude deviation between the measured distribution characteristics and the standard curve, and the closer their shapes match; conversely, the smaller the difference value, the larger the amplitude deviation. Specifically, when the minimum sum of absolute values ​​exceeds the maximum permissible difference, i.e., the difference calculation result is negative, it indicates that the current amplitude difference has exceeded the acceptable range. In this case, the first similarity component is forcibly set to zero to avoid negative similarity values.

[0085] S202: When the target shift phase amount falls within the preset shift amount range, the second similarity component is set to 1; otherwise, it is set to 0.

[0086] The preset shift range is an angle interval predetermined for each specific partial discharge mode. This range defines the maximum allowable offset during phase matching correction. This range is set based on the physical characteristics of the discharge and the engineering understanding of the inherent phase delay of the system. For example, the phase position of a certain type of discharge is relatively fixed in the power frequency cycle, and the allowable measurement phase deviation is only within ±Δφ.

[0087] The target phase shift represents the amount of phase shift that needs to be applied to the measured curve to achieve optimal morphological matching between the two curves. This is then compared to a preset shift range. If the target phase shift value is entirely within this preset range, the phase correction is deemed reasonable and acceptable, and the second similarity component is set to "1". Otherwise, even if a large forced shift makes the two curves morphologically similar, this shift does not conform to the inherent physical characteristics of this discharge type and should be considered a false match; in this case, the second similarity component is set to "0".

[0088] S203: The product of the first similarity component and the second similarity component is taken as the similarity score.

[0089] Specifically, the first similarity component and the second similarity component are directly multiplied. If the second similarity component is 0, the product result will always be 0, regardless of the size of the first similarity component, indicating that the preset mode is completely excluded due to unreasonable phase shift; only when the second similarity component is 1 will the similarity retain the actual value of the first similarity component, that is, the similarity level is completely determined by the degree of amplitude matching.

[0090] This product-based combination strategy decouples and recouples the continuous evaluation of amplitude differences with the rigid criterion for phase rationality. This allows the similarity score to not only finely distinguish different degrees of amplitude matching but also to reject physically unacceptable matches. The resulting similarity score is a non-negative real number between 0 and the maximum permissible difference. The higher the value, the more consistent the measured distribution characteristics are with the standard phase distribution characteristics of the preset partial discharge mode at both the morphological and phase levels, and the higher the reliability of the mode as the final identification result.

[0091] For example, suppose the maximum allowable difference is set to 10. If the minimum sum of absolute values ​​obtained in a certain calculation is 3, the target shift phase is +5°, and the allowable shift range for this preset pattern is [-8°, +8°]. Then the first similarity component is 10-3=7, the second similarity component is 1, and the final similarity is 7. If another calculation shows a target shift phase of +15°, exceeding the range, then even if the minimum sum of absolute values ​​is only 2, the second similarity component is 0, and the final similarity is also 0, thus the pattern is excluded. This mechanism makes the recognition decision more physically meaningful and engineering-practical.

[0092] Optionally, as one implementation, after determining the current partial discharge mode of the ring main unit, the following steps may also be included: S105: Compare the maximum value among multiple similarities with a preset recognition threshold.

[0093] Each preset partial discharge pattern corresponds to a similarity value, among which there must be a maximum value. This step compares this maximum value with a preset identification threshold. The preset identification threshold is an empirical calibration constant used to define the minimum confidence level at which a pattern can be identified as belonging to a known pattern. If the maximum value is greater than or equal to the threshold, it indicates that the currently measured distribution characteristics highly match a standard pattern, and the system can rely on the identification result. Conversely, if the maximum value is less than the threshold, it means that even the preset pattern with the highest similarity does not meet the confidence standard for matching the current distribution characteristics. In this case, it is likely that a new discharge type not included in the standard library has appeared, or the interference signal is too severe to form an effective match.

[0094] S106: When the comparison result is that the maximum value is less than the preset recognition threshold, output a labeling request containing the current distribution features.

[0095] When the comparison results indicate that the maximum similarity is insufficient to trust any known pattern, this step does not opt ​​to output a low-confidence recognition result as a compromise. Instead, it proactively issues a request for human collaboration, namely a labeling request. This labeling request includes the distribution features that cannot be effectively identified at present, which is the target curve or its data representation reflecting the change of the average bit error rate with the power frequency phase in step S103. The labeling request can be presented to domain experts or maintenance personnel through human-computer interaction interfaces, backend monitoring terminals, or mobile applications, requesting them to review the attached distribution features based on their own experience and auxiliary information and provide a correct judgment on the discharge type.

[0096] S107: Retrieve the pattern tag returned in response to the tagging request.

[0097] After a labeling request is sent, the distribution characteristics attached to the request can be manually reviewed. Based on the morphology, possible on-site inspection results, or historical records, the true category of the discharge phenomenon can be determined, and this determination is returned as a pattern label. The pattern label is usually a predefined category name or a newly added custom category name that reflects the physical nature of the discharge or the source of interference. The system receives the returned pattern label through the communication interface, thereby establishing the correct category labeling relationship for previously unidentifiable distribution characteristics.

[0098] S108: Add the current distribution features and mode labels as new standard phase distribution features to the standard phase distribution features of multiple preset partial discharge modes.

[0099] After obtaining the returned pattern labels, the previously unrecognizable distribution features become a set of valuable data with authoritative category labels. This step directly adds this set of paired data, i.e., the current distribution features, along with their pattern labels, as a new standard phase distribution feature record to the standard phase distribution feature library stored internally by the system.

[0100] This process expands the standard library from containing only a few preset patterns to now also cover this newly emerging discharge type. In the future, when similar partial discharge events are detected again, the system can accurately identify them as this pattern directly by comparing them with the newly added standard features during the similarity matching stage, without requiring manual intervention again.

[0101] Specifically, the system can obtain the existing set of standard phase distribution features corresponding to the pattern label. If the set is empty, the current distribution feature and the pattern label are added to the set. Otherwise, the similarity between the current distribution feature and each existing standard phase distribution feature in the set is determined. When all similarities are less than a preset addition threshold, the current distribution feature and the pattern label are added to the set.

[0102] A deduplication strategy based on similarity thresholds ensures the compactness and representativeness of the standard library. Without screening new samples, long-term operation could lead to a large accumulation of highly similar curves for the same discharge pattern, consuming storage resources, increasing similarity calculation time, and potentially reducing matching accuracy due to the introduction of outliers. By setting an appropriate addition threshold, only distribution features with sufficient novelty are added to the library, maintaining its diversity and clear classification boundaries, which is beneficial for the long-term stable operation of the system.

[0103] To further illustrate the technical effects of this invention, a field verification case is provided: In a 10kV ring main unit, the method of this invention was implemented using an existing HPLC communication module to continuously monitor artificially set air gap discharges. Before averaging, the single-cycle bit error rate fluctuations were irregular, and the signal-to-noise ratio was only about 3dB. After averaging for 100 power frequency cycles, the phase distribution curve showed a clear "double-peak" characteristic, with the double-peak phases located near 60° on the voltage rise edge and 240° on the fall edge, respectively, which highly matched the PRPD spectrum of typical internal air gap discharges. At this point, the signal-to-noise ratio improved to about 19dB. The identification results were consistent with those recorded by a temporarily installed high-frequency CT partial discharge detector.

[0104] Please see Figure 2 Based on the same inventive concept, embodiments of this application also propose a ring network enclosure partial discharge pattern recognition system, including a control terminal and a ring network enclosure, wherein a carrier communication node is installed inside the ring network enclosure, and the system includes: The first acquisition module is used to acquire test data transmitted between carrier communication nodes and determine the fluctuation data of bit error rate over time based on the test data. The second acquisition module is used to acquire the phase synchronization relationship between the fluctuation data and the power frequency voltage; The determination module is used to determine the average bit error rate by aligning the fluctuation data of multiple power frequency cycles according to the power frequency phase based on the phase synchronization relationship and then superimposing and averaging them, and to determine the distribution characteristics of the average bit error rate as the power frequency phase changes. The third acquisition module is used to acquire the similarity between the distribution characteristics and the standard phase distribution characteristics of multiple preset partial discharge modes, and to determine the current partial discharge mode of the ring network box based on the multiple similarities.

[0105] Based on the same inventive concept, embodiments of this application also propose an electronic device, which includes: At least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the ring network box partial discharge pattern recognition method of the present application embodiments.

[0106] In addition, to achieve the above objectives, embodiments of this application also propose a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the ring network box partial discharge pattern recognition method of embodiments of this application.

[0107] The following is a detailed introduction to each component of the processing equipment: The processor is the control center of the processing device. It can be a single processor or a collective term for multiple processing elements. For example, the processor can be one or more central processing units (CPUs), application-specific integrated circuits (ASICs), or one or more integrated circuits configured to implement embodiments of the present invention, such as one or more digital signal processors (DSPs), or one or more field-programmable gate arrays (FPGAs).

[0108] Alternatively, the processor can perform various functions of the processing device by running or executing software programs stored in memory and by calling data stored in memory.

[0109] The memory is used to store the software program that executes the solution of the present invention, and the execution is controlled by the processor. The specific implementation method can be referred to the above method embodiment, and will not be repeated here.

[0110] Optionally, the memory can be read-only memory (ROM) or other types of static storage devices capable of storing static information and instructions, random access memory (RAM) or other types of dynamic storage devices capable of storing information and instructions, or electrically erasable programmable read-only memory (EEPROM), compact disc read-only memory (CD-ROM) or other optical disc storage, optical disc storage (including compressed optical discs, laser discs, optical discs, digital universal optical discs, Blu-ray discs, etc.), magnetic disk storage media or other magnetic storage devices, or any other medium capable of carrying or storing desired program code in the form of instructions or data structures and accessible by a computer, but not limited thereto. The memory can be integrated with the processor or exist independently and coupled to the processor through the interface circuit of the processing device; the embodiments of the present invention do not specifically limit this.

[0111] A transceiver is used to communicate with network devices or with terminal devices.

[0112] Optionally, the transceiver may include a receiver and a transmitter. The receiver is used to implement the receiving function, and the transmitter is used to implement the sending function.

[0113] Optionally, the transceiver can be integrated with the processor or exist independently and coupled to the processor through the router's interface circuit. This embodiment of the invention does not specifically limit this.

[0114] Figure 3 This is a schematic diagram of the structure of a processing device provided in an embodiment of the present invention. Exemplarily, this processing device may be a network device, or a chip (system) or other component or assembly that can be disposed in a network device. Figure 3 As shown, the processing device 300 may include a processor 301. Optionally, the processing device 300 may also include a memory 302 and / or a transceiver 303. The processor 301 is coupled to the memory 302 and the transceiver 303, for example, via a communication bus.

[0115] The following is combined Figure 3 A detailed description of each component of the processing equipment 300 is provided below: The processor 301 is the control center of the processing device 300. It can be a single processor or a collective term for multiple processing elements. For example, the processor 301 can be one or more central processing units (CPUs), application-specific integrated circuits (ASICs), or one or more integrated circuits configured to implement embodiments of the present invention, such as one or more digital signal processors (DSPs), or one or more field-programmable gate arrays (FPGAs).

[0116] Optionally, the processor 301 can perform various functions of the processing device 300 by running or executing software programs stored in the memory 302 and by calling data stored in the memory 302.

[0117] In a specific implementation, as one example, processor 301 may include one or more CPUs, for example... Figure 3 CPU0 and CPU1 are shown in the diagram.

[0118] In a specific implementation, as one embodiment, the processing device 300 may also include multiple processors. Each of these processors may be a single-core processor (single-CPU) or a multi-core processor (multi-CPU). Here, a processor may refer to one or more devices, circuits, and / or processing cores used for processing data (e.g., computer program instructions).

[0119] The memory 302 is used to store the software program that executes the solution of the present invention, and the processor 301 controls its execution. The specific implementation method can be referred to the above method embodiment, and will not be repeated here.

[0120] Optionally, the memory 302 may be a read-only memory (ROM) or other type of static storage device capable of storing static information and instructions, random access memory (RAM) or other type of dynamic storage device capable of storing information and instructions, or electrically erasable programmable read-only memory (EEPROM), compact disc read-only memory (CD-ROM) or other optical disc storage, optical disc storage (including compressed optical discs, laser discs, optical discs, digital universal optical discs, Blu-ray discs, etc.), magnetic disk storage media or other magnetic storage devices, or any other medium capable of carrying or storing desired program code in the form of instructions or data structures and accessible by a computer, but not limited thereto. The memory 302 may be integrated with the processor 301 or exist independently, and may be connected via the interface circuit of the processing device 300. Figure 3 (Not shown in the image) is coupled to processor 301, and this embodiment of the invention does not specifically limit this.

[0121] Transceiver 303 is used for communication with other processing devices. For example, if processing device 300 is a terminal, transceiver 303 can be used to communicate with a network device or with another terminal device. As another example, if processing device 300 is a network device, transceiver 303 can be used to communicate with a terminal or with another network device.

[0122] Alternatively, transceiver 303 may include a receiver and a transmitter. Figure 3 (Not shown separately). The receiver is used to implement the receiving function, and the transmitter is used to implement the transmitting function.

[0123] Optionally, the transceiver 303 can be integrated with the processor 301, or it can exist independently and be connected via the interface circuit of the processing device 300. Figure 3 (Not shown in the image) is coupled to processor 301, and this embodiment of the invention does not specifically limit this.

[0124] Understandable Figure 3 The structure of the processing device 300 shown does not constitute a limitation on the processing device. The actual processing device may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0125] Furthermore, the technical effects of the processing device 300 can be referred to the technical effects of the methods in the above-described method embodiments, and will not be repeated here.

[0126] It should be understood that the processor in the embodiments of the present invention can be a central processing unit (CPU), or it can be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor.

[0127] It should also be understood that the memory in the embodiments of the present invention can be volatile memory or non-volatile memory, or may include both volatile and non-volatile memory. The non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. The volatile memory can be random access memory (RAM), which is used as an external cache. By way of example, but not limitation, many forms of random access memory (RAM) are available, such as static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDRSDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous linked dynamic random access memory (SLDRAM), and direct rambus RAM (DRRAM). The above embodiments can be implemented, in whole or in part, by software, hardware (such as circuitry), firmware, or any other combination thereof. When implemented in software, the above embodiments can be implemented, in whole or in part, as a computer program product. A computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the flow or function according to embodiments of the present invention is produced. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. Computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, computer instructions can be transmitted from one website, computer, server, or data center to another via wired (e.g., infrared, wireless, microwave, etc.) means. A computer-readable storage medium can be any available medium that a computer can access, or a data storage device such as a server or data center that includes one or more sets of available media. Available media can be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., DVDs), or semiconductor media. Semiconductor media can be solid-state drives (SSDs).

[0128] It should be understood that the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. A and B can be singular or plural. Additionally, the character " / " in this article generally indicates an "or" relationship between the preceding and following related objects, but it can also represent an "and / or" relationship. Please refer to the context for a more accurate understanding.

[0129] In this invention, "at least one" refers to one or more items, and "more than one" refers to two or more items. "At least one of the following" or similar expressions refer to any combination of these items, including any combination of a single item or a plurality of items. For example, at least one of a, b, or c can represent: a, b, c, ab, ac, bc, or abc, where a, b, and c can be single or multiple. It should be understood that in various embodiments of this invention, the sequence number of the above processes does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this invention.

[0130] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

Claims

1. A method for partial discharge pattern recognition in a ring network box, characterized in that, A partial discharge pattern recognition system for a ring network enclosure is provided. The system includes a control terminal and a ring network enclosure, wherein a carrier communication node is installed inside the ring network enclosure. The method is executed by the control terminal and includes: Acquire test data transmitted between the carrier communication nodes, and determine the fluctuation data of the bit error rate over time based on the test data; Obtain the phase synchronization relationship between the fluctuation data and the power frequency voltage; Based on the phase synchronization relationship, the fluctuation data of multiple power frequency cycles are aligned according to the power frequency phase and then superimposed and averaged to determine the average bit error rate, and the distribution characteristics of the average bit error rate as the power frequency phase changes are determined. The similarity between the distribution characteristics and the standard phase distribution characteristics of multiple preset partial discharge modes is obtained, and the current partial discharge mode of the ring network box is determined based on the multiple similarities.

2. The method for partial discharge pattern recognition of a ring network box according to claim 1, characterized in that, The carrier communication node includes a first carrier communication node and a second carrier communication node. Test data transmitted between the carrier communication nodes is acquired, and based on the test data, fluctuation data of the bit error rate over time is determined, including: Instruct the first carrier communication node to send a test frame containing the target bit sequence to the second carrier communication node; The bit sequence obtained by the second carrier communication node is compared bit by bit with the target bit sequence stored locally to determine the number of erroneous bits within a preset time window, and the bit error rate of the preset time window is obtained. The fluctuation data is determined based on the bit error rate of multiple preset time windows.

3. The method for partial discharge pattern recognition of a ring network box according to claim 1, characterized in that, Obtaining the phase synchronization relationship between the fluctuation data and the power frequency voltage includes: The power frequency voltage signal is obtained based on the carrier communication node; Zero-crossing detection is performed on the power frequency voltage signal, and the pulse sequence at the start of the power frequency cycle is determined; The timestamps of the fluctuation data are compared with the pulse sequence; Based on the comparison results, the correspondence between the bit error rate and the power frequency phase at each moment is determined.

4. The method for partial discharge pattern recognition of a ring network box according to claim 3, characterized in that, Based on the phase synchronization relationship, the fluctuation data of multiple power frequency cycles are aligned according to the power frequency phase and then superimposed and averaged to determine the average bit error rate, and the distribution characteristics of the average bit error rate as a function of the power frequency phase are determined, including: Each power frequency cycle is divided into M phase intervals, where M is an integer greater than 1; Based on the correspondence, the bit error rate values ​​of multiple power frequency cycles that fall into the same phase interval are averaged to determine the average bit error rate of the corresponding phase interval; A target curve is constructed with the power frequency phase as the horizontal axis and the average bit error rate as the vertical axis, and the target curve is used as the distribution feature.

5. The method for partial discharge pattern recognition of a ring network box according to claim 4, characterized in that, Obtaining the similarity between the distribution characteristics and the standard phase distribution characteristics of multiple preset partial discharge modes, and determining the current partial discharge mode of the ring network box based on the multiple similarities, includes: The target curve is cyclically shifted along the power frequency phase axis with a preset phase step to obtain multiple shift curves, each of which corresponds to a shift phase amount; The sum of the absolute values ​​of the differences between the standard curve and the shift curve of each preset partial discharge mode in the M phase intervals is determined, and the shift curve with the smallest sum of absolute values ​​is determined as the target shift curve. Obtain the shift phase amount corresponding to the target shift curve as the target shift phase amount; Based on the minimum value of the sum of the absolute values ​​and the matching of the target shift phase amount with the preset shift range corresponding to the preset partial discharge mode, the similarity between the distribution characteristics and the standard phase distribution characteristics of the preset partial discharge mode is determined.

6. The method for partial discharge pattern recognition of a ring network box according to claim 5, characterized in that, Based on the minimum sum of the absolute values ​​and the matching between the target shift phase amount and the preset shift range corresponding to the preset partial discharge mode, the similarity between the distribution characteristics and the standard phase distribution characteristics of the preset partial discharge mode is determined, including: Obtain a preset maximum allowable difference, determine the difference between the maximum allowable difference and the minimum value of the sum of the absolute values, and use the difference as the first similarity component, wherein when the difference is less than zero, the first similarity component is set to zero; When the target shift phase amount falls within the preset shift range, the second similarity component is set to 1; otherwise, it is set to 0. The product of the first similarity component and the second similarity component is taken as the similarity.

7. The method for partial discharge pattern recognition of a ring network box according to claim 1, characterized in that, After determining the current partial discharge mode of the ring network box based on multiple similarities, the method further includes: The maximum value among the multiple similarities is compared with a preset recognition threshold; When the comparison result shows that the maximum value is less than the preset recognition threshold, a labeling request containing the current distribution features is output. Obtain the pattern label returned in response to the labeling request; The current distribution features and the mode label are added as new standard phase distribution features to the standard phase distribution features of the multiple preset partial discharge modes.

8. The method for partial discharge pattern recognition of a ring network box according to claim 7, characterized in that, The current distribution characteristics and the mode label are used as new standard phase distribution characteristics to supplement the standard phase distribution characteristics of the multiple preset partial discharge modes, including: Obtain the existing standard phase distribution feature set corresponding to the pattern label; If the set is empty, then the current distribution features and the pattern labels are added to the set; Otherwise, determine the similarity between the current distribution feature and each existing standard phase distribution feature in the set; When all the similarities are less than the preset addition threshold, the current distribution features and the pattern labels are added to the set.

9. A partial discharge pattern recognition system for a ring network box, characterized in that, The system includes a control terminal and a ring network enclosure, wherein a carrier communication node is installed within the ring network enclosure. The first acquisition module is used to acquire test data transmitted between the carrier communication nodes and determine the fluctuation data of the bit error rate over time based on the test data. The second acquisition module is used to acquire the phase synchronization relationship between the fluctuation data and the power frequency voltage; Based on the phase synchronization relationship, the fluctuation data of multiple power frequency cycles are aligned according to the power frequency phase and then superimposed and averaged to determine the average bit error rate, and the distribution characteristics of the average bit error rate as the power frequency phase changes are determined. The third acquisition module is used to acquire the similarity between the distribution characteristics and the standard phase distribution characteristics of multiple preset partial discharge modes, and to determine the current partial discharge mode of the ring network box based on the multiple similarities.

10. A processing apparatus, characterized in that, include: At least one processor; And, a memory communicatively connected to at least one of the processors; The memory stores instructions that can be executed by at least one of the processors, which are executed by at least one of the processors to enable at least one of the processors to perform a partial discharge pattern recognition method for a ring network box as described in any one of claims 1-8.

Images