A protection method and system for a split-type on-load tap-changer

Abstract

The present application discloses a protection method and system for a split type on-load tap-changer. The method includes extracting spectral feature data by extracting the signal amplitude change data on the secondary side of a current transformer; extracting abnormal fluctuation feature data by extracting the switching delay data of a secondary oil tank; obtaining a current anomaly eigenvalue based on the spectral feature data and the abnormal fluctuation feature data; comparing the current anomaly eigenvalue with a first anomaly feature threshold. If the result is a preliminary disconnection fault, the peak short-circuit current is obtained; a correction coefficient for the first anomaly feature threshold is obtained based on the peak short-circuit current, and the first anomaly feature threshold is corrected based on the correction coefficient. A secondary determination is made on the current anomaly eigenvalue based on the obtained second anomaly feature threshold to obtain a second judgment result, and a fault handling strategy generated by the second judgment result is executed. The method of the present application improves the accuracy of fault detection for the tap-changer.

Description

Technical Field

[0001] This application relates to the technical field of fault detection, and particularly to a protection method and system for a split type on-load tap-changer. Background Art

[0002] The on-load tap-changer of the UHV DC transmission converter transformer compensates for the AC system voltage change by dynamically adjusting the tap position to ensure stable power transmission of the DC system. Its structure is divided into an integral type and a split type, both of which are composed of a switch selection part and a switching part. Among them, the switch selection part of the split type on-load tap-changer is installed in the main tank of the converter transformer, and the switch switching part is installed in the auxiliary tank of the converter transformer. The main and auxiliary tanks are physically connected through a flange, and electrically connected through the bushing installed in the flange.

[0003] In the prior art, the protection method for the split type on-load tap-changer is to preset a fixed abnormal feature threshold to detect specific parameters in real time, and it cannot be adjusted according to the real-time operating conditions of the converter transformer. This solution in the prior art is difficult to accurately distinguish normal operation fluctuations from real fault features, and is prone to false positives or false negatives, thus affecting the accurate detection of tap-changer faults. Summary of the Invention

[0004] This application provides a protection method and system for a split type on-load tap-changer to solve the technical problem of how to improve the existing protection method for tap-changers, and achieve the effect of improving the accuracy of tap-changer fault detection.

[0005] To solve the above technical problems, an embodiment of this application provides a protection method for a split type on-load tap-changer, which is applied to a split type on-load tap-changer including a current transformer and an auxiliary tank. The protection method includes:

[0006] Performing feature extraction on the signal amplitude change data on the secondary side of the obtained current transformer to obtain spectral feature data reflecting the signal frequency characteristics; performing feature extraction on the obtained switching delay data of the auxiliary tank to obtain abnormal fluctuation feature data reflecting the change trend of the delay time;

[0007] Performing weighted processing on the spectral feature data and the abnormal fluctuation feature data to obtain a current abnormal feature value reflecting the open-circuit fault of the current transformer;

[0008] Comparing the current abnormal feature value with a preset first abnormal feature threshold respectively to obtain a first judgment result, wherein the first abnormal feature threshold is designed to be dynamically adjusted according to the operating conditions of the split type on-load tap-changer;

[0009] If the first judgment result is a preliminary disconnection fault of the current transformer, obtain the peak short-circuit current flowing through the split on-load tap-changer corresponding to the current anomaly characteristic value;

[0010] Obtain a correction coefficient for the first anomaly characteristic threshold according to the ratio relationship between the peak short-circuit current and the rated current of the split on-load tap-changer, and correct the first anomaly characteristic threshold based on the correction coefficient to obtain a second anomaly characteristic threshold;

[0011] Based on the second anomaly characteristic threshold, perform a secondary determination of the current transformer disconnection fault on the current anomaly characteristic value to obtain a second judgment result reflecting the true fault of the current transformer, and execute the fault handling strategy generated by the second judgment result.

[0012] As one of the preferred solutions, the feature extraction of the acquired signal amplitude change data on the secondary side of the current transformer to obtain spectral feature data reflecting the signal frequency characteristics includes:

[0013] Perform segmentation processing on the signal amplitude change data through a sliding time window method to obtain the signal amplitude extreme values within a preset time window;

[0014] Process the signal amplitude extreme values of adjacent time windows through a differential calculation method to obtain the signal amplitude change rate per unit time;

[0015] Perform Fourier transform on the signal amplitude change data through a frequency domain conversion method to obtain spectral feature data reflecting the signal frequency characteristics.

[0016] As one of the preferred solutions, the feature extraction of the acquired switching delay data of the auxiliary oil tank to obtain abnormal fluctuation feature data reflecting the change trend of the delay time includes:

[0017] Perform an average calculation based on the acquired historical switching delay data of the auxiliary oil tank to obtain the historical average switching delay time;

[0018] Perform a deviation calculation on the switching delay time and the historical average switching delay time through a deviation calculation method to obtain a switching delay deviation rate;

[0019] Analyze the switching delay deviation rates of consecutive multiple switching cycles through a trend analysis method to obtain abnormal fluctuation feature data reflecting the change trend of the delay time.

[0020] As one of the preferred solutions, the weighted processing of the spectral feature data and the abnormal fluctuation feature data to obtain a current anomaly characteristic value reflecting the disconnection fault of the current transformer includes:

[0021] Calculate a first correlation coefficient between the spectrum feature data and the current transformer open - circuit fault, and calculate a second correlation coefficient between the abnormal fluctuation feature data and the current transformer open - circuit fault. The correlation coefficient reflects the degree of association between the spectrum feature data, the abnormal fluctuation feature data and the current transformer open - circuit fault;

[0022] Allocate corresponding weights to the spectrum feature data and the abnormal fluctuation feature data according to the calculation results of the correlation coefficient;

[0023] Multiply the spectrum feature data and the abnormal fluctuation feature data by their corresponding weights respectively, and add the obtained calculation results to generate a current anomaly feature value reflecting the current transformer open - circuit fault.

[0024] As one of the preferred solutions, before comparing the current anomaly feature value with a preset first anomaly feature threshold respectively, it further includes:

[0025] Obtain the operating parameters of the split - type on - load tap - changer, where the operating parameters include load rate, operating time and ambient temperature;

[0026] Analyze the operating parameters to obtain the operating conditions of the split - type on - load tap - changer, where the operating conditions include light - load condition, heavy - load condition and high - temperature condition;

[0027] Match the operating conditions with a pre - constructed abnormal feature threshold database, and obtain the first anomaly feature threshold according to the matching result.

[0028] As one of the preferred solutions, the obtaining of the short - circuit current peak value flowing through the split - type on - load tap - changer corresponding to the current anomaly feature value includes:

[0029] Based on the current sensor to collect the current data flowing through the split - type on - load tap - changer, obtain an original data sequence containing short - circuit current information;

[0030] Perform filtering processing on the original data sequence to generate a smooth current data curve;

[0031] Detect the peak value of the smooth current data curve according to the difference method, identify the peak points in the curve, analyze multiple consecutive peak points, and generate the short - circuit current peak value based on the analysis result.

[0032] As one of the preferred solutions, the obtaining of the correction coefficient of the first anomaly feature threshold according to the proportional relationship between the short - circuit current peak value and the rated current of the split - type on - load tap - changer includes:

[0033] Obtain the historical peak short-circuit current flowing through the split on-load tap-changer, and calculate the first proportional relationship data between the historical peak short-circuit current and the rated current of the split on-load tap-changer;

[0034] Divide the first proportional relationship data into intervals to obtain multiple proportional intervals, and construct a corresponding correction coefficient mapping table for each proportional interval, where the correction coefficient mapping table reflects the correction coefficients corresponding to different proportional intervals;

[0035] Match the proportional relationship with the correction coefficient mapping table to obtain the correction coefficient corresponding to the proportional relationship.

[0036] As one of the preferred solutions, implementing the fault handling strategy generated by the second judgment result includes:

[0037] Generate a fault level according to the amplitude by which the current abnormal characteristic value in the second judgment result exceeds the second abnormal characteristic threshold, where the types of the fault level include minor fault, moderate fault, and severe fault;

[0038] If the fault level is the minor fault, detect the contact resistance of the secondary terminal of the current transformer by the loop self-check method, and simultaneously send a fault warning to the local monitoring system;

[0039] If the fault level is the moderate fault, trigger the single-pole locking of the split on-load tap-changer, cut off the fault-phase current path, and simultaneously start the standby current transformer signal switching logic;

[0040] If the fault level is the severe fault, immediately cut off the power supply of the split on-load tap-changer and send an emergency signal to the dispatching center.

[0041] As one of the preferred solutions, after generating the fault level according to the amplitude by which the current abnormal characteristic value in the second judgment result exceeds the second abnormal characteristic threshold, it further includes:

[0042] Calculate the aging coefficient according to the factory parameters and the cumulative operation time of the current transformer. If the aging coefficient is greater than the preset threshold and the moderate fault or the severe fault appears continuously, perform secondary calibration of the fault level;

[0043] Among them, the secondary calibration includes: synchronously collecting the vibration signal data at the connection flange of the split on-load tap-changer and the waveform data of the oil-gap insulation breakdown, analyzing the vibration signal data and the waveform data, and performing secondary calibration of the fault level according to the analysis results.

[0044] Compared with the prior art, the beneficial effects of the embodiments of the present application are at least one of the following:

[0045] 1) Through multi-dimensional feature fusion and dynamic threshold adjustment, this application weights parameters such as the signal frequency characteristics of current transformers and the trend of auxiliary oil tank switching delay, and combines the operating conditions and short-circuit current to real-time correct the judgment threshold, effectively solving the problem of misjudgment of single parameters in traditional protection methods and greatly improving the accuracy of disconnection fault judgment.

[0046] 2) Based on the adaptive correction mechanism of short-circuit current ratio, this application can dynamically optimize the protection threshold according to the actual operating impact load of the equipment, avoiding protection failure or overreaction in high-load or short-circuit scenarios, significantly enhancing the operating reliability of the split-type on-load tap-changer under complex conditions, extending the equipment life and reducing the operation and maintenance costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 is a schematic flow chart of the protection method for the split-type on-load tap-changer in one embodiment of this application;

[0048] Figure 2 is a schematic structural diagram of the split-type on-load tap-changer in one embodiment of this application;

[0049] Figure 3 is a schematic structural diagram of the connecting flange in one embodiment of this application;

[0050] Figure 4 is a schematic diagram of the protection system for the split-type on-load tap-changer in one embodiment of this application;

[0051] REFERENCE SIGNS:

[0052] Among them, 1, switch switching part; 2, switch selection part; 3, oil-oil bushing; 4, connecting flange; 5, current transformer; 6, terminal post, among which, H0, H1, H2, H3, H4 are different terminal posts; 11, acquisition module; 12, fusion module; 13, comparison module; 14, judgment module; 15, correction module; 16, execution module. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0053] Next, the technical solutions in the embodiments of this application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of this application. Obviously, the described embodiments are only a part of the embodiments of this application, rather than all the embodiments. The purpose of providing these embodiments is to make the disclosure content of this application more thorough and comprehensive. Based on the embodiments in this application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of this application.

[0054] In the description of this application, the terms "first", "second", "third", etc. are only used for descriptive purposes and cannot be construed as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, features defined with "first", "second", "third", etc. may explicitly or implicitly include one or more of such features. In the description of this application, unless otherwise stated, the meaning of "a plurality" is two or more than two.

[0055] In the description of this application, it should be noted that, unless otherwise clearly specified and limited, the terms "installed", "connected", "coupled" should be understood in a broad sense. For example, it may be a fixed connection, a detachable connection, or an integral connection; it may be a mechanical connection or an electrical connection; it may be directly connected or indirectly connected through an intermediate medium, and it may be the communication inside two components. The terms "vertical", "horizontal", "left", "right", "upper", "lower" and similar expressions used herein are only for the purpose of illustration and do not indicate or imply that the indicated device or element must have a specific orientation, be constructed and operated in a specific orientation, and thus cannot be construed as a limitation to this application. The term "and / or" used herein includes any and all combinations of one or more of the related listed items. For those of ordinary skill in the art, the specific meanings of the above terms in this application can be understood according to specific circumstances.

[0056] In the description of this application, it should be noted that, unless otherwise defined, all technical and scientific terms used in this application have the same meaning as commonly understood by those of ordinary skill in the technical field to which this application belongs. The terms used in the specification of this application are only for the purpose of describing specific embodiments and are not intended to limit this application. For those of ordinary skill in the art, the specific meanings of the above terms in this application can be understood according to specific circumstances.

[0057] An embodiment of this application provides a protection method for a split on-load tap-changer. Specifically, please refer to Figure 1 , Figure 1 which shows a schematic flowchart of the protection method for the split on-load tap-changer in one of the embodiments of this application, applied to a split on-load tap-changer including a current transformer and a secondary oil tank, and it includes steps S1 - S6:

[0058] S1: Extract features from the signal amplitude change data on the secondary side of the obtained current transformer to obtain spectral feature data reflecting the signal frequency characteristics; extract features from the obtained switching delay data of the secondary oil tank to obtain abnormal fluctuation feature data reflecting the change trend of the delay time;

[0059] Preferably, in an embodiment of the present application, the feature extraction of the signal amplitude change data obtained from the secondary side of the current transformer to obtain spectral feature data reflecting the signal frequency characteristics includes:

[0060] Segment the signal amplitude change data by the sliding time window method to obtain the signal amplitude extreme values within a preset time window;

[0061] Process the signal amplitude extreme values of adjacent time windows by the differential calculation method to obtain the signal amplitude change rate per unit time;

[0062] Perform Fourier transform on the signal amplitude change data by the frequency domain conversion method to obtain spectral feature data reflecting the signal frequency characteristics.

[0063] Among them, the sliding time window method means that the continuous signal amplitude change data is segmented into multiple overlapping time windows according to a fixed time length, so that the data in each window contains both the characteristics of the current moment and is associated with the data in the front and back windows. The time-domain signal is converted into a frequency-domain representation through Fourier transform, and the amplitude and phase information of different frequency components in the signal are extracted to form spectral feature data (such as the energy distribution of each frequency component).

[0064] It should be noted that the disconnection fault characteristics of the current signal may have short-term mutability (such as high-frequency noise) or persistence (such as signal disappearance). The sliding time window can capture the change trend of the signal at different time scales, avoid the one-sidedness of the data at a single moment, and improve the comprehensiveness of feature extraction. For example, the signal amplitude extreme value within the window can reflect the signal fluctuation range and provide a basis for subsequent frequency characteristic analysis.

[0065] Furthermore, the disconnection fault of the current transformer will cause abnormal frequency components to appear in the signal (such as high-frequency oscillation at the moment of disconnection or DC offset after signal disappearance). Frequency domain analysis can identify these characteristic frequencies and can locate the fault type more accurately compared with time domain analysis. For example, the disconnection fault may have abnormal energy concentration at specific frequencies (such as 100Hz, 200Hz), and the spectral characteristics can effectively distinguish it from normal load fluctuations.

[0066] In this embodiment, first, the continuous signal on the secondary side of the current transformer is segmented into multiple consecutive "time windows" according to a fixed time length of 100 milliseconds, and a 50% overlap rate is set between adjacent windows (for example, the first window covers 0-100ms, the second window covers 50-150ms, and so on) to ensure the continuity of signal characteristics.

[0067] For the signal data within each time window, two key parameters are extracted: the maximum signal amplitude: the highest instantaneous value of the current signal within the window, reflecting the peak fluctuation of the signal; and the minimum signal amplitude: the lowest instantaneous value of the current signal within the window, reflecting the trough fluctuation of the signal. In this way, the continuous signal is converted into multiple segmented data containing amplitude fluctuation information, providing a basis for subsequent analysis.

[0068] Perform Fourier transform on the signal data within each time window, converting the time-domain signal (with the horizontal axis being time and the vertical axis being the current amplitude) into a frequency-domain signal (with the horizontal axis being frequency and the vertical axis being the energy amplitude of the corresponding frequency component).

[0069] Specifically in the operation, the following frequency characteristics are mainly extracted:

[0070] 1) Fundamental frequency component (such as 50Hz or 60Hz, depending on the power grid frequency): The main frequency component during normal operation, and the change in its amplitude reflects the load size;

[0071] 2) Harmonic frequency components (such as 100Hz, 150Hz, etc.): Abnormal frequency components that may be caused by disconnection faults, and the sudden increase in their energy can be used as a fault warning signal;

[0072] 3) DC component: The component with a constant amplitude in the signal. If a significant DC offset appears (such as signal loss caused by disconnection), it can directly reflect signal abnormality. Finally, spectral characteristic data containing the energy values of each frequency component is generated (for example, "amplitude of 50Hz is 0.8A, amplitude of 100Hz is 0.3A").

[0073] Preferably, in an embodiment of the present application, the feature extraction of the obtained switching delay data of the auxiliary fuel tank to obtain abnormal fluctuation feature data reflecting the change trend of the delay time includes:

[0074] Based on the obtained historical switching delay data of the auxiliary fuel tank, average calculation is performed to obtain the historical average switching delay time;

[0075] Perform deviation calculation on the switching delay time and the historical average switching delay time through a deviation calculation method to obtain the switching delay deviation rate;

[0076] Analyze the switching delay deviation rates of consecutive multiple switching cycles through a trend analysis method to obtain abnormal fluctuation feature data reflecting the change trend of the delay time.

[0077] Among them, the deviation calculation method is to calculate the deviation ratio between the current switching delay time and the historical average delay time, so as to quantify the deviation degree of the current delay relative to the normal operation state of the device. The trend analysis method (abnormal fluctuation characteristics) is to statistically analyze the delay deviation rates of multiple consecutive switching cycles (such as moving average, linear regression), identify the change trend (rising, falling, fluctuating) of the deviation rate, and judge whether there is a systematic abnormality in the delay time.

[0078] It should be noted that the switching delay of the auxiliary fuel tank is affected by factors such as mechanical wear and lubrication state. It is difficult to judge whether it is abnormal with a single delay time (for example, the normal delay times in different gears are different). Through the calculation of the deviation rate, the influence of gear differences and device individual differences can be eliminated, and only the relative change of the delay time is concerned, effectively identifying progressive mechanical failures (such as the delay gradually increasing due to contact wear).

[0079] Furthermore, a single delay deviation may be caused by instantaneous interference (such as voltage fluctuation), while a trend change in multiple consecutive cycles (such as the deviation rate continuously rising by more than 5%) is more likely to indicate a device failure (such as mechanical component aging). Through trend analysis, random noise can be filtered, progressive failure characteristics can be captured, and early warning can be realized.

[0080] In this embodiment, first, a historical database is established to store the historical switching delay times of the auxiliary fuel tank in the same gear (for example, record the delay times of the past 100 gear switches in this gear). Calculate the historical average switching delay time of this gear: sum the historical data and divide it by the number of records, which is used as the reference value during normal operation. After collecting the current switching delay time in real time, calculate the delay deviation rate: that is, the degree of difference between the current delay time and the historical average value. For example, if the historical average delay is 200ms and the current delay is 220ms, the deviation rate is "(220 - 200) / 200×100% = 10%", indicating that the current delay is 10% longer than the historical average.

[0081] Continuously collect the delay deviation rates of the recent 10 switching cycles to form a time series (such as "5%, 7%, 9%, 11%, 13%"). Conduct trend analysis on this series:

[0082] 1) If the deviation rate shows a continuous upward trend (such as increasing continuously for more than 3 times, and the increase amplitude each time exceeds 2%), it indicates that the delay time is gradually getting longer, and there may be progressive mechanical failures such as contact wear and insufficient lubrication;

[0083] 2) If the deviation rate shows violent fluctuations (such as suddenly jumping from 5% to 20% and then falling back to 8%), it may be caused by instantaneous mechanical shock or signal interference, and other characteristics (such as the current signal spectrum) need to be combined for further judgment.

[0084] Finally, abnormal fluctuation characteristic data such as "continuous increase in delay deviation rate" and "abnormal fluctuation in delay deviation rate" are generated as auxiliary basis for fault determination.

[0085] S2: Perform weighted processing on the spectral characteristic data and the abnormal fluctuation characteristic data to obtain a current abnormal characteristic value reflecting the open-circuit fault of the current transformer;

[0086] Preferably, in an embodiment of the present application, the performing weighted processing on the spectral characteristic data and the abnormal fluctuation characteristic data to obtain a current abnormal characteristic value reflecting the open-circuit fault of the current transformer includes:

[0087] Calculate a first correlation coefficient between the spectral characteristic data and the open-circuit fault of the current transformer, and calculate a second correlation coefficient between the abnormal fluctuation characteristic data and the open-circuit fault of the current transformer. The correlation coefficient reflects the degree of association between the spectral characteristic data and the abnormal fluctuation characteristic data and the open-circuit fault of the current transformer;

[0088] Assign corresponding weights to the spectral characteristic data and the abnormal fluctuation characteristic data according to the calculation result of the correlation coefficient;

[0089] Multiply the spectral characteristic data and the abnormal fluctuation characteristic data by their corresponding weights respectively, and add the obtained calculation results to generate a current abnormal characteristic value reflecting the open-circuit fault of the current transformer.

[0090] Among them, the correlation coefficient refers to the Pearson correlation coefficient, which is an index to measure the degree of linear correlation between two variables, and its value range is [-1, 1]. A positive value indicates a positive correlation, a negative value indicates a negative correlation, and the closer the absolute value is to 1, the stronger the correlation. In this step, the "first correlation coefficient" reflects the degree of association between the spectral characteristic data and the "open-circuit fault of the current transformer" (for example, the higher the energy of a certain frequency component, the higher the probability of failure); the "second correlation coefficient" reflects the degree of association between the abnormal fluctuation characteristic data (such as the delay deviation rate of the auxiliary fuel tank) and the fault.

[0091] The weighted processing assigns corresponding weights (such as 0.7, 0.3) to each characteristic according to the degree of association (i.e., the correlation coefficient) between different characteristics and the fault, and generates a comprehensive index (current abnormal characteristic value) through weighted summation, so that this index can more comprehensively reflect the probability of failure.

[0092] It should be noted that the disconnection fault of the current transformer may simultaneously exhibit abnormal electrical signals (spectrum feature changes) and abnormal mechanical states (auxiliary tank switching delay changes). A single feature may be affected by noise interference or have blind spots (such as normal spectrum features but abnormal delays caused by mechanical jamming). For example, the spectrum feature shows high-frequency noise (possibly caused by external interference), but the auxiliary tank delay is normal. In this case, a single spectrum feature may lead to misjudgment; the auxiliary tank delay slightly increases (initial contact wear), but the spectrum feature has no obvious abnormality, and a single delay feature may miss early faults. Through weighted processing, the complementary information of the two types of features can be fused to form a more reliable comprehensive judgment basis.

[0093] Under different operating conditions, the characteristics and faults may change (such as spectrum features being more sensitive under high load and delay features being more sensitive under low load). By calculating the correlation coefficient in real time and assigning weights, different scenarios can be adapted to avoid the limitations of fixed weights.

[0094] In this embodiment, historical operation data is collected, including at least 1000 groups of spectrum features (such as the energy of each frequency component) and auxiliary tank delay deviation rates when there is no disconnection fault; and at least 200 groups of corresponding feature data within 10 seconds before the disconnection fault occurs (obtained by artificially injecting faults or historical fault recordings).

[0095] For each frequency component in the spectrum feature (such as the energy of 50Hz and 100Hz), calculate its correlation coefficient with "fault occurrence" (0 - normal, 1 - fault) respectively, and take the 3 frequency components with the strongest correlation with the fault as key spectrum features (for example, the correlation coefficient of 100Hz energy is 0.85 and that of 200Hz is 0.78); for the auxiliary tank delay deviation rate, calculate its correlation coefficient with "fault occurrence" (for example, 0.65). Data analysis tools (such as the pandas library in Python) can be used to calculate the correlation between the current feature and historical fault data in real time.

[0096] Normalize the correlation coefficients of the key spectrum features so that their sum is 1. The correlation coefficient (0.65) of the auxiliary tank delay deviation rate is taken as a single type of feature and forms the second type of weight with the total weight of the spectrum feature. To reflect the complementarity of electrical and mechanical features, the proportion of the total weight of the spectrum feature can be set to 60% - 80%, and the weight of the delay feature can be set to 20% - 40% (for example, the total spectrum weight is 0.7 and the delay weight is 0.3).

[0097] Multiply the key spectrum features (such as 100Hz energy and 200Hz energy) by their corresponding weights and sum them to obtain the spectrum comprehensive value; multiply the auxiliary tank delay deviation rate by its weight to obtain the delay feature value; the final current anomaly feature value = spectrum comprehensive value + delay feature value.

[0098] S3: Compare the current anomaly eigenvalue with a preset first anomaly feature threshold respectively to obtain a first judgment result, where the first anomaly feature threshold is designed to be dynamically adjusted according to the operating conditions of the split on-load tap-changer;

[0099] Preferably, in an embodiment of the present application, before comparing the current anomaly eigenvalue with the preset first anomaly feature threshold respectively, it further includes:

[0100] Obtain the operating parameters of the split on-load tap-changer, where the operating parameters include load rate, operating time, and ambient temperature;

[0101] Analyze the operating parameters to obtain the operating conditions of the split on-load tap-changer, where the operating conditions include light load condition, heavy load condition, and high temperature condition;

[0102] Match the operating conditions with a pre-constructed anomaly feature threshold database, and obtain the first anomaly feature threshold according to the matching result.

[0103] The operating condition refers to the comprehensive state of the split on-load tap-changer during operation, which is jointly determined by key parameters such as load rate (the percentage of the current load to the rated load, reflecting the working intensity), operating time (the cumulative power-on duration of the equipment, indirectly reflecting the aging degree), and ambient temperature (the real-time temperature at the equipment installation location, affecting the stability of electrical parameters). For example:

[0104] Light load condition: load rate < 30%, the equipment is operating at a low load, and the current fluctuation is small;

[0105] Heavy load condition: load rate > 80%, the equipment is operating at a high load, and the current changes violently and the heat generation increases;

[0106] High temperature condition: ambient temperature > 40 degrees, which may cause a decline in the performance of insulating materials or abnormal thermal expansion of mechanical components.

[0107] This application abandons the fixed threshold in the traditional protection method and instead retrieves the corresponding threshold from the pre-constructed threshold database according to the real-time operating conditions (light load / heavy load / high temperature, etc.), so that the judgment standard adapts to the current working state of the equipment.

[0108] Traditional protection systems use fixed thresholds (such as "Alarm when the signal change rate > 5A / s"), but in reality: during heavy loads, the natural current fluctuations are large, and the fixed threshold may lead to frequent misjudgments; during light loads, the fault characteristics are weak, and the fixed threshold may miss early faults; in high-temperature environments, the equipment parameters drift, and the fixed threshold cannot adapt to the performance changes. The dynamic adjustment of this application solves the "one-size-fits-all" problem of fixed thresholds through "operating condition - threshold matching". Exclusive thresholds are set for different operating conditions. For example, when the load is heavy, the signal change rate threshold is increased from 5A / s to 8A / s to avoid false alarms triggered by normal fluctuations under high loads.

[0109] In this embodiment, the load current is collected in real time through a current transformer, the power-on duration of the equipment is accumulated through the built-in timer of the equipment, accurate to the hour, and the ambient temperature is collected in real time through a temperature sensor installed near the equipment. And various collected data are preprocessed to filter out obvious outliers.

[0110] A three-level operating condition classification system is established as shown in Table 1.

[0111] Table 1

[0112]

[0113] When the parameters cross categories (such as a load rate of 75% and a temperature of 38°C), fuzzy logic is used for determination (such as "close to heavy load condition") to avoid threshold jumps caused by absolute classification.

[0114] Through laboratory simulations (such as high-temperature chamber tests, load bench tests) and on-site measurements, normal characteristic data (current anomaly characteristic values without faults) under different operating conditions are collected, and their distribution ranges (mean ± 3 times the standard deviation) are statistically analyzed as the basis for thresholds; the threshold setting principles are as follows:

[0115] Light load condition: Threshold = normal characteristic mean × 0.9 (lower the threshold to improve sensitivity);

[0116] Heavy load condition: Threshold = normal characteristic mean × 1.1 (raise the threshold to filter out normal fluctuations);

[0117] High-temperature condition: On the basis of the threshold for the corresponding load condition, multiply by a temperature correction factor (such as for every 10°C increase, the threshold increases by 5%);

[0118] Aging condition: According to historical data, for equipment with an operating time exceeding 10,000 hours, the threshold is increased by 10% to adapt to parameter drift.

[0119] Specifically, the current load rate, temperature, and operating time are obtained in real time; according to the preset operating condition classification rules, it is determined whether the current belongs to the "light load", "heavy load", "high temperature" or composite operating condition (such as "high temperature + heavy load"); the thresholds corresponding to the operating conditions are retrieved from the database. If there are multiple matching operating conditions (such as meeting both heavy load and high temperature at the same time), the maximum value of the thresholds is taken (to ensure that the protection system is not overly sensitive). Through the three-level architecture of "parameter acquisition - operating condition classification - threshold matching", the protection system can adjust the judgment criteria "according to local conditions", solving the problem of insufficient accuracy of traditional fixed thresholds under complex operating conditions.

[0120] S4: If the first judgment result is a preliminary disconnection fault of the current transformer, obtain the peak short-circuit current flowing through the split type on-load tap-changer corresponding to the current abnormal characteristic value;

[0121] Preferably, in an embodiment of the present application, the obtaining the peak short-circuit current flowing through the split type on-load tap-changer corresponding to the current abnormal characteristic value includes:

[0122] Based on the current sensor collecting the current data flowing through the split type on-load tap-changer, obtain the original data sequence containing short-circuit current information;

[0123] Perform filtering processing on the original data sequence to generate a smooth current data curve;

[0124] According to the difference method, perform peak detection on the smooth current data curve, identify the peak points in the curve, analyze multiple consecutive peak points, and generate the peak short-circuit current based on the analysis results.

[0125] Among them, when a short-circuit fault occurs, the current flowing through the split type on-load tap-changer will increase instantaneously, and the maximum value reached by the current during this instantaneous increase process is the peak short-circuit current, which reflects the maximum impact degree of the current in the circuit when the short-circuit fault occurs.

[0126] The original data sequence is a set of data formed by arranging the current data collected by the current sensor for the current flowing through the split type on-load tap-changer in chronological order. These data contain normal operating current and possible short-circuit current information, but there may be noise interference. The filtering processing includes processing the original data sequence to remove the noise and interference signals therein, making the data smoother, highlighting the useful current information, and facilitating subsequent analysis. The difference method is a mathematical analysis method that discovers the change trend and characteristics of the data by calculating the differences between adjacent data points in the data sequence. In current data processing, it can be used to detect the peak points in the current curve.

[0127] When the first judgment result is a preliminary disconnection fault of the current transformer, the peak short-circuit current is an important reference index. Different degrees of disconnection faults may be accompanied by different magnitudes of peak short-circuit currents. By obtaining the peak short-circuit current, the severity of the fault can be more accurately evaluated, providing a basis for subsequent fault handling strategies.

[0128] In this embodiment, a high-precision current sensor is used to collect the current data flowing through the split on-load tap-changer in real time. The current sensor needs to have the characteristics of fast response and high sensitivity to ensure that it can accurately capture the instantaneous changes of the short-circuit current. The collected data is arranged in chronological order to form an original data sequence containing short-circuit current information. During the collection process, it is necessary to ensure that the data collection frequency is high enough to completely record the change process of the current.

[0129] The collected original data sequence is filtered to remove the noise and interference signals therein. Specifically, common filtering algorithms such as moving average filtering or median filtering can be used. Moving average filtering calculates the average value of the data within a certain time window and replaces the data point at the center of the window with this average value to smooth the data curve; median filtering sorts the data within the window and takes the median value as the data point at the center of the window. Through the filtering process, a smooth current data curve is generated, making the characteristics of the short-circuit current more obvious.

[0130] The differential method is used to detect the peak value of the smoothed current data curve. The specific method is to calculate the difference between adjacent data points. When the difference changes from positive to negative, it indicates that a peak point may appear. Analyze the detected multiple consecutive peak points to determine whether these peak points conform to the characteristics of the peak short-circuit current. For example, check the magnitude of the peak, the occurrence time, and the correlation with the preliminary disconnection fault. By comprehensively analyzing these peak points, the peak short-circuit current is finally determined. If the detected peak point is extremely large and coincides with the time of the preliminary disconnection fault, then it can be determined as the peak short-circuit current.

[0131] S5: Obtain a correction coefficient of the first abnormal feature threshold according to the proportional relationship between the peak short-circuit current and the rated current of the split on-load tap-changer, and correct the first abnormal feature threshold based on the correction coefficient to obtain a second abnormal feature threshold;

[0132] Preferably, in an embodiment of the present application, the obtaining a correction coefficient of the first abnormal feature threshold according to the proportional relationship between the peak short-circuit current and the rated current of the split on-load tap-changer includes:

[0133] Obtain the historical short - circuit current peak value flowing through the split - type on - load tap - changer, and calculate the first proportional relationship data between the historical short - circuit current peak value and the rated current of the split - type on - load tap - changer;

[0134] Divide the range of the first proportional relationship data to obtain multiple proportional ranges, and construct a corresponding correction coefficient mapping table for each proportional range, where the correction coefficient mapping table reflects the correction coefficients corresponding to different proportional ranges;

[0135] Match the proportional relationship with the correction coefficient mapping table to obtain the correction coefficient corresponding to the proportional relationship.

[0136] Among them, the rated current is the current value that the split - type on - load tap - changer can withstand for a long time under normal operating conditions, and it is an important parameter for equipment design and selection. The first abnormal feature threshold is a critical value initially set according to the equipment operating conditions for judging whether the current transformer has a disconnection fault. The correction coefficient is a coefficient for adjusting the first abnormal feature threshold, which is determined according to the proportional relationship between the short - circuit current peak value and the rated current, and the purpose is to make the abnormal feature threshold more in line with the actual fault situation.

[0137] The second abnormal feature threshold is the new abnormal feature threshold obtained by correcting the first abnormal feature threshold through the correction coefficient, which is used for more accurate secondary determination of the current transformer disconnection fault. The proportional range is the different range intervals obtained by dividing the proportional relationship data between the historical short - circuit current peak value and the rated current, and each interval corresponds to a different correction coefficient. The correction coefficient mapping table is a table recording the relationship between different proportional ranges and the corresponding correction coefficients, and through this table, the corresponding correction coefficient can be quickly found according to the proportional relationship between the current short - circuit current peak value and the rated current.

[0138] It should be noted that the proportional relationship between the short - circuit current peak value and the rated current can reflect the severity of the fault and the actual working conditions. After initially determining that it is a current transformer disconnection fault, different short - circuit current situations may mean different degrees of complexity of the fault, and the original first abnormal feature threshold may no longer be applicable. Therefore, it is necessary to correct the first abnormal feature threshold according to this proportional relationship to improve the accuracy of fault determination.

[0139] In this embodiment, collect the historical short - circuit current peak value data when the split - type on - load tap - changer had short - circuit faults in the past. These data can be obtained from channels such as the equipment operation records and the fault monitoring system. Then, divide each historical short - circuit current peak value by the rated current of the split - type on - load tap - changer to obtain a series of proportional values, and these proportional values constitute the first proportional relationship data.

[0140] Analyze the first proportional relationship data and divide it into multiple proportional intervals according to the data distribution. For example, the proportional values can be divided into intervals such as 0 - 1 times, 1 - 2 times, 2 - 3 times, etc. in ascending order. For each proportional interval, set a corresponding correction coefficient based on experience or a large amount of experimental data. Organize the proportional intervals and the corresponding correction coefficients into a correction coefficient mapping table for convenient subsequent query.

[0141] Calculate the proportional relationship between the currently obtained peak short - circuit current and the rated current of the split - type on - load tap - changer. Compare this proportional relationship with the correction coefficient mapping table to find the proportional interval where this proportional relationship is located, so as to obtain the corresponding correction coefficient. For example, if the current proportional relationship is 1.5 times and it is in the 1 - 2 times proportional interval, find the correction coefficient corresponding to this interval from the mapping table. Multiply the obtained correction coefficient by the first abnormal feature threshold to get the second abnormal feature threshold. This new threshold takes into account the actual situation of the current short - circuit fault and can be used more accurately for the secondary determination of current transformer open - circuit faults.

[0142] S6: Based on the second abnormal feature threshold, perform a secondary determination of the current transformer open - circuit fault on the current abnormal feature value to obtain a second judgment result reflecting the true fault of the current transformer, and execute the fault handling strategy generated by the second judgment result.

[0143] Preferably, in an embodiment of the present application, the execution of the fault handling strategy generated by the second judgment result includes:

[0144] Generate a fault level according to the amplitude by which the current abnormal feature value exceeds the second abnormal feature threshold in the second judgment result, where the types of the fault levels include mild faults, moderate faults, and severe faults;

[0145] If the fault level is the mild fault, detect the contact resistance of the secondary terminal of the current transformer by the loop self - inspection method and simultaneously send a fault warning to the local monitoring system;

[0146] If the fault level is the moderate fault, trigger the single - pole locking of the split - type on - load tap - changer to cut off the current path of the faulty phase, and at the same time start the standby current transformer signal switching logic;

[0147] If the fault level is the severe fault, immediately cut off the power supply of the split - type on - load tap - changer and send an emergency signal to the dispatching center.

[0148] Specifically, compare the previously obtained current anomaly eigenvalue with the second anomaly threshold. If the current anomaly eigenvalue is less than or equal to the second anomaly threshold, it is determined that the current transformer has not experienced a real disconnection fault; if the current anomaly eigenvalue is greater than the second anomaly threshold, enter the fault level classification step.

[0149] Determine the fault level based on the extent to which the current anomaly eigenvalue exceeds the second anomaly threshold. For example, if the excess extent is small (such as exceeding 10% - 30%), it is determined as a minor fault; if the excess extent is moderate (such as exceeding 30% - 70%), it is determined as a medium fault; if the excess extent is large (such as exceeding 70%), it is determined as a severe fault.

[0150] When the fault level is a minor fault, use the loop self - inspection method to detect the contact resistance of the secondary - side connection terminals of the current transformer. A professional resistance measuring instrument can be used to detect whether the contact resistance between the connection terminals is within the normal range. At the same time, send a fault warning message to the local monitoring system to remind the operation and maintenance personnel to pay attention to the equipment status.

[0151] When the fault level is a medium fault, trigger the single - pole locking function of the split - type on - load tap - changer to cut off the current path of the faulty phase. This can be achieved by controlling the operating mechanism of the switch. At the same time, start the signal switching logic of the standby transformer to switch the measurement and protection signals to the standby transformer to ensure the normal operation of the system.

[0152] When the fault level is a severe fault, immediately cut off the power supply of the split - type on - load tap - changer to prevent the fault from further expanding and causing serious damage to the equipment and the system. At the same time, send an emergency signal to the dispatching center to notify the dispatching personnel to take corresponding measures.

[0153] Preferably, in an embodiment of the present application, after generating the fault level according to the extent to which the current anomaly eigenvalue exceeds the second anomaly threshold in the second judgment result, it further includes:

[0154] Calculate the aging coefficient based on the factory parameters and the cumulative operation time of the current transformer. If the aging coefficient is greater than the preset threshold and the medium fault or the severe fault appears continuously, perform a secondary calibration of the fault level;

[0155] Wherein, the secondary calibration includes: synchronously collecting the vibration signal data at the connection flange of the split - type on - load tap - changer and the waveform data of oil - gap insulation breakdown, analyzing the vibration signal data and the waveform data, and performing a secondary calibration of the fault level according to the analysis result.

[0156] The circuit self-checking method is a method for detecting the secondary-side wiring terminals of current transformers, and it determines whether the wiring is good by detecting the contact resistance. Single-pole locking means triggering a pole of the split-type on-load tap-changer for locking operation to cut off the current path of the faulty phase and prevent the further expansion of the fault. The spare current transformer signal switching logic means that when the main current transformer fails, the signal is automatically switched to the spare current transformer to ensure the normal operation of the system. The aging coefficient is a coefficient calculated based on the factory parameters and the cumulative operating time of the current transformer, and it is used to reflect the aging degree of the current transformer.

[0157] Secondary calibration includes further correcting and adjusting the fault level by collecting and analyzing the vibration signal data and oil-gap insulation breakdown waveform data of the split-type on-load tap-changer after considering the aging coefficient and the continuous occurrence of faults.

[0158] In this embodiment, the aging coefficient is calculated through a preset calculation formula according to the factory parameters of the current transformer (such as the type of insulating material, design life, etc.) and the cumulative operating time. For example, the aging coefficient can be in a proportional relationship with the cumulative operating time. If the aging coefficient is greater than the preset threshold and moderate or severe faults occur continuously, secondary calibration of the fault level is required.

[0159] Specifically, synchronously collect the vibration signal data at the connection flange of the split-type on-load tap-changer and the waveform data of the oil-gap insulation breakdown. Vibration sensors and waveform recorders can be used to collect this data. Analyze the collected data, such as through spectrum analysis, waveform feature extraction, etc., to judge the actual fault situation of the equipment. Secondary calibration of the previously divided fault level is performed according to the analysis results, which may increase or decrease the fault level, and then the corresponding fault handling strategy is re-executed according to the calibrated fault level.

[0160] To further explain the effectiveness and feasibility of this application, the following embodiments are used to further illustrate this application. As Figure 2 、 Figure 3 shown, where Figure 2 is a structural diagram of a split-type on-load tap-changer provided in an embodiment of this application, Figure 3 is a structural schematic diagram of a connection flange provided in an embodiment of this application.

[0161] As Figure 2 shown, the split-type on-load tap-changer mainly consists of a switch switching part 1 and a switch selection part 2. The switch switching part 1 contains several switching switches, and its function is to cut off or connect the load switch to achieve the on-off control of the circuit. The switch selection part 2 contains several tap selectors. During the tap position switching process, the tap selector will pre-connect the tap to be switched to ensure the smooth progress of the tap position switching.

[0162] The top of the switch switching part 1 is connected to the top of the switch selection part 2 through an oil-oil bushing 3. The bottom of the switch switching part 1 is connected to one end of a connection flange 4 through a current-carrying conductor connection, and the other end of the connection flange 4 is connected to the switch selection part 2 through a current-carrying conductor. A number of current transformers 5 are arranged inside the connection flange 4.

[0163] The switch switching part 1 is installed in the secondary oil tank of the converter transformer, and the switch selection part 2 is installed in the main oil tank of the converter transformer. The secondary oil tank and the main oil tank of the converter transformer are completely isolated, and this design helps to improve the safety and stability of the equipment operation.

[0164] As Figure 3 shown, a number of terminal posts 6 are arranged on the connection flange 4. In this embodiment, a total of five terminal posts 6 are arranged on the connection flange 4, namely H0, H1, H2, H3, and H4. Among them, the terminal post H0 is not connected to the current transformer 5, while each of the terminal posts H1, H2, H3, and H4 is equipped with a current transformer 5. The terminal posts H1 and H2 respectively correspond to the odd and even terminals of the first voltage regulating column of the converter transformer; the terminal posts H3 and H4 respectively correspond to the odd and even terminals of the second voltage regulating column of the converter transformer; the terminal post H0 is the neutral point lead-out terminal. It should be noted that the current directions of the terminal posts H1, H2, H3, and H4 are the same, the current direction of the terminal post H0 is opposite to that of other terminal posts, and the current magnitude of the terminal post H0 is the sum of the currents of the terminal posts H1 and H3 or H2 and H4.

[0165] The connection flange 4 is composed of a flange plate and a flange cover connected to the flange plate. The flange cover is made of epoxy resin, and an epoxy resin protective sleeve is arranged on each terminal post 6, which can effectively improve the insulation performance of the equipment. The current transformer 5 adopts an electromagnetic current transformer. Each of the terminal posts H1, H2, H3, and H4 is configured with 3 or 4 electromagnetic current transformers 5, and these transformers respectively send signals to three sets of control protection and measurement devices A, B, and C for comprehensive monitoring and protection.

[0166] Based on the four CT measuring points H0, H1, H2, H3, and H4 in the tap-changer, this application configures overcurrent protection for the tap-changer to quickly detect the inter-stage short-circuit fault of the tap-changer and prevent the expansion of the fault range. The protection scope of this protection method covers the switching switch area inside the switch oil tank, and can effectively prevent the direct occurrence of inter-stage short-circuit faults in the switching switch.

[0167] Another embodiment of this application provides a protection system for a split-type on-load tap-changer. Specifically, please refer to Figure 4 , Figure 4 which shows a schematic diagram of the protection system for a split-type on-load tap-changer in one of the embodiments of this application, applied to a split-type on-load tap-changer including a current transformer and a secondary oil tank, and it includes:

[0168] An acquisition module 11 is configured to extract features from the acquired signal amplitude change data of the secondary side of the current transformer to obtain spectral feature data reflecting the signal frequency characteristics; extract features from the acquired switching delay data of the auxiliary tank to obtain abnormal fluctuation feature data reflecting the change trend of the delay time.

[0169] A fusion module 12 is configured to perform weighted processing on the spectral feature data and the abnormal fluctuation feature data to obtain a current anomaly eigenvalue reflecting the open-circuit fault of the current transformer.

[0170] A comparison module 13 is configured to compare the current anomaly eigenvalue with a preset first anomaly feature threshold respectively to obtain a first judgment result, wherein the first anomaly feature threshold is designed to be dynamically adjusted according to the operating conditions of the split-type on-load tap-changer.

[0171] A judgment module 14 is configured to, if the first judgment result is a preliminary open-circuit fault of the current transformer, obtain the short-circuit current peak value flowing through the split-type on-load tap-changer corresponding to the current anomaly eigenvalue.

[0172] A correction module 15 is configured to obtain a correction coefficient of the first anomaly feature threshold according to the proportional relationship between the short-circuit current peak value and the rated current of the split-type on-load tap-changer, and correct the first anomaly feature threshold based on the correction coefficient to obtain a second anomaly feature threshold.

[0173] An execution module 16 is configured to perform a secondary determination of the open-circuit fault of the current transformer on the current anomaly eigenvalue based on the second anomaly feature threshold to obtain a second judgment result reflecting the true fault of the current transformer, and execute a fault handling strategy generated by the second judgment result.

[0174] Compared with the prior art, the beneficial effects of the embodiments of the present application are at least one of the following:

[0175] 1) Through multi-dimensional feature fusion and dynamic threshold adjustment, the present application performs weighted processing on parameters such as the signal frequency characteristics of the current transformer and the switching delay trend of the auxiliary tank, and combines the operating conditions and short-circuit current to correct the judgment threshold in real time, effectively solving the problem of misjudgment of single parameters in traditional protection methods and greatly improving the accuracy of open-circuit fault judgment.

[0176] 2) The adaptive correction mechanism based on the short-circuit current ratio of the present application can dynamically optimize the protection threshold according to the actual operating impact load of the equipment, avoid protection failure or overreaction in high-load or short-circuit scenarios, significantly enhance the operating reliability of the split-type on-load tap-changer under complex working conditions, extend the equipment life and reduce the operation and maintenance cost.

[0177] The above-described embodiments merely represent several implementation manners of the present application. The description thereof is relatively specific and detailed, but it should not be construed as a limitation on the patent scope of the present application. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present application, several modifications and improvements can still be made, and these all fall within the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the appended claims.

Claims

1. A protection method for a split type on-load tap-changer, characterized in that, Applied to a split-type on-load tap-changer including a current transformer and a secondary oil tank, the protection method includes: Performing feature extraction on the obtained signal amplitude change data on the secondary side of the current transformer to obtain spectral feature data reflecting the signal frequency characteristics; performing feature extraction on the obtained switching delay data of the secondary oil tank to obtain abnormal fluctuation feature data reflecting the change trend of the delay time; Performing weighted processing on the spectral feature data and the abnormal fluctuation feature data to obtain a current abnormal feature value reflecting the open-circuit fault of the current transformer; Comparing the current abnormal feature value with a preset first abnormal feature threshold respectively to obtain a first judgment result, wherein the first abnormal feature threshold is designed to be dynamically adjusted according to the operating conditions of the split-type on-load tap-changer; If the first judgment result is a preliminary open-circuit fault of the current transformer, obtaining the short-circuit current peak value flowing through the split-type on-load tap-changer corresponding to the current abnormal feature value; Obtaining a correction coefficient of the first abnormal feature threshold according to the proportional relationship between the short-circuit current peak value and the rated current of the split-type on-load tap-changer, and correcting the first abnormal feature threshold based on the correction coefficient to obtain a second abnormal feature threshold; Performing a secondary determination of the open-circuit fault of the current transformer on the current abnormal feature value based on the second abnormal feature threshold to obtain a second judgment result reflecting the true fault of the current transformer, and executing the fault handling strategy generated by the second judgment result.

2. The protection method of the split type on-load tap-changer according to claim 1, characterized in that, The performing feature extraction on the obtained signal amplitude change data on the secondary side of the current transformer to obtain spectral feature data reflecting the signal frequency characteristics includes: Performing segmentation processing on the signal amplitude change data by means of a sliding time window method to obtain signal amplitude extreme values within a preset time window; Processing the signal amplitude extreme values of adjacent time windows by means of a differential calculation method to obtain the signal amplitude change rate per unit time; Performing Fourier transform on the signal amplitude change data by means of a frequency domain conversion method to obtain spectral feature data reflecting the signal frequency characteristics.

3. The protection method of the split type on-load tap-changer according to claim 1, characterized in that, The performing feature extraction on the obtained switching delay data of the secondary oil tank to obtain abnormal fluctuation feature data reflecting the change trend of the delay time includes: Performing average calculation based on the obtained historical switching delay data of the secondary oil tank to obtain the historical average switching delay time; Performing deviation calculation on the switching delay time and the historical average switching delay time by means of a deviation calculation method to obtain the switching delay deviation rate; Analyzing the switching delay deviation rates of continuous multiple switching cycles by means of a trend analysis method to obtain abnormal fluctuation feature data reflecting the change trend of the delay time.

4. The protection method of the split type on-load tap-changer according to claim 1, characterized in that, The performing weighted processing on the spectral feature data and the abnormal fluctuation feature data to obtain a current abnormal feature value reflecting the open-circuit fault of the current transformer includes: Calculate a first correlation coefficient between the spectrum feature data and the current transformer open - circuit fault, and calculate a second correlation coefficient between the abnormal fluctuation feature data and the current transformer open - circuit fault. The correlation coefficient reflects the degree of association between the spectrum feature data and the abnormal fluctuation feature data and the current transformer open - circuit fault; Allocate corresponding weights to the spectrum feature data and the abnormal fluctuation feature data according to the calculation results of the correlation coefficients; Multiply the spectrum feature data and the abnormal fluctuation feature data by their corresponding weights respectively, and add the obtained calculation results to generate a current anomaly feature value reflecting the current transformer open - circuit fault.

5. The protection method of the split on-load tap-changer according to claim 1, characterized in that, Before comparing the current anomaly feature value with a preset first anomaly feature threshold respectively, it further includes: Obtain the operating parameters of the split - type on - load tap - changer, where the operating parameters include load rate, operating time, and ambient temperature; Analyze the operating parameters to obtain the operating condition of the split - type on - load tap - changer, where the operating condition includes light - load condition, heavy - load condition, and high - temperature condition; Match the operating condition with a pre - constructed abnormal feature threshold database, and obtain the first abnormal feature threshold according to the matching result.

6. The protection method of the split type on-load tap-changer according to claim 1, characterized in that The obtaining of the short - circuit current peak value flowing through the split - type on - load tap - changer corresponding to the current anomaly feature value includes: Based on a current sensor, collect the current data flowing through the split - type on - load tap - changer to obtain an original data sequence containing short - circuit current information; Perform filtering processing on the original data sequence to generate a smooth current data curve; Detect the peak value of the smooth current data curve according to the difference method, identify the peak points in the curve, analyze multiple consecutive peak points, and generate the short - circuit current peak value based on the analysis result.

7. The protection method of the split type on-load tap-changer according to claim 1, characterized in that, The obtaining of the correction coefficient of the first abnormal feature threshold according to the proportional relationship between the short - circuit current peak value and the rated current of the split - type on - load tap - changer includes: Obtain the historical short - circuit current peak value flowing through the split - type on - load tap - changer, and calculate the first proportional relationship data between the historical short - circuit current peak value and the rated current of the split - type on - load tap - changer; Divide the first proportional relationship data into multiple proportional intervals, and construct a corresponding correction coefficient mapping table for each proportional interval, where the correction coefficient mapping table reflects the correction coefficients corresponding to different proportional intervals; Match the proportional relationship with the correction coefficient mapping table to obtain the correction coefficient corresponding to the proportional relationship.

8. The protection method of the split type on-load tap-changer according to claim 1, characterized in that, The execution of the fault handling strategy generated by the second judgment result includes: Generate a fault level according to the amplitude by which the current anomaly feature value exceeds the second abnormal feature threshold in the second judgment result, where the types of the fault level include mild fault, moderate fault, and severe fault; If the fault level is the mild fault, detect the contact resistance of the current transformer secondary - side wiring terminal by the loop self - inspection method, and simultaneously send a fault warning to the local monitoring system; If the fault level is the medium fault, trigger the single-pole locking of the split type on-load tap-changer, cut off the current path of the faulty phase, and at the same time start the signal switching logic of the standby current transformer; If the fault level is the severe fault, immediately cut off the power supply of the split type on-load tap-changer and send an emergency signal to the dispatching center.

9. The protection method for the split type on-load tap-changer according to claim 8, characterized in that, After generating the fault level according to the amplitude by which the current abnormal characteristic value in the second judgment result exceeds the second abnormal characteristic threshold, it further includes: Calculate the aging coefficient according to the factory parameters and the cumulative operation time of the current transformer. If the aging coefficient is greater than the preset threshold and the medium fault or the severe fault occurs continuously, perform secondary calibration of the fault level; Wherein, the secondary calibration includes: synchronously collecting the vibration signal data at the connection flange of the split type on-load tap-changer and the waveform data of the oil gap insulation breakdown, analyzing the vibration signal data and the waveform data, and performing secondary calibration of the fault level according to the analysis result.

10. A protection system for a split type on-load tap-changer, characterized in that, Applied to a split type on-load tap-changer including a current transformer and a secondary oil tank, the protection system includes: An acquisition module, configured to extract features from the acquired signal amplitude change data on the secondary side of the current transformer to obtain spectral feature data reflecting the signal frequency characteristics; extract features from the acquired switching delay data of the secondary oil tank to obtain abnormal fluctuation feature data reflecting the change trend of the delay time; A fusion module, configured to perform weighted processing on the spectral feature data and the abnormal fluctuation feature data to obtain a current abnormal characteristic value reflecting the open-circuit fault of the current transformer; A comparison module, configured to compare the current abnormal characteristic value with a preset first abnormal characteristic threshold respectively to obtain a first judgment result, wherein the first abnormal characteristic threshold is designed to be dynamically adjusted according to the operating conditions of the split type on-load tap-changer; A judgment module, configured to, if the first judgment result is a preliminary open-circuit fault of the current transformer, obtain the short-circuit current peak value flowing through the split type on-load tap-changer corresponding to the current abnormal characteristic value; A correction module, configured to obtain a correction coefficient of the first abnormal characteristic threshold according to the ratio relationship between the short-circuit current peak value and the rated current of the split type on-load tap-changer, and correct the first abnormal characteristic threshold based on the correction coefficient to obtain a second abnormal characteristic threshold; An execution module, configured to perform secondary determination of the open-circuit fault of the current transformer on the current abnormal characteristic value based on the second abnormal characteristic threshold to obtain a second judgment result reflecting the true fault of the current transformer, and execute the fault handling strategy generated by the second judgment result.

Images