Combined mutual inductor remote monitoring and failure early warning system

Abstract

The application relates to the technical field of electrical fault positioning, and discloses a combined mutual inductor remote monitoring and fault early warning system which comprises a collection module, a feature module, a segmentation module, an anchor point module, a marking module and a warning module; the collection module synchronously collects current data and voltage data and forms a transient observation section, the feature module extracts a waveform top cutting position, a phase distortion direction, a recovery lag section, an amplitude swing section, an oscillation continuation section and an overvoltage rising section, the segmentation module forms a transient event section, the anchor point module determines a first time sequence anchor point and a second time sequence anchor point, the marking module generates current risk marking and voltage risk marking, and the warning module generates a composite risk warning in the same transient event section; the application can perform fault early warning based on the time sequence correlation between current waveforms, voltage waveforms, phase distortion and overvoltage rising, and reduces the misjudgment caused by single electrical measurement features.

Description

Technical Field

[0001] This application relates to the field of electrical fault location technology, specifically to a remote monitoring and fault early warning system for combined instrument transformers. Background Technology

[0002] Combined instrument transformers typically include both current and voltage transformers. In substation, distribution line, and power equipment monitoring scenarios, the current and voltage signals output from the secondary side of the combined instrument transformer are frequently used for current measurement, voltage measurement, electrical quantity measurement, relay protection, metering data acquisition, and remote fault monitoring. Existing monitoring systems generally obtain current and voltage data through waveform sampling on the secondary side of the instrument transformer, and then determine abnormal states such as overcurrent, overvoltage, frequency deviation, and waveform distortion based on amplitude measurement results, frequency measurement results, harmonic measurement results, or over-limit duration.

[0003] Under operating conditions such as ground faults, unloaded bus switching, line reclosing, and sudden load changes, current transformers may enter transient saturation due to transient DC components and core magnetization, resulting in peak clipping, phase shift, and recovery lag in secondary current measurement waveforms. Voltage transformers may exhibit ferroresonant precursors due to system capacitance, inductance, and nonlinear core excitation, leading to amplitude fluctuations, low-frequency oscillations, and overvoltage rise in secondary voltage measurement waveforms. These current and voltage anomalies may originate from the same grid disturbance or occur intermittently within adjacent time ranges. It is difficult to determine the timing relationship between anomalies based solely on amplitude or frequency exceeding limits of a single channel.

[0004] Existing devices for electrical quantity measurement and fault early warning typically focus on separate current or voltage measurement channels, using threshold judgments or independent alarms for transient saturation of current transformers and precursors of ferroresonance in voltage transformers, respectively. When the current measurement waveform first shows clipping and phase distortion, followed by low-frequency oscillations or overvoltage rise in the voltage measurement waveform, existing systems tend to treat voltage-side anomalies as independent overvoltage events. Conversely, when the voltage measurement waveform first shows amplitude fluctuations, followed by distortion in the current measurement waveform, existing systems may treat current-side anomalies as load disturbances or short-circuit transients. Due to the lack of a unified event boundary and a common acquisition time reference between current and voltage data, fault early warning results output from remote monitoring interfaces are prone to attribution confusion.

[0005] Furthermore, existing systems typically rely on a single number of over-limit occurrences, a single duration, or a single waveform characteristic as the basis for risk level when generating fault warnings, lacking the ability to identify the sequential relationship between the current-side recovery process and the voltage-side continuation process. If the transient saturation of the current transformer has ended, but the ferroresonant precursor of the voltage transformer is still ongoing, existing systems struggle to separate the complex risk links related to the combined transformers from ordinary transient disturbances. This makes it impossible for maintenance personnel to determine the priority of subsequent actions based on the current and voltage measurement characteristics within the same event. Therefore, existing remote monitoring technologies for combined transformers still need to address the problem of how to prevent current and voltage measurement anomalies from mutually masking each other and causing misjudgments of fault warnings under the same grid disturbance. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this application aims to provide a remote monitoring and fault early warning system for combined instrument transformers. This system solves the problem that in existing remote monitoring of combined instrument transformers, the transient saturation characteristics of current transformers and the ferroresonant precursor characteristics of voltage transformers are judged based on single current or single voltage measurement results. This leads to a lack of unified temporal correlation between current waveform clipping, phase distortion, abnormal harmonic components, low-frequency oscillations, and overvoltage rise, resulting in fragmented fault early warning criteria and inaccurate identification of composite risks. This application synchronously collects current and voltage data, extracts current and voltage characteristics under the same acquisition time number, and generates composite risk warnings based on transient event segmentation, a first time-series anchor point, a second time-series anchor point, current risk markers, and voltage risk markers. This enables remote monitoring and fault early warning of combined instrument transformers to be based on the temporal relationship between current measurement, voltage measurement, waveform analysis, phase analysis, and overvoltage analysis.

[0007] To achieve the above objectives, this application adopts the following technical solution:

[0008] This application provides a combined instrument transformer remote monitoring and fault early warning system, including:

[0009] The acquisition module is used to synchronously acquire current and voltage data and determine the current reference state; determine the disturbance start point based on the current reference state; and read current and voltage data backward from the disturbance start point to form a transient observation segment.

[0010] The feature module is used to extract current and voltage features within the transient observation segment. The current features include waveform clipping position, phase distortion direction, and recovery hysteresis segment. The voltage features include amplitude swing segment, oscillation continuation segment, and overvoltage rise segment.

[0011] The segmentation module is used to form transient event segments based on the later end time of the disturbance start point to the recovery hysteresis segment and the overvoltage rise segment.

[0012] Anchor point module is used to take the earliest time range in which the waveform clipping position and phase distortion direction coexist as the first time anchor point, and the earliest time range in which the oscillation continuation segment and the overvoltage rise segment coexist as the second time anchor point.

[0013] The marking module is used to generate a current risk mark when the start time of the first time-series anchor point is earlier than the start time of the second time-series anchor point, and the end time of the oscillation continuation section or the overvoltage rise section is later than the end time of the recovery hysteresis section; and to generate a voltage risk mark when the start time of the amplitude swing section is earlier than the first occurrence time of the phase distortion direction, and the first occurrence time of the current characteristic is later than the start time of the amplitude swing section.

[0014] The warning module is used to generate a composite risk warning when current risk markers and voltage risk markers are generated in the same transient event segment.

[0015] Preferably, the method for extracting the waveform clipping position and phase distortion direction includes:

[0016] A current data of one rated power frequency cycle length is intercepted before the start of the disturbance, and the current amplitude boundary and the current reference half-cycle waveform are determined based on the intercepted current data.

[0017] When the difference between the current data and the current amplitude boundary at multiple acquisition times within the transient observation segment does not exceed the preset clipping difference, the acquisition time range formed by multiple acquisition times is determined as the waveform clipping position.

[0018] Compare the timing of the initial acquisition of the waveform clipping position with the timing of the acquisition of the peak of the same polarity in the current reference half-cycle waveform, and determine the direction of phase distortion based on the timing relationship.

[0019] Preferably, the method for determining the recovery lag segment includes:

[0020] The current data of one rated power frequency cycle length is intercepted before the start of the disturbance to determine the allowable range of peak phase offset before the start of the disturbance.

[0021] Harmonic component separation is performed on the current data within the transient observation segment to obtain the high-order harmonic concentration section;

[0022] Starting from the acquisition time after the end of the waveform clipping position, determine one by one whether the acquisition time does not belong to the high-order harmonic concentration section, and determine whether the sampling point interval between the current half-cycle same polarity peak acquisition time and the current half-cycle start acquisition time falls within the allowable range of peak phase offset before the disturbance start point.

[0023] The earliest acquisition time that simultaneously meets both judgment conditions is determined as the recovery completion time;

[0024] The interval between the end of the waveform clipping position and the completion of the recovery is defined as the recovery lag segment.

[0025] Preferably, the method for determining the oscillation continuation section and the overvoltage rise section includes:

[0026] Voltage data of a rated power frequency cycle length is intercepted before the start of the disturbance, and the voltage reference frequency range and voltage reference amplitude range are determined based on the intercepted voltage data.

[0027] Within the transient observation period, when the frequencies corresponding to the voltage data at multiple acquisition moments are all below the lower boundary of the voltage reference frequency range, the range of acquisition moments formed by the multiple acquisition moments is defined as the oscillation continuation segment.

[0028] Within the transient observation period, when the voltage amplitude at multiple acquisition moments is higher than the upper boundary of the voltage reference amplitude range, and the voltage amplitude at the later acquisition moment is greater than or equal to the voltage amplitude at the previous acquisition moment, the acquisition moment range formed by the multiple acquisition moments is defined as the overvoltage rise zone.

[0029] Preferably, the method for determining the amplitude swing range includes:

[0030] Based on the zero-crossing acquisition time within the transient observation period, the voltage data within the transient observation period is divided into multiple half-cycle voltage data.

[0031] Read the voltage data with the highest absolute value in each half-cycle voltage data, and determine the voltage data with the highest absolute value as the half-cycle voltage amplitude of the corresponding half-cycle voltage data.

[0032] The range of acquisition times corresponding to half-cycle voltage data whose half-cycle voltage amplitude is higher than the upper boundary of the voltage reference amplitude range is marked as the upper boundary half-cycle range.

[0033] The range of acquisition times corresponding to half-cycle voltage data whose half-cycle voltage amplitude is lower than the lower boundary of the voltage reference amplitude range is marked as the lower out-of-bounds half-cycle range.

[0034] The order in which the upper and lower bounded half-circle ranges appear is determined according to the order of data collection time.

[0035] When the upper and lower bound half-cycle ranges alternate within multiple acquisition time ranges, these multiple acquisition time ranges are defined as amplitude swing segments.

[0036] Preferably, the method for determining the starting point of the disturbance includes:

[0037] In the sequence of acquisition time numbers, for each current acquisition time, the current data and voltage data of the rated power frequency cycle length before the current acquisition time are read, and the current reference state is determined based on the read current data and voltage data;

[0038] If the current data at the current acquisition time does not belong to the current reference state, or the voltage data at the current acquisition time does not belong to the current reference state, the current acquisition time will be determined as the disturbance start point.

[0039] Preferably, the method for forming a transient observation segment includes:

[0040] The current reference state corresponding to one rated power frequency cycle length before the start of the disturbance is determined as the reference state before the disturbance.

[0041] Read current and voltage data starting from the point of disturbance.

[0042] When the current data within the next rated power frequency cycle is all in the reference state before the disturbance, and the voltage data within the next rated power frequency cycle is all in the reference state before the disturbance, the start time of the next rated power frequency cycle is determined as the end time of the observation.

[0043] The interval between the start of the disturbance and the end of the observation is defined as the transient observation segment.

[0044] Preferably, the method for forming transient event segments includes:

[0045] After extracting current and voltage characteristics within the transient observation period, the interval between the disturbance start point and the later end time in the recovery hysteresis section and the overvoltage rise section is determined as the transient event segment.

[0046] When the transient observation segments corresponding to two adjacent disturbance initiations overlap in time, the two overlapping transient observation segments are merged into a merged transient observation segment with the start time preceding the end time, and transient event segments are re-formed based on the merged transient observation segment.

[0047] When the transient observation segments corresponding to two adjacent disturbance initiations do not overlap in time, two transient observation segments are retained respectively, and two transient event segments are formed respectively.

[0048] Preferably, the method for determining the first time-series anchor point and the second time-series anchor point includes:

[0049] According to the order of the acquisition time numbers, read the waveform clipping position with the determined phase distortion direction, take the acquisition time number covered by the waveform clipping position as the first candidate acquisition time number, merge multiple acquisition time numbers that are adjacent in number and time sequence in the first candidate acquisition time number to form the first coexistence interval, and take the first coexistence interval with the earliest start time as the first time sequence anchor point.

[0050] According to the order of the acquisition time numbers, the acquisition time numbers that belong to both the oscillation continuation segment and the overvoltage rise segment are selected. Multiple acquisition time numbers that are adjacent in number and time sequence are merged to form a second coexisting interval. The second coexisting interval with the earliest start time is used as the second time sequence anchor point.

[0051] Preferably, the method for generating current risk markers includes:

[0052] The time difference between the start time of the second timing anchor point and the start time of the first timing anchor point is calculated to obtain the current leader interval;

[0053] The voltage duration interval is obtained by calculating the time difference between the later end time of the oscillation duration interval and the end time of the overvoltage rise interval, and the end time of the recovery lag interval.

[0054] A current risk marker is generated when both the current leader interval and the voltage duration interval are positive.

[0055] Preferably, the method for generating voltage risk markers includes:

[0056] The voltage leader interval is obtained by calculating the time difference between the initial acquisition time of the waveform clipping position in the first generation phase distortion direction and the initial acquisition time of the amplitude swing segment.

[0057] The time difference between the first appearance of the current characteristic and the start of the amplitude swing segment is calculated to obtain the current hysteresis interval.

[0058] A voltage risk marker is generated when both the voltage leader interval and the current hysteresis interval are positive.

[0059] Preferably, the method for generating a composite risk warning includes:

[0060] Within the same transient event segment, it was confirmed that both current risk markers and voltage risk markers had been generated.

[0061] The composite risk level is determined based on the time difference between the end of the oscillation continuation segment and the end of the recovery lag segment, and the time difference between the end of the overvoltage rise segment and the end of the recovery lag segment.

[0062] Generate composite risk warnings based on composite risk levels.

[0063] Compared with the prior art, the beneficial effects achieved by this application are as follows:

[0064] This application enables the simultaneous acquisition of current and voltage data, and the determination of the disturbance start point and the formation of a transient observation segment based on the current reference state. This allows the secondary current measurement data and voltage measurement data of the combined instrument transformer to enter the subsequent waveform analysis process under the same acquisition time number, thereby reducing the timing mismatch caused by the inconsistency of the time base between the current and voltage characteristics and improving the consistency of electrical measurement data input in remote monitoring.

[0065] This application extracts waveform clipping location, phase distortion direction, recovery lag section, amplitude swing section, oscillation continuation section, and overvoltage rise section within the transient observation section. It transforms the amplitude, phase, and harmonic recovery characteristics related to transient saturation of current transformers, and the frequency, amplitude, and overvoltage characteristics related to ferroresonant precursors of voltage transformers, into comparable structured sections. As a result, fault warning no longer relies on a single current overrun, a single voltage overrun, or a single abnormal duration, which can reduce the possibility of ordinary transient disturbances being misjudged as composite risks.

[0066] This application segments transient events by using the later end time of the disturbance start point to the recovery lag section and the overvoltage rise section, and determines the first and second time-series anchor points within the transient event segments, so that the current-side determineable state and the voltage-side determineable state can be compared sequentially within the same time range. As a result, the problem of the separation between the current transformer transient saturation judgment and the voltage transformer ferroresonance precursor judgment in the prior art is improved, and the judgment basis for composite fault early warning is more complete.

[0067] This application generates current risk markers through current leader intervals and voltage duration intervals, and generates voltage risk markers through voltage leader intervals and current lag intervals. Then, within the same transient event segment, a composite risk warning is generated based on the current risk markers and voltage risk markers, thus forming a continuous timing judgment chain between early voltage amplitude fluctuations, current waveform clipping and phase distortion, low-frequency voltage oscillations, and overvoltage rise. As a result, the combined instrument transformer remote monitoring system can distinguish between unilateral anomalies, short-term disturbances, and composite risk chains formed by the mutual influence of current and voltage.

[0068] This application determines the composite risk level based on the time difference between the end of the oscillation continuation section and the end of the recovery lag section, as well as the time difference between the end of the overvoltage rise section and the end of the recovery lag section. This allows the composite risk warning to not only include the judgment of whether a composite risk exists, but also to reflect the difference between the low-frequency voltage oscillation continuation and the overvoltage rise continuation relative to the current recovery process. This improves the interpretability of the fault warning results of the combined instrument transformer and the pertinence of subsequent operation and maintenance. Attached Figure Description

[0069] Figure 1This is the overall logic diagram of a combined instrument transformer remote monitoring and fault early warning system;

[0070] Figure 2 Logic diagram for current feature extraction;

[0071] Figure 3 Logic diagram for voltage feature extraction;

[0072] Figure 4 This is a logic diagram for generating and applying composite risk warnings. Detailed Implementation

[0073] The technical solution of this application will be described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the embodiments and specific features in the embodiments are detailed descriptions of the technical solution of this application, rather than limitations thereof.

[0074] Example 1:

[0075] like Figures 1-4 As shown, this embodiment provides a combined instrument transformer remote monitoring and fault early warning system, including a data acquisition module for synchronously acquiring current data and voltage data and determining the current reference state; determining the disturbance start point based on the current reference state; and reading current data and voltage data backward from the disturbance start point to form a transient observation segment.

[0076] The feature module is used to extract current and voltage features within the transient observation segment. The current features include waveform clipping position, phase distortion direction, and recovery hysteresis segment. The voltage features include amplitude swing segment, oscillation continuation segment, and overvoltage rise segment.

[0077] The segmentation module is used to form transient event segments based on the later end time of the disturbance start point to the recovery hysteresis segment and the overvoltage rise segment.

[0078] The anchor point module is used to take the earliest time range in which the waveform clipping position and phase distortion direction coexist as the first timing anchor point, and the earliest time range in which the oscillation continuation segment and the overvoltage rise segment coexist as the second timing anchor point.

[0079] The marking module is used to generate a current risk mark when the start time of the first timing anchor point is earlier than the start time of the second timing anchor point, and the end time of the oscillation continuation section or the overvoltage rise section is later than the end time of the recovery hysteresis section; and to generate a voltage risk mark when the start time of the amplitude swing section is earlier than the first occurrence time of the phase distortion direction, and the first occurrence time of the current characteristic is later than the start time of the amplitude swing section.

[0080] The warning module is used to generate a composite risk warning when current risk markers and voltage risk markers are generated in the same transient event segment.

[0081] The modules are connected via wired or wireless means and transmit data, as follows:

[0082] Synchronously acquire current and voltage data and determine the current reference state; determine the disturbance start point based on the current reference state; read current and voltage data backward from the disturbance start point to form a transient observation segment; extract current and voltage features within the transient observation segment. Current features include waveform clipping position, phase distortion direction, and recovery lag segment, while voltage features include amplitude swing segment, oscillation continuation segment, and overvoltage rise segment.

[0083] The current reference state is a pre-disturbance control state determined based on current and voltage data from a rated power frequency cycle prior to the current acquisition time. The current reference state includes the current amplitude reference range, current harmonic reference range, voltage amplitude reference range, and voltage frequency reference range. First, the current and voltage data from a rated power frequency cycle prior to the current acquisition time are read and arranged according to the acquisition time number. The positive current peak value and negative current peak value are read from the current data, and the current amplitude reference range is determined by combining the upper limit of current measurement error and the minimum resolution of analog-to-digital conversion in the calibration record of the acquisition device.

[0084] The fundamental component is extracted from the current data, and the fundamental component corresponding to the same acquisition time is subtracted from the instantaneous sampled value of the current at each acquisition time to obtain the higher harmonic components corresponding to each acquisition time. Then, the upper boundary of the amplitude of the higher harmonic components is read to obtain the current harmonic reference range. The peak values ​​of positive and negative voltages are read from the voltage data, and the voltage measurement error upper limit and the minimum resolution of analog-to-digital conversion in the calibration record of the acquisition device are combined to determine the voltage amplitude reference range. The zero-crossing acquisition times in the same direction are found from the voltage data, and the voltage reference frequency is determined based on the number of sampling points and the sampling time interval between the zero-crossing acquisition times in the same direction. Then, the voltage frequency reference range is determined by combining the rated frequency allowable deviation in the system parameter table and the sampling time error in the calibration record of the acquisition device. The current amplitude reference range, current harmonic reference range, voltage amplitude reference range, and voltage frequency reference range are jointly determined as the current reference state.

[0085] In an optional embodiment, the method for forming a transient observation segment includes: in the acquisition time number sequence, for each current acquisition time, reading current data and voltage data for a rated power frequency cycle length preceding the current acquisition time, and determining the current reference state based on the read current data and voltage data; if the current data corresponding to the current acquisition time does not belong to the current reference state, or the voltage data corresponding to the current acquisition time does not belong to the current reference state, the current acquisition time is determined as the disturbance start point; after determining the disturbance start point, the current reference state corresponding to a rated power frequency cycle length preceding the disturbance start point is frozen as the pre-disturbance reference state, and the transient observation segment end judgment is performed based on the frozen pre-disturbance reference state; starting from the disturbance start point, reading current data and voltage data backwards, when the current data within the subsequent rated power frequency cycle length all belong to the pre-disturbance reference state, and the voltage data within the subsequent rated power frequency cycle length all belong to the pre-disturbance reference state, the starting acquisition time of the subsequent rated power frequency cycle length is determined as the observation end time; the interval between the disturbance start point and the observation end time is determined as the transient observation segment.

[0086] In this embodiment, when determining the disturbance start point based on the current reference state, the current data and voltage data corresponding to the current acquisition time are read sequentially according to the acquisition time number. First, the current data corresponding to the current acquisition time is compared with the current amplitude reference range. When the current data corresponding to the current acquisition time is positive and higher than the positive boundary of the current amplitude reference range, or when the current data corresponding to the current acquisition time is negative and lower than the negative boundary of the current amplitude reference range, the current acquisition time is recorded as an abnormal acquisition time of current amplitude. Then, the higher harmonic components corresponding to the current acquisition time are extracted, and the higher harmonic components corresponding to the current acquisition time are compared with the current harmonic reference range. When the higher harmonic components corresponding to the current acquisition time exceed the current harmonic reference range, the current acquisition time is recorded as an abnormal acquisition time of current harmonics.

[0087] Subsequently, the voltage amplitude anomaly acquisition time is determined based on the half-cycle voltage amplitude; taking the current acquisition time as the end position, the two most recent adjacent zero-crossing acquisition times are read backwards; if no two adjacent zero-crossing acquisition times have been formed before the current acquisition time, voltage amplitude anomaly judgment is not performed on the current acquisition time; if two adjacent zero-crossing acquisition times have been formed before the current acquisition time, the voltage data between the two adjacent zero-crossing acquisition times is determined as the half-cycle voltage data to be judged; the absolute value of the voltage data corresponding to each acquisition time in the half-cycle voltage data to be judged is read, and the item with the highest absolute value of the voltage data is determined as the half-cycle voltage amplitude; the acquisition time where the half-cycle voltage amplitude is located is determined. The half-cycle voltage amplitude is determined as the acquisition time. When the half-cycle voltage amplitude is higher than the upper boundary of the voltage amplitude reference range, or lower than the lower boundary of the voltage amplitude reference range, the half-cycle voltage amplitude acquisition time is recorded as an abnormal voltage amplitude acquisition time. When the half-cycle voltage amplitude is between the lower boundary and the upper boundary of the voltage amplitude reference range, no acquisition time in the half-cycle voltage data to be judged is recorded as an abnormal voltage amplitude acquisition time. The half-cycle voltage amplitude is used to characterize the voltage amplitude level between two adjacent zero-crossing acquisition times. The absolute value of the voltage data corresponding to a single acquisition time is not used as the basis for judging the abnormal voltage amplitude acquisition time.

[0088] Then, using the current acquisition time as the end position, read forward voltage data that can cover one voltage cycle, and determine the voltage frequency corresponding to the current acquisition time based on the number of sampling points and the sampling time interval between acquisition times with zero crossing in the same direction. When the voltage frequency corresponding to the current acquisition time is not within the voltage frequency reference range, record the current acquisition time as an abnormal voltage frequency acquisition time. The abnormal acquisition times of current amplitude, current harmonics, voltage amplitude, and voltage frequency are all taken as candidate disturbance starting points. Read the candidate disturbance starting points in the order of acquisition time number, and determine the earliest candidate disturbance starting point in time sequence as the disturbance starting point. Read current and voltage data backward from the disturbance starting point to form a transient observation segment.

[0089] In one optional embodiment, the method for synchronously acquiring current data and voltage data includes: writing the same acquisition time number for each acquisition time; pairing current data and voltage data according to the same acquisition time number; and removing the data pair corresponding to the same acquisition time number when current data or voltage data is missing for the same acquisition time number.

[0090] In this embodiment, the current data is a data sequence formed by arranging instantaneous current sampled values ​​obtained from the secondary side of the current transformer according to the acquisition time number; the voltage data is a data sequence formed by arranging instantaneous voltage sampled values ​​obtained from the secondary side of the voltage transformer according to the acquisition time number; both are connected to the secondary side of the current transformer and the secondary side of the voltage transformer respectively; the output current signal of the secondary side of the current transformer is isolated and sampled and converted from analog to digital, and recorded as instantaneous current sampled values ​​according to the acquisition time number; the output voltage signal of the secondary side of the voltage transformer is isolated and sampled and converted from analog to digital, and recorded as instantaneous voltage sampled values ​​according to the same acquisition time number; when both instantaneous current sampled values ​​and instantaneous voltage sampled values ​​exist under the same acquisition time number, the instantaneous current sampled values ​​and instantaneous voltage sampled values ​​are retained as data pairs that can participate in subsequent processing; when there are no instantaneous current sampled values ​​or instantaneous voltage sampled values ​​under the same acquisition time number, the data pairs corresponding to the same acquisition time number are discarded, so that both current and voltage features are extracted based on the same time reference.

[0091] In an optional embodiment, the method for extracting the waveform clipping position and phase distortion direction includes: intercepting current data for a rated power frequency cycle length before the disturbance start point; determining the current amplitude boundary and the current reference half-cycle waveform based on the intercepted current data; when the difference between the current data at multiple acquisition times within the transient observation segment and the current amplitude boundary does not exceed a preset clipping difference, determining the acquisition time range formed by the multiple acquisition times as the waveform clipping position; comparing the initial acquisition time of the waveform clipping position with the acquisition time of the peak of the same polarity in the current reference half-cycle waveform, and determining the phase distortion direction based on the sequential relationship.

[0092] In this embodiment, the rated power frequency cycle length is the time length corresponding to the monitored power grid completing one current waveform or voltage waveform cycle at the rated frequency; the rated frequency of the monitored power grid is read from the system parameter table, and the number of sampling points corresponding to one power frequency cycle is determined based on the rated frequency; when intercepting current data of one rated power frequency cycle length before the disturbance start point, the instantaneous current sampling value with the number of sampling points meeting the requirements of one power frequency cycle is read from before the disturbance start point, and the read instantaneous current sampling value is used as the basic data for determining the current amplitude boundary and the current reference half-cycle waveform.

[0093] In this embodiment, when determining the positive current amplitude boundary, negative current amplitude boundary, and current reference half-cycle waveform based on the intercepted current data, the intercepted current data are first arranged according to the acquisition time number; the current data with values ​​greater than zero in the intercepted current data are read, and the highest value among the current data with values ​​greater than zero is determined as the positive current amplitude boundary; the current data with values ​​less than zero in the intercepted current data are read, and the lowest value among the current data with values ​​less than zero is determined as the negative current amplitude boundary; the half-cycle waveform corresponding to the positive current amplitude boundary is determined as the positive current reference half-cycle waveform, and the half-cycle waveform corresponding to the negative current amplitude boundary is determined as the negative current reference half-cycle waveform; the positive current reference half-cycle waveform and the negative current reference half-cycle waveform are used together as the current reference half-cycle waveform.

[0094] In this embodiment, the preset clipping difference is used to limit whether the instantaneous current sampling value has approached the current amplitude boundary. The preset clipping difference is set during the commissioning calibration phase. During the commissioning calibration phase, multiple sets of current data are collected during the operation period without disturbance, and the positive current amplitude boundary and negative current amplitude boundary in each set of current data are read respectively. Then, the upper limit of current measurement error and the minimum resolution of analog-to-digital conversion in the sampling device calibration record are read. The upper limit of current measurement error, the minimum resolution of analog-to-digital conversion, and the amplitude difference between the sampling point near the peak in multiple sets of current data and the corresponding current amplitude boundary are compared, and the one with the highest value is selected as the preset clipping difference. When the sampling device is replaced, the sampling frequency is adjusted, or the rated ratio of the current transformer is adjusted, the commissioning calibration phase setting process is re-executed, and the newly obtained preset clipping difference is written into the parameter table.

[0095] In this embodiment, when determining the waveform clipping position, the current data corresponding to each acquisition time within the transient observation segment is read according to the acquisition time number. When the current data corresponding to the acquisition time is positive, only the current data corresponding to the acquisition time is compared with the positive current amplitude boundary, and the absolute value of the difference between the current data corresponding to the acquisition time and the positive current amplitude boundary is taken as the positive boundary difference. When the current data corresponding to the acquisition time is negative, only the current data corresponding to the acquisition time is compared with the negative current amplitude boundary, and the absolute value of the difference between the current data corresponding to the acquisition time and the negative current amplitude boundary is taken as the negative boundary difference. When the positive boundary differences corresponding to multiple acquisition times do not exceed the preset clipping difference, or the negative boundary differences corresponding to multiple acquisition times do not exceed the preset clipping difference, the acquisition time range formed by multiple acquisition times is determined as the waveform clipping position. By comparing the positive and negative polarities of the current data with the positive and negative current amplitude boundaries respectively, the misjudgment of the waveform clipping position caused by comparing positive current data with the negative current amplitude boundary or negative current data with the positive current amplitude boundary can be avoided.

[0096] It should be noted that when a current transformer experiences transient saturation, the secondary output current is limited by magnetization characteristics near its peak value. The instantaneous current sampling value no longer follows the half-cycle waveform before the disturbance, but instead forms multiple adjacent sampling moments with closely spaced amplitudes near the positive or negative current amplitude boundaries. Therefore, using the fact that the difference between the current data at multiple sampling moments within the transient observation period and the current amplitude boundary does not exceed the preset clipping difference as the condition for determining the waveform clipping position can eliminate misjudgments caused by sampling errors at a single sampling moment. Determining the sampling moment range formed by multiple sampling moments as the waveform clipping position allows us to record the start, end, and continuous sampling range of the current transformer's transient saturation in the half-cycle waveform, providing a time basis for subsequently determining the first timing anchor point.

[0097] In this embodiment, when comparing the order of acquisition between the initial acquisition time of the waveform clipping position and the acquisition time of the peak value of the same polarity in the current reference half-cycle waveform, the polarity of the instantaneous current sample value corresponding to the waveform clipping position is first determined. When the instantaneous current sample value corresponding to the waveform clipping position is positive, the acquisition time corresponding to the positive polarity peak value is read in the current reference half-cycle waveform, and the acquisition time corresponding to the positive polarity peak value is taken as the acquisition time of the peak value of the same polarity. When the instantaneous current sample value corresponding to the waveform clipping position is negative, the acquisition time corresponding to the negative polarity peak value is read in the current reference half-cycle waveform, and the acquisition time corresponding to the negative polarity peak value is taken as the acquisition time of the peak value of the same polarity. Then, the initial acquisition time of the same polarity half-cycle, which includes the initial acquisition time of the waveform clipping position, is read, and the initial acquisition time of the half-cycle corresponding to the current reference half-cycle waveform is read.

[0098] The sampling point interval is calculated relative to the starting acquisition time of the waveform clipping position and the starting acquisition time of the same polarity half-cycle. The sampling point interval is also calculated relative to the starting acquisition time of the same polarity peak and the half-cycle corresponding to the current reference half-cycle waveform. When the previous sampling point interval is less than the next sampling point interval, the sequence indicates that the waveform clipping position appears earlier than the same polarity peak position before the disturbance, and the phase distortion direction is determined as the phase advance direction. When the previous sampling point interval is greater than the next sampling point interval, the sequence indicates that the waveform clipping position appears later than the same polarity peak position before the disturbance, and the phase distortion direction is determined as the phase lag direction. When the previous sampling point interval is equal to the next sampling point interval, no phase distortion direction is generated, and the corresponding acquisition time is recorded as a phase-no-offset state. The phase-no-offset state does not participate in the generation of the first timing anchor point, current risk marker, and voltage risk marker.

[0099] In an optional embodiment, the method for determining the recovery lag segment includes: intercepting current data for a rated power frequency cycle length before the disturbance initiation point to determine the allowable range of peak phase offset before the disturbance initiation point; performing harmonic component separation on the current data within the transient observation segment to obtain the high-order harmonic concentration segment; starting from the acquisition time after the end of the waveform clipping position, determining one by one whether the acquisition time does not belong to the high-order harmonic concentration segment, and determining whether the sampling point interval between the current half-cycle same polarity peak acquisition time and the current half-cycle start acquisition time falls within the allowable range of peak phase offset before the disturbance initiation point; determining the earliest acquisition time that simultaneously meets the two judgment conditions as the recovery completion time; and determining the interval between the end of the waveform clipping position and the recovery completion time as the recovery lag segment.

[0100] In this embodiment, when determining the allowable range of peak phase offset before the disturbance start point, current data of one rated power frequency cycle length is first intercepted before the disturbance start point, and the intercepted current data is arranged according to the acquisition time number; in the intercepted current data, the two adjacent acquisition times where the instantaneous current sample value changes from a non-positive value to a positive value are found, and the latter acquisition time is determined as the phase start acquisition time; when no two adjacent acquisition times where the instantaneous current sample value changes from a non-positive value to a positive value are found in the intercepted current data, the two adjacent acquisition times where the instantaneous current sample value changes from a non-negative value to a negative value are found, and the latter acquisition time is taken as the reverse zero-crossing acquisition time, and the phase start acquisition time is obtained by backward calculation from the reverse zero-crossing acquisition time according to the number of sampling points corresponding to half the rated power frequency cycle length; the phase start acquisition time is taken as the zero phase position, and each acquisition time from the phase start acquisition time to the next phase start acquisition time is converted into a phase value according to the number of sampling points corresponding to one rated power frequency cycle length.

[0101] Read the sampling point interval between the peak sampling time of the same polarity before the disturbance start time and the start sampling time of the half-cycle; read the sampling time error from the sampling device calibration record and read the number of sampling points corresponding to one rated power frequency cycle; determine the phase deviation corresponding to the sampling time error based on the sampling time error and the number of sampling points corresponding to one rated power frequency cycle; read the upper limit of the current measurement error from the sampling device calibration record, and read the sampling time range covered by the upper limit of the current measurement error before and after the peak sampling time of the same polarity from the current data intercepted for one rated power frequency cycle before the disturbance start time; and compare the current measurement data before and after the peak sampling time of the same polarity. The number of sampling points corresponding to the range of acquisition times covered by the upper limit of error is determined as the phase deviation corresponding to the current measurement error; the phase deviation corresponding to the sampling time error and the phase deviation corresponding to the current measurement error are added together to obtain the allowable phase deviation; the value after deducting the allowable phase deviation from the sampling point interval of the same polarity peak acquisition time before the disturbance start time relative to the half-cycle start acquisition time is determined as the lower boundary of the allowable range of peak phase offset before the disturbance start time; the value after superimposing the allowable phase deviation on the sampling point interval of the same polarity peak acquisition time before the disturbance start time relative to the half-cycle start acquisition time is determined as the upper boundary of the allowable range of peak phase offset before the disturbance start time.

[0102] When determining whether the sampling point interval between the current half-cycle peak acquisition time and the current half-cycle start acquisition time falls within the allowable range of peak phase offset before the disturbance start point, the same phase start acquisition time determination method and the same phase value conversion method are used to ensure that the current phase within the transient event segment and the allowable range of peak phase offset before the disturbance start point are under the same phase reference.

[0103] In this embodiment, when obtaining the high-order harmonic concentration section, the current data intercepted before the disturbance start point for one rated power frequency cycle length is first used as the harmonic reference data; the sampling time interval and the rated power frequency cycle length are read, and the number of sampling points corresponding to one rated power frequency cycle length is determined based on the sampling time interval and the rated power frequency cycle length; the harmonic reference data are arranged according to the acquisition time number, and the two adjacent acquisition times in the harmonic reference data where the current data changes from a non-positive value to a positive value are found, and the latter acquisition time of the two adjacent acquisition times is taken as the fundamental wave start acquisition time; starting from the fundamental wave start acquisition time, the acquisition times within one rated power frequency cycle length are divided according to the acquisition time number order. The sampling point sequence is determined; a fundamental frequency reference sequence corresponding one-to-one with the sampling point sequence is generated according to the sinusoidal change sequence of the rated power frequency cycle; and the fundamental frequency reference sequence is aligned in amplitude and phase according to the positive peak, negative peak and fundamental frequency start acquisition time in the harmonic reference data to obtain the fundamental frequency component corresponding to each acquisition time in the harmonic reference data; the current data at each acquisition time in the harmonic reference data is subtracted from the fundamental frequency component corresponding to the same acquisition time to obtain the reference higher harmonic component corresponding to the same acquisition time; the amplitude of the reference higher harmonic component corresponding to each acquisition time is read, and the highest value among the amplitudes of the reference higher harmonic components is determined as the upper boundary of the amplitude of the reference higher harmonic component.

[0104] In an optional embodiment, the fundamental component and higher harmonic components are determined according to the following formula:

[0105] ;

[0106] ;

[0107] ;

[0108] ;

[0109] ;

[0110] in, Indicates the center value of the fundamental frequency; Indicates the fundamental frequency amplitude; This represents the positive peak value in the harmonic reference data; This represents the negative peak value in the harmonic reference data; n represents the acquisition time number. Indicates the fundamental frequency acquisition start time number. The value is determined by the second acquisition time among two adjacent acquisition times when the current data changes from a non-positive value to a positive value. This indicates the number of sampling points corresponding to one rated power frequency cycle length. Determined by the rated power frequency cycle length and the sampling time interval; This represents the fundamental frequency component corresponding to the acquisition time number n; This represents the current data corresponding to the acquisition time number n; This represents the higher harmonic component corresponding to the acquisition time number n; This indicates the range of acquisition times corresponding to the current data intercepted one rated power frequency cycle length before the start of the disturbance; This indicates the upper boundary of the amplitude of the reference higher harmonic component.

[0111] During the transient observation period, when judging the higher harmonic components at each acquisition time to be judged, the combined instrument transformer remote monitoring unit follows the same procedure. and The calculation method yields the current higher harmonic components corresponding to the acquisition time to be judged; when When the acquisition time is n, the acquisition time number is marked as the higher harmonic acquisition time; the acquisition times with adjacent numbers and both marked as higher harmonic acquisition times are merged into a higher harmonic concentration section.

[0112] Next, according to the acquisition time numbering order, the higher harmonic components are judged for each acquisition time to be judged within the transient observation segment; taking the acquisition time to be judged as the end position, the current data for one rated power frequency cycle length is read forward to form the current harmonic separation range corresponding to the acquisition time to be judged; when the current data before the acquisition time to be judged is less than one rated power frequency cycle length, the acquisition time required for the current harmonic separation range is supplemented from the current data intercepted before the disturbance start point; the fundamental wave start acquisition time is determined in the current harmonic separation range, and the sampling point sequence position is assigned to each acquisition time in the current harmonic separation range according to the acquisition time numbering order.

[0113] According to the sinusoidal change sequence of the rated power frequency period, a fundamental reference sequence is generated, corresponding one-to-one with the sequential position of each sampling point within the current harmonic separation range. Based on the positive peak value, negative peak value, and fundamental wave start acquisition time in the current harmonic separation range, the fundamental reference sequence is aligned in amplitude and phase to obtain the fundamental wave component corresponding to each acquisition time within the current harmonic separation range. The current data corresponding to the acquisition time to be judged is subtracted from the fundamental wave component corresponding to the acquisition time to be judged to obtain the current higher harmonic component corresponding to the acquisition time to be judged. The amplitude of the current higher harmonic component corresponding to the acquisition time to be judged is compared with the upper boundary of the amplitude of the reference higher harmonic component. When the amplitude of the current higher harmonic component corresponding to the acquisition time to be judged exceeds the upper boundary of the amplitude of the reference higher harmonic component, the acquisition time to be judged is marked as a higher harmonic acquisition time. According to the acquisition time numbering order, acquisition times with adjacent numbers and both marked as higher harmonic acquisition times are merged into the same acquisition time range, and the merged acquisition time range is determined as the higher harmonic concentration segment.

[0114] It should be noted that the earliest acquisition time that simultaneously meets both judgment conditions is determined as the recovery completion time because after the current transformer exits transient saturation, the current data needs to simultaneously meet two conditions: the higher harmonic components must return to the state corresponding to the transient event segment, and the current phase must return to the allowable range of peak phase offset before the disturbance start point. Only then can it be said that the current data has escaped the distortion state after waveform clipping. The acquisition time not belonging to the higher harmonic concentration segment means that the higher harmonic components in the current data corresponding to the same acquisition time no longer exceed the judgment benchmark before the transient event segment. The current phase corresponding to the acquisition time falling into the allowable range of peak phase offset before the disturbance start point means that the current phase corresponding to the same acquisition time has returned to the phase benchmark range before the disturbance start point. Only when both judgment conditions are met simultaneously is the corresponding acquisition time taken as the end position of the recovery lag segment. Selecting the earliest acquisition time that simultaneously meets both judgment conditions allows the recovery lag segment to end when the current data first meets the recovery judgment requirements, avoiding including subsequent acquisition times that have already met the recovery judgment requirements in the recovery lag segment.

[0115] It should be noted that the interval between the end of the waveform clipping position and the completion of the recovery is defined as the recovery lag segment because the end of the waveform clipping position only indicates that the instantaneous current sample value is no longer close to the current amplitude boundary, and cannot alone indicate that the higher harmonic components and current phase in the current data have recovered to the state before the disturbance start point. After the end of the waveform clipping position and before the completion of the recovery, although the current data has left the clipping state near the current amplitude boundary, there are still high harmonic concentration segments or situations where the current phase has not fallen into the allowable range of peak phase offset before the disturbance start point within the same time period. Therefore, by taking the end of the waveform clipping position as the starting position of the recovery lag segment and the completion of the recovery as the ending position of the recovery lag segment, the lag process between the disappearance of clipping and the recovery of the waveform phase of the current transformer transient saturation can be preserved, providing a time basis for subsequent judgment on whether the oscillation continuation segment or the overvoltage rise segment is later than the end of the recovery lag segment.

[0116] In an optional embodiment, the method for determining the oscillation continuation segment and the overvoltage rise segment includes: intercepting voltage data for a rated power frequency cycle length before the disturbance initiation point, and determining a voltage reference frequency range and a voltage reference amplitude range based on the intercepted voltage data; within the transient observation segment, when the frequencies corresponding to the voltage data at multiple acquisition times are all lower than the lower boundary of the voltage reference frequency range, the acquisition time range formed by the multiple acquisition times is determined as the oscillation continuation segment; within the transient observation segment, when the voltage amplitudes at multiple acquisition times are all higher than the upper boundary of the voltage reference amplitude range, and the voltage amplitude at the later acquisition time is greater than or equal to the voltage amplitude at the previous acquisition time, the acquisition time range formed by the multiple acquisition times is determined as the overvoltage rise segment.

[0117] In this embodiment, when determining the voltage reference frequency range and voltage reference amplitude range based on the intercepted voltage data, the intercepted voltage data is first arranged according to the acquisition time number, and the instantaneous voltage sample value corresponding to each acquisition time number is read. The intercepted voltage data is then searched for two adjacent acquisition times where the instantaneous voltage sample value changes from non-positive to positive, and the latter acquisition time is determined as the positive zero-crossing acquisition time. Next, two adjacent acquisition times where the instantaneous voltage sample value changes from non-negative to negative are searched, and the latter acquisition time is determined as the reverse zero-crossing acquisition time. When there are two positive zero-crossing acquisition times in the intercepted voltage data, the number of adjacent acquisition time intervals between the two positive zero-crossing acquisition times is first determined based on the number of sampling points between them. Then, the sampling time intervals corresponding to the number of adjacent acquisition time intervals are accumulated to obtain one voltage cycle. The corresponding time length is then calculated. The time length corresponding to one voltage cycle is then converted into the number of voltage cycles that can be completed per unit time, and the number of voltage cycles that can be completed per unit time is determined as the voltage reference frequency. When the intercepted voltage data contains only one positive zero-crossing acquisition moment and one negative zero-crossing acquisition moment, the number of adjacent acquisition moment intervals between the positive and negative zero-crossing acquisition moments is first determined based on the number of sampling points between them. Then, the sampling time intervals corresponding to the number of adjacent acquisition moment intervals are accumulated to obtain the time length corresponding to half a voltage cycle. Subsequently, the time length corresponding to half a voltage cycle is converted into the time length corresponding to one voltage cycle, and the time length corresponding to one voltage cycle is converted into the number of voltage cycles that can be completed per unit time to obtain the voltage reference frequency.

[0118] The rated frequency allowable deviation is read from the system parameter table, and the sampling time error is read from the calibration record of the acquisition device. The sampling time error is the allowable error of the sampling time interval between two adjacent acquisition moments. When the voltage reference frequency is determined by two positive zero-crossing acquisition moments, the number of adjacent acquisition moment intervals between the two positive zero-crossing acquisition moments is multiplied by the sampling time error to obtain the periodic time error corresponding to one voltage cycle. The periodic time error is subtracted from the time length corresponding to one voltage cycle to obtain the lower limit of the periodic time. The periodic time error is then added to the time length corresponding to one voltage cycle to obtain the upper limit of the periodic time. The lower limit of the periodic time is converted into the number of voltage cycles that can be completed per unit time to obtain the upper boundary of the frequency corresponding to the sampling time error. The upper limit of the periodic time is converted into the number of voltage cycles that can be completed per unit time to obtain the lower boundary of the frequency corresponding to the sampling time error.

[0119] The absolute value of the difference between the voltage reference frequency and the lower boundary of the frequency corresponding to the sampling time error is determined as the first frequency deviation; the absolute value of the difference between the upper boundary of the frequency corresponding to the sampling time error and the voltage reference frequency is determined as the second frequency deviation; the higher of the first and second frequency deviations is determined as the frequency deviation corresponding to the sampling time error; when the voltage reference frequency is determined by one positive zero-crossing acquisition moment and one negative zero-crossing acquisition moment, the number of adjacent acquisition moment intervals between the positive and negative zero-crossing acquisition moments is multiplied by the sampling time error to obtain the half-cycle time error corresponding to half a voltage cycle; the half-cycle time error is converted into the cycle time error corresponding to one voltage cycle; the cycle time error is subtracted from the time length corresponding to one voltage cycle to obtain the lower limit of the cycle time, and then... The upper limit of the period time is obtained by adding the period time error to the time length corresponding to a voltage cycle. The lower limit of the period time is converted into the number of voltage cycles that can be completed per unit time, which gives the upper boundary of the frequency corresponding to the sampling time error. The upper limit of the period time is converted into the number of voltage cycles that can be completed per unit time, which gives the lower boundary of the frequency corresponding to the sampling time error. The frequency deviation corresponding to the sampling time error is obtained again according to the determination method of the first frequency deviation, the second frequency deviation, and the frequency deviation corresponding to the sampling time error. The frequency boundary deviation is obtained by adding the allowable deviation of the rated frequency and the frequency deviation corresponding to the sampling time error. The value of the voltage reference frequency minus the frequency boundary deviation is determined as the lower boundary of the voltage reference frequency range. The value of the voltage reference frequency plus the frequency boundary deviation is determined as the upper boundary of the voltage reference frequency range.

[0120] In an optional embodiment, the voltage reference frequency range is determined as follows: when the voltage reference frequency is determined by two positive zero-crossing acquisition times, the time length corresponding to one voltage cycle and the cycle time error corresponding to one voltage cycle are determined according to the following formula:

[0121] ;

[0122] ;

[0123] When the voltage reference frequency is determined by a positive zero-crossing acquisition time and a reverse zero-crossing acquisition time, the time length corresponding to a voltage cycle and the cycle time error corresponding to a voltage cycle are determined according to the following formula:

[0124] ;

[0125] ;

[0126] The voltage reference frequency range is then determined using the following formula:

[0127] ;

[0128] ;

[0129] ;

[0130] ;

[0131] ;

[0132] in, This indicates the number of adjacent acquisition time intervals between two positive zero-crossing acquisition times; This indicates the number of adjacent acquisition time intervals between positive zero-crossing acquisition times and negative zero-crossing acquisition times; Indicates the sampling time interval. Read from the data acquisition device configuration record; This indicates the sampling time error. Read from the calibration record of the data acquisition device; This indicates the time length corresponding to one voltage cycle; This represents the periodic time error corresponding to one voltage cycle. Indicates the voltage reference frequency; This indicates the frequency deviation corresponding to the sampling time error; Indicates the allowable deviation of the rated frequency. Read from the system parameter table; Indicates frequency boundary deviation; Indicates the lower boundary of the voltage reference frequency range; Indicates the upper boundary of the voltage reference frequency range; when When the voltage is less than or equal to zero, the remote monitoring unit of the combined instrument transformer does not use the corresponding voltage data to determine the voltage reference frequency range, and rereads the next set of voltage data that meets the requirements of the zero-crossing acquisition time.

[0133] The peak values ​​of positive and negative voltages are read from the captured voltage data, and the absolute values ​​of the positive and negative voltage peak values ​​are used as the voltage amplitude reference before the disturbance start point. The upper limit of voltage measurement error and the minimum resolution of analog-to-digital conversion are read from the calibration record of the acquisition device. The lowest value among the absolute values ​​of positive and negative voltage peak values ​​is subtracted from the amplitude corresponding to the upper limit of voltage measurement error and the minimum resolution of analog-to-digital conversion, and this value is determined as the lower boundary of the voltage reference amplitude range. The highest value among the absolute values ​​of positive and negative voltage peak values ​​is superimposed with the amplitude corresponding to the upper limit of voltage measurement error and the minimum resolution of analog-to-digital conversion, and this value is determined as the upper boundary of the voltage reference amplitude range.

[0134] It should be noted that within the transient observation period, when the frequencies corresponding to voltage data at multiple acquisition moments are all below the lower boundary of the voltage reference frequency range, the range of acquisition moments formed by these multiple acquisition moments is defined as the oscillation continuation section. This is because the precursor to the ferroresonance of the voltage transformer usually causes the voltage data to exhibit oscillation components below the frequency reference before the start of the disturbance after the disturbance. When determining multiple acquisition moments, each acquisition moment within the transient observation period is first taken as the current acquisition moment, and voltage data covering one voltage waveform cycle is read backward from the current acquisition moment as the end position. The zero-crossing acquisition moments in the same direction are then searched among the read voltage data, and the sampling time is determined based on the number of sampling points between the zero-crossing acquisition moments in the same direction and the sampling interval. The frequency corresponding to the current acquisition time is obtained by inter-interval conversion; when the frequency corresponding to the current acquisition time is lower than the lower boundary of the voltage reference frequency range, the current acquisition time is recorded as an acquisition time that meets the low-frequency judgment condition; according to the acquisition time number order, acquisition times with adjacent acquisition time numbers and consecutive acquisition times that meet the low-frequency judgment condition are merged into the same acquisition time range; when the merged acquisition time range contains more than two acquisition times, the merged acquisition time range is determined as the oscillation continuation segment; using the acquisition time range formed by multiple acquisition times for determination can eliminate misjudgment caused by the zero-crossing identification deviation of a single acquisition time, and retain the start and end positions of the low-frequency oscillation within the transient observation segment.

[0135] It should be noted that, within the transient observation period, when the voltage amplitude at multiple acquisition moments is higher than the upper boundary of the voltage reference amplitude range, and the voltage amplitude at the later acquisition moment is greater than or equal to the voltage amplitude at the previous acquisition moment, the acquisition moment range formed by the multiple acquisition moments is defined as the overvoltage rise section. This is because when the precursor of the voltage transformer ferroresonance enters the enhancement stage, the voltage data will not only exceed the upper limit of the voltage amplitude reference before the disturbance start point, but will also form an upward process in time sequence with no decrease in amplitude. When determining multiple acquisition moments, the instantaneous voltage sample value of each acquisition moment within the transient observation period is read according to the acquisition moment number, and the absolute value of the instantaneous voltage sample value is determined as the voltage amplitude of the corresponding acquisition moment. When the voltage amplitude corresponding to the acquisition moment is higher than the upper boundary of the voltage reference amplitude range, the acquisition moment is recorded as the acquisition moment that meets the boundary judgment condition.

[0136] Next, adjacent acquisition times that meet the boundary judgment conditions are compared according to the acquisition time number order. When the voltage amplitude of the later acquisition time is greater than or equal to the voltage amplitude of the previous acquisition time, the adjacent acquisition times are kept within the same acquisition time range. When the voltage amplitude of the later acquisition time is less than the voltage amplitude of the previous acquisition time, the current acquisition time range ends at the previous acquisition time, and the judgment starts again from the later acquisition time. When the acquisition time range contains more than two acquisition times, the acquisition time range is determined as the overvoltage rise segment. The overvoltage rise segment is determined by using two conditions: the voltage amplitude exceeds the upper boundary of the overvoltage reference amplitude range and the amplitude does not decrease under the acquisition time sequence. This can distinguish between a single amplitude boundary violation and an overvoltage rise process with time extension, providing an amplitude basis and a time basis for determining the second time sequence anchor point.

[0137] In an optional embodiment, the method for determining the amplitude swing segment includes: dividing the voltage data within the transient observation segment into multiple half-cycle voltage data according to the zero-crossing acquisition time within the transient observation segment; reading the item with the highest absolute value of voltage data in each half-cycle voltage data, and determining the item with the highest absolute value of voltage data as the half-cycle voltage amplitude of the corresponding half-cycle voltage data; marking the acquisition time range corresponding to the half-cycle voltage data whose half-cycle voltage amplitude is higher than the upper boundary of the voltage reference amplitude range as the upper boundary half-cycle range; marking the acquisition time range corresponding to the half-cycle voltage data whose half-cycle voltage amplitude is lower than the lower boundary of the voltage reference amplitude range as the lower boundary half-cycle range; counting the occurrence order of the upper boundary half-cycle range and the lower boundary half-cycle range according to the acquisition time order; when the upper boundary half-cycle range and the lower boundary half-cycle range alternate in multiple acquisition time ranges, determining the multiple acquisition time ranges as amplitude swing segments.

[0138] In this embodiment, when determining the amplitude swing segment, voltage data is first read according to the acquisition time number within the transient observation segment. Then, the positive zero-crossing acquisition time when the voltage data changes from a non-positive value to a positive value, and the reverse zero-crossing acquisition time when the voltage data changes from a non-negative value to a negative value, are searched within the transient observation segment. According to the acquisition time sequence, the voltage data between two adjacent zero-crossing acquisition times is divided into a half-cycle voltage data. The absolute value of the voltage data corresponding to each acquisition time in each half-cycle voltage data is read, and the item with the highest absolute value is determined as the half-cycle voltage amplitude of the corresponding half-cycle voltage data. The half-cycle voltage amplitude is used to represent the amplitude level of the corresponding half-cycle voltage data; the absolute value of the voltage data at a single acquisition time near the zero crossing is not used as the basis for amplitude swing judgment.

[0139] In this embodiment, the half-cycle voltage amplitude is compared with the upper and lower boundaries of the voltage reference amplitude range. When the half-cycle voltage amplitude is higher than the upper boundary of the voltage reference amplitude range, the acquisition time range covered by the corresponding half-cycle voltage data is marked as the upper out-of-bounds half-cycle range. When the half-cycle voltage amplitude is lower than the lower boundary of the voltage reference amplitude range, the acquisition time range covered by the corresponding half-cycle voltage data is marked as the lower out-of-bounds half-cycle range. When the half-cycle voltage amplitude is between the lower boundary and the upper boundary of the voltage reference amplitude range, the acquisition time range covered by the corresponding half-cycle voltage data is not marked as either the upper or lower out-of-bounds half-cycle range. The upper out-of-bounds half-cycle range and the lower out-of-bounds half-cycle range are arranged according to the acquisition time order. The lower bound half-cycle range is determined, and when the bounds of two adjacent marked half-cycle ranges are different, the two adjacent marked half-cycle ranges are included in the same amplitude swing candidate range; when the same amplitude swing candidate range contains more than two marked half-cycle ranges, and the bounds of the marked half-cycle ranges alternate between the upper bound half-cycle range and the lower bound half-cycle range according to the acquisition time sequence, the range between the start acquisition time and the end acquisition time of the same amplitude swing candidate range is determined as the amplitude swing segment; by judging the upper bound half-cycle range and the lower bound half-cycle range by the half-cycle voltage amplitude, it is possible to avoid mistakenly marking the acquisition time when the absolute value of the voltage data near the zero crossing of the sine wave is lower than the lower boundary of the voltage reference amplitude range as the lower bound acquisition time.

[0140] It should be noted that by extracting current and voltage characteristics, the criteria for determining the transient saturation side of the current transformer and the precursor side of the ferroresonance of the voltage transformer can be formed separately within the same transient observation period. The waveform clipping position in the current characteristics is used to record the acquisition time range where the current data forms a clipping state near the current amplitude boundary, providing a time position for determining whether transient saturation of the current transformer has occurred. The phase distortion direction in the current characteristics is used to record the advance, lag, or no offset relationship between the waveform clipping position and the same polarity peak in the current reference half-cycle waveform, providing a phase basis for determining the direction of current phase change caused by transient saturation of the current transformer. The recovery lag section in the current characteristics is used to record the acquisition time range between the end of the waveform clipping position and the completion of recovery, providing an end basis for determining whether the influence of transient saturation of the current transformer has subsided. The amplitude swing section in the voltage characteristics is used to record the voltage data around the upper boundary of the voltage reference amplitude range and the voltage reference... The range of acquisition times where the lower boundary of the amplitude range alternately crosses the limit provides an amplitude basis for judging whether there is an amplitude reciprocating change in the precursor of the voltage transformer ferroresonance. The oscillation continuation segment in the voltage characteristics is used to record the acquisition time range where the frequency corresponding to the voltage data is lower than the lower boundary of the voltage reference frequency range, and provides a frequency basis for judging whether there is a low-frequency oscillation process in the precursor of the voltage transformer ferroresonance. The overvoltage rise segment in the voltage characteristics is used to record the acquisition time range where the voltage amplitude is higher than the upper boundary of the voltage reference amplitude range and has not decreased according to the acquisition time sequence, and provides an amplitude change basis for judging whether the precursor of the voltage transformer ferroresonance has entered the overvoltage enhancement process. The first time sequence anchor point is determined based on the waveform clipping position and the phase distortion direction, and the second time sequence anchor point is determined based on the oscillation continuation segment and the overvoltage rise segment. Combined with the start and end times of the recovery lag segment, amplitude swing segment, oscillation continuation segment and overvoltage rise segment, the time judgment basis for current risk mark, voltage risk mark and composite risk warning is formed.

[0141] Transient events are segmented based on the later end time of the disturbance in the recovery lag section and the overvoltage rise section.

[0142] In this embodiment, a transient observation segment is first formed, then current and voltage features are extracted within the transient observation segment, and finally, transient event segments are formed based on the extracted recovery hysteresis segment and overvoltage rise segment. The transient observation segment is used to provide the initial processing range required for feature extraction. The transient event segments are used to provide the timing comparison range of the first timing anchor point, the second timing anchor point, the current risk marker, the voltage risk marker, and the composite risk warning. By setting the transient observation segment before feature extraction and the transient event segments after feature extraction, it is possible to avoid requiring the identification of the above segments in advance within the transient observation segment before the waveform clipping position, recovery hysteresis segment, oscillation continuation segment, and overvoltage rise segment have been determined.

[0143] It should be noted that the disturbance start point is the acquisition time at which the current or voltage data first deviates from the current reference state within the same acquisition time sequence. The disturbance start point is used to indicate the time position at which the same power grid event begins to affect the current or voltage data. Using the disturbance start point as the starting boundary of the transient event segmentation allows subsequent recovery lag segments, oscillation continuation segments, overvoltage rise segments, first time-series anchor points, and second time-series anchor points to be compared within the same time range. When the current characteristic appears first, the disturbance start point corresponds to the initial influence time of the transient saturation side of the current transformer; when the voltage characteristic appears first, the disturbance start point corresponds to the initial influence time of the ferroresonant precursor side of the voltage transformer. By retaining the first occurrence time of any characteristic as the disturbance start point, it is possible to avoid the exclusion of early anomalies in another channel from the transient event segmentation caused by determining the start time solely from the current data or solely from the voltage data.

[0144] In an optional embodiment, the method for forming transient event segments includes: extracting current and voltage characteristics within a transient observation segment, and determining the transient event segment as the interval between the disturbance start point and the later end time of the recovery hysteresis segment and the overvoltage rise segment; when the transient observation segments corresponding to two adjacent disturbance start points overlap in time, merging the two overlapping transient observation segments into a merged transient observation segment with the start time preceding the end time, and re-forming transient event segments based on the merged transient observation segment; when the transient observation segments corresponding to two adjacent disturbance start points do not overlap in time, retaining two transient observation segments respectively, and forming two transient event segments respectively.

[0145] In this embodiment, when forming a transient event segment based on the later end time of the disturbance start point to the recovery lag section and the overvoltage rise section, the sampling time number corresponding to the disturbance start point is first read, and this sampling time number is determined as the starting sampling time number of the transient event segment; then, the end sampling time number of the recovery lag section and the end sampling time number of the overvoltage rise section are read; when the end time corresponding to the end sampling time number of the recovery lag section is later than the end time corresponding to the end sampling time number of the overvoltage rise section, the end sampling time number of the recovery lag section is determined as the end sampling time number of the transient event segment; when the overvoltage rise section... When the end time of the overvoltage rise segment is later than the end time of the recovery lag segment, the end time of the overvoltage rise segment is determined as the end time of the transient event segment. When the end time of the recovery lag segment is the same as the end time of the overvoltage rise segment, either end time is determined as the end time of the transient event segment. The current and voltage data between the start and end time of the transient event segment are included in the transient event segment.

[0146] It should be noted that the transient event segmentation is determined by the above method because the recovery process of the current transformer after transient saturation still exists before the end of the recovery lag segment, and the overvoltage rise process in the pre-resonance of the voltage transformer still exists before the end of the overvoltage rise segment. Using the later of the two end times as the end boundary of the transient event segmentation can simultaneously cover the current-side recovery process and the voltage-side overvoltage rise process, avoiding premature interruption of the same power grid event before the abnormal process on either side has ended.

[0147] It should be noted that by segmenting transient events, current and voltage data are first confined to the same disturbance start point to the same end boundary. Within the same transient observation segment, the waveform clipping position, phase distortion direction, recovery lag segment, amplitude swing segment, oscillation continuation segment, and overvoltage rise segment all correspond to the same acquisition time number sequence. Therefore, the first and second time sequence anchor points can be compared based on the same time reference to determine the order of their start times. The end times of the recovery lag segment, oscillation continuation segment, and overvoltage rise segment can also be compared based on the same time reference to determine the order of their end times. Based on the above comparisons, it is possible to distinguish between cases where transient saturation of the current transformer occurs first and the voltage characteristics subsequently increase, and cases where the amplitude swing segment occurs first and the current characteristics subsequently increase. As a result, current risk markers, voltage risk markers, and composite risk warnings are no longer generated based on the boundary crossing results of a single channel, but rather based on the temporal relationship between the two types of characteristics within the same transient observation segment, thereby reducing the confusion between ordinary transient disturbances and composite risk links.

[0148] The earliest time range in which the waveform clipping position and phase distortion direction coexist is taken as the first time series anchor point, and the earliest time range in which the oscillation continuation segment and the overvoltage rise segment coexist is taken as the second time series anchor point.

[0149] In an optional embodiment, the method for determining the first timing anchor point and the second timing anchor point includes: reading the waveform clipping position with the determined phase distortion direction according to the order of the acquisition time numbers; taking the acquisition time number covered by the waveform clipping position as the first candidate acquisition time number; merging multiple acquisition time numbers that are adjacent in number and time sequence among the first candidate acquisition time numbers to form a first coexistence interval; and taking the first coexistence interval with the earliest start time as the first timing anchor point; and selecting acquisition time numbers that simultaneously belong to the oscillation continuation segment and the overvoltage rise segment according to the order of the acquisition time numbers; merging multiple acquisition time numbers that are adjacent in number and time sequence among the selected acquisition time numbers to form a second coexistence interval; and taking the second coexistence interval with the earliest start time as the second timing anchor point.

[0150] It should be noted that the waveform clipping position is a range of acquisition times formed by multiple acquisition times; the phase distortion direction is determined by the sequential relationship between the initial acquisition time of the waveform clipping position and the acquisition time of the peak of the same polarity in the current reference half-cycle waveform; the phase distortion direction corresponds to the entire waveform clipping position, rather than being generated separately for each acquisition time number within the waveform clipping position; therefore, when determining the first timing anchor point, it is first confirmed that the phase distortion direction of the waveform clipping position has been determined, and then the acquisition time number covered by the waveform clipping position is used as the first candidate acquisition time number, and multiple acquisition time numbers that are adjacent in number and time sequence in the first candidate acquisition time number are merged into the first coexisting interval; in this way, the first timing anchor point indicates that the current-side time range that simultaneously has the waveform clipping position and phase distortion direction is not required to generate the phase distortion direction separately for each acquisition time number.

[0151] It should be noted that the earliest time range in which the oscillation continuation segment and the overvoltage rise segment coexist is used as the second time series anchor point because the oscillation continuation segment only indicates that the frequency corresponding to the voltage data is below the lower boundary of the voltage reference frequency range, and cannot alone indicate that the voltage amplitude has entered the overvoltage rise process; the overvoltage rise segment only indicates that the voltage amplitude is above the upper boundary of the voltage reference amplitude range and has not decreased according to the acquisition time sequence, and cannot alone indicate that the voltage data has entered a low-frequency oscillation process; only when the oscillation continuation segment and the overvoltage rise segment coexist at the same acquisition time number can the timing be confirmed as the same. The voltage data corresponding to a single acquisition time number simultaneously meets the requirements for judging low-frequency oscillation and overvoltage rise. Therefore, the acquisition time numbers that simultaneously exist in the oscillation continuation segment and the overvoltage rise segment are merged to form a second coexisting interval. The second coexisting interval with the earliest start time is used as the second time sequence anchor point. This allows the enhancement process of the voltage transformer ferroresonance precursor to be mapped to a specific acquisition time range. In addition, if there is no acquisition time number that simultaneously belongs to the oscillation continuation segment and the overvoltage rise segment, the second time sequence anchor point is not generated, and current risk markers are not generated based on the segmentation of this transient event.

[0152] It should be noted that after determining the first and second time-series anchor points respectively, the start time of the first time-series anchor point can represent the transient saturation side of the current transformer, and simultaneously possess the earliest time position of the clipping state and phase change; the start time of the second time-series anchor point can represent the precursor side of the ferroresonance of the voltage transformer, and simultaneously possess the earliest time position of the low-frequency oscillation process and the overvoltage rise process; when the start time of the first time-series anchor point is earlier than the start time of the second time-series anchor point, it can be deduced that the transient saturation side of the current transformer enters the identifiable state first according to the same acquisition time number sequence; when the start time of the second time-series anchor point is earlier than the start time of the first time-series anchor point, it can be deduced that the precursor side of the ferroresonance of the voltage transformer enters the identifiable state first according to the same acquisition time number sequence; therefore, when generating current risk markers and voltage risk markers subsequently, the judgment is no longer based on the occurrence time of a single feature, but on the start time of the coexistence interval of the two types of features, so that the cross-channel time-series correlation results can reflect the sequential relationship between current features and voltage features.

[0153] When the start time of the first time-series anchor point is earlier than the start time of the second time-series anchor point, and the end time of the oscillation continuation section or the overvoltage rise section is later than the end time of the recovery lag section, a current risk marker is generated.

[0154] In an optional embodiment, the method for generating a current risk marker includes: calculating the time difference between the start time of the second timing anchor point and the start time of the first timing anchor point to obtain a current leader interval; calculating the time difference between the later end time of the oscillation continuation segment and the end time of the overvoltage rise segment and the end time of the recovery hysteresis segment to obtain a voltage continuation interval; and generating a current risk marker when both the current leader interval and the voltage continuation interval are positive.

[0155] In this embodiment, the current risk marker is the risk identification result generated within the same transient event segment. The current risk marker indicates that the current characteristic enters the determinate state before the voltage characteristic, and the voltage characteristic is still retained after the recovery hysteresis segment ends. When generating the current risk marker, the transient event segment identifier, the start time of the first time-series anchor point, the start time of the second time-series anchor point, the end time of the recovery hysteresis segment, the end time of the oscillation continuation segment, the end time of the overvoltage rise segment, the current leader interval, and the voltage continuation interval are all recorded in the current risk marker. The current risk marker is not used to represent the current transformer fault conclusion alone, but to represent the time relationship between the current transformer transient saturation side leader and the voltage transformer ferroresonance precursor side delay within the same transient event segment. When generating a composite risk warning later, the current risk marker is read, and the composite risk level is determined by combining the voltage risk marker and the composite risk parameters.

[0156] It should be noted that the current leader interval is obtained by calculating the time difference between the start time of the second time-series anchor point and the start time of the first time-series anchor point. This allows the sequential relationship between the first and second time-series anchor points to be converted into a comparable time quantity. When the current leader interval is positive, it indicates that the start time of the first time-series anchor point is earlier than the start time of the second time-series anchor point. In other words, the current data first shows the waveform clipping position and phase distortion direction simultaneously, while the voltage data shows the oscillation continuation section and the overvoltage rise section simultaneously. Therefore, the current leader interval is used to confirm that the current characteristic enters the determineable state first, excluding the case where the voltage characteristic enters the determineable state first.

[0157] It should be noted that by calculating the time difference between the later end time of the oscillation continuation segment and the end time of the overvoltage rise segment and the end time of the recovery lag segment, the voltage continuation interval can be obtained, which can convert the lag relationship of the voltage characteristic relative to the current characteristic recovery process into a comparable time quantity. When both the oscillation continuation segment and the overvoltage rise segment exist, the later end time of the oscillation continuation segment and the overvoltage rise segment is read for calculation. When the voltage continuation interval is positive, it indicates that the oscillation continuation segment or the overvoltage rise segment has not ended after the recovery lag segment ends. That is, after the transient saturation side of the current transformer has entered the recovery completion state, the ferroresonant side of the voltage transformer still has a low-frequency oscillation process or an overvoltage rise process.

[0158] It should be noted that when both the current leader interval and the voltage continuation interval are positive, the following time relationship can be derived from the sequence of acquisition time numbers within the same transient observation segment: the first time-series anchor point appears first, followed by the second time-series anchor point, the recovery lag segment ends first, and the oscillation continuation segment or overvoltage rise segment ends later. This time relationship indicates that within the same transient observation segment, the clipping state and phase change of the transient saturation side of the current transformer are formed first, followed by the low-frequency oscillation process and overvoltage rise process of the ferroresonant side of the voltage transformer, and the voltage characteristics are still retained after the current characteristics recover. Therefore, generating a current risk marker when both the current leader interval and the voltage continuation interval are positive can separate the risk link with current characteristic leader and accompanying voltage characteristic delay from the ordinary transient event segmentation, providing a current-side leader basis for the subsequent generation of composite risk warnings.

[0159] A voltage risk marker is generated when the start time of the amplitude swing section is earlier than the first appearance time of the phase distortion direction, and the first appearance time of the current characteristic is later than the start time of the amplitude swing section.

[0160] In this embodiment, the voltage risk marker is a risk identification result generated within the same transient observation segment. The voltage risk marker indicates that the amplitude swing segment enters the identifiable state before the current characteristic, and the current characteristic appears only after the amplitude swing segment appears. When generating the voltage risk marker, the transient event segment identifier, the start time of the amplitude swing segment, the first appearance time of the phase distortion direction, the first appearance time of the current characteristic, the voltage leader interval, and the current lag interval are all recorded in the voltage risk marker. The voltage risk marker is not used to represent the voltage transformer fault conclusion alone, but rather to represent the time relationship between the voltage transformer ferroresonant precursor side leader and the current transformer transient saturation side lag within the same transient observation segment. When generating a composite risk warning subsequently, the voltage risk marker is read, and the composite risk level is determined by combining the current risk marker and the composite risk parameters.

[0161] In an optional embodiment, the method for generating voltage risk markers includes: calculating the time difference between the initial acquisition time of the waveform clipping position in the first generation of phase distortion direction and the initial time of the amplitude swing segment to obtain the voltage leader interval; calculating the time difference between the first appearance time of the current characteristic and the initial time of the amplitude swing segment to obtain the current lag interval; and generating voltage risk markers when both the voltage leader interval and the current lag interval are positive.

[0162] It should be noted that when calculating the voltage leader interval, the start time of the amplitude swing segment is read first, followed by the first occurrence time of the phase distortion direction. The time difference between the first occurrence time of the phase distortion direction and the start time of the amplitude swing segment is then calculated. When the voltage leader interval is positive, it indicates that the start time of the amplitude swing segment is earlier than the first occurrence time of the phase distortion direction. In other words, the voltage data first appears during the process of exceeding the upper and lower limits of the voltage reference amplitude range, and the current data appears later with the phase distortion direction. Therefore, the voltage leader interval is used to confirm that the amplitude swing process in the voltage characteristics occurs before the current phase distortion, thus eliminating the situation where the current phase distortion occurs first and then causes the voltage side judgment.

[0163] It should be noted that when calculating the current lag interval, the start time of the amplitude swing segment is read first, followed by the first occurrence time of the current characteristic. The time difference between the first occurrence time of the current characteristic and the start time of the amplitude swing segment is then calculated. The first occurrence time of the current characteristic is determined by selecting the earliest acquisition time from the following three: the start time of waveform clipping position, the first occurrence time of phase distortion direction, and the start time of recovery lag segment. When the current lag interval is positive, it indicates that the first occurrence time of the current characteristic is later than the start time of the amplitude swing segment. In other words, the voltage data enters the amplitude swing process first, and the current data enters the characteristic occurrence process on the transient saturation side of the current transformer later. Therefore, the current lag interval is used to confirm that the current characteristic is not the first abnormal characteristic to appear within the same transient observation segment.

[0164] It should be noted that both the voltage leader interval and the current lag interval are positive. The following time relationship can be derived from the numbering sequence of acquisition times within the same transient observation segment: the amplitude swing segment appears first, the phase distortion direction appears later, and the first appearance of the current characteristic is also later than the start time of the amplitude swing segment. The above time relationship indicates that within the same transient observation segment, the amplitude swing process of the precursor side of the voltage transformer ferroresonance is formed first, followed by the phase change and current characteristics of the transient saturation side of the current transformer. Therefore, generating voltage risk markers when both the voltage leader interval and the current lag interval are positive can separate the risk link with voltage characteristic leader and current characteristic lag from the ordinary transient event segmentation, providing a voltage-side leader basis for the subsequent generation of composite risk warnings.

[0165] When current risk markers and voltage risk markers are generated in segments during the same transient event, a composite risk warning is generated.

[0166] In an optional embodiment, the method for generating a composite risk warning includes: confirming that both current risk markers and voltage risk markers have been generated within the same transient event segment; determining the composite risk level based on the time difference between the end time of the oscillation continuation segment and the end time of the recovery hysteresis segment, and the time difference between the end time of the overvoltage rise segment and the end time of the recovery hysteresis segment, and writing the composite risk parameters into the composite risk warning.

[0167] It should be noted that limiting the data to the same transient event segment is to ensure that both the current risk marker and the voltage risk marker correspond to the current and voltage data generated by the same power grid event. When forming the transient event segment, the data between the start of the disturbance and the end time of the later end of the current recovery lag segment and the overvoltage rise segment are included in the same time range. The current risk marker is used to record the time relationship where the start time of the first time-series anchor point is earlier than the start time of the second time-series anchor point, and the oscillation continuation segment or the overvoltage rise segment is later than the end of the current recovery lag segment. The voltage risk marker is used to record the time relationship where the start time of the amplitude swing segment is earlier than the first occurrence time of the phase distortion direction, and the first occurrence time of the current characteristic is later than the start time of the amplitude swing segment.

[0168] Therefore, the current risk marker and voltage risk marker do not represent the simultaneous existence of two mutually exclusive conclusions, a current-side leader and a voltage-side leader, within the same transient event segment. Instead, they represent two different levels of temporal relationships: the amplitude swing segment precedes the current distortion, and the current-side anchor point precedes the voltage enhancement anchor point. When both the current risk marker and voltage risk marker are generated within the same transient event segment, it confirms the simultaneous existence of an evolutionary chain within the same disturbance link, including early voltage amplitude swing, the current transformer transient saturation entering a determinable state, and the delayed enhancement process of the voltage transformer ferroresonance precursor. When the current risk marker and voltage risk marker are located in different transient event segments, they correspond to different disturbance start points or different end boundaries. The current risk markers and voltage risk markers in different transient event segments are not merged into a composite risk warning. By limiting the same transient event segment, it is possible to avoid merging current-side anomalies and voltage-side anomalies in two independent grid events into a composite risk warning.

[0169] In this embodiment, when determining the composite risk level based on composite risk parameters, the current risk marker and voltage risk marker within the same transient event segment are read first. If neither the current risk marker nor the voltage risk marker is read simultaneously within the same transient event segment, the composite risk level determination process is not initiated. If both the current risk marker and the voltage risk marker are read simultaneously within the same transient event segment, the start time of the first time-series anchor point and the start time of the second time-series anchor point are read, and the time difference between the start time of the second time-series anchor point and the start time of the first time-series anchor point is written into the composite risk parameter record. The end time of the oscillation continuation segment and the end time of the recovery lag segment are read, and the time difference between the end time of the oscillation continuation segment and the end time of the recovery lag segment is written into the composite risk parameter record. The end time of the overvoltage rise segment and the end time of the recovery lag segment are read, and the time difference between the end time of the overvoltage rise segment and the end time of the recovery lag segment is written into the composite risk parameter record. The first occurrence time of the current characteristic and the start time of the amplitude swing segment are read, and the time difference between the first occurrence time of the current characteristic and the start time of the amplitude swing segment is written into the composite risk parameter record.

[0170] In this embodiment, a pre-established order of composite risk levels is defined, including a first composite risk level, a second composite risk level, and a third composite risk level. The first composite risk level corresponds to the state where both current risk markers and voltage risk markers exist simultaneously within the same transient event segment, the end time of the oscillation continuation segment is later than the end time of the recovery lag segment, and the end time of the overvoltage rise segment is not later than the end time of the recovery lag segment. The second composite risk level corresponds to the state where both current risk markers and voltage risk markers exist simultaneously within the same transient event segment, the end time of the overvoltage rise segment is later than the end time of the recovery lag segment, and the end time of the oscillation continuation segment is not later than the end time of the recovery lag segment. The third composite risk level corresponds to the state where both current risk markers and voltage risk markers exist simultaneously within the same transient event segment, the end time of the oscillation continuation segment is later than the end time of the recovery lag segment, and the end time of the overvoltage rise segment is later than the end time of the recovery lag segment.

[0171] After reading the composite risk parameter records, first determine whether the time difference between the end time of the oscillation continuation segment and the end time of the recovery lag segment is positive, and then determine whether the time difference between the end time of the overvoltage rise segment and the end time of the recovery lag segment is positive. If the time difference between the end time of the oscillation continuation segment and the end time of the recovery lag segment is positive, and the time difference between the end time of the overvoltage rise segment and the end time of the recovery lag segment is not positive, the composite risk level is determined as the first composite risk level. If the time difference between the end time of the oscillation continuation segment and the end time of the recovery lag segment is not positive, and the time difference between the end time of the overvoltage rise segment and the end time of the recovery lag segment is not positive, the composite risk level is determined as the first composite risk level. When the time difference between the end of the subsequent segment and the end of the recovery lag segment is positive, the composite risk level is determined as the second composite risk level. When the time difference between the end of the oscillation continuation segment and the end of the recovery lag segment is positive, and the time difference between the end of the overvoltage rise segment and the end of the recovery lag segment is positive, the composite risk level is determined as the third composite risk level. When the time difference between the end of the oscillation continuation segment and the end of the recovery lag segment is not positive, and the time difference between the end of the overvoltage rise segment and the end of the recovery lag segment is not positive, no composite risk level is generated, and the corresponding transient event is segmented and recorded as a continuation relationship mismatch record.

[0172] In this embodiment, when generating a composite risk warning based on a composite risk level, the determined composite risk level is first read, and the transient event segment identifier corresponding to the composite risk level is also read. Then, the time difference between the first and second time-series anchor points, the time difference between the end of the oscillation continuation segment and the end of the recovery lag segment, the time difference between the end of the overvoltage rise segment and the end of the recovery lag segment, and the time difference between the first occurrence of the current characteristic and the start of the amplitude swing segment are read from the composite risk parameter record. Subsequently, the transient event segment identifier, composite risk level, time difference between the first and second time-series anchor points, time difference between the end of the oscillation continuation segment and the end of the recovery lag segment, time difference between the end of the overvoltage rise segment and the end of the recovery lag segment, time difference between the first occurrence of the current characteristic and the start of the amplitude swing segment, start time of the first time-series anchor point, start time of the second time-series anchor point, end time of the recovery lag segment, end time of the oscillation continuation segment, and end time of the overvoltage rise segment are written into the composite risk warning.

[0173] When the composite risk level is the first composite risk level, a composite risk warning is output to the event list of the remote monitoring interface; when the composite risk level is the second composite risk level, a composite risk warning is output to the event list of the remote monitoring interface, and a review prompt is sent to the operation and maintenance terminal; when the composite risk level is the third composite risk level, a composite risk warning is output to the event list of the remote monitoring interface, a review prompt is sent to the operation and maintenance terminal, and the corresponding combined current transformer identifier is written into the operation and maintenance handling queue; the composite risk warning is used to enable the remote monitoring interface and the operation and maintenance terminal to read the current-side time relationship, voltage-side time relationship, and composite risk level within the same transient event segment.

[0174] In summary, this application, by synchronously acquiring current and voltage data and extracting current and voltage features under the same acquisition time number, enables the waveform clipping position, phase distortion direction, and recovery lag section of the transient saturation side of the current transformer to be compared with the amplitude swing section, oscillation continuation section, and overvoltage rise section of the ferroresonant side of the voltage transformer within the same transient event segment. Furthermore, this application determines the time position when the current and voltage features enter a determinable state through a first and second time-series anchor point, and records the current-side leader relationship and voltage-side leader relationship through current and voltage risk markers, respectively. When both the current and voltage risk markers are in the same transient event segment, this application determines the composite risk level based on composite risk parameters and generates a composite risk warning. This ensures that remote monitoring and fault early warning of the combined transformer no longer relies solely on a single current out-of-bounds, single voltage out-of-bounds, or single abnormal duration, but outputs a composite risk warning based on the time-series correlation between current and voltage features.

[0175] The above description is merely a preferred embodiment of this application. The scope of protection of this application is not limited to the above embodiments. All technical solutions falling within the scope of this application's concept are within the scope of protection of this application. It should be noted that for those skilled in the art, any improvements and modifications made without departing from the principles of this application should also be considered within the scope of protection of this application.

Claims

1. A combined instrument transformer remote monitoring and fault early warning system, characterized in that, include: The acquisition module is used to simultaneously acquire current and voltage data and determine the current reference state; Determine the start point of the disturbance based on the current reference state; Starting from the point of disturbance, current and voltage data are read sequentially to form a transient observation segment; The feature module is used to extract current and voltage features within the transient observation segment. The current features include waveform clipping position, phase distortion direction, and recovery hysteresis segment. The voltage features include amplitude swing segment, oscillation continuation segment, and overvoltage rise segment. The segmentation module is used to form transient event segments based on the later end time of the disturbance start point to the recovery hysteresis segment and the overvoltage rise segment. Anchor point module is used to take the earliest time range in which the waveform clipping position and phase distortion direction coexist as the first time anchor point, and the earliest time range in which the oscillation continuation segment and the overvoltage rise segment coexist as the second time anchor point. The marking module is used to generate a current risk mark when the start time of the first time-series anchor point is earlier than the start time of the second time-series anchor point, and the end time of the oscillation continuation section or the overvoltage rise section is later than the end time of the recovery hysteresis section; and to generate a voltage risk mark when the start time of the amplitude swing section is earlier than the first occurrence time of the phase distortion direction, and the first occurrence time of the current characteristic is later than the start time of the amplitude swing section. The warning module is used to generate a composite risk warning when current risk markers and voltage risk markers are generated in the same transient event segment.

2. The combined instrument transformer remote monitoring and fault early warning system according to claim 1, characterized in that, Methods for extracting waveform clipping position and phase distortion direction include: A current data of one rated power frequency cycle length is intercepted before the start of the disturbance, and the current amplitude boundary and the current reference half-cycle waveform are determined based on the intercepted current data. When the difference between the current data and the current amplitude boundary at multiple acquisition times within the transient observation segment does not exceed the preset clipping difference, the acquisition time range formed by multiple acquisition times is determined as the waveform clipping position. Compare the timing of the initial acquisition of the waveform clipping position with the timing of the acquisition of the peak of the same polarity in the current reference half-cycle waveform, and determine the direction of phase distortion based on the timing relationship.

3. The combined instrument transformer remote monitoring and fault early warning system according to claim 1, characterized in that, Methods for determining the recovery lag segment include: The current data of one rated power frequency cycle length is intercepted before the start of the disturbance to determine the allowable range of peak phase offset before the start of the disturbance. Harmonic component separation is performed on the current data within the transient observation segment to obtain the high-order harmonic concentration section; Starting from the acquisition time after the end of the waveform clipping position, determine one by one whether the acquisition time does not belong to the high-order harmonic concentration section, and determine whether the sampling point interval between the current half-cycle same polarity peak acquisition time and the current half-cycle start acquisition time falls within the allowable range of peak phase offset before the disturbance start point. The earliest acquisition time that simultaneously meets both judgment conditions is determined as the recovery completion time; The interval between the end of the waveform clipping position and the completion of the recovery is defined as the recovery lag segment.

4. The combined instrument transformer remote monitoring and fault early warning system according to claim 1, characterized in that, Methods for determining the oscillation continuation section and the overvoltage rise section include: Voltage data of a rated power frequency cycle length is intercepted before the start of the disturbance, and the voltage reference frequency range and voltage reference amplitude range are determined based on the intercepted voltage data. Within the transient observation period, when the frequencies corresponding to the voltage data at multiple acquisition moments are all below the lower boundary of the voltage reference frequency range, the range of acquisition moments formed by the multiple acquisition moments is defined as the oscillation continuation segment. Within the transient observation period, when the voltage amplitude at multiple acquisition moments is higher than the upper boundary of the voltage reference amplitude range, and the voltage amplitude at the later acquisition moment is greater than or equal to the voltage amplitude at the previous acquisition moment, the acquisition moment range formed by the multiple acquisition moments is defined as the overvoltage rise zone.

5. The combined instrument transformer remote monitoring and fault early warning system according to claim 4, characterized in that, Methods for determining amplitude swing ranges include: Based on the zero-crossing acquisition time within the transient observation period, the voltage data within the transient observation period is divided into multiple half-cycle voltage data. Read the voltage data with the highest absolute value in each half-cycle voltage data, and determine the voltage data with the highest absolute value as the half-cycle voltage amplitude of the corresponding half-cycle voltage data. The range of acquisition times corresponding to half-cycle voltage data whose half-cycle voltage amplitude is higher than the upper boundary of the voltage reference amplitude range is marked as the upper boundary half-cycle range. The range of acquisition times corresponding to half-cycle voltage data whose half-cycle voltage amplitude is lower than the lower boundary of the voltage reference amplitude range is marked as the lower out-of-bounds half-cycle range. The order in which the upper and lower bounded half-circle ranges appear is determined according to the order of data collection time. When the upper and lower bound half-cycle ranges alternate within multiple acquisition time ranges, these multiple acquisition time ranges are defined as amplitude swing segments.

6. The combined instrument transformer remote monitoring and fault early warning system according to claim 1, characterized in that, Methods for determining the initiation point of a disturbance include: In the sequence of acquisition time numbers, for each current acquisition time, the current data and voltage data of the rated power frequency cycle length before the current acquisition time are read, and the current reference state is determined based on the read current data and voltage data; If the current data at the current acquisition time does not belong to the current reference state, or the voltage data at the current acquisition time does not belong to the current reference state, the current acquisition time will be determined as the disturbance start point.

7. The combined instrument transformer remote monitoring and fault early warning system according to claim 6, characterized in that, Methods for forming transient observation segments include: The current reference state corresponding to one rated power frequency cycle length before the start of the disturbance is determined as the reference state before the disturbance. Read current and voltage data starting from the point of disturbance. When the current data within the next rated power frequency cycle is all in the reference state before the disturbance, and the voltage data within the next rated power frequency cycle is all in the reference state before the disturbance, the start time of the next rated power frequency cycle is determined as the end time of the observation. The interval between the start of the disturbance and the end of the observation is defined as the transient observation segment.

8. The combined instrument transformer remote monitoring and fault early warning system according to claim 1, characterized in that, Methods for forming transient event segments include: After extracting current and voltage characteristics within the transient observation period, the interval between the disturbance start point and the later end time in the recovery hysteresis section and the overvoltage rise section is determined as the transient event segment. When the transient observation segments corresponding to two adjacent disturbance initiations overlap in time, the two overlapping transient observation segments are merged into a merged transient observation segment with the start time preceding the end time, and transient event segments are re-formed based on the merged transient observation segment. When the transient observation segments corresponding to two adjacent disturbance initiations do not overlap in time, two transient observation segments are retained respectively, and two transient event segments are formed respectively.

9. The combined instrument transformer remote monitoring and fault early warning system according to claim 1, characterized in that, The methods for determining the first and second time series anchor points include: According to the order of the acquisition time numbers, read the waveform clipping position with the determined phase distortion direction, take the acquisition time number covered by the waveform clipping position as the first candidate acquisition time number, merge multiple acquisition time numbers that are adjacent in number and time sequence in the first candidate acquisition time number to form the first coexistence interval, and take the first coexistence interval with the earliest start time as the first time sequence anchor point. According to the order of the acquisition time numbers, the acquisition time numbers that belong to both the oscillation continuation segment and the overvoltage rise segment are selected. Multiple acquisition time numbers that are adjacent in number and time sequence are merged to form a second coexisting interval. The second coexisting interval with the earliest start time is used as the second time sequence anchor point.

10. The combined instrument transformer remote monitoring and fault early warning system according to claim 1, characterized in that, Methods for generating current risk markers include: The time difference between the start time of the second timing anchor point and the start time of the first timing anchor point is calculated to obtain the current leader interval; The voltage duration interval is obtained by calculating the time difference between the later end time of the oscillation duration interval and the end time of the overvoltage rise interval, and the end time of the recovery lag interval. A current risk marker is generated when both the current leader interval and the voltage duration interval are positive.

11. The combined instrument transformer remote monitoring and fault early warning system according to claim 1, characterized in that, Methods for generating voltage risk markers include: The voltage leader interval is obtained by calculating the time difference between the initial acquisition time of the waveform clipping position in the first generation phase distortion direction and the initial acquisition time of the amplitude swing segment. The time difference between the first appearance of the current characteristic and the start of the amplitude swing segment is calculated to obtain the current hysteresis interval. A voltage risk marker is generated when both the voltage leader interval and the current hysteresis interval are positive.

12. The combined instrument transformer remote monitoring and fault early warning system according to claim 1, characterized in that, Methods for generating compound risk warnings include: Within the same transient event segment, it was confirmed that both current risk markers and voltage risk markers had been generated. The composite risk level is determined based on the time difference between the end of the oscillation continuation segment and the end of the recovery lag segment, and the time difference between the end of the overvoltage rise segment and the end of the recovery lag segment. Generate composite risk warnings based on composite risk levels.

Images