High-resistance grounding fault detection method and system based on zero-sequence voltage periodic differential energy
By calculating the zero-sequence voltage periodic differential energy and adaptively setting the threshold, the problem of high-resistance grounding faults being difficult to identify in the distribution network is solved, and reliable detection of high-resistance grounding faults is achieved.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- STATE GRID HENAN ELECTRIC POWER COMPANY ZHENGZHOU POWER SUPPLY CO
- Filing Date
- 2026-02-10
- Publication Date
- 2026-05-12
AI Technical Summary
In distribution networks where the neutral point is not effectively grounded, the fault current of high-resistance grounding faults is weak and its characteristics are not obvious. It is easily overwhelmed by normal load fluctuations and noise in the system, resulting in insufficient sensitivity of traditional protection devices and difficulty in reliable identification.
By calculating the zero-sequence voltage periodic differential energy of the three-phase voltage, a periodic differential sequence is generated, and squared and accumulated operations are performed. Combined with smoothing filtering, a periodic differential energy value characterizing the fault is obtained. A high-resistance grounding fault judgment signal is generated by continuously comparing the adaptively set dynamic threshold.
It amplifies the subtle distortion characteristics of high-resistance grounding faults, avoids noise interference, improves the stability and reliability of fault diagnosis, and provides reliable identification of high-resistance grounding faults in low-current grounding systems.
Smart Images
Figure CN122017455A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the field of power distribution network protection technology, and in particular to a high-resistivity grounding fault detection method and system based on zero-sequence voltage periodic differential energy. Background Technology
[0002] In the operation of distribution networks with non-effectively grounded neutral points, single-phase grounding faults are the most frequent type of fault. Their reliable detection is crucial for ensuring power supply safety and continuity. Among them, the identification of high-resistance grounding faults with weak fault currents and indistinct characteristics has been a long-standing technical challenge.
[0003] Existing detection methods mainly rely on steady-state analysis of the amplitude, phase, or harmonic components of zero-sequence voltage or zero-sequence current. However, the electrical signal changes generated by high-resistance grounding faults are extremely weak and easily drowned out by normal load fluctuations, three-phase imbalances, and measurement noise. This leads to technical problems such as insufficient sensitivity and difficulty in setting traditional protection devices, making it difficult to reliably identify high-resistance grounding faults in low-current grounding systems. Summary of the Invention
[0004] In view of this, the present invention provides a high-resistivity grounding fault detection method and system based on zero-sequence voltage periodic differential energy to solve the problems mentioned in the background art.
[0005] In a first aspect, this application provides a high-resistivity grounding fault detection method based on zero-sequence voltage periodic differential energy, comprising: The instantaneous values of the three-phase voltages are acquired in real time, and a continuous zero-sequence voltage discrete-time sequence is calculated based on the instantaneous values of the three-phase voltages. The zero-sequence voltage discrete-time sequence is periodically truncated and aligned, and the point-by-point difference between adjacent period zero-sequence voltage waveforms after alignment is calculated to generate a periodic difference sequence. The periodic difference sequence is squared and summed, and then smoothed and filtered to obtain the periodic difference energy value that characterizes the fault features. The periodic differential energy value is continuously compared with an adaptively set dynamic threshold. When the periodic differential energy value continuously exceeds the dynamic threshold for a preset time, a high-resistance grounding fault judgment signal and a corresponding control command are generated.
[0006] Secondly, this application provides a high-resistivity grounding fault detection system based on zero-sequence voltage periodic differential energy, comprising: The acquisition module is used to acquire the instantaneous values of the three-phase voltage in real time and calculate a continuous zero-sequence voltage discrete-time sequence based on the instantaneous values of the three-phase voltage. The first calculation module is used to perform periodic truncation and alignment processing on the zero-sequence voltage discrete time sequence, calculate the point-by-point difference of the zero-sequence voltage waveforms of adjacent periods after alignment, and generate a periodic difference sequence. The second calculation module is used to perform square and sum operations on all differences in the periodic difference sequence, and after smoothing and filtering, obtain the periodic difference energy value that characterizes the fault features. The comparison module is used to continuously compare the periodic differential energy value with an adaptively set dynamic threshold. When the periodic differential energy value continuously exceeds the dynamic threshold for a preset time, a high-resistance grounding fault judgment signal and a corresponding control command are generated.
[0007] This application provides a method and system for detecting high-resistance grounding faults based on the periodic differential energy of zero-sequence voltage. This method, on the one hand, amplifies the subtle distortion characteristics caused by high-resistance grounding faults, manifested as inconsistencies in waveforms between adjacent periods, by calculating the point-by-point difference of the zero-sequence voltage waveforms of adjacent periods and converting it into energy values, thus revealing the weak fault characteristics. On the other hand, by employing a dynamic threshold adaptively set based on the normal operating state of the system, the judgment criteria can be aligned with the actual background noise level of the system, avoiding misjudgments caused by environmental changes. Furthermore, by requiring the fault characteristic signal (periodic differential energy value) to continuously exceed the threshold for a preset duration, instantaneous interference pulses are filtered out, making the fault judgment conclusion more stable and reliable. In summary, this method solves the problems of insufficient sensitivity and difficulty in setting traditional steady-state quantity analysis methods due to weak fault signals and susceptibility to noise, providing a reliable way to identify high-resistance grounding faults in low-current grounding systems. Attached Figure Description
[0008] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.
[0009] Figure 1 A schematic flowchart illustrating the high-resistivity grounding fault detection method based on zero-sequence voltage periodic differential energy provided in this application embodiment; Figure 2 A schematic block diagram of the structure of a high-resistivity grounding fault detection system based on zero-sequence voltage periodic differential energy provided in an embodiment of this application. Detailed Implementation
[0010] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0011] The flowchart shown in the attached diagram is for illustrative purposes only and does not necessarily include all content and operations / steps, nor does it require execution in the described order. For example, some operations / steps can be broken down, combined, or partially merged, so the actual execution order may change depending on the actual situation.
[0012] It should also be understood that the terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the scope of the application. As used in this specification and the appended claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise.
[0013] It should also be further understood that the term “and / or” as used in this application specification and the appended claims means any combination of one or more of the relevant listed items and all possible combinations, and includes such combinations.
[0014] The following detailed description of some embodiments of this application is provided in conjunction with the accompanying drawings. Unless otherwise specified, the following embodiments and features can be combined with each other.
[0015] Please see Figure 1 , Figure 1 This is a flowchart illustrating the high-resistivity grounding fault detection method based on zero-sequence voltage periodic differential energy provided in an embodiment of this application, as shown below. Figure 1 As shown, the high-resistivity grounding fault detection method based on zero-sequence voltage periodic differential energy provided in this application includes steps S1 to S4.
[0016] Step S1: Real-time acquisition of the instantaneous values of the three-phase voltage, and calculation of the continuous zero-sequence voltage discrete time sequence based on the instantaneous values of the three-phase voltage.
[0017] In practice, voltage transformers are used to synchronously acquire instantaneous voltage signals from phases A, B, and C of the power system. At each fixed, very short time interval (i.e., sampling moment), the three acquired instantaneous voltage values are summed, and the sum is divided by three to obtain the zero-sequence voltage instantaneous value at that moment. This process is repeated continuously, and the zero-sequence voltage instantaneous values calculated at each sampling moment are arranged in chronological order to form a continuously changing digital sequence over time, i.e., the zero-sequence voltage discrete-time sequence.
[0018] Step S2: Perform periodic truncation and alignment processing on the zero-sequence voltage discrete time sequence, calculate the point-by-point difference of the zero-sequence voltage waveforms of adjacent periods after alignment, and generate a periodic difference sequence.
[0019] In practice, from the continuous zero-sequence voltage discrete-time sequence generated in step S1, the data segment of the latest complete power frequency cycle length is selected and marked as the current analysis cycle sequence. Simultaneously, the data segment of the immediately preceding complete power frequency cycle length is selected and marked as the reference cycle sequence. To accurately compare the waveforms of these two cycles, the current analysis cycle sequence needs to be phase-adjusted to align with the reference cycle sequence at its time starting point. After alignment, the values at the same time position in the two sequences are subtracted one by one to obtain a new sequence composed of the differences. This new sequence reflects the difference in the zero-sequence voltage waveform of two adjacent cycles at each sampling point, i.e., the periodic difference sequence.
[0020] Step S3: Perform a square and sum operation on all differences in the periodic difference sequence, and then perform a smoothing filter to obtain the periodic difference energy value that characterizes the fault features.
[0021] In practice, each difference in the periodic difference sequence generated in step S2 is multiplied (squared) to make all differences positive. Then, all these squared values are summed to obtain a total. This total represents the overall magnitude of the difference between adjacent period waveforms, i.e., the original difference energy. To eliminate instantaneous fluctuations in energy values caused by random interference, the original difference energy values calculated for multiple consecutive periods are averaged, for example, by taking the average of the most recent few periods. This stable value obtained after averaging is the periodic difference energy value.
[0022] Step S4: Continuously compare the periodic differential energy value with an adaptively set dynamic threshold. When the periodic differential energy value continuously exceeds the dynamic threshold for a preset time, a high-resistance grounding fault judgment signal and a corresponding control command are generated.
[0023] In practical implementation, a judgment criterion, namely a dynamic threshold, needs to be set first. This threshold is not a fixed value, but is automatically calculated based on multiple cycle differential energy values recorded during normal system operation. In subsequent real-time detection, each latest cycle differential energy value calculated in step S3 is compared with this dynamic threshold. If the cycle differential energy value is observed to be higher than the dynamic threshold for multiple consecutive cycles, it is judged that a high-resistance grounding fault has occurred. Once the fault is determined to have occurred, an electrical signal or logic command indicating the presence of the fault is generated. This signal or command can be used to trigger an alarm or drive the protective equipment to operate.
[0024] The method provided in this embodiment, on the one hand, amplifies the subtle distortion characteristics caused by high-resistance grounding faults, which manifest as inconsistent waveforms between adjacent periods, by calculating the point-by-point difference of the zero-sequence voltage waveforms of adjacent periods and converting it into energy values, thus revealing the weak fault characteristics. On the other hand, by employing a dynamic threshold adaptively set based on the normal operating state of the system, the judgment criteria can be aligned with the actual background noise level of the system, avoiding misjudgments caused by environmental changes. Furthermore, by requiring the fault characteristic signal (periodic differential energy value) to continuously exceed the threshold for a preset duration, instantaneous interference pulses are filtered out, making the fault judgment conclusion more stable and reliable. In summary, this method solves the problems of insufficient sensitivity and difficulty in setting traditional steady-state quantity analysis methods due to weak fault signals and susceptibility to noise, providing a reliable way to identify high-resistance grounding faults in low-current grounding systems.
[0025] In some embodiments, the real-time acquisition of instantaneous values of three-phase voltages and the calculation of a continuous zero-sequence voltage discrete-time sequence based on the instantaneous values of the three-phase voltages include: Step S11: Synchronously acquire the instantaneous signals of the three-phase phase voltages at a sampling frequency no less than twice the number of sampling points corresponding to the system's fundamental frequency period.
[0026] In practical implementation, a sampling frequency should be set, which should be at least twice the power system frequency. For example, for a 50 Hz system, the sampling frequency should be no less than 100 points per second. Using three voltage measurement channels, under the control of the same sampling clock, the instantaneous voltage values of phases A, B, and C are measured and recorded simultaneously to ensure that the three-phase data are completely synchronized in time.
[0027] Step S12: At each sampling moment, the instantaneous values of the three-phase voltages are algebraically summed and averaged to obtain the instantaneous sampled value of the zero-sequence voltage. In specific implementation, whenever the sampling clock triggers a new sampling, the readings of the three voltage measurement channels (phase A, phase B, and phase C) are acquired at that moment. These three readings are directly added together, and then the sum is divided by three. The calculated result is the instantaneous value of the zero-sequence voltage at the current sampling moment.
[0028] Step S13: The calculated continuous instantaneous zero-sequence voltage samples are stored in a first-in-first-out (FIFO) data buffer in chronological order, forming the zero-sequence voltage discrete-time series. In practice, a storage area with a fixed length, i.e., a data buffer, is established. Each time a new instantaneous zero-sequence voltage sample is calculated, it is stored at one end (the write end) of this buffer. When the buffer is full, the oldest stored data is automatically overwritten or removed to ensure that the latest continuous data is always stored in the buffer. This chronologically ordered and dynamically updated data set is the zero-sequence voltage discrete-time series used for subsequent analysis. The method provided in this embodiment, on the one hand, by setting a sampling rate no lower than the Nyquist frequency, fully preserves the key information of the power frequency signal and its fault transient components, providing a data foundation for subsequent accurate analysis; on the other hand, by synchronously acquiring the three-phase voltage and calculating the zero-sequence voltage in real time at each sampling moment, the timeliness of the fault characteristics is ensured, meeting the real-time requirements of the protection system; furthermore, by using a first-in-first-out data buffer to manage continuous sampling data, a stable and controllable-length analysis data window is formed, ensuring the continuity and feasibility of periodic interception operations.
[0029] In some embodiments, the step of periodically truncating and aligning the zero-sequence voltage discrete-time sequence, calculating the point-by-point difference of the aligned adjacent period zero-sequence voltage waveforms, and generating a periodic difference sequence includes: Step S21: Extract the data of the latest complete power frequency cycle from the data buffer of the zero-sequence voltage discrete time sequence as the current cycle sequence, and extract the data of the immediately preceding complete power frequency cycle as the reference cycle sequence.
[0030] In practice, based on the known power frequency cycle length and sampling frequency, the number of sampling points required to accurately encompass a complete power frequency cycle is calculated. Starting from the end of the data buffer maintained in step S13 (the latest data end), backtracking is performed to extract continuous data that meets the above point requirement, which serves as the current cycle sequence. Then, the same number of sampling points are backtracked to extract continuous data from the previous cycle, which serves as the reference cycle sequence.
[0031] Step S22: Perform cross-correlation operation on the current periodic sequence and the reference periodic sequence to determine the phase offset that maximizes the waveform similarity between the two, and perform cyclic shift alignment on the current periodic sequence based on the phase offset.
[0032] In practice, the cross-correlation coefficients of the current periodic sequence and the reference periodic sequence under different relative shifts are calculated. The shift amount corresponding to the maximum cross-correlation coefficient is the phase offset when the two sequence waveforms are most matched. Then, the data points of the current periodic sequence are cyclically shifted according to this offset (for example, moving some data from the beginning of the sequence to the end, or vice versa), so that the shifted current periodic sequence and the reference periodic sequence achieve optimal alignment in the initial phase of the waveform.
[0033] Step S23: Calculate the numerical difference between the aligned current periodic sequence and the reference periodic sequence at each corresponding sampling point, thereby generating the periodic difference sequence.
[0034] In practice, after waveform alignment, the first data point of the reference periodic sequence is subtracted from the first data point of the aligned current periodic sequence to obtain the first difference; the second data point is subtracted from the second data point to obtain the second difference; and so on, subtracting each corresponding data point with the same index from the two sequences. This series of differences, arranged in order, constitutes the periodic difference sequence.
[0035] The method provided in this embodiment, on the one hand, provides clear, paired input objects for subsequent differential comparison by accurately extracting data segments of adjacent periods from a continuous data stream; on the other hand, it achieves precise phase alignment of the waveform through cross-correlation and cyclic shifting, eliminating the influence of fundamental phase deviation caused by minor fluctuations in system frequency or asynchronous sampling on the differential results, ensuring that the differential operation reflects the true waveform distortion rather than phase error; furthermore, by calculating the point-by-point difference of the aligned sequence, it directly and losslessly extracts all the subtle differences in shape between adjacent period waveforms, forming a pure input for subsequent energy calculation.
[0036] In some embodiments, performing a cross-correlation operation on the current periodic sequence and the reference periodic sequence to determine the phase shift that maximizes the waveform similarity between the two includes: Step S221: Based on the current periodic sequence and the reference periodic sequence, calculate the cross-correlation function values under different phase offsets.
[0037] In practice, the reference periodic sequence is fixed, and the current periodic sequence is slid relative to the reference periodic sequence, sliding a distance of one sampling point at a time. At each sliding position, the sum of the products of the corresponding data points in the overlapping portion of the two sequences is calculated; this sum is the cross-correlation function value at that specific phase offset. By traversing all possible sliding positions, a series of cross-correlation function values are obtained.
[0038] Step S222: Determine the phase offset that maximizes the cross-correlation function value, and use this phase offset as the optimal phase calibration parameter.
[0039] In practice, among all the calculated cross-correlation function values, the one with the largest value is identified. The number of sampling points that the current periodic sequence slides relative to the reference periodic sequence when this maximum value is achieved is recorded. This number of sliding points is the optimal phase offset, indicating the amount of phase adjustment needed to achieve the best waveform matching.
[0040] Step S223: Based on the optimal phase calibration parameters, perform cyclic shifting on the current periodic sequence to achieve precise alignment of the zero-sequence voltage waveforms of adjacent power frequency cycles.
[0041] In practice, the data points of the current periodic sequence are reordered according to the optimal phase offset determined in step S222. If the offset is positive, the corresponding number of data points at the beginning of the sequence are moved to the end; if the offset is negative, the opposite operation is performed. After such a cyclic shift operation, the starting phase of the new current periodic sequence is optimally aligned with the reference periodic sequence.
[0042] The method provided in this embodiment, on the one hand, provides a complete search space for finding the optimal phase matching point by systematically calculating the cross-correlation coefficients at all possible offset positions; on the other hand, by establishing the offset corresponding to the maximum cross-correlation coefficient as the calibration parameter, the optimal solution for waveform alignment is determined in a quantitative and objective manner; furthermore, by performing phase calibration through cyclic shifting operations, the phase starting point of the sequence is precisely adjusted without losing any sampling point information, laying a precise time reference for subsequent point-by-point differential comparison.
[0043] In some embodiments, the step of squaring and summing all differences in the periodic difference sequence and then performing smoothing filtering to obtain the periodic difference energy value characterizing the fault features includes: Step S31: Perform a squaring operation on each difference element in the periodic difference sequence.
[0044] In practice, the periodic difference sequence generated in step S123 is read, and each value in the sequence (i.e., the difference between each sampling point) is multiplied by itself. This operation converts all differences, whether they were originally positive or negative, into non-negative squared values.
[0045] Step S32: Sum all the results after squaring, and multiply the sum by the sampling time interval to obtain the original differential energy value within one power frequency cycle.
[0046] In practice, all the squared values obtained in step S31 are summed together to obtain a total. This total is then multiplied by the time length between two adjacent sampling points (i.e., the sampling time interval). This product, in a physical sense, approximately represents the amount of energy contained in the differential signal within one power frequency cycle, i.e., the original differential energy value.
[0047] Step S33: Using the moving average filtering method, process the multiple original differential energy values obtained by continuous calculation, and output the smoothed periodic differential energy value.
[0048] In practice, a fixed-length queue is established to store multiple recently calculated raw differential energy values (e.g., values from the last 5 periods). Whenever a new raw differential energy value is calculated, it is added to the queue, while the oldest value is removed. Then, the arithmetic mean of all raw differential energy values in the current queue is calculated. This average is the more stable periodic differential energy value output after smoothing and filtering.
[0049] The method provided in this embodiment, on the one hand, eliminates the positive and negative signs of the differences by squaring them, transforming all waveform differences into positive contributions and amplifying the impact of larger differences; on the other hand, by summing the squared values and multiplying them by the time interval, the discrete difference sequence is quantized into a scalar energy value, which intuitively represents the overall intensity of the inconsistency between adjacent period waveforms; furthermore, by performing a moving average on the original energy values of multiple periods, the drastic jumps in energy values caused by random noise or instantaneous interference are effectively suppressed, outputting a stable fault characteristic quantity that reflects trend changes, thereby improving the reliability of subsequent threshold comparisons.
[0050] In some embodiments, the continuous comparison of the periodic differential energy value with an adaptively set dynamic threshold, and the generation of a high-resistance grounding fault determination signal and a corresponding control command when the periodic differential energy value continuously exceeds the dynamic threshold for a preset time, includes: Step S41: During the fault-free operation phase of the system, continuously record the cycle differential energy values within multiple power frequency cycles, and calculate their statistical average and standard deviation.
[0051] In practice, after the system starts up or during a period of confirmed normal operation, the periodic differential energy values output in step S33 are continuously collected. After collecting a sufficient number of values (e.g., values corresponding to hundreds of cycles), the arithmetic mean of these values is calculated as the average level (statistical mean) of the background differences of the system under fault-free conditions. At the same time, the dispersion of these values relative to this mean is calculated, i.e., the standard deviation.
[0052] Step S42: Set the initial dynamic threshold based on the statistical mean and standard deviation.
[0053] In practice, the statistical average and standard deviation calculated in step S41 are substituted into a preset calculation rule. For example, the average is added to a multiple of the standard deviation, and the result is set as the initial dynamic threshold. This threshold represents the boundary for judging whether the energy value is abnormal after considering normal fluctuations.
[0054] Step S43: In the real-time detection phase, each newly calculated periodic differential energy value is compared with the currently valid dynamic threshold.
[0055] In practice, after entering the continuous fault monitoring state, each time a new cycle differential energy value is generated, it is immediately compared with the currently used dynamic threshold value to determine whether the energy value is greater than or less than or equal to the threshold.
[0056] Step S44: If the duration for which the periodic differential energy value continuously exceeds the dynamic threshold reaches or exceeds the preset number of power frequency cycles, a high-resistance grounding fault is determined to have occurred, and a high-resistance grounding fault determination signal containing a fault alarm or protection trip command is generated.
[0057] In practical implementation, a counting mechanism is set up. Whenever the comparison result in step S43 indicates that the energy value exceeds the threshold, the counter increments; if the energy value is lower than or equal to the threshold, the counter is reset to zero. Only when the number of consecutive over-limit cycles recorded by this counter reaches or exceeds a preset value (e.g., 3 cycles) is a high-resistance grounding fault finally confirmed. At this point, a signal with a specific format or level is generated to indicate the fault occurrence and can be associated with subsequent operations such as triggering an audible and visual alarm or sending a trip command to the circuit breaker.
[0058] The method provided in this embodiment, on the one hand, sets a personalized baseline that conforms to the actual operating conditions by learning the background energy level and its fluctuation characteristics of the system during the fault-free phase; on the other hand, it realizes continuous and online monitoring of fault characteristics by comparing the current feature quantity with the dynamic threshold on a cycle-by-cycle basis during real-time detection; and furthermore, by introducing the duration criterion of "continuous over-limit", it requires that the fault characteristic must last for a certain period of time before it is confirmed, effectively distinguishing between true continuous faults and occasional instantaneous interference, and preventing maloperation of the protection device.
[0059] In some embodiments, setting the initial dynamic threshold based on the statistical average and standard deviation includes: step S421, calculating the arithmetic mean of the periodic differential energy values recorded in multiple power frequency cycles during the fault-free operation phase of the system as a background energy benchmark.
[0060] In practice, all the periodic differential energy values recorded in step S41 are summed, and then divided by the total number of these energy values. The quotient is the arithmetic mean. This average value represents the typical level of periodic waveform difference energy caused by factors such as load fluctuations and measurement noise when the system is running without faults, and serves as the background energy benchmark.
[0061] Step S422: Calculate the standard deviation of the periodic differential energy values within the multiple power frequency cycles to characterize the fluctuation range of the background energy.
[0062] In practice, the difference between the differential energy value of each cycle and the background energy benchmark obtained in step S421 is first calculated. These differences are then squared, and the average of all squared values is calculated. Finally, the square root of this average is taken. The result is the standard deviation, which quantitatively describes the amplitude of fluctuation of the differential energy value around its average level when there are no faults.
[0063] Step S423: Based on the background energy benchmark and the standard deviation, calculate and set the initial dynamic threshold according to the preset reliability coefficient.
[0064] In practice, a reliability coefficient greater than zero is chosen (e.g., 3 or 4). The result of adding the background energy benchmark to (standard deviation multiplied by the reliability coefficient) is set as the initial dynamic threshold. The formula is: Dynamic Threshold = Background Energy Benchmark + (Reliability Coefficient × Standard Deviation). This threshold is set above the background energy benchmark, leaving a margin several times greater than the normal fluctuation range.
[0065] The method provided in this embodiment, on the one hand, establishes a background energy benchmark by calculating the arithmetic mean, which objectively reflects the background signal strength when the system is operating normally; on the other hand, it quantifies the normal random fluctuation range of the background signal by calculating the standard deviation, providing a scientific basis for setting a reasonable detection threshold; furthermore, by adding the background energy benchmark to a margin based on the standard deviation to set a threshold, the threshold can be adapted to the noise level of different systems, ensuring that there are no false alarms within the normal fluctuation range, and ensuring sufficient detection sensitivity for abnormal features that exceed the normal fluctuation range.
[0066] Please see Figure 2 , Figure 2 A schematic block diagram of the structure of a high-resistivity grounding fault detection system 100 based on zero-sequence voltage periodic differential energy provided in this application embodiment is shown below. Figure 2 As shown in the figure, the high-resistivity grounding fault detection system 100 based on zero-sequence voltage periodic differential energy provided in this application includes: The acquisition module 110 is used to acquire the multi-source sensing data stream of the heterogeneous roadside sensing terminal.
[0067] The acquisition module 110 is used to acquire the instantaneous values of the three-phase voltage in real time and calculate a continuous zero-sequence voltage discrete-time sequence based on the instantaneous values of the three-phase voltage.
[0068] The first calculation module 120 is used to perform periodic truncation and alignment processing on the zero-sequence voltage discrete time sequence, calculate the point-by-point difference of the zero-sequence voltage waveforms of adjacent periods after alignment, and generate a periodic difference sequence.
[0069] The second calculation module 130 is used to perform square and sum operations on all differences in the periodic difference sequence, and after smoothing and filtering, obtain the periodic difference energy value that characterizes the fault features.
[0070] The comparison module 140 is used to continuously compare the periodic differential energy value with an adaptively set dynamic threshold. When the periodic differential energy value continuously exceeds the dynamic threshold for a preset time, a high-resistance grounding fault judgment signal and a corresponding control command are generated.
[0071] It should be noted that those skilled in the art will understand that, for the sake of convenience and brevity, the specific working process of the system and each module described above can be referred to the process in the aforementioned embodiment of the high-resistivity grounding fault detection method based on zero-sequence voltage periodic differential energy, and will not be repeated here.
[0072] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in this application, and these modifications or substitutions should all be covered within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.
Claims
1. A high-resistivity grounding fault detection method based on zero-sequence voltage periodic differential energy, characterized in that, include: The instantaneous values of the three-phase voltages are acquired in real time, and a continuous zero-sequence voltage discrete-time sequence is calculated based on the instantaneous values of the three-phase voltages. The zero-sequence voltage discrete-time sequence is periodically truncated and aligned, and the point-by-point difference between adjacent period zero-sequence voltage waveforms after alignment is calculated to generate a periodic difference sequence. The periodic difference sequence is squared and summed, and then smoothed and filtered to obtain the periodic difference energy value that characterizes the fault features. The periodic differential energy value is continuously compared with an adaptively set dynamic threshold. When the periodic differential energy value continuously exceeds the dynamic threshold for a preset time, a high-resistance grounding fault judgment signal and a corresponding control command are generated.
2. The high-resistivity grounding fault detection method based on zero-sequence voltage periodic differential energy according to claim 1, characterized in that, The real-time acquisition of instantaneous values of three-phase voltages, and the calculation of a continuous zero-sequence voltage discrete-time sequence based on these instantaneous values, includes: The instantaneous signals of the three-phase phase voltages are synchronously acquired at a sampling frequency of not less than twice the number of sampling points corresponding to the system's fundamental frequency period. At each sampling moment, the instantaneous values of the three-phase voltages are algebraically summed and averaged to obtain the instantaneous sampled value of the zero-sequence voltage. The calculated continuous zero-sequence voltage instantaneous sample values are stored in a first-in-first-out data buffer in chronological order to form the zero-sequence voltage discrete time sequence.
3. The high-resistivity grounding fault detection method based on zero-sequence voltage periodic differential energy according to claim 1, characterized in that, The process of periodically truncating and aligning the zero-sequence voltage discrete-time sequence, calculating the point-by-point difference of the aligned adjacent period zero-sequence voltage waveforms, and generating a periodic difference sequence includes: From the data buffer of the zero-sequence voltage discrete time sequence, the data of the latest complete power frequency cycle is extracted as the current cycle sequence, and the data of the immediately preceding complete power frequency cycle is extracted as the reference cycle sequence; A cross-correlation operation is performed on the current periodic sequence and the reference periodic sequence to determine the phase offset that maximizes the waveform similarity between the two, and the current periodic sequence is cyclically shifted and aligned based on the phase offset. The numerical difference between the aligned current periodic sequence and the reference periodic sequence at each corresponding sampling point is calculated to generate the periodic difference sequence.
4. The high-resistivity grounding fault detection method based on zero-sequence voltage periodic differential energy according to claim 3, characterized in that, The step of performing a cross-correlation operation between the current periodic sequence and the reference periodic sequence to determine the phase shift that maximizes the waveform similarity between the two includes: Based on the current periodic sequence and the reference periodic sequence, calculate the cross-correlation function values under different phase offsets; Determine the phase offset that maximizes the cross-correlation function value, and use this phase offset as the optimal phase calibration parameter; Based on the optimal phase calibration parameters, the current periodic sequence is cyclically shifted to achieve precise alignment of the zero-sequence voltage waveforms of adjacent power frequency cycles.
5. The high-resistivity grounding fault detection method based on zero-sequence voltage periodic differential energy according to claim 1, characterized in that, The step of squaring and summing all differences in the periodic difference sequence and then smoothing and filtering them to obtain the periodic difference energy value characterizing the fault features includes: Perform a square operation on each difference element in the periodic difference sequence; Sum all the results after squaring, and multiply the sum by the sampling time interval to obtain the original differential energy value within one power frequency cycle; The moving average filtering method is used to process the multiple original differential energy values obtained by continuous calculation, and the smoothed periodic differential energy value is output.
6. The high-resistivity grounding fault detection method based on zero-sequence voltage periodic differential energy according to claim 1, characterized in that, The process involves continuously comparing the periodic differential energy value with an adaptively set dynamic threshold. When the periodic differential energy value continuously exceeds the dynamic threshold for a preset duration, a high-resistance grounding fault determination signal and a corresponding control command are generated, including: During the fault-free operation phase of the system, the cycle differential energy values within multiple power frequency cycles are continuously recorded, and their statistical average and standard deviation are calculated. The initial dynamic threshold is set based on the statistical mean and standard deviation; During the real-time detection phase, each newly calculated periodic differential energy value is compared with the currently valid dynamic threshold. If the duration for which the periodic differential energy value continuously exceeds the dynamic threshold reaches or exceeds the preset number of power frequency cycles, a high-resistance grounding fault is determined to have occurred, and a high-resistance grounding fault determination signal containing a fault alarm or protection trip command is generated.
7. The high-resistivity grounding fault detection method based on zero-sequence voltage periodic differential energy according to claim 6, characterized in that, The step of setting the initial dynamic threshold based on the statistical average and standard deviation includes: calculating the arithmetic mean of the periodic differential energy values recorded in multiple power frequency cycles during the fault-free operation phase of the system as a background energy benchmark. Based on the periodic differential energy values within the multiple power frequency cycles, their standard deviation is calculated to characterize the fluctuation range of the background energy. Based on the background energy benchmark and the standard deviation, the initial dynamic threshold is calculated and set according to the preset reliability coefficient.
8. A high-resistivity grounding fault detection system based on zero-sequence voltage periodic differential energy, characterized in that, include: The acquisition module is used to acquire the instantaneous values of the three-phase voltage in real time and calculate a continuous zero-sequence voltage discrete-time sequence based on the instantaneous values of the three-phase voltage. The first calculation module is used to perform periodic truncation and alignment processing on the zero-sequence voltage discrete time sequence, calculate the point-by-point difference of the zero-sequence voltage waveforms of adjacent periods after alignment, and generate a periodic difference sequence. The second calculation module is used to perform square and sum operations on all differences in the periodic difference sequence, and after smoothing and filtering, obtain the periodic difference energy value that characterizes the fault features. The comparison module is used to continuously compare the periodic differential energy value with an adaptively set dynamic threshold. When the periodic differential energy value continuously exceeds the dynamic threshold for a preset time, a high-resistance grounding fault judgment signal and a corresponding control command are generated.