High resistance grounding distance protection method based on modal reactance imaginary frequency band equivalent quantity

Abstract

The application discloses a high-resistance grounding distance protection method based on modal electric reactance imaginary part frequency band equivalent quantity. The method collects line voltage and current signals, carries out synchronous preprocessing, frequency domain analysis and modal transformation, combines a frequency related line parameter library to determine a target mode, calculates an impedance imaginary part, and carries out frequency band statistics according to a preset weight in an effective frequency point to obtain a comprehensive modal electric reactance imaginary part characteristic quantity. Then, the method combines a local compensation electric reactance equivalent value and a unit length modal electric reactance equivalent value to estimate a fault distance and outputs a distance protection result. The method weakens the influence of transition resistance, grounding medium nonlinearity and arc fluctuation, and improves the discrimination stability of high-resistance grounding faults in a complex line scene.

Description

Technical Field

[0001] This invention relates to the field of power system relay protection technology, and in particular to a distance protection method for high-resistance grounding faults or weak feeders from new energy sources, and more specifically to a high-resistance grounding distance protection method based on the equivalent value of the imaginary part of modal reactance in the frequency band. Background Technology

[0002] High-resistance grounding faults are typically caused by conductors touching trees, conductors breaking and falling to the ground, contact with high-resistance media such as asphalt or sand, and short-gap electric arcs. They are characterized by small fault currents, large fault path resistances, and complex transient processes. Existing distance protection systems generally rely on complete complex impedances for section discrimination. However, after a high-resistance grounding fault occurs, the transition resistance and arc conductivity fluctuations cause a significant shift in the real part of the measured impedance, resulting in distortion of the circular or polygonal impedance characteristics of traditional distance protection systems. This leads to problems such as failure to operate, over-range operation, or decreased sensitivity.

[0003] On the other hand, in scenarios involving frequency-dependent parameters, multi-conductor coupling, and a high proportion of new energy sources, the performance of traditional power frequency distance protection further degrades. Therefore, there is an urgent need to propose a novel distance protection scheme that does not rely on the real part of impedance and can more stably reflect the line structure and propagation characteristics.

[0004] Existing distance protection typically uses fault measurement impedance as the criterion, determining whether a fault is located within the protection zone based on whether the measured impedance at the protection installation point falls within a preset impedance characteristic area. In engineering practice, impedance circle characteristics, offset circle characteristics, quadrilateral characteristics, or polygonal characteristics are usually constructed based on the impedance parameters per unit length of the line, the system operating mode, and the compensation coefficient, and these are used to complete the distance protection discrimination for the first, second, and backup sections.

[0005] The technical characteristics of this type of method are: it uses complete complex impedance as the basis for fault section identification; it mainly utilizes power frequency steady-state quantities or power frequency components within a short time window; the action criteria are closely related to factors such as line equivalent impedance, transition resistance, system power supply strength, and CT / PT error; the technology is mature and widely used in engineering.

[0006] However, this type of method faces significant shortcomings in high-resistance grounding fault scenarios. Because high-resistance grounding fault channels typically exhibit large transition resistance, arc nonlinearity, and grounding medium nonlinearity, these factors significantly alter the resistive component of the measured impedance, causing the impedance trajectory to shift towards the real axis in the complex plane. Specifically, during faults within the designated area, the measured impedance may deviate from the preset operating zone, leading to protection failure; under certain system operating modes, impedance deviation may cause erroneous out-of-zone operation; and in weak feeder lines of new energy sources, due to the limited amplitude of the fault current, the power frequency phasor relationship of voltage and current further deviates from the assumptions under traditional synchronous power supply conditions, further degrading the distance protection performance.

[0007] Therefore, although the traditional complex impedance distance protection method is the closest existing technology to this invention, its core criterion depends on the joint participation of the real and imaginary parts of the impedance in the discrimination. The core improvement of this invention is that it no longer uses the complete complex impedance as the main distance quantity, but only extracts the imaginary part of the target mode impedance and performs equivalent statistics within the frequency band, thereby reducing the adverse effect of resistive offset on protection performance in high impedance faults.

[0008] To improve the performance of traditional impedance distance protection under ground fault conditions, an improved distance protection method exists in existing technologies. This type of method typically does not abandon the complex impedance criterion, but rather corrects the original measured impedance by introducing zero-sequence compensation coefficients, fault component compensation, power supply side compensation, polarization correction, memory voltage, and additional impedance correction, in order to improve the accuracy of distance measurement and the stability of section operation under ground fault conditions.

[0009] These technologies typically follow these approaches: Zero-sequence compensation methods: By introducing a zero-sequence compensation coefficient, the protective current during a single-phase ground fault is equivalently corrected to reduce the impact of the zero-sequence network on the measured impedance. For example, in grounding distance protection, a fault loop impedance expression with compensation current is used to improve the distance measurement performance under single-phase ground faults. Fault component methods: By extracting the incremental components, abrupt changes, or power frequency fault components of voltage and current before and after the fault, fault component impedance or incremental impedance criteria are constructed to weaken the impact of normal load current on distance protection. Polarization and memory voltage improvement methods: By introducing positive-sequence polarization, negative-sequence polarization, and memory voltage polarization, the direction and distance discrimination capabilities under voltage drop or weak feeder conditions are improved. Compensation reactance or system equivalent correction methods: In the process of measuring impedance calculation, compensation amounts for the local system, compensation amounts for the influence of the power supply at the opposite end, and correction amounts for line distributed parameters are added to enhance the accuracy of distance measurement. The above methods are improvements over traditional distance protection, but they are essentially still technical routes that involve correction or compensation based on complete complex impedance. In other words, regardless of the compensation method, the final fault segment identification is still based on the "measured impedance defined by the real part and the imaginary part". For high-resistance grounding faults, especially when the transition resistance is large, arc extinguishing and reignition are frequent, and the conductivity characteristics of the grounding medium are significantly nonlinear, the above correction amount is often difficult to accurately characterize the dynamic changes of the fault channel, leading to the following problems: the correction model is more dependent on the system equivalent conditions and has a limited scope of application; it still cannot fundamentally get rid of the direct influence of the transition resistance on the real part of the impedance; under the coupled conditions of high-resistance faults and weak feeder conditions, the compensation parameter setting is difficult and the protection robustness is insufficient; when the line parameters have obvious frequency correlation, the distance protection based solely on power frequency steady-state compensation cannot fully reflect the actual line characteristics.

[0010] In contrast, this invention does not simply add various compensation terms within the complex impedance framework, but rather changes the criteria at the level of: first, it no longer directly relies on the real part of the impedance; second, it is not limited to single-point or limited steady-state characteristics at power frequency; and third, it incorporates frequency-related line parameters, mode transformation results, and frequency band statistics into the ranging chain simultaneously. Therefore, this invention differs significantly from other improved distance protection methods in its technical approach.

[0011] To address the difficulty of reliably identifying high-resistance grounding faults using traditional distance protection systems, existing technologies have largely focused on high-resistance grounding fault detection, fault perception, fault line selection, and fault location. These technologies typically treat high-resistance faults as "weak faults," "weak-feature faults," or "nonlinear faults," primarily addressing the following issues from the perspective of fault feature extraction: determining whether a high-resistance grounding fault has occurred in the system; identifying the faulty feeder or branch; and providing coarse location of the faulty section in certain scenarios.

[0012] The existing technologies of this type can generally be divided into the following technical branches: (1) Detection / line selection methods based on zero-sequence current, zero-sequence voltage or zero-sequence admittance: These methods realize high-resistance fault detection and fault line identification by analyzing the direction, amplitude, phase, harmonic or admittance changes of the zero-sequence current of each feeder before and after the fault. Its advantage is that the implementation structure is relatively clear, but its disadvantage is that when the arc suppression coil compensation is strong, the fault current is extremely weak or the system background imbalance is large, the zero-sequence quantity characteristics may be weakened or even submerged. (2) High-resistance fault detection methods based on transient characteristics: These methods use the frequency band energy distribution, abrupt change, transient waveform integral area, transient slope change, cross wavelet characteristics, etc. in the transient quantity in the early stage of the fault to identify high-resistance faults. Its advantage is that it avoids the problem of insufficient power frequency steady-state characteristics to a certain extent, but it also often has the problem of being sensitive to sampling rate, noise level, threshold setting and system operation mode. (3) Methods based on mode decomposition, multimodal fusion, or multicriteria fusion: These methods typically decompose high-impedance fault signals into multiple modes or frequency band components using VMD, EMD, MVMD, wavelet packet decomposition, etc., and then extract indicators such as energy, correlation, kurtosis, entropy, and geometric features, and achieve fault detection through voting, fusion, or machine learning classification. Its advantage lies in the full utilization of information, but its main goal is still "fault detection / fault identification" rather than "distance measurement / segment distance protection". (4) Methods based on nonlinearity of grounding medium or arc mechanism modeling: These methods model the physical mechanism of fault channels such as sand, red soil, asphalt, trees, and short-gap arcs, analyze the nonlinear distortion, M-shaped change, zero-dimensional characteristics, or energy distribution law of fault current, and construct detection criteria accordingly. These methods have a deeper understanding of the nature of faults, but most results are still used for fault perception, detection, and fire risk assessment, rather than forming a distance protection-type segment discrimination scheme.

[0013] From a patent comparison perspective, the relationship between this type of prior art and the present invention is as follows: they both belong to the broad category of "high-resistance grounding fault handling technology"; they may also involve similar methods such as frequency domain analysis, modal analysis, feature statistics, and frequency band weighting; however, their ultimate goal is usually detection, sensing, line selection, or classification, rather than section protection action based on distance. In other words, although this type of technology is partially similar to the present invention in terms of "signal processing methods," it does not disclose the following key technology combination: starting from modal impedance, only extracting the imaginary part of the target modal impedance; after effective frequency point screening and weighted statistics, forming a comprehensive modal reactance imaginary part feature quantity; then coupling this statistical quantity with the frequency band equivalent value of the unit length modal reactance and the local compensation reactance to construct a clear fault distance estimation formula; finally used for distance protection stage one, stage two, and blocking decisions. Therefore, although this type of prior art is close to the present invention in the local feature extraction stage, it differs fundamentally from the present invention in terms of the final technical goal, criterion system, and protection logic.

[0014] In addition to the high-resistivity fault detection methods mentioned above, there is another type of fast protection method based on transient quantities, traveling wave quantities, line-mode / zero-mode propagation differences, boundary high-frequency characteristics, or reflected wave features. This type of technology is mainly aimed at rapid fault identification in transmission lines, DC lines, or new energy transmission lines, and its goal is usually to solve the problems of slow operation, poor adaptability to weak feeders, or difficulty in distinguishing between forward and reverse directions in traditional power frequency quantity protection.

[0015] Common technical approaches for this type of method include: ranging or direction protection based on the time difference of arrival of traveling waves; protection criteria based on the difference in wave velocity between the linear mode and the zero mode; fault identification inside and outside the zone based on the polarity of reflected waves, secondary anti-traveling waves, or boundary high-frequency components; and rapid protection based on transient energy, transient similarity, correlation coefficient, or pattern recognition.

[0016] These methods share the following characteristics: they use high-frequency information in the early stages of a fault; they typically emphasize speed; they focus on solving problems such as distinguishing between inside and outside the fault zone, direction identification, or traveling wave ranging; and they are often used for long AC lines, flexible DC lines, or new energy transmission lines. Summary of the Invention

[0017] Technical Problem: The purpose of this invention is to provide a high-resistivity grounding distance protection method based on the equivalent value of the imaginary part of modal reactance, so as to solve the problems of poor ability to withstand transition resistance, unstable discrimination between inside and outside the zone, and decreased ranging accuracy under frequency-related parameter scenarios in existing distance protection under high-resistivity grounding fault conditions.

[0018] Technical solution: To achieve the above objectives, the present invention provides a high-resistance grounding distance protection method based on the equivalent value of the imaginary part of modal reactance, comprising the following steps:

[0019] Step 1: Collect voltage and current signals at the installation point of the protected line, and perform synchronous preprocessing on the voltage and current signals to obtain discrete time series voltage and current quantities;

[0020] Step 2: Perform frequency domain analysis on the discrete time series voltage and current quantities to obtain the frequency domain voltage vector U(ω) and frequency domain current vector I(ω) corresponding to each frequency point within the analysis frequency band.

[0021] Step 3: Call the preset frequency-related line parameter library to obtain the mode transformation matrix T(ω) and the unit length mode reactance parameter x corresponding to each frequency point. m (ω) and the equivalent compensation reactance parameter X on this side c (ω);

[0022] Step four: Perform mode transformation on the frequency domain voltage vector U(ω) and frequency domain current vector I(ω) according to the mode transformation matrix T(ω) to obtain the mode voltage U corresponding to each candidate mode.m (ω) and modal current I m (ω), and determine the target mode based on the preset modal optimization index. ;

[0023] Step 5, for the target mode The set of frequency points Ω that meets the effective frequency point selection criteria e Calculate the target modal impedance. And extract its imaginary part. ;

[0024] Step 6: Apply the preset weighting function w(ω) to the imaginary part. By performing frequency band weighted statistics, the imaginary part characteristic of the integrated modal reactance, X, is obtained. eq And obtain the equivalent value of modal reactance per unit length x eq Equivalent value X of the equivalent compensation reactance on this side ceq ;

[0025] Step 7: Based on the imaginary part characteristic quantity X of the comprehensive modal reactance eq The equivalent value X of the local equivalent compensation reactance ceq and the equivalent value of the modal reactance per unit length x eq Calculate the estimated fault distance =(X eq -X ceq ) / x eq ;

[0026] Step 8: Compare the estimated fault distance with the preset protection section setting, and output the action result of the first protection section, the action result of the second protection section, or the blocking result.

[0027] The synchronization preprocessing includes one or more of the following: DC component removal, anti-aliasing filtering, time alignment, noise suppression, and window function weighting; the frequency domain analysis employs one of the following: Discrete Fourier Transform, Fast Fourier Transform, Short-Time Fourier Transform, Continuous Wavelet Transform, Discrete Wavelet Transform, S-Transform, or Hilbert-Huang Transform.

[0028] The frequency-related line parameter library includes at least: mode transformation matrix T(ω), inverse mode transformation matrix T -1 (ω), modal reactance parameter per unit length x m (ω), Equivalent Compensation Reactance Parameter X of this Side c (ω), one or more of the modal propagation parameters and modal impedance parameters.

[0029] The target mode The modality selection index is determined based on the modality selection index, which is used to characterize one or more of the following: sensitivity, signal-to-noise ratio, monotonicity, or coherence of each candidate mode to grounding faults; the target mode is the mode that makes the modality selection index achieve the optimal value.

[0030] The set of frequency points Ω of the effective frequency point selection criteria e Determined by the following conditions: , where Ω b For the set of frequency points corresponding to the preset analysis frequency band, I th γ is the modal current amplitude threshold, and γ(ω) is the coherence index of voltage and current at frequency ω. th This is the coherence threshold.

[0031] The weighting function w(ω) is determined based on the modal current signal-to-noise ratio, frequency coherence, modal energy, or a combination thereof, and satisfies any of the following forms: ,or ,or , where σ n ^2 represents the noise power estimate.

[0032] The imaginary part characteristic of the integrated modal reactance, the equivalent value of the modal reactance per unit length, and the equivalent value of the local side's equivalent compensation reactance are obtained using the following frequency band equivalent form: , , Alternatively, it can be obtained using a discrete summation form equivalent to the above integral form.

[0033] The preset protection segment setpoints include the upper limit d1 of the first protection segment and the upper limit d2 of the second protection segment, wherein: when 0 ≤ Output the action result of the first protected section when d1 ≤ d1; when d1 < d1 When ≤d2, output the action result of the second protected section; when >d2 or Output the locking result when <0.

[0034] The sequence of protection devices executing this method includes: sampling module, preprocessing module, frequency domain analysis module, parameter calling module, mode transformation and optimization module, imaginary part extraction module, frequency band statistics module, distance estimation module, and protection decision module.

[0035] Compared with the prior art, the present invention has the following significant differences:

[0036] The key differences lie in the following aspects: **Focusing on different objects:** This type of method targets metallic faults, short-circuit faults, or general asymmetrical faults, rather than specifically addressing the resistive offset problem of high-resistance grounding faults. **Different physical quantities for judgment:** This type of method relies on the traveling wave front, arrival time, wave polarity, transient energy, or wave process characteristics, while this invention relies on the band equivalent of the imaginary part of the target modal impedance. **Different distance construction mechanisms:** Traveling wave methods often identify fault locations through propagation time, reflection time, boundary reflection-reflection relationships, or transient processes; this invention establishes fault distance estimates through the equivalent relationship between modal reactance and unit-length modal reactance. **Different technical problems:** This type of method mainly solves the problems of speed and boundary discrimination; this invention mainly solves the problem of impedance real part offset caused by transition resistance under high-resistance faults. **Different implementation constraints:** Traveling wave protection typically has high requirements for sampling rate, wavefront detection accuracy, and synchronization conditions, while this invention can be implemented in medium-to-high sampling rate digital protection devices through band analysis and statistics, resulting in different engineering implementation paths.

[0037] Based on the above closest existing technologies, the differences between this invention and existing technologies can be summarized as follows:

[0038] Difference 1: Different core quantities of the criteria: Existing technologies are based on complete complex impedance, or on zero-sequence quantities, transient characteristic quantities or classification characteristic quantities, or on traveling wave quantities, wave arrival time or transient boundary characteristics; this invention is based on the equivalent quantity of the imaginary part of the target modal impedance.

[0039] Difference 2: Different fault distance construction paths: In existing technologies, traditional distance protection achieves section discrimination by comparing complex impedance with setting impedance; high-impedance detection methods mostly do not output distance values; traveling wave methods construct distance through propagation time; the present invention directly constructs the fault distance estimate through the frequency band equivalent relationship.

[0040] Difference 3: Different ways of involving frequency-related parameters: Existing technologies usually use line parameters at power frequency or a few characteristic frequencies, or only use parameters as auxiliary correction terms; This invention directly calls the frequency-related line parameter library and incorporates the mode transformation matrix, unit length mode reactance parameters and compensation reactance parameters into the complete distance calculation chain.

[0041] Difference 4: Different frequency band processing methods: Existing technologies either use a single frequency point or a few transient features; this invention constructs a comprehensive quantity by selecting effective frequency points, weighting with a weighting function, and using equivalent frequency band statistics to suppress single frequency point fluctuations and noise amplification.

[0042] Under high-resistivity grounding fault conditions, this invention aims to construct a distance protection method that is insensitive to impedance real part deviation, highly adaptable to frequency-varying parameters, and capable of stably estimating fault distance and achieving section protection discrimination, especially given conditions such as large transition resistance, significant nonlinearity of the grounding medium, significant arc fluctuations, non-negligible line frequency-related parameters, and weak feeder operation of new energy sources. Furthermore, this invention also addresses: how to automatically select the most sensitive and stable target mode for grounding faults from multiple modes; how to avoid impedance calculation divergence or noise amplification under weak current conditions of high-resistivity faults; and how to effectively couple frequency-related line parameters with the distance measurement formula.

[0043] Beneficial effects:

[0044] (1) Reduce the influence of transition resistance. This invention does not directly use the complete complex impedance for distance discrimination, but only uses the imaginary part of the target modal impedance to construct the fault distance, thereby reducing the adverse effects of fault path resistance, arc fluctuation and grounding medium nonlinearity on distance protection performance in high-resistance grounding faults.

[0045] (2) Improve adaptability in frequency-dependent scenarios. This invention reduces the impact of multi-conductor coupling and line frequency-varying parameters on ranging results by using a preset frequency-dependent line parameter library for mode transformation and parameter calling, thereby improving the protection adaptability in complex line scenarios.

[0046] (3) Improve stability in weak feature scenarios. This invention constructs the imaginary part characteristic of the comprehensive modal reactance by effective frequency point screening and frequency band weighted statistics, avoiding the random deviation caused by directly using the impedance of a single frequency point, and improving the protection stability in high-resistance grounding, weak feeder operation and noise interference scenarios.

[0047] (4) Easy to implement in engineering. The frequency domain analysis, mode transformation, frequency band statistics and segment decision process adopted in this invention can all be implemented in digital protection devices, which has good engineering application value. Attached Figure Description

[0048] Figure 1 This is a schematic diagram of high-resistance grounding distance protection deployment based on the change of the imaginary part of modal reactance;

[0049] Figure 2 This is a flowchart of a high-resistance grounding distance protection method based on the variation of the imaginary part of modal reactance;

[0050] Figure 3 This is a block diagram of a high-resistance grounding distance protection device based on the variation of the imaginary part of modal reactance;

[0051] Figure 4 This is a schematic diagram showing the mapping relationship between the imaginary part characteristics of the integrated modal reactance and the fault distance;

[0052] Figure 5 This is a schematic diagram of the protection action logic based on the fault distance estimation value. Detailed Implementation

[0053] 5.1 Overall Process

[0054] The basic idea of ​​this invention is as follows: For high-resistance grounding faults, the transition resistance, grounding dielectric nonlinearity, and arc conductivity fluctuations in the fault path are mainly reflected in the resistive component of the measured impedance, while the characteristics of the line body, mode propagation characteristics, and compensation branches in terms of fault distance are more reflected in the reactive component. Therefore, this invention does not directly use the complete complex impedance as the distance criterion, but instead extracts the imaginary part from the target modal impedance through frequency domain analysis and mode transformation, and constructs the imaginary part characteristic quantity of the comprehensive modal reactance within the effective frequency band, thereby establishing a mapping relationship with the fault distance.

[0055] The protection device is installed on one side of the protected line, collecting voltage and current signals at the beginning of the line. After preprocessing, the signals are converted to the frequency domain. Using a preset frequency-related line parameter library, the voltage and current at each frequency point are transformed to obtain the modal voltage and modal current of multiple candidate modes. According to the mode selection rules, the target mode with the most significant response to ground faults and with a high signal-to-noise ratio or monotonicity is selected. A modal impedance is constructed on the target mode and its imaginary part is extracted. Then, weighted statistics are performed within the effective frequency point set that meets the threshold condition to obtain the characteristic quantity of the imaginary part of the comprehensive modal reactance. Combined with the equivalent value of the equivalent compensation reactance on this side and the equivalent value of the modal reactance per unit length, the fault distance estimate is calculated, and finally, the section protection decision is realized.

[0056] Specifically, the steps include the following:

[0057] Step 1: Collect voltage and current signals at the installation point of the protected line, and perform synchronous preprocessing on the voltage and current signals to obtain discrete time series voltage and current quantities;

[0058] Step 2: Perform frequency domain analysis on the discrete time series voltage and current quantities to obtain the frequency domain voltage vector U(ω) and frequency domain current vector I(ω) corresponding to each frequency point within the analysis frequency band.

[0059] Step 3: Call the preset frequency-related line parameter library to obtain the mode transformation matrix T(ω), unit length mode reactance parameter xm(ω), and local equivalent compensation reactance parameter Xc(ω) corresponding to each frequency point;

[0060] Step four: Perform mode transformation on the frequency domain voltage vector U(ω) and frequency domain current vector I(ω) according to the mode transformation matrix T(ω) to obtain the mode voltage Um(ω) and mode current Im(ω) corresponding to each candidate mode, and determine the target mode according to the preset mode optimization index. ;

[0061] Step 5, for the target mode Within the set of frequency points Ωe that meet the effective frequency point selection criteria, calculate the target modal impedance. And extract its imaginary part. ;

[0062] Step 6: Apply the preset weighting function w(ω) to the imaginary part. By performing frequency band weighted statistics, the imaginary part characteristic quantity Xeq of the comprehensive modal reactance is obtained, and the equivalent value of modal reactance per unit length xeq and the equivalent value of local compensation reactance Xceq are obtained.

[0063] Step 7: Calculate the fault distance estimate d = (Xeq - Xceq) / xeq based on the imaginary part characteristic of the integrated modal reactance Xeq, the equivalent value of the local equivalent compensation reactance Xceq, and the equivalent value of the modal reactance per unit length xeq.

[0064] Step 8: Compare the estimated fault distance with the preset protection section setting, and output the action result of the first protection section, the action result of the second protection section, or the blocking result.

[0065] 5.2 Parameter Library and Mode Selection

[0066] In one embodiment, the sampling frequency can be set to 20kHz, 50kHz, or higher. To balance the effectiveness of high-impedance fault feature extraction with the cost of device implementation, 50kHz is preferred. The preprocessed discrete voltage and current sequences are then subjected to a discrete Fourier transform or a short-time Fourier transform to obtain the analysis band Ω. b The frequency domain voltage vector U(ω) and frequency domain current vector I(ω) correspond to each frequency point. The analysis frequency band can be preset from 300Hz to 3kHz, or it can be adaptively selected according to the line length, fault type, system operation mode and sampling frequency.

[0067] The frequency-dependent line parameter library can be calculated offline based on line structure parameters, conductor parameters, and grounding parameters. It can also be pre-identified using an electromagnetic transient simulation platform, or the offline parameters can be corrected using online operational data. The parameter library should at least include the mode transformation matrix T(ω) and the unit-length modal reactance parameter x for each frequency point. m (ω) and the equivalent compensation reactance parameter X on this side c (ω).

[0068] Under high-resistance grounding faults, different modes respond differently to the fault. To improve the stability of subsequent distance estimation, this invention does not fixate on a single mode, but instead sets mode selection rules. The mode selection rules can determine the target mode based on mode energy, signal-to-noise ratio, coherence, fault distance monotonicity, or a combination thereof.

[0069]

[0070] 5.3 Effective Frequency Point Screening and Frequency Band Statistics

[0071] Because some frequency points in high-impedance grounding fault scenarios may have excessively low modal currents or noise dominance, direct calculation across the entire frequency band can lead to impedance divergence or increased deviation. Therefore, this invention introduces an effective frequency point screening mechanism to construct an effective frequency point set Ω. e .

[0072]

[0073] In the effective frequency set Ω e Within this framework, a pre-defined weighting function w(ω) is used to perform frequency band weighted statistics on the imaginary part of the target modal impedance, yielding the characteristic quantity of the imaginary part of the comprehensive modal reactance. The weighting function can be determined based on the modal current signal-to-noise ratio, modal energy, frequency coherence, or a combination thereof. For example, it can be implemented using... Furthermore, the same weighting function is used to perform frequency band equivalence on the unit length modal reactance parameters and the local equivalent compensation reactance parameters, so that the fault distance estimation is established under a unified frequency band equivalence framework, avoiding ranging deviations caused by fluctuations in parameters at a single frequency point.

[0074] 5.4 Segment Decision Logic

[0075] Based on the obtained imaginary part characteristic quantity X of the integrated modal reactance eq Equivalent value of the equivalent compensation reactance on this side X ceq and the equivalent value of modal reactance per unit length x eq Calculate the estimated fault distance. =(X eq -X ceq ) / x eq The protection device compares the estimated fault distance with the preset protection zone setting. Assuming the upper limit of the first protection zone is d1 and the upper limit of the second protection zone is d2, then when 0 ≤ When d1 ≤ d1, the fault is determined to be located in the first protected section and an action command is issued; when d1 < When ≤d2, the fault is determined to be located in the second protection zone; when >d2 or When <0, output the locking result.

[0076] 5.5 Example 1: High-resistance grounding fault scenario in conventional overhead lines

[0077] A protection device is deployed at the beginning of the line, with a sampling frequency of 50kHz and an analysis frequency band of 300Hz to 3kHz. The protection device collects voltage and current signals in real time. After DC removal, filtering, and synchronization alignment, a discrete Fourier transform is used to obtain the frequency domain voltage vector U(ω) and frequency domain current vector I(ω) corresponding to each frequency point in the analysis frequency band.

[0078] The frequency-dependent line parameter library is used to read the mode transformation matrix T(ω) and the unit length mode reactance parameter x corresponding to each frequency point. m (ω) and the equivalent compensation reactance parameter X on this side c (ω), after completing the mode transformation, each candidate mode is evaluated according to the preset mode selection index, and the target mode with the most significant response to ground faults is selected. In satisfying And γ(ω)≥γ th Effective frequency set Ω e Within this, the imaginary part of the target modal impedance is calculated, and then based on the weighting function... The imaginary part characteristic of the integrated modal reactance X is obtained. eq .

[0079] The upper limit of the first protection section is set at 80% of the total line length, and the upper limit of the second protection section is set at 120% of the total line length. The protection device outputs the corresponding section protection result based on the calculated fault distance estimate. Compared with the traditional complex impedance distance criterion, this embodiment can still maintain a smaller distance measurement deviation and a more stable in-zone operation capability even when the transition resistance is large.

[0080] 5.6 Example 2: New Energy Weak Feeder Line Scenario

[0081] In weak feeder lines containing inverter-type renewable energy power sources, the same overall process as in Example 1 is adopted. The difference is that the mode selection index further considers the monotonicity and signal-to-noise ratio of the target mode in the frequency band, and adaptively selects the target mode that is more suitable for the current operating mode; a noise suppression term is added to the weighting function to adapt to weak fault current scenarios.

[0082] For example, it can be adopted As a weighting function, where σ n ^2 represents the noise power estimate. Even with limited fault current amplitude in new energy sources and significant impedance deviation in traditional distance protection measurements, this embodiment still maintains good fault distance estimation performance by utilizing the equivalent value of the imaginary part of the modal reactance.

[0083] 5.7 Optional Implementation Methods

[0084] This invention is not limited to the specific formulas described above. For the frequency band statistics module, a discrete summation form can be used instead of an integral form; for the frequency domain analysis method, wavelet domain decomposition can be used to complete the statistics within the characteristic frequency band; the target mode selection method, weighting function, and parameter library establishment method can also be replaced or extended according to different line types and hardware conditions. As long as the equivalent distance related to the imaginary part of the target mode impedance can be obtained and the segment decision can be made accordingly, it should fall within the protection scope of this invention.

Claims

1. A high-resistance grounding distance protection method based on the equivalent value of the imaginary part of modal reactance, characterized in that, The steps include the following: Step 1: Collect voltage and current signals at the installation point of the protected line, and perform synchronous preprocessing on the voltage and current signals to obtain discrete time series voltage and current quantities; Step 2: Perform frequency domain analysis on the discrete time series voltage and current quantities to obtain the frequency domain voltage vector U(ω) and frequency domain current vector I(ω) corresponding to each frequency point within the analysis frequency band. Step 3: Call the preset frequency-related line parameter library to obtain the mode transformation matrix T(ω) and the unit length mode reactance parameter x corresponding to each frequency point. m (ω) and the equivalent compensation reactance parameter X on this side c (ω); Step four: Perform mode transformation on the frequency domain voltage vector U(ω) and frequency domain current vector I(ω) according to the mode transformation matrix T(ω) to obtain the mode voltage U corresponding to each candidate mode. m (ω) and modal current I m (ω), and determine the target mode based on the preset modal optimization index. ; Step 5, for the target mode The set of frequency points Ω that meets the effective frequency point selection criteria e Internally, calculate the target modal impedance. And extract its imaginary part. ; Step 6: Apply a weighted weight function w(ω) to the imaginary part. By performing frequency band weighted statistics, the imaginary part characteristic of the integrated modal reactance, X, is obtained. eq And obtain the equivalent value of modal reactance per unit length x eq Equivalent value X of the equivalent compensation reactance on this side ceq ; Step 7: Based on the imaginary part characteristic quantity X of the comprehensive modal reactance eq The equivalent value X of the local equivalent compensation reactance ceq and the equivalent value of the modal reactance per unit length x eq Calculate the estimated fault distance =(X eq -X ceq ) / x eq ; Step 8: Compare the estimated fault distance with the preset protection section setting, and output the action result of the first protection section, the action result of the second protection section, or the blocking result.

2. The high-resistance grounding distance protection method based on the equivalent value of the imaginary part of modal reactance frequency band as described in claim 1, characterized in that, The synchronization preprocessing includes one or more of the following: DC component removal, anti-aliasing filtering, time alignment, noise suppression, and window function weighting; the frequency domain analysis employs one of the following: Discrete Fourier Transform, Fast Fourier Transform, Short-Time Fourier Transform, Continuous Wavelet Transform, Discrete Wavelet Transform, S-Transform, or Hilbert-Huang Transform.

3. The high-resistance grounding distance protection method based on the equivalent value of the imaginary part of modal reactance frequency band as described in claim 2, characterized in that, The frequency-related line parameter library includes at least: mode transformation matrix T(ω), inverse mode transformation matrix T -1 (ω), modal reactance parameter per unit length x m (ω), Equivalent Compensation Reactance Parameter X of this Side c (ω), one or more of the modal propagation parameters and modal impedance parameters.

4. The high-resistance grounding distance protection method based on the equivalent value of the imaginary part of modal reactance frequency band as described in claim 3, characterized in that, The target mode m× is determined based on the mode optimization index, which is used to characterize one or more of the following: sensitivity, signal-to-noise ratio, monotonicity, or coherence of each candidate mode to ground faults; the target mode is the mode that makes the mode optimization index achieve the optimal value.

5. The high-resistance grounding distance protection method based on the equivalent value of the imaginary part of modal reactance frequency band according to claim 4, characterized in that, The set of frequency points Ω of the effective frequency point selection criteria e Determined by the following conditions: , where Ω b For the set of frequency points corresponding to the preset analysis frequency band, I th γ is the modal current amplitude threshold, and γ(ω) is the coherence index of voltage and current at frequency ω. th This is the coherence threshold.

6. The high-resistance grounding distance protection method based on the equivalent value of the imaginary part of modal reactance frequency band according to claim 5, characterized in that, The weighting function w(ω) is determined based on the modal current signal-to-noise ratio, frequency coherence, modal energy, or a combination thereof, and satisfies any of the following forms: ,or ,or , where σ n ^2 represents the noise power estimate.

7. The high-resistance grounding distance protection method based on the equivalent value of the imaginary part of modal reactance frequency band according to claim 6, characterized in that, The imaginary part characteristic of the integrated modal reactance, the equivalent value of the modal reactance per unit length, and the equivalent value of the local side's equivalent compensation reactance are obtained using the following frequency band equivalent form: , , Alternatively, it can be obtained using a discrete summation form equivalent to the integral form described above.

8. The high-resistance grounding distance protection method based on the equivalent value of the imaginary part of modal reactance frequency band according to claim 7, characterized in that, The preset protection segment setpoints include the upper limit d1 of the first protection segment and the upper limit d2 of the second protection segment, wherein: when 0 ≤ Output the action result of the first protected section when d1 ≤ d1; when d1 < d1 When ≤d2, output the action result of the second protected section; when >d2 or Output the locking result when <0.

9. The high-resistance grounding distance protection method based on the equivalent value of the imaginary part of modal reactance frequency band according to claim 8, characterized in that, The sequence of protection devices executing this method includes: sampling module, preprocessing module, frequency domain analysis module, parameter calling module, mode transformation and optimization module, imaginary part extraction module, frequency band statistics module, distance estimation module, and protection decision module.

Images