Power distribution network high-resistance fault positioning method and positioning system based on multistage joint filtering

Abstract

The invention discloses a power distribution network high-resistance fault positioning method and positioning system based on multistage joint filtering, and the method comprises the steps: collecting the power data of a power distribution network in real time, including a zero-sequence voltage signal and a zero-sequence current signal; constructing a signal flow graph strategy to process the power data by using a multi-stage joint filtering technology, and generating output data; the extraction of fault features is completed while the signal descrambling is carried out; establishing a dynamic threshold value, and performing comparative analysis based on the output data and the dynamic threshold value to determine a fault occurrence moment; based on a fault occurrence moment, generating a fault measure by using a heuristic method and feature polarity accumulation; based on two-dimensional characteristics of fault measurement amplitude and polarity and according to distribution characteristics of fault measurement of each measurement point relative to a fault point, a fault section is positioned through a recursion positioning algorithm containing an adaptive threshold value, a fault feeder line is quickly positioned, recursion positioning is carried out by using the adaptive threshold value, abnormal values caused by interference are eliminated, and the fault feeder line is quickly positioned. Therefore, accurate positioning under complex working conditions is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of power system relay protection, and in particular to a distribution network high-resistance fault locating method and locating system based on multi-stage joint filtering. Background Art

[0002] Resonant grounding systems are widely used in distribution networks to suppress single-phase ground fault currents. However, the detection and location of high-impedance faults (HIFs) remain technically challenging due to their large transition resistance (hundreds to thousands of ohms) and weak fault currents (several amperes to a few tenths of an ampere). Traditional methods rely on the amplitude or phase characteristics of the zero-sequence current, but the signal amplitude in high-impedance faults approaches the background noise level, making direct identification difficult. With the large-scale integration of distributed generation (DGs), the topology complexity of distribution networks has increased, exacerbating problems such as line parameter asymmetry, harmonic pollution, and load fluctuations, further amplifying the challenges of locating high-impedance faults.

[0003] Existing positioning technologies are mostly based on transient signal analysis, such as wavelet transforms or time-frequency analysis. These require multi-scale signal decomposition and frequency domain feature extraction, resulting in computational complexity as high as O(n log n). This places stringent computing demands on edge devices (such as feeder terminals), making it difficult to achieve millisecond-level real-time response. Deep learning-based models, while performing well in the laboratory, are parameter-intensive and rely on high-performance computing resources, resulting in high deployment costs and limited generalization capabilities. Some optimization algorithms, such as least-squares parameter identification, improve accuracy through iteration, but this significantly increases latency, making them incapable of meeting the requirements for rapid fault isolation. Inadequate interference immunity is another core bottleneck. For example, the zero-sequence mutation amplitude comparison method is susceptible to interference from steady-state unbalance components caused by three-phase asymmetry in line parameters. This is particularly true in DG access scenarios, where inverter harmonics and power fluctuations exacerbate background noise, leading to increased false positives. Existing algorithms are typically designed for single interference sources and lack the ability to comprehensively mitigate combined interference (harmonics, noise, and imbalance), limiting their practical application in engineering.

[0004] Existing high-resistance fault location methods suffer from computational complexity and interference immunity limitations, limiting their practical application in complex distribution networks. Faced with the demands of new power systems for rapid fault isolation and high reliability, a lightweight, highly robust location method and system is urgently needed to address core issues such as weak signal extraction, multi-source interference suppression, and real-time response. Summary of the Invention

[0005] The purpose of the present invention is to provide a distribution network high-resistance fault location method and location system based on multi-stage joint filtering. By jointly filtering the zero-sequence voltage and zero-sequence current measurement values, the multi-stage filtering architecture is layered to suppress interference and highlight fault characteristics, the lightweight algorithm is used to reduce the computational load, and combined with an adaptive recursive algorithm, the method and location system can achieve rapid and accurate location of high-resistance faults in coordination with the master station and distributed measurement devices under complex working conditions.

[0006] The present invention adopts the following technical solution: a method for locating high-resistance faults in a distribution network based on multi-stage joint filtering, comprising the following steps:

[0007] Real-time collection of power data from the distribution network, including zero-sequence voltage and current signals;

[0008] Using multi-stage joint filtering technology, a signal flow graph strategy is constructed to process power data and generate output data;

[0009] Establish dynamic thresholds and compare and analyze the output data with the dynamic thresholds to determine the time when the fault occurs;

[0010] Based on the fault occurrence time, a fault measure is generated using heuristics and characteristic polarity accumulation;

[0011] Based on the dual-dimensional characteristics of fault measurement amplitude and polarity, and according to the distribution characteristics of the fault measurement of each measuring point relative to the fault point, the fault section is located through a recursive location algorithm with adaptive threshold.

[0012] Preferably, the sampling frequency f is fixed by distributing the detection points in the distribution network. s The zero-sequence voltage and zero-sequence current signals at the detection point are collected in real time.

[0013] Preferably, at least the latest two power grid frequency cycles 2T are saved based on rolling storage technology n Zero sequence voltage u m (k) and zero-sequence current signal i m (k), k is the detection point number.

[0014] Preferably, the signal flow graph strategy includes:

[0015] For zero sequence voltage u m (k) Use a differential filter with a window length of N to process and obtain the processed zero-sequence voltage output value u out (k);

[0016] For zero sequence current i m (k) Use a sliding cumulative filter with a window length of N to process and obtain the processed zero-sequence current output value I sum (k);

[0017] The zero-sequence voltage output value u out (k) and zero-sequence current output value I sum (k) multiply and fuse to get the fused output value K fu (k);

[0018] For the output value K fu(k) Use a sliding cumulative filter with a window length of N to process and obtain the processed output value K out (k).

[0019] Preferably, N is equal to one power frequency period T n The sampling frequency f s The number of sampling points; N = f s T n .

[0020] Preferably, the process of determining the fault occurrence time includes:

[0021] Using the zero-sequence voltage output value u out (k) Construct the difference value ΔV(k):

[0022] ΔV(k)=|u out (k)-u out (k-1)|;

[0023] Calculate the dynamic threshold T(k) based on the sliding window N:

[0024] T(k)=μ ΔV +α·σ ΔV ;

[0025] Among them, μ ΔV is the mean of the difference values within the window, σ ΔV The standard deviation of the difference value within the window, α is the sensitivity coefficient;

[0026] When ΔV(k)>T(k), it is marked as a candidate mutation point; and when the number of points that exceed the threshold continuously M≥2, it is determined to be the sampling point k at the start of the fault. start .

[0027] Preferably, the fault measure generation includes:

[0028] Judge K out (k) The polarity at the initial moment of the fault is determined by the trial method to obtain the trial value TCV:

[0029]

[0030] According to the polarity of the trial value TCV, K in the fixed window out (k) Generate fault measure FM using characteristic polarity accumulation:

[0031] If the test value TCV>0, calculate K within one power frequency cycle after the fault out (k) The sum of all positive values gives the fault measure FM:

[0032]

[0033] If the test value TCV≤0, calculate K within one power frequency cycle after the fault out (k) The sum of all negative values gives the fault measure FM:

[0034]

[0035] Preferably, the fault section is located by a recursive location algorithm with an adaptive threshold, including:

[0036] Determine whether the signs of the fault measures FM at the head end of each feeder are all positive. If they are all positive, it is determined to be a busbar fault. If there are negative values, it is determined to be a feeder fault.

[0037] Select min{FM s The feeder where the measuring point is located is directly determined to be the faulty feeder, and is compared with the measuring points downstream of the measuring point in sequence:

[0038] If there is a measurement point that satisfies FM j ≤0.5min{FM s}, and the adjacent measuring points downstream of the measuring point all meet FM p >0.5min{FM s}, then the fault location is determined to be in the area between the measuring point and its downstream adjacent measuring point;

[0039] If there is FM at the measuring point at the end of the feeder j <0.5min{FM s}, the fault location is determined to be in the area downstream of the measuring point and at the end of the line.

[0040] The present invention also discloses a distribution network high-resistance fault location system based on multi-stage joint filtering, comprising:

[0041] Signal acquisition module: collects power data of the distribution network in real time, including zero-sequence voltage signal and zero-sequence current signal;

[0042] Signal processing module: built-in signal flow graph strategy, combined with multi-stage joint filtering technology to process power data and generate output data, including zero-sequence voltage output value u out (k) and zero-sequence current output value I sum (k), and the output value K after the fusion and processing of the two out (k);

[0043] Time determination module: using zero sequence voltage output value u out (k) Calculate the difference value ΔV(k) and the dynamic threshold value T(k), and compare the difference value ΔV(k) and the dynamic threshold value T(k), determine and output the sampling point k at the start of the fault start ;

[0044] Measurement generation module: generates fault measurements using heuristics and characteristic polarity accumulation, and uploads the fault measurements to the master station;

[0045] The master station includes a fault location module with a built-in recursive location algorithm with adaptive thresholds to locate the fault section based on fault measurement.

[0046] Beneficial effects: The present invention processes power data by utilizing multi-stage joint filtering technology and constructing a signal flow graph strategy to effectively remove the influence of measurement noise and zero-sequence unbalanced components, realize the fusion and enhancement of fault characteristics, and can adapt to more complex environments. The algorithm has better robustness, and completes the extraction of fault characteristics while removing signal interference, compresses the step process, and makes the algorithm process more efficient. Among them, the signal flow graph strategy is used to realize the stream processing of the sampled signal point by sample point. The amount of calculation is proportional to the data scale. The processing of a single sampling point only requires 5 additions and subtractions and 1 multiplication. The algorithm complexity is only O(n), so it can run in real time in edge devices with limited computing power, and the computational complexity is lower.

[0047] In the present invention, only the fault measurement can be uploaded to the main station, and the fault measurement is uploaded to the main station only as a signed number, which makes feature extraction more comprehensive and reduces the amount of data transmission with the main station to a minimum, greatly improving the efficiency of the positioning process.

[0048] During the fault location process, the present invention adopts an adaptive threshold recursive positioning algorithm, utilizes the dual-dimensional characteristics of the amplitude and polarity of the fault measurement, and quickly locates the fault feeder according to the distribution characteristics of the measurement of each measuring point relative to the fault point. It also uses an adaptive threshold for recursive positioning to eliminate outliers caused by interference, thereby achieving accurate positioning under complex working conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] The accompanying drawings are used to provide further understanding of the present invention and constitute a part of the specification. They are used to explain the present invention together with the embodiments of the present invention and do not constitute a limitation of the present invention.

[0050] In the attached figure:

[0051] Figure 1 This is a flow chart of the method and system for locating high-resistance faults in a distribution network according to the present invention;

[0052] Figure 2 It is a multi-stage joint filtering signal flow graph of the present invention;

[0053] Figure 3 This is a schematic diagram of a 10kV radial resonant grounding system of the present invention;

[0054] Figure 4 It is the effect diagram of the present invention under complex working conditions. DETAILED DESCRIPTION

[0055] The following describes the embodiments of the present invention in conjunction with the accompanying drawings. The terms used in the embodiments of the present invention are only used to explain the specific embodiments of the present invention and are not intended to limit the present invention. The following describes the embodiments of the present application in conjunction with the accompanying drawings.

[0056] Example 1: Distribution network high-resistance fault location method based on multi-stage joint filtering, refer to Figure 1 As shown: It includes the following steps:

[0057] Step 1: Real-time collection of power data from the distribution network, including zero-sequence voltage and current signals;

[0058] By distributing the detection points in the distribution network with a fixed sampling frequency f s Real-time collection of zero-sequence voltage and zero-sequence current signals at the detection point;

[0059] And based on the rolling storage technology, at least the latest two power grid frequency cycles (2T) are saved n Zero sequence voltage u m (k) and zero-sequence current signal i m (k), k is the detection point number.

[0060] Step 2: Utilize multi-stage joint filtering technology to construct a signal flow graph strategy to process power data and generate output data. This effectively removes unbalanced components, suppresses noise interference, and achieves the fusion and enhancement of fault characteristics.

[0061] Among them, the signal flow graph strategy includes multiple associated processing modules, such as Figure 2 As shown, each module is configured to perform the following functions:

[0062] 1) Data filtering and feature extraction module:

[0063] For zero sequence voltage u m (k) Use a differential filter with a window length of N to process and obtain the processed zero-sequence voltage output value u out (k):

[0064] u out (k)=u m (k)-u m (kN);

[0065] The amplitude-frequency characteristic curve of the differential filter is:

[0066] |H D (f)|=2|sin(πf / f n )|;

[0067] Where f is the frequency, fn =1 / T n is the system operating frequency.

[0068] Therefore, |H D (f)|at f0=af n There is a zero point at (a=0,1,...,N-1), which means the frequency is af n The signal is completely suppressed, so the power frequency unbalanced component and high-order harmonic interference can be removed;

[0069] For zero sequence current i m (k) Use a sliding cumulative filter with a window length of N to process and obtain the processed zero-sequence current output value I sum (k):

[0070]

[0071] It has a recursive form that reduces the computational effort:

[0072] I sum (k)=I sum (k-1)+i m (k)-i m (kN);

[0073] Among them, the amplitude-frequency characteristic curve of the sliding accumulation filter is:

[0074]

[0075] Same as the differential filter, |H I (f)|also at f0=af n There is a zero point at the , which can remove the unbalanced power frequency component and high-order harmonic interference. However, the difference is that for low-frequency signals, the sliding accumulator filter has a gain of nearly N times. As the frequency increases, the gain decreases rapidly, thereby achieving the purpose of extracting low-frequency signal characteristics, eliminating the influence of fixed frequencies, and suppressing high-frequency noise.

[0076] 2) Feature fusion module:

[0077] The zero-sequence voltage output value u out (k) and zero-sequence current output value I sum (k) multiply and fuse to get the fused output value K fu (k):

[0078] K fu (k)=u out (k)I sum (k);

[0079] 3) Feature enhancement and result output module:

[0080] The fused output value K fu (k) Use a sliding cumulative filter with a window length of N to process and obtain the processed output value K out (k):

[0081]

[0082] It has a recursive form that reduces the computational effort:

[0083] K out (k) = K out (k-1)+K fu (k)-K fu (kN);

[0084] Through K fu (k) Using the high-frequency attenuation characteristics of the sliding cumulative filter to suppress the zero-sequence voltage output value u out (k) brings noise interference, and the integral characteristics of the sliding accumulator filter can enhance the initial fault moment K fu (k) characteristics, the delay characteristics of the sliding accumulator filter also delays the fault path (the path through which the arc suppression coil zero-sequence current flows) of the measuring point K fu The zero crossing point of (k) enables the heuristic method in step 4 to extract the characteristic polarity more accurately;

[0085] In this embodiment, N is equal to one power frequency period T n The sampling frequency f s The number of sampling points; N = f s T n ;

[0086] In this embodiment, the multi-stage filtering module is combined with Figure 2 The signal flow graph shown in the figure is executed by processing the sample points one by one. m (k) and i m (k) as input, K out (k) As output, stream processing of sampled data is realized, so it is not restricted by the data scale and can run in real time in the terminal device.

[0087] Step 3: Establish a dynamic threshold and compare and analyze the output data with the dynamic threshold to determine the time when the fault occurs. Specifically:

[0088] Using the zero-sequence voltage output value u out (k) Construct the difference value ΔV(k):

[0089] ΔV(k)=|u out (k)-u out (k-1)|;

[0090] Calculate the dynamic threshold T(k) based on the sliding window N:

[0091] T(k)=μ ΔV +α·σ ΔV ;

[0092] Among them, μ ΔV is the mean of the difference values within the window, σ ΔV The standard deviation of the difference value within the window, α is the sensitivity coefficient, and the value of α in this embodiment is 3-5;

[0093] When ΔV(k)>T(k), it is marked as a candidate mutation point; and when the number of points that exceed the threshold continuously M≥2, it is determined to be the sampling point k at the start of the fault. start

[0094] Step 4: Based on the fault occurrence time, generate the fault measure using heuristics and characteristic polarity accumulation, specifically:

[0095] Judge K out (k) The polarity at the initial moment of the fault is determined by the trial method to obtain the trial value TCV:

[0096]

[0097] According to the polarity of the trial value TCV, K in the fixed window out (k) Generate fault measure FM using characteristic polarity accumulation:

[0098] If the test value TCV>0, calculate K within one power frequency cycle after the fault out (k) The sum of all positive values gives the fault measure FM:

[0099]

[0100] If the test value TCV≤0, calculate K within one power frequency cycle after the fault out (k) The sum of all negative values gives the fault measure FM:

[0101]

[0102] Among them, the generated fault measurement can be uploaded to the main station for comprehensive analysis.

[0103] Step 5: Based on the dual-dimensional characteristics of fault measurement amplitude and polarity, and the distribution characteristics of the fault measurement of each measurement point relative to the fault point, the fault section is located using a recursive location algorithm with adaptive thresholds. Specifically:

[0104] Determine whether the signs of the fault measures FM at the head end of each feeder are all positive. If they are all positive, it is determined to be a busbar fault. If there are negative values, it is determined to be a feeder fault.

[0105] Select min{FM s The feeder where the measuring point is located is directly determined to be the faulty feeder, and is compared with the measuring points downstream of the measuring point in sequence:

[0106] If there is a measurement point that satisfies FM j ≤0.5min{FM s}, and the adjacent measuring points downstream of the measuring point all meet FM p >0.5min{FM s}, then the fault location is determined to be in the area between the measuring point and its downstream adjacent measuring point;

[0107] If there is FM at the measuring point at the end of the feeder j <0.5min{FM s}, the fault location is determined to be in the area downstream of the measuring point and at the end of the line.

[0108] Example 2, distribution network high resistance fault location system based on multi-stage joint filtering, refer to Figure 1 As shown, the positioning method based on Example 1 includes:

[0109] Signal acquisition module: collects power data of the distribution network in real time, including zero-sequence voltage signal and zero-sequence current signal;

[0110] Signal processing module: built-in signal flow graph strategy, combined with multi-stage joint filtering technology to process power data and generate output data, including zero-sequence voltage output value u out (k) and zero-sequence current output value I sum (k), and the output value K after the fusion and processing of the two out (k);

[0111] Time determination module: using zero sequence voltage output value u out (k) Calculate the difference value ΔV(k) and the dynamic threshold value T(k), and compare the difference value ΔV(k) and the dynamic threshold value T(k), determine and output the sampling point k at the start of the fault start ;

[0112] Measurement generation module: generates fault measurements using heuristics and characteristic polarity accumulation, and uploads the fault measurements to the master station;

[0113] The master station includes a fault location module with a built-in recursive location algorithm with adaptive thresholds to locate the fault section based on the fault measurement.

[0114] In a specific case, based on the method of Example 1 or the system of Example 2, a 10.5 kV resonant grounding system simulation model is built using electromagnetic transient simulation software PSCAD / EMTDC, such as Figure 3As shown in the figure. The cable line uses the YJV22-3*400 model, and the overhead line uses the JKLYJ-150 model. The arc suppression coil is overcompensated by 5%, and the arc suppression coil inductance L = 0.58H. f1 to f3 represent different single-phase ground fault fault points. Measuring points m1 and m4 are set at the beginning of feeders l1 and l2, respectively, for fault line selection; measuring points m2 and m3 are set at the beginning of the branch feeder of feeder l1 for fault location. The simulation sampling frequency is 10 kHz, so the window length N = 200.

[0115] In order to explore the robustness of this method in the presence of unbalanced components and measurement noise, the three-phase parameters of feeder l3 are adjusted to artificially create an unbalanced state of the three-phase parameters. Figure 3 At f2, a single-phase ground fault with a transition resistance of 3500Ω and an initial fault phase angle of 288° occurred in phase A. The fault time was 3.016s. A white noise with a signal-to-noise ratio of 53.7dB within one power frequency cycle at the initial time of the fault was added to the zero-sequence voltage, as shown in the following example: Figure 4 As shown in (a), white noise with an average noise energy of 1.05A2 per unit sampling point is added to the zero-sequence current, as shown in Figure 4 As shown in (b), the signal-to-noise ratios of each measuring point within one power frequency cycle at the initial moment of the fault are m1: 1.99dB, m2: -83.29dB, m3: -22.75dB, and m4: -7.73dB. Figure 4 In (f), the fault measurement of each measuring point is negative at the measuring point m1 upstream of the fault point, and positive at the measuring points downstream of the fault point and on the healthy feeder. The final positioning result is between measuring points m1 and m2, indicating that the positioning is correct.

[0116] The above describes the embodiments of the present invention in detail with reference to the accompanying drawings. However, the present invention is not limited to the above embodiments. After knowing the contents described in the present invention, ordinary technicians in this technical field can make several equivalent changes and substitutions without departing from the principles of the present invention. These equivalent changes and substitutions should also be regarded as falling within the scope of protection of the present invention.

Claims

1. A method for locating high-resistance faults in distribution networks based on multi-stage joint filtering, characterized by: The steps include: Real-time collection of power data from the distribution network, including zero-sequence voltage and current signals; Using multi-stage joint filtering technology, a signal flow graph strategy is constructed to process power data and generate output data; Establish dynamic thresholds and compare and analyze the output data with the dynamic thresholds to determine the time when the fault occurs; Based on the fault occurrence time, a fault measure is generated using heuristics and characteristic polarity accumulation; Based on the dual-dimensional characteristics of fault measurement amplitude and polarity, and according to the distribution characteristics of the fault measurement of each measuring point relative to the fault point, the fault section is located through a recursive location algorithm with adaptive threshold.

2. The method for locating high-resistance faults in distribution networks based on multi-stage joint filtering according to claim 1, characterized in that: By distributing the detection points in the distribution network with a fixed sampling frequency f s The zero-sequence voltage and zero-sequence current signals at the detection point are collected in real time.

3. The method for locating high-resistance faults in a distribution network based on multi-stage joint filtering according to claim 1 or 2, characterized in that: Based on rolling storage technology, at least the latest two power grid frequency cycles are saved Zero sequence voltage u m (k) and zero-sequence current signal i m (k).

4. The method for locating high-resistance faults in a distribution network based on multi-stage joint filtering according to claim 1, characterized in that: The signal flow graph strategy includes: For zero sequence voltage u m (k) Use a differential filter with a window length of N to process and obtain the processed zero-sequence voltage output value u out (k); For zero sequence current i m (k) Use a sliding cumulative filter with a window length of N to process and obtain the processed zero-sequence current output value I sum (k); The zero-sequence voltage output value u out (k) and zero-sequence current output value I sum (k) multiply and fuse to get the fused output value K fu (k); For the output value K fu (k) Use a sliding cumulative filter with a window length of N to process and obtain the processed output value K out (k).

5. The method for locating high-resistance faults in distribution networks based on multi-stage joint filtering according to claim 4, characterized in that: N is equal to one power frequency period T n The sampling frequency f s The number of sampling points, that is: N = f s T n .

6. The method for locating high-resistance faults in a distribution network based on multi-stage joint filtering according to claim 4 or 5, characterized in that: The process of determining the fault occurrence time includes: Using the zero-sequence voltage output value u out (k) Construct the difference value ΔV(k): ΔV(k)=|u out (k)-u out (k-1)|; Calculate the dynamic threshold T(k) based on the window length N: T(k)=μ ΔV +a·s ΔV ; Among them, μ ΔV is the mean of the difference values within the window, σ ΔV The standard deviation of the difference value within the window, α is the sensitivity coefficient; When ΔV(k)>T(k), it is marked as a candidate mutation point; and when the number of points that exceed the threshold continuously M≥2, it is determined to be the sampling point k at the start of the fault. start .

7. The method for locating high-resistance faults in a distribution network based on multi-stage joint filtering according to claim 6, characterized in that: The fault measure generation includes: Judge K out (k) The polarity at the initial moment of the fault is determined by the trial method to obtain the trial value TCV: According to the polarity of the trial value TCV, K in the fixed window out (k) Generate fault measure FM using characteristic polarity accumulation: If the test value TCV>0, calculate K within one power frequency cycle after the fault out (k) The sum of all positive values gives the fault measure FM: If the test value TCV≤0, calculate K within one power frequency cycle after the fault out (k) The sum of all negative values gives the fault measure FM:

8. The method for locating high-resistance faults in a distribution network based on multi-stage joint filtering according to claim 7, characterized in that: The fault section is located using a recursive location algorithm with adaptive thresholds, including: Determine whether the signs of the fault measures FM at the head end of each feeder are all positive. If they are all positive, it is determined to be a busbar fault. If there are negative values, it is determined to be a feeder fault. Select min{FM s The feeder where the measuring point is located is directly determined to be the faulty feeder, and is compared with the measuring points downstream of the measuring point in sequence: If there is a measurement point that satisfies FM j ≤0.5min{FM s }, and the adjacent measuring points downstream of the measuring point all meet FM p >0.5min{FM s }, then the fault location is determined to be in the area between the measuring point and its downstream adjacent measuring point; If there is FM at the measuring point at the end of the feeder j <0.5min{FM s }, the fault location is determined to be in the area downstream of the measuring point and at the end of the line.

9. A distribution network high-resistance fault location system based on multi-stage joint filtering, characterized by: include: Signal acquisition module: collects power data of the distribution network in real time, including zero-sequence voltage signal and zero-sequence current signal; Signal processing module: built-in signal flow graph strategy, combined with multi-stage joint filtering technology to process power data and generate output data, including zero-sequence voltage output value u out (k) and zero-sequence current output value I sum (k), and the output value K after the fusion and processing of the two out (k); Time determination module: using zero sequence voltage output value u out (k) Calculate the difference value ΔV(k) and the dynamic threshold value T(k), and compare the difference value ΔV(k) and the dynamic threshold value T(k), determine and output the sampling point k at the start of the fault start ; Measurement generation module: generates fault measurements using heuristics and characteristic polarity accumulation, and uploads the fault measurements to the master station; The master station includes a fault location module with a built-in recursive location algorithm with adaptive thresholds to locate the fault section based on the fault measurement.

Images