Improved traveling wave distance measurement method based on power distribution network fault location analysis

Abstract

The application discloses an improved method for traveling wave distance measurement based on power distribution network fault location analysis, and relates to the technical field of traveling wave distance measurement, solves the technical problem that fault related paths and irrelevant branches cannot be effectively distinguished, and path identification errors are prone to occur, and the application realizes accurate identification of fault related paths, improves irrelevant branch elimination rate and multi-branch line path matching degree, reduces positioning error, through the construction of a dynamic topology directed graph + GIS spatial correlation mechanism, through terminal time difference consistency verification and physical path constraint, realizes accurate identification of fault related paths, improves irrelevant branch elimination rate and multi-branch line path matching degree, reduces positioning error, and solves the problem of wave speed uncertainty, constructs a multi-dimensional weighted model, adapts different fault scenes comprehensively, improves correction accuracy, reduces positioning error in complex scenes such as mixed lines and new energy grid connection, and meets the engineering operation and maintenance requirements.

Description

Technical Field

[0001] This invention relates to the field of traveling wave ranging technology, specifically to an improved traveling wave ranging method based on distribution network fault location analysis. Background Technology

[0002] Traveling wave ranging technology has become one of the core technologies for fault location in power distribution networks due to its fast positioning speed and high theoretical accuracy. Its core principle is to calculate the fault distance by detecting the time difference between the arrival of the transient traveling wave generated at the fault point to different monitoring terminals and combining the traveling wave propagation speed.

[0003] According to patent application number CN202510774950.8, an improved traveling wave ranging method for distribution network fault location analysis is disclosed. This method uses waveform translation and coincidence verification to calculate the average amplitude difference under different translation amounts, automatically identifying the optimal time alignment point and locking the associated time difference t1. This process effectively corrects time errors caused by signal distortion, asynchronous sampling, etc., improving the time difference accuracy from microseconds to nanoseconds, laying the foundation for subsequent precise location. Multiple associated points are randomly selected within the locked range, and a corrected waveform is constructed based on the attenuation ratio and repeatedly verified. The optimal solution is selected through a feature mean minimization algorithm. This mechanism effectively suppresses single-point errors and improves the robustness of the location results through statistical averaging effects.

[0004] However, the multi-branch structure of the distribution network can cause traveling waves to reflect multiple times at the branch points, forming false wavefronts. Existing technologies lack a dynamic correlation mechanism with the physical topology of the distribution network, making it impossible to effectively distinguish fault-related paths from irrelevant branches, which can easily lead to path identification errors. Secondly, existing technologies mostly output fault locations based on line mileage, without deep integration with GIS spatial coordinates and lack physical constraint verification, which may cause the positioning results to deviate from the actual line path, resulting in poor operational and maintenance practicality. Although some technologies use multi-terminal data correction, the weight allocation only depends on a single factor and does not consider key indicators such as wavefront confidence and signal amplitude, resulting in insufficient correction accuracy. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides an improved traveling wave ranging method based on distribution network fault location analysis, which solves the problem of being unable to effectively distinguish between fault-related paths and irrelevant branches, and the tendency to make path identification errors.

[0006] To achieve the above objectives, the present invention provides the following technical solution: an improved traveling wave ranging method based on distribution network fault location analysis, which specifically includes the following steps: Step 1: Collect geographic information data of the distribution network. Using the physical nodes of the distribution network as vertices and the line segments between adjacent vertices as directed edges, construct a dynamic topology directed graph of the distribution network. Deploy monitoring terminals to collect traveling wave signals and perform noise reduction processing on the traveling wave signals to obtain preprocessed traveling wave signals. Step 2: Acquire the preprocessed traveling wave signal and calculate the corresponding TEO energy. Select wavefronts with the maximum TEO energy greater than the judgment threshold as valid candidate wavefronts. Select a reference wavefront and mark its polarity. Select valid wavefronts based on polarity. Count the number of criteria satisfied by the valid wavefronts and calculate the confidence level. Classify the valid wavefronts as high-reliability wavefronts, medium-reliability wavefronts, and low-reliability wavefronts. Step 3: Extract high-reliability wavefronts to form a standardized time series, call the distribution network dynamic topology directed graph, search for the shortest propagation path that satisfies physical constraints, calculate the actual traveling wave propagation time difference and theoretical time difference for any terminal pair, if the difference is less than the time window threshold for scenario adaptation, mark it as a candidate path, construct a set of positioning equations based on the candidate path to calculate the initial location of the fault point, and correct the initial location of the fault point based on the optimal wave velocity to obtain the fault location. Step 4: Convert the fault location into initial spatial coordinates and associate it with the corresponding line segment. Extract the physical constraints of the line segment. If the physical constraints are not met, trigger the correction mechanism. Construct a multi-dimensional weighted coefficient model and a weighted positioning equation set. Iteratively solve the equations to obtain the precise location of the fault point and generate fault location information.

[0007] As a further aspect of the present invention, the method for constructing a dynamic topology directed graph of the distribution network is as follows: Using physical nodes of the distribution network as vertices and line segments between adjacent vertices as directed edges, a dynamic topology directed graph of the distribution network is constructed. The vertices include busbars, branch points, line terminals, and segment boundary points. The core attributes associated with the directed edges include line segment length, wave impedance, and wave velocity. The attributes associated with the vertices include node type, number of connected branches, and terminal deployment status.

[0008] As a further aspect of the present invention, the effective wavefront selection method is as follows: Obtain the preprocessed traveling wave signal x(t), and according to formula E TEO (t)=x(t) 2 The corresponding TEO energy is calculated using -x(t-1)·x(t+1), where t is the sampling time, and the energy is determined according to the formula E. y =k1×E avg The judgment threshold is calculated, where k1 is the energy coefficient, and E avg The average TEO energy of the signal 2.5 seconds before the fault is obtained, followed by the maximum TEO energy E corresponding to the wavefront. max And compare it with the judgment threshold, if E max ≥Ey If so, the corresponding wavefront is marked as a valid candidate wavefront; The valid candidate wavefronts are labeled as i, and i = 1, 2, ..., j, where j represents the number of valid candidate wavefronts. Then, the first wavefront that satisfies the energy criterion is extracted as the reference wavefront, and its polarity is marked as P0. For subsequent candidate wavefronts i, their polarity P is obtained. i If P i =-P0, denoted as the effective wavefront, if P i =P0, then analyze in conjunction with the amplitude criterion, if the amplitude of the current valid candidate wavefront i is A i The reference wavefront amplitude A0 satisfies 0.3 ≤ A i If / A0≤0.7, it is determined to be a reflection wavefront of the opposite bus and is removed. If the amplitude criterion is not met, it is determined to be an interference wavefront and is removed.

[0009] As a further aspect of the present invention, the effective wavefronts are classified into high-reliability wavefronts, medium-reliability wavefronts, and low-reliability wavefronts as follows: The number of valid wavefronts that meet the criteria is counted. The criteria include energy-related criteria and consistency-related criteria. Energy-related criteria include TEO peak energy criteria and energy duration criteria. Consistency-related criteria include time difference consistency criteria and amplitude attenuation consistency criteria. The confidence level is calculated according to the formula: confidence level = (number of criteria met / total number of criteria) × 100%. Wavefronts with a confidence level ≥ 90% are classified as high-reliability wavefronts, those with a confidence level between 70% and 90% are classified as medium-reliability wavefronts, and those with a confidence level < 70% are classified as low-reliability wavefronts and removed.

[0010] As a further aspect of the present invention, the candidate path selection method is as follows: Based on the obtained effective wavefronts, the high-reliability wavefronts are identified, and their arrival time series are extracted to form a standardized time series T=[t1, t2, ..., t]. n ], where t n Given the arrival time of the nth high-reliability wavefront, starting from the first terminal that detects the high-reliability wavefront, traverse all nodes and directed edges to search for the physical constraint v×. ≈ Shortest propagation path for line length, for any pair of terminals (a, b), calculate the actual traveling wave propagation time difference of the pair. =|t a -t b | Extracting the shortest path length L between terminal a and terminal b based on a dynamic directed graph. ab Combined with the corrected wave velocity v1, the theoretical time difference t is calculated. theo =L AB / v1, sets the time window threshold for scene adaptation. If | -t theoIf |≤ the time window threshold, then the path is determined to be a fault-related path, and the corresponding branch is retained and recorded as a candidate path; otherwise, irrelevant branches are removed.

[0011] As a further aspect of the present invention, the specific method for obtaining the fault location is as follows: Next, candidate paths are obtained, and the effective wavefront times of at least three terminals in the candidate paths are selected to construct a set of positioning equations. The three terminals are denoted as A1, A2, and A3, respectively. L A1F L A1F and L A1F t represents the distance from fault point F to terminals A1, A2, and A3, respectively. F Given the time of the fault occurrence, the initial location L0 of the fault point is obtained by solving the system of equations using the least squares method. The initial fault location L0 is corrected to obtain the actual time difference of the terminal. And calculate the theoretical time difference based on the initial location L0 of the fault point, according to the formula = The theoretical time difference was calculated. , where v 初 This represents the initial corrected wave velocity, according to the formula. =| - |Calculate the residual If residual The preset value is then used to solve for the optimal wave velocity v using the least squares method. 优 , specific v 优 = v is the corrected wave velocity, and the optimal wave velocity v is calculated accordingly. 优 Recalculate the fault location L1, specifically L1 = v 优 × / 2, where This is the effective time difference.

[0012] As a further aspect of the present invention, the method for generating fault location information is as follows: The calculated fault location L1 is mapped to a pre-constructed directed graph of the distribution network dynamic topology, and the mileage value corresponding to L1 is converted into initial spatial coordinates (X1, Y1, Z). 1） This is then associated with the corresponding line segment in the directed graph of the topology. Simultaneously, the physical constraints of the corresponding line segment are extracted, including the azimuth of the line direction, the path buffer, and the precise coordinates of the physical nodes at both ends. Next, based on the physical constraints, the rationality of the fault location is determined, judging (X1, Y... 1） Whether it falls within the path buffer; if it exceeds the buffer, it is determined to be a path deviation, triggering the correction mechanism.

[0013] As a further aspect of the present invention, the specific method for triggering the correction mechanism is as follows: The initial spatial coordinates (X1, Y1, Z) of the initial location L1 of the fault point are given. 1） Centered on the line, a search area with radius R is defined, and R is dynamically adjusted according to the line length, typically taking 5%-10% of the total line length. Terminal location information corresponding to all high-reliability wavefronts within this area is extracted. Next, a multi-dimensional weighted coefficient model is constructed, including spatial distance weights w1= , where d a The spatial distance from terminal a to the initial spatial coordinates, and the wavefront confidence weight w1 = C a For the confidence level of the high-reliability wavefront of terminal a, the signal amplitude weight w2 = A a Let w_total be the traveling wave amplitude of terminal a. The final weight is calculated by combining multi-dimensional weights: w_total = w1 + w2 + w3.

[0014] As a further aspect of the present invention, based on the obtained final weights, combined with the optimal wave velocity v 优 Reconstruct the weighted positioning equations ,in For the effective time difference of the high-reliability wavefront of terminal a, L 基准 Let b be the mileage reference value of the physical node where the terminal is located, and b represent the number of terminals. By iteratively solving the optimized system of equations, the precise location L of the fault point after spatial calibration can be obtained. 终 At the same time, fault location information is generated.

[0015] This invention provides an improved traveling wave ranging method based on distribution network fault location analysis. Compared with existing technologies, it has the following advantages: This invention constructs a dynamic topological directed graph and a GIS spatial association mechanism. By verifying the consistency of time difference and physical path constraints at the terminal, it achieves accurate identification of fault-related paths, improves the rate of irrelevant branch removal and the matching degree of multi-branch line paths, and reduces positioning errors. Secondly, it optimizes wave velocity through residual verification and least squares iterative optimization, combined with dynamic correction of temperature and load current, to reduce wave velocity errors and solve the problem of wave velocity uncertainty.

[0016] This invention constructs a multi-dimensional weighted model, which integrates weights to adapt to different fault scenarios, improves correction accuracy, reduces positioning errors in complex scenarios such as mixed lines and new energy grid connection, and meets the needs of engineering operation and maintenance. At the same time, it adopts a design that integrates primary and secondary terminals with existing GIS platforms, which does not require large-scale modification of existing equipment, can achieve full branch coverage, reduce deployment costs, and has an anomaly handling mechanism to improve system reliability and adapt to distribution networks of different voltage levels. Attached Figure Description

[0017] Figure 1 This is a flowchart of the improved traveling wave ranging method of the present invention. Detailed Implementation

[0018] 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 embodiments of the present invention, and not all embodiments. 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.

[0019] Please see Figure 1 This application provides an improved traveling wave ranging method based on distribution network fault location analysis, which specifically includes the following steps: Step 1: Collect GIS data of the distribution network, including line length, branch point location, conductor type, wave impedance parameters, and overhead / cable segment information. Using physical nodes of the distribution network as vertices, and defining fixed points covering busbars, branch points, line terminals, and segment boundary points, and using line segments between adjacent vertices as directed edges, construct a dynamic topology directed graph of the distribution network. The core attributes associated with directed edges include line segment length, wave impedance, and wave velocity. Vertex associated attributes include node type, number of connected branches, and terminal deployment status. Simultaneously, establish a basic parameter database to store the structured information of each branch. Next, deploy monitoring terminals at both ends of the main line and the beginning of each branch, ensuring at least one terminal covers each branch. Simultaneously, use the set monitoring terminals to collect traveling wave signals and perform noise reduction processing on the obtained traveling wave signals to obtain pre-processed traveling wave signals. The specific processing method is as follows: First, a hardware filtering circuit is used to filter out the 50Hz power frequency fundamental wave and harmonic interference above the 20th order. Then, a variational mode decomposition algorithm is used for signal denoising. The number of decomposed modes K=5 is set to eliminate low-frequency noise modes with a frequency <500kHz and retain the effective frequency band components of the traveling wave from 500kHz to 2MHz. Finally, a pre-processed traveling wave signal with a signal-to-noise ratio ≥35dB and complete wavefront characteristics is output.

[0020] Step 2: Obtain the preprocessed traveling wave signal x(t), and apply it according to formula E. TEO (t)=x(t) 2 The corresponding TEO energy is calculated using -x(t-1)·x(t+1), where t is the sampling time, and the energy is determined according to the formula E. y =k1×E avg The judgment threshold is calculated, where k1 is the energy coefficient, and E avg The average TEO energy of the signal 2.5 seconds before the fault is obtained, followed by the maximum TEO energy E corresponding to the wavefront. max And compare it with the judgment threshold, if E max ≥Ey If so, the corresponding wavefront is marked as a valid candidate wavefront; All valid candidate wavefronts are obtained and labeled as i, where i = 1, 2, ..., j, and j represents the number of valid candidate wavefronts. Then, the first wavefront that satisfies the energy criterion is extracted as the reference wavefront, and its polarity is marked as P0. For subsequent candidate wavefronts i, their polarities P are obtained. i If P i =-P0, where "-" indicates a negative direction, which is determined to be the reflected wavefront at the fault point and recorded as a valid wavefront. If P i =P0, then analyze in conjunction with the amplitude criterion, if the amplitude of the current valid candidate wavefront i is A i The reference wavefront amplitude A0 satisfies 0.3 ≤ A i If / A0≤0.7, it is determined to be a reflection wavefront of the opposite bus and is removed. If the amplitude criterion is not met, it is determined to be an interference wavefront and is removed. Finally, all valid wavefronts are selected. Next, all valid wavefronts are obtained, along with the number of corresponding criteria. These criteria include energy-related criteria, such as TEO peak energy criterion and energy duration criterion, and consistency-related criteria, such as time difference consistency criterion and amplitude attenuation consistency criterion. Simultaneously, the confidence level of each valid wavefront is calculated using the formula: confidence level = (number of satisfied criteria / total number of criteria) × 100%. Based on this confidence level, valid wavefronts are classified as follows: those with a confidence level ≥ 90% are classified as high-reliability wavefronts and directly used for ranging; those with a confidence level between 70% and 90% are classified as medium-reliability wavefronts; and those with a confidence level < 70% are classified as low-reliability wavefronts and removed after cross-validation with other wavefronts.

[0021] Step 3: Based on the obtained effective wavefronts, acquire the high-reliability wavefronts and form a standardized time series T=[t1, t2, ..., t] based on the arrival time series of the high-reliability wavefronts. n ], where tn is the arrival time of the nth high-reliability wavefront. The system calls the distribution network dynamic topology directed graph and basic parameter database, starting from the first terminal that detects a high-reliability wavefront, traversing all nodes and directed edges to search for the physical constraint v×. ≈ Shortest propagation path for line length, for any pair of terminals (a, b), calculate the actual traveling wave propagation time difference of the pair. =|t a -t b | Extracting the shortest path length L between terminal a and terminal b based on a dynamic directed graph. ab Combined with the corrected wave velocity v1, the theoretical time difference t is calculated. theo =L AB / v1, sets the time window threshold for scene adaptation. If | -t theoIf |≤ the time window threshold, then the path is determined to be a fault-related path, and the corresponding branch is retained and recorded as a candidate path. If all terminals corresponding to the same branch do not meet the above constraints in terms of time difference, or only a single terminal detects a high-reliability wavefront, then the branch is determined to be an irrelevant branch and is removed. Next, candidate paths are obtained, and the effective wavefront times of at least three terminals in the candidate paths are selected to construct a set of positioning equations. The three terminals are denoted as A1, A2, and A3, respectively. L A1F L A1F and L A1F t represents the distance from fault point F to terminals A1, A2, and A3, respectively. F Given the time of the fault occurrence, the initial location L0 of the fault point is obtained by solving the system of equations using the least squares method. The initial fault location L0 is corrected to obtain the actual time difference of the terminal. And calculate the theoretical time difference based on the initial location L0 of the fault point, according to the formula = The theoretical time difference was calculated. , where v 初 This represents the initial corrected wave velocity, according to the formula. =| - |Calculate the residual If residual The preset value is then used to solve for the optimal wave velocity v using the least squares method. 优 , specific v 优 = v is the corrected wave velocity, and the optimal wave velocity v is calculated accordingly. 优 Recalculate the fault location L1, specifically L1 = v 优 × / 2, where This is the effective time difference.

[0022] Step 4: Map the calculated fault location L1 to the pre-constructed directed graph of the distribution network dynamic topology, and convert the mileage value corresponding to L1 into initial spatial coordinates (X1, Y1, Z). 1） This is then associated with the corresponding line segment in the directed graph of the topology. Simultaneously, the physical constraints of the corresponding line segment are extracted, including the azimuth of the line direction, the path buffer, and the precise coordinates of the physical nodes at both ends. Next, based on the physical constraints, the rationality of the fault location is determined, judging (X1, Y... 1） Whether it falls within the path buffer; if it exceeds the buffer, it is determined to be a path deviation, triggering the correction mechanism. First, the initial spatial coordinates (X1, Y1, Z1) of the initial location L1 of the fault point are used. 1）Centered on the line, a search area with radius R is defined, and R is dynamically adjusted according to the line length, typically taking 5%-10% of the total line length. Terminal location information corresponding to all high-reliability wavefronts within this area is extracted. Next, a multi-dimensional weighted coefficient model is constructed, including spatial distance weights w1= , where d a The spatial distance from terminal a to the initial spatial coordinates, and the wavefront confidence weight w1 = C a For the confidence level of the high-reliability wavefront of terminal a, the signal amplitude weight w2 = A a Given the traveling wave amplitude of terminal a, the final weight is calculated by combining multi-dimensional weights: w_total = w1 + w2 + w3; Based on the obtained final weights, combined with the optimal wave speed v 优 Reconstruct the weighted positioning equations ,in For the effective time difference of the high-reliability wavefront of terminal a, L 基准 Let b be the mileage reference value of the physical node where the terminal is located, and b represent the number of terminals. The optimized system of equations is solved iteratively using the Gauss-Newton iterative method. The iteration termination condition is set as follows: the positional deviation between two iterations ≤ 0.5m or the number of iterations ≥ 5, ensuring convergence. This yields the precise location L of the fault point after spatial calibration. 终 At the same time, fault location information is generated.

[0023] Some of the data in the above formulas are numerical calculations with dimensions removed, and the contents not described in detail in this specification are all prior art known to those skilled in the art.

[0024] The above embodiments are only used to illustrate the technical methods of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical methods of the present invention without departing from the spirit and scope of the technical methods of the present invention.

Claims

1. An improved traveling wave ranging method based on distribution network fault location analysis, characterized in that, The method specifically includes the following steps: Step 1: Collect geographic information data of the distribution network. Using the physical nodes of the distribution network as vertices and the line segments between adjacent vertices as directed edges, construct a dynamic topology directed graph of the distribution network. Deploy monitoring terminals to collect traveling wave signals and perform noise reduction processing on the traveling wave signals to obtain preprocessed traveling wave signals. Step 2: Acquire the preprocessed traveling wave signal and calculate the corresponding TEO energy. Select wavefronts with the maximum TEO energy greater than the judgment threshold as valid candidate wavefronts. Select a reference wavefront and mark its polarity. Select valid wavefronts based on polarity. Count the number of criteria satisfied by the valid wavefronts and calculate the confidence level. Classify the valid wavefronts as high-reliability wavefronts, medium-reliability wavefronts, and low-reliability wavefronts. Step 3: Extract high-reliability wavefronts to form a standardized time series, call the distribution network dynamic topology directed graph, search for the shortest propagation path that satisfies physical constraints, calculate the actual traveling wave propagation time difference and theoretical time difference for any terminal pair, if the difference is less than the time window threshold for scenario adaptation, mark it as a candidate path, construct a set of positioning equations based on the candidate path to calculate the initial location of the fault point, and correct the initial location of the fault point based on the optimal wave velocity to obtain the fault location. Step 4: Convert the fault location into initial spatial coordinates and associate it with the corresponding line segment. Extract the physical constraints of the line segment. If the physical constraints are not met, trigger the correction mechanism. Specifically, take the initial spatial coordinates (X1, Y1, Z1) of the initial fault location L1 as the center and set a search area with a radius of R. R is dynamically adjusted according to the line length, taking 5%-10% of the total line length. Extract the terminal location information corresponding to all high-reliability wavefronts within this area. Then, construct a multi-dimensional weighted coefficient model, including spatial distance weight w1= , where d a The spatial distance from terminal a to the initial spatial coordinates, and the wavefront confidence weight w1 = C a For the confidence level of the high-reliability wavefront of terminal a, the signal amplitude weight w2 = A a Given the traveling wave amplitude of terminal a, the final weight is calculated by combining multi-dimensional weights: w_total = w1 + w2 + w3; A multidimensional weighted coefficient model and a weighted localization equation system are constructed, and the precise location of the fault point is obtained by iterative solution. Specifically, based on the obtained final weights, the optimal wave velocity v is combined with the equations. 优 Reconstruct the weighted positioning equations ,in For the effective time difference of the high-reliability wavefront of terminal a, L 基准 Let b be the mileage reference value of the physical node where the terminal is located, and b represent the number of terminals. By iteratively solving the optimized system of equations, the precise location L of the fault point after spatial calibration can be obtained. 终 This generates fault location information.

2. The improved traveling wave ranging method based on distribution network fault location analysis according to claim 1, characterized in that, The method for constructing a dynamic topology directed graph of a distribution network is as follows: Using physical nodes of the distribution network as vertices and line segments between adjacent vertices as directed edges, a dynamic topology directed graph of the distribution network is constructed. The vertices include busbars, branch points, line terminals, and segment boundary points. The core attributes associated with the directed edges include line segment length, wave impedance, and wave velocity. The attributes associated with the vertices include node type, number of connected branches, and terminal deployment status.

3. The improved traveling wave ranging method based on distribution network fault location analysis according to claim 1, characterized in that, The selection method for valid wavefronts is as follows: Obtain the preprocessed traveling wave signal x(t), and according to formula E TEO (t)=x(t) 2 The corresponding TEO energy is calculated using -x(t-1)·x(t+1), where t is the sampling time, and the energy is determined according to the formula E. y =k1×E avg The judgment threshold is calculated, where k1 is the energy coefficient, and E avg The average TEO energy of the signal 2.5 seconds before the fault is obtained, followed by the maximum TEO energy E corresponding to the wavefront. max And compare it with the judgment threshold, if E max ≥E y If so, the corresponding wavefront is marked as a valid candidate wavefront; The valid candidate wavefronts are labeled as i, and i = 1, 2, ..., j, where j represents the number of valid candidate wavefronts. Then, the first wavefront that satisfies the energy criterion is extracted as the reference wavefront, and its polarity is marked as P0. For subsequent candidate wavefronts i, their polarity P is obtained. i If P i =-P0, denoted as the effective wavefront, if P i =P0, then analyze in conjunction with the amplitude criterion, if the amplitude of the current valid candidate wavefront i is A i The reference wavefront amplitude A0 satisfies 0.3 ≤ A i If / A0≤0.7, it is determined to be a reflection wavefront of the opposite bus and is removed. If the amplitude criterion is not met, it is determined to be an interference wavefront and is removed.

4. The improved traveling wave ranging method based on distribution network fault location analysis according to claim 1, characterized in that, The effective wavefronts are classified into high-reliability wavefronts, medium-reliability wavefronts, and low-reliability wavefronts as follows: The number of valid wavefronts that meet the criteria is counted. The criteria include energy-related criteria and consistency-related criteria. Energy-related criteria include TEO peak energy criteria and energy duration criteria. Consistency-related criteria include time difference consistency criteria and amplitude attenuation consistency criteria. The confidence level is calculated according to the formula: confidence level = (number of criteria met / total number of criteria) × 100%. Wavefronts with a confidence level ≥ 90% are classified as high-reliability wavefronts, those with a confidence level between 70% and 90% are classified as medium-reliability wavefronts, and those with a confidence level < 70% are classified as low-reliability wavefronts and removed.

5. The improved traveling wave ranging method based on distribution network fault location analysis according to claim 1, characterized in that, The candidate path selection method is as follows: Based on the obtained effective wavefronts, the high-reliability wavefronts are identified, and their arrival time series are extracted to form a standardized time series T=[t1, t2, ..., t]. n ], where t n Given the arrival time of the nth high-reliability wavefront, starting from the first terminal that detects the high-reliability wavefront, traverse all nodes and directed edges to search for the physical constraint v×. ≈ Shortest propagation path for line length, for any pair of terminals (a, b), calculate the actual traveling wave propagation time difference of the pair. =|t a -t b | Extracting the shortest path length L between terminal a and terminal b based on a dynamic directed graph. ab Combined with the corrected wave velocity v1, the theoretical time difference t is calculated. theo =L AB / v1, sets the time window threshold for scene adaptation. If | -t theo If |≤ the time window threshold, then the path is determined to be a fault-related path, and the corresponding branch is retained and recorded as a candidate path; otherwise, irrelevant branches are removed.

6. The improved traveling wave ranging method based on distribution network fault location analysis according to claim 1, characterized in that, The specific method for obtaining the fault location is as follows: Next, candidate paths are obtained, and the effective wavefront times of at least three terminals in the candidate paths are selected to construct a set of positioning equations. The three terminals are denoted as A1, A2, and A3, respectively. L A1F L A1F and L A1F t represents the distance from fault point F to terminals A1, A2, and A3, respectively. F Given the time of the fault occurrence, the initial location L0 of the fault point is obtained by solving the system of equations using the least squares method. The initial fault location L0 is corrected to obtain the actual time difference of the terminal. And calculate the theoretical time difference based on the initial location L0 of the fault point, according to the formula = The theoretical time difference was calculated. , where v 初 This represents the initial corrected wave velocity, according to the formula. =| - |Calculate the residual If residual The preset value is then used to solve for the optimal wave velocity v using the least squares method. 优 , specific v 优 = v is the corrected wave velocity, and the optimal wave velocity v is calculated accordingly. 优 Recalculate the fault location L1, specifically L1 = v 优 × / 2, where This is the effective time difference.

7. The improved traveling wave ranging method based on distribution network fault location analysis according to claim 1, characterized in that, The method for generating fault location information is as follows: The calculated fault location L1 is mapped to a pre-constructed directed graph of the dynamic topology of the distribution network. The mileage value corresponding to L1 is converted into initial spatial coordinates (X1, Y1, Z1) and associated with the corresponding line segment in the directed graph of the topology. At the same time, the physical constraints of the corresponding line segment are extracted, including the azimuth angle of the line direction, the path buffer, and the precise coordinates of the physical nodes at both ends. Then, the rationality of the fault location is judged based on the physical constraints. It is determined whether (X1, Y1) falls within the path buffer. If it exceeds the buffer, it is determined as a path deviation and the correction mechanism is triggered.

Images