A method for identifying a key path of metro stray current intrusion into a city power grid

Abstract

The application discloses a subway stray current invasion city power grid key path identification method, relates to the field of rail transit and power system analysis, and establishes a subway-city power grid space coupling graph network topology model according to the electrical connection relationship among a subway traction power supply system, a rail return system, a drainage network, soil medium, underground pipelines and a city power grid grounding system, and constructs a whole network node potential transfer impedance matrix based on edge weight; the injection source node set is determined according to the possible running position of a train, the return flow sink node set is determined according to the corresponding position of a negative pole of a traction substation, and the alternating current power grid side target interval is determined according to the grounding position of a transformer neutral point and the associated area; under the condition of multi-source concurrency, the injection source position distribution is exhausted, a plurality of injection current sets are formed, and the return current shared by each return flow sink node is solved through an improved node voltage method; the node potential and the edge signal strength of each edge are further calculated, and the average wandering betweenness of each edge is obtained through statistics; finally, based on the local reference threshold value of the target node in the target interval, the key path of the subway stray current invasion city power grid is identified and output. The method can improve the accuracy and pertinence of stray current invasion path identification under a complex coupling scenario.

Description

Technical Field

[0001] This application relates to the field of rail transit and power system analysis technology, and in particular to a method for identifying critical paths of stray current intrusion into urban power grids from subways. Background Technology

[0002] With the rapid development of urban rail transit, subway systems generally adopt DC traction power supply. During train operation, because the rails, which serve as the return circuit, cannot achieve complete insulation from the ground, some traction current leaks from the rails into the surrounding soil, forming stray currents. These stray currents traveling in the underground multi-medium environment can cause distortion of the nearby ground potential distribution. When subway lines overlap with the urban AC power grid in geographical space, stray currents are easily driven by the ground potential difference and flow into the transformer neutral point through the substation grounding point, causing DC bias in the transformer. This leads to distortion of the transformer excitation current, increased noise and vibration, and localized overheating, seriously affecting the safe and stable operation of the AC power grid.

[0003] Current research primarily relies on constructing refined pure physical field models (such as electromagnetic field simulation based on the finite element method and CDEGS grounding analysis software) or equivalent circuit models with distributed parameters. These methods discretize subway rails, drainage networks, soil stratification structures, and underground pipeline networks to establish a three-dimensional spatial DC resistance network containing a massive number of nodes, thereby calculating the static distribution of stray currents and the neutral point potential of transformers.

[0004] However, existing research methods and evaluation systems still have significant limitations in practical engineering applications and path identification: First, traditional physical field modeling methods are highly dependent on boundary conditions and system parameters. These methods require extremely detailed input of physical parameters such as soil resistivity, three-dimensional pipe network topology, and precise contact resistance. At the macroscopic city-level scale, these parameters are difficult to obtain and dynamically change with the environment (e.g., humidity, season), resulting in large-scale model calculations and long solution times, making it difficult to meet the timeliness requirements of online evaluation. Second, existing research mainly focuses on the numerical calculation of the overall system leakage level, lacking methods for identifying stray current propagation paths. Therefore, there is an urgent need for an identification method based on network topology characteristics that can accurately reflect the diffusion law of stray currents under multi-source, multi-path conditions, thereby quickly and accurately identifying the critical paths of stray current intrusion into urban power grids in complex coupled systems. Summary of the Invention

[0005] This invention addresses the technical problems existing in the background art by proposing a method for identifying critical paths of stray current intrusion into urban power grids from subways. This method introduces complex network theory and considers the physical reality of multiple trains running concurrently and multiple substations operating in parallel. By constructing an improved nodal voltage method matrix model with multiple sources and sinks, it achieves an accurate solution for the global stray current superposition and distribution, and ultimately accurately identifies and locates the critical intrusion path on the AC power grid side.

[0006] To solve the technical problem, the technical solution of the present invention is as follows:

[0007] A method for identifying critical paths of stray current intrusion into urban power grids from subways includes the following steps:

[0008] Based on the electrical connection relationships between the metro traction power supply system, rail return system, drainage network, soil medium and urban power grid grounding system, a graph network topology model of metro-urban power grid spatial coupling is established, and the potential transfer impedance matrix of all network nodes is obtained based on the graph network topology model.

[0009] The injection source node set is determined based on the possible operating location of the train, the return sink node set is determined based on the rail return location corresponding to the negative pole of the traction substation, and the target section on the AC power grid side is determined based on the grounding correlation location of the urban power grid.

[0010] Under multi-source concurrent conditions, the distribution of injection source locations in the injection source node set is exhaustively enumerated to form multiple injection current sets;

[0011] For each set of injected currents, the equivalent open-circuit potential at the return current sink is calculated based on the network node potential transfer impedance matrix, and the return current shared by each return current sink is solved by the improved node voltage method.

[0012] Based on the injected current set, the return current, and the network node potential transfer impedance matrix, calculate the node potential of each node in the graph network topology model, and calculate the edge signal strength of each edge based on the node potential difference and edge weight.

[0013] The average wander betweenness of each edge is obtained by statistically analyzing the edge signal strengths corresponding to all injected current sets.

[0014] For a target node within the target interval of the AC power grid, a local reference threshold is determined based on the average betweenness of the target associated edges connected to the target node, and the target associated edges with an average betweenness greater than the local reference threshold are identified as critical paths for stray currents from the subway to invade the urban power grid.

[0015] Furthermore, when establishing the graph network topology model, the subway contact network, rail stations, drainage network, soil medium, pipelines, and transformer neutral grounding locations are mapped as nodes, branches with electrical connections are mapped as edges, and the equivalent admittance of the branches is used as the edge weight of the corresponding edges.

[0016] Furthermore, when obtaining the full network node potential transfer impedance matrix, the full network node admittance matrix is ​​first constructed based on the edge weights in the graph network topology model. Then, a reference node is selected and the full network node admittance matrix is ​​reduced in order. Subsequently, the reduced admittance matrix is ​​inverted and filled in at the corresponding position of the reference node to obtain the full network node potential transfer impedance matrix.

[0017] Furthermore, when forming multiple injection current sets, first count the number of nodes in the injection source node set and the number of injection sources when multiple sources are concurrent. Then, use a combination method without replacement to exhaustively enumerate the distribution of different injection source locations, and determine the positive or negative sign of the corresponding injection current based on the physical direction of the current injection or extraction network at each injection source node.

[0018] Furthermore, when solving for the return current shared by each return current sink node, the impedance relationship corresponding to the set of return current sink nodes is first extracted from the impedance matrix of the entire network node potential transfer. Then, the equivalent open circuit potential at the set of return current sink nodes is calculated based on the injected current set. Subsequently, the return current of each return current sink node is solved jointly through the common equivalent potential constraint and the return current conservation constraint.

[0019] Furthermore, when calculating the node potential of each node, the equivalent open-circuit potential generated by the injected current set in the graph network topology model is superimposed with the potential influence generated by the return current of each return sink node in the graph network topology model to obtain the node potential distribution under the corresponding injected current set; when calculating the edge signal strength of each edge, for any two interconnected nodes in the graph network topology model, the edge signal strength of the corresponding edge is obtained according to the node potential difference between the two nodes and the edge weight between the two nodes, and the effective edge signal strength belonging to the target interval of the AC power grid side is extracted from it.

[0020] Furthermore, when calculating the average betweenness of each edge, the absolute values ​​of the edge signal strength of the same edge under the entire set of injected currents are accumulated and averaged; when determining the local reference threshold, the average betweenness of the target associated edges of the same target node within the target interval of the AC grid side is averaged, and this average value is used as the local reference threshold corresponding to the target node; when outputting the critical path, the same target associated edge repeatedly identified by different target nodes is deduplicated, and the critical path sorting results are output in descending order of average betweenness.

[0021] A device for identifying critical paths of stray current intrusion into urban power grids from subways, comprising:

[0022] The model building module is used to establish a graph network topology model of spatial coupling between the subway and the urban power grid, and to obtain the potential transfer impedance matrix of all network nodes.

[0023] The injection source combination module is used to generate multiple injection current sets under multi-source concurrent conditions based on the possible operating positions of the train;

[0024] The return current calculation module is used to solve the return current shared by each return sink node based on the potential transfer impedance matrix of all network nodes and the set of injected currents, using an improved node voltage method.

[0025] The edge signal calculation module is used to calculate the edge signal strength of each edge based on the injected current set, the return current, and the potential transfer impedance matrix of all network nodes.

[0026] The critical path identification module is used to identify and output the critical path of stray current from the subway into the urban power grid based on the average betweenness of each side and the local reference threshold corresponding to the target node in the target interval of the AC power grid.

[0027] An electronic device is characterized by comprising a processor and a memory, wherein the memory stores a computer program, and when the computer program is executed by the processor, it implements the above-described method for identifying critical paths of stray current intrusion into urban power grids in subways.

[0028] A computer-readable storage medium storing a computer program, which, when executed by a processor, implements the aforementioned method for identifying critical paths of stray current intrusion into urban power grids in subways.

[0029] The present invention has the following advantages:

[0030] First, this invention maps the subway traction power supply system, rail return system, drainage network, soil medium, underground pipelines, and urban power grid grounding system into a unified graph network topology model, which can comprehensively characterize the direct electrical connection and indirect conduction relationship between the subway system and the urban power grid, avoiding the one-sidedness of identification caused by analyzing stray current intrusion problems from only a single system or local node.

[0031] Secondly, this invention characterizes the potential influence relationship between different nodes through the potential transfer impedance matrix of the entire network nodes, which can quickly calculate the node potential and edge signal strength under various injection source location distributions, thus improving the computational efficiency of stray current conduction analysis in complex networks.

[0032] Third, this invention considers the concurrent operation of trains with multiple sources. By exhaustively listing the distribution of injection source locations, multiple sets of injection currents are constructed, so that the critical path identification results can reflect the comprehensive impact under different combinations of train operating positions, rather than being limited to a single operating condition.

[0033] Fourth, this invention solves the return current shared by multiple return current junctions by improving the node voltage method, which is more consistent with the actual physical process of multiple return current junctions working together in the rail return current network and improves the rationality of the return current distribution calculation.

[0034] Fifth, this invention introduces the average wandering betweenness number to characterize the degree of participation of each edge in stray current conduction under all injection conditions, and combines it with the local benchmark threshold of the target node in the target interval of the AC power grid to identify the critical path. This can highlight the conduction edge that is more sensitive to the risk of intrusion into the urban power grid, and improve the pertinence and stability of critical path screening.

[0035] Sixth, the critical path sequencing results output by this invention can provide a basis for stray current protection in subways, drainage measures, grounding risk assessment of urban power grids, and selection of key monitoring points, and have good engineering application value. Attached Figure Description

[0036] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0037] Figure 1 This is a flowchart of the method for identifying critical paths of stray current intrusion into the urban power grid in the present invention.

[0038] Figure 2 The graph network topology model of spatial coupling between the subway and the urban power grid is established for step 1. Detailed Implementation

[0039] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0040] Example 1:

[0041] This embodiment provides a method for identifying critical paths of stray current intrusion into urban power grids from subway systems. Based on a graph network topology model of spatial coupling between the subway and urban power grid, this method simulates the distribution patterns of stray currents under multi-source concurrent operating conditions to achieve quantitative identification of critical paths of intrusion into the AC power grid. The specific steps of this embodiment include:

[0042] Step 1:

[0043] S101: Based on the actual physical structure of the subway and power grid, each station, drainage network, soil, and transformer neutral point of the subway catenary, rails, is mapped as nodes in a graph network model. Lines with electrical connections are mapped as edges in the graph network model. The equivalent branch admittance parameters between each node are extracted as weights on the edges in the graph network model. Finally, the total number of network nodes is counted. .

[0044] S102: Obtain all possible running position nodes of the train on the rails, defined as the injection source node set. Define the rail return node corresponding to the negative pole of the traction substation as the set of sink nodes. Simultaneously, the neutral grounding point of the transformer and its closely associated soil and pipeline nodes are selected to form the target section on the AC power grid side. ;

[0045] S103: Construction based on equivalent branch admittance parameters The global node admittance matrix is ​​calculated using a reference node. By removing the row and column corresponding to the reference node, and performing matrix reduction, a full-rank reduced admittance matrix is ​​obtained. ;

[0046] S104: For the reduced-order admittance matrix The reduced-order impedance matrix is ​​obtained by performing an inversion operation, and zero elements are padded to the rows and columns corresponding to the original reference nodes to obtain the desired impedance. 3D network node potential transfer impedance matrix .

[0047] Step 2:

[0048] S201: Based on the graph network topology model obtained in step 1, statistically analyze the set of injection source nodes. The total number of nodes included is And at this time, the number of injection source nodes in multi-source concurrent operation is A combination algorithm without replacement is used to exhaustively enumerate all injection source location distributions for multi-source concurrency, forming a set of injection source location distributions. Based on the principles of probability theory and combinatorics, the formula for calculating the total number of combinations contained in the injection source location distribution set is as follows: The total number of combinations of injection source location distribution sets is calculated. Among them, This represents the number of combinations, that is, the total number of all possible choices. Indicates the set of injection source nodes The total number of nodes included, that is, the number of all possible positions where the train can stop on the rails. This represents the number of injection source nodes that exist simultaneously under multi-source concurrency conditions, i.e., the number of trains running concurrently. The exclamation mark (!) represents a factorial operation, for example... .

[0049] S202: Based on the graph network topology model obtained in step 1, statistically analyze the set of backflow sink nodes. Total of One return flow sink node; impedance matrix for potential transfer from all network nodes. Extract the set corresponding to the backflow sink node. of 3D impedance matrix .

[0050] S203: For each injection source location distribution in the injection source location distribution set, extract the corresponding... There are several injection source nodes; based on the physical direction of the current injection or extraction network at each node, the sign of the injected current at the corresponding node is determined, thus forming the set of injected currents corresponding to that location; the set of injected currents is used as the signal input after normalization of the graph network topology model, based on the superposition theorem of linear circuits, The normalized signal of each injection source node is in The equivalent open-circuit potential column vector generated at each return junction. Numerically equivalent to the total network transfer impedance matrix The formula for directly summing the corresponding column vectors is:

[0051]

[0052] in, The impedance matrix for potential transfer of all nodes in the network The middle column corresponds to the first injection source nodes , corresponding backflow sink set Local extraction of column vectors. For the first injection source nodes The injected current; For the backflow sink node set The equivalent open-circuit potential column vector at the location.

[0053] S204: Establish the improved node voltage method matrix equation, and jointly solve the return current column vector shared by each return current sink node under the normalized signal input. This process iterates through all injected current sets, as shown in the following equation:

[0054]

[0055] Among them, 1 is A column vector of all 1s; Let be the common equivalent potential of the return current sink set. In the improved node voltage method matrix equations, the first row of equations ensures that the actual potentials of each return current sink in the physically connected return current network remain equal, and the second row of equations ensures that the sum of the return currents of all return current sinks is strictly equal to . The algebraic sum of the signal inputs from the injection source nodes.

[0056] Step 3:

[0057] S301: Based on each normalized signal obtained in step 2, input the corresponding return current column vector of the return current sink node. Combined with the potential transfer impedance matrix of all network nodes Calculate the nodal potential column vector of each node. , The actual potential of each node in the network is numerically equal to the difference between the equivalent open-circuit potential generated by the normalized signal input in the graph network topology model and the potential drop generated by the corresponding return current of each return sink node in the entire network. Its calculation formula is as follows:

[0058]

[0059] in, The impedance matrix for potential transfer of all nodes in the network The middle column corresponds to the first injection source nodes The entire network column vector; For the first injection source nodes The injected current; The impedance matrix for potential transfer of all nodes in the network The middle column corresponds to the backflow sink set. of Dimensional submatrix. The set of backflow sink nodes obtained in step 2 The corresponding return current column vector;

[0060] S302: Combining the weight parameters of each edge in the graph network topology model with the node potential column vectors of each node. Calculate the signal distribution of each edge in the model under the injected current set; construct the edge signal distribution matrix. The product of the actual potential difference between any two connected nodes in a graph network topology model and the weight of the edge connecting these two nodes is defined as the signal distribution intensity of the corresponding edge, and its matrix elements are... The calculation formula is:

[0061]

[0062] in, For nodes and nodes Signal distribution intensity between; For nodes and nodes The weight of the edges connecting them; and These are the nodal potential column vectors. Middle node and nodes The potential.

[0063] S303: Extract the side signal distribution matrix The signal distribution data of all valid edges within the target interval are recorded; then, the edge signal distribution data of the injected current set are recorded independently until the complete traversal of all injected current sets of the model is completed.

[0064] Step 4:

[0065] S401: For any node in the entire network and nodes For each valid connection between them, the absolute values ​​of the edge signal distribution intensity under each injected current set are summed and then divided by the total number of combinations. This is used to calculate the average traversal betweenness of all connected edges in the entire network. The calculation formula is:

[0066]

[0067] in, For extracted nodes and nodes The edge between them is in the 1st The intensity of the edge signal distribution under the condition of injection source location distribution.

[0068] S402: For the AC power grid side target interval defined in step 1 Each node contained within Extract the nodes corresponding to the target node. The total number of connected valid edges is denoted as . This constitutes the target-related edge set corresponding to the target node. ; For the target associated edge set Average walking betweenness of each side Calculate the arithmetic mean and use this mean as the target node. Corresponding local reference threshold The calculation formula is:

[0069]

[0070] in, For the target section of the AC power grid The first in Target nodes, To reach the target node Connected target associated edge set, For the target associated edge set The number of valid edges contained therein;

[0071] S403: Traverse the target interval on the AC power grid side Each target node within and for each target node Corresponding target associated edge set The process involves evaluating each edge in the graph and calculating the average betweenness of each edge. Compared with the baseline threshold Perform size comparison; identify the edges whose average walk betweenness number is strictly greater than the baseline threshold as critical paths; retain only one edge for repeated identification; finally, sort all the identified critical paths in descending order according to the size of the average walk betweenness number and output the critical paths.

[0072] For example, the "spatial coupling" described in this invention refers to the electrical influence relationship formed between the subway traction power supply system, the rail return system, the drainage network, the soil medium, underground pipelines, and the urban power grid grounding system due to physical proximity, electrical connection, grounding association, or conduction through conductive media. This spatial coupling relationship is not limited to direct metallic connection, but also includes indirect electrical associations formed through the soil medium, grounding electrode, or pipelines.

[0073] For example, the "metro-urban power grid spatial coupling graph network topology model" described in this invention refers to a graph network model that abstracts key electrical locations in the metro traction power supply system, rail return system, drainage network, soil medium, pipelines, and urban power grid grounding system as nodes, abstracts the electrical connection relationships or equivalent conduction relationships between different nodes as edges, and uses the equivalent admittance of the corresponding branch as the edge weight.

[0074] For example, the "injection source node" in this invention refers to a node that may inject stray current into the rail or related return network during train operation; the "return sink node" refers to the rail return node corresponding to the negative pole of the traction substation, or an equivalent return node capable of bearing the return current; the "AC grid side target section" refers to the area in the urban power grid where the risk of stray current intrusion needs to be assessed, including the transformer neutral point grounding location and its associated soil nodes, pipeline nodes, and grounding nodes.

[0075] For example, the "whole network node potential transfer impedance matrix" described in this invention refers to a matrix used to characterize the impact of a unit current injection at any node in a graph network topology model on the potentials of all nodes in the network. Using this matrix, the potentials of each node and the edge signal strengths can be quickly calculated under different injection source location distributions.

[0076] For example, the "edge signal strength" mentioned in this invention refers to the current conduction strength between adjacent nodes in a graph network topology model under a specific set of injected currents, which is determined by the node potential difference and edge weights, and is used to characterize the strength of stray current propagation along the corresponding edge.

[0077] For example, the "average wandering betweenness number" mentioned in this invention refers to the average statistical result of the absolute value of the edge signal strength under all injection source location distribution conditions. It is used to characterize the comprehensive degree to which the edge participates in stray current conduction under various train operating conditions. The larger the average wandering betweenness number, the more likely the edge is to become the main conduction channel for stray current intrusion into the urban power grid under different injection conditions.

[0078] For example, the "local benchmark threshold" described in this invention refers to a comparison benchmark determined based on the average betweenness of each target associated edge connected to a target node within a target interval on the AC power grid side. By comparing the average betweenness of the target associated edges with the corresponding local benchmark threshold, critical paths with higher conduction contributions relative to the same target node can be identified.

[0079] Example 2:

[0080] Taking a certain subway line as an example, this line includes 5 stations, 2 traction substations, and 2 trains operating simultaneously. Based on the actual physical connection between this line and the city's power grid, establish a system as follows: Figure 2 The diagram shows a network topology model of the spatial coupling between the subway and the urban power grid. In the model, node 2 represents the contact wire node, nodes 3, 4, 5, 6, and 7 represent rail layer nodes, nodes 8, 9, 10, 11, and 12 represent through ground wire nodes, nodes 13, 14, 15, 16, 17, and 18 represent target section nodes on the AC power grid side, and reference nodes are used to represent grounding or earth reference points. In this embodiment, the distribution of 10 injection source locations under the condition of simultaneous operation of two trains is traversed and calculated to obtain the edge signal distribution matrix corresponding to each injection current set. In this embodiment, the distribution of 10 injection source locations under the condition of simultaneous operation of two trains is traversed and calculated according to the above method to obtain the average betweenness of each target associated edge, and further judged based on the local reference threshold corresponding to each target node in the target section, and the critical path identification results are shown in Table 1.

[0081] Table 1. Critical path identification results for a certain subway line.

[0082] Sort Target associated edge set Average walking betweenness Local reference threshold 1 (12,15) 2.19E-09 1.46E-09 2 (9.14) 2.03E-09 1.36E-09 3 (8,13) 1.72E-09 1.14E-09

[0083] Although preferred embodiments of this application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of this application.

[0084] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include such modifications and variations.

Claims

1. A method for identifying critical paths of stray current intrusion into urban power grids from subways, characterized in that, Includes the following steps: Based on the electrical connection relationships between the metro traction power supply system, rail return system, drainage network, soil medium and urban power grid grounding system, a graph network topology model of metro-urban power grid spatial coupling is established, and the potential transfer impedance matrix of all network nodes is obtained based on the graph network topology model. The injection source node set is determined based on the possible operating location of the train, the return sink node set is determined based on the rail return location corresponding to the negative pole of the traction substation, and the target section on the AC power grid side is determined based on the grounding correlation location of the urban power grid. Under multi-source concurrent conditions, the distribution of injection source locations in the injection source node set is exhaustively enumerated to form multiple injection current sets; For each set of injected currents, the equivalent open-circuit potential at the return current sink is calculated based on the network node potential transfer impedance matrix, and the return current shared by each return current sink is solved by the improved node voltage method. Based on the injected current set, the return current, and the network node potential transfer impedance matrix, calculate the node potential of each node in the graph network topology model, and calculate the edge signal strength of each edge based on the node potential difference and edge weight. The average wander betweenness of each edge is obtained by statistically analyzing the edge signal strengths corresponding to all injected current sets. For a target node within the target interval of the AC power grid, a local reference threshold is determined based on the average betweenness of the target associated edges connected to the target node, and the target associated edges with an average betweenness greater than the local reference threshold are identified as critical paths for stray currents from the subway to invade the urban power grid.

2. The method for identifying critical paths of stray current intrusion into urban power grids from subways according to claim 1, characterized in that, When establishing the graph network topology model, the subway contact network, rail stations, drainage network, soil medium, pipelines, and transformer neutral grounding locations are mapped as nodes, branches with electrical connections are mapped as edges, and the equivalent admittance of the branches is used as the edge weight of the corresponding edges.

3. The method for identifying critical paths of stray current intrusion into urban power grids from subways according to claim 1, characterized in that, To obtain the full network node potential transfer impedance matrix, first construct the full network node admittance matrix based on the edge weights in the graph network topology model, then select a reference node and reduce the order of the full network node admittance matrix, then invert the reduced admittance matrix and fill in the corresponding positions at the reference node to obtain the full network node potential transfer impedance matrix.

4. The method for identifying critical paths of stray current intrusion into urban power grids from subways according to claim 1, characterized in that, When forming multiple injection current sets, first count the number of nodes in the injection source node set and the number of injection sources when multiple sources are concurrent. Then, use the no-replacement combination method to exhaustively enumerate the distribution of different injection source locations, and determine the positive or negative of the corresponding injection current based on the physical direction of the current injection or extraction network at each injection source node.

5. The method for identifying critical paths of stray current intrusion into urban power grids from subways according to claim 1, characterized in that, When solving for the return current shared by each return current sink node, first extract the impedance relationship corresponding to the set of return current sink nodes from the impedance matrix of the entire network node potential transfer, then calculate the equivalent open circuit potential at the set of return current sink nodes based on the injected current set, and then solve for the return current of each return current sink node by means of common equivalent potential constraints and return current conservation constraints.

6. The method for identifying critical paths of stray current intrusion into urban power grids from subways according to claim 1, characterized in that, When calculating the node potential of each node, the equivalent open-circuit potential generated by the injected current set in the graph network topology model is superimposed with the potential influence generated by the return current of each return sink node in the graph network topology model to obtain the node potential distribution under the corresponding injected current set. When calculating the edge signal strength of each edge, for any two interconnected nodes in the graph network topology model, the edge signal strength of the corresponding edge is obtained according to the node potential difference between the two nodes and the edge weight between the two nodes, and the effective edge signal strength belonging to the target interval of the AC power grid side is extracted from it.

7. The method for identifying critical paths of stray current intrusion into urban power grids from subways according to claim 1, characterized in that, When calculating the average wander betweenness of each edge, the absolute values ​​of the edge signal strength of the same edge under the set of all injected currents are accumulated and averaged. When determining the local benchmark threshold, the average walk betweenness of the target associated edges of the same target node within the target interval of the AC power grid is averaged, and this average value is used as the local benchmark threshold corresponding to the target node. When outputting the critical path, the same target associated edges repeatedly identified by different target nodes are deduplicated, and the critical path sorting results are output in descending order of average walk betweenness.

8. A device for identifying critical paths of stray current intrusion into urban power grids from subways, characterized in that, include: The model building module is used to establish a graph network topology model of spatial coupling between the subway and the urban power grid, and to obtain the potential transfer impedance matrix of all network nodes. The injection source combination module is used to generate multiple injection current sets under multi-source concurrent conditions based on the possible operating positions of the train; The return current calculation module is used to solve the return current shared by each return sink node based on the potential transfer impedance matrix of all network nodes and the set of injected currents, using an improved node voltage method. The edge signal calculation module is used to calculate the edge signal strength of each edge based on the injected current set, the return current, and the potential transfer impedance matrix of all network nodes. The critical path identification module is used to identify and output the critical path of stray current from the subway into the urban power grid based on the average betweenness of each side and the local reference threshold corresponding to the target node in the target interval of the AC power grid.

9. An electronic device, characterized in that, It includes a processor and a memory, wherein the memory stores a computer program, and when the computer program is executed by the processor, it implements the critical path identification method for stray current intrusion into urban power grids as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the critical path identification method for stray current intrusion into urban power grids as described in any one of claims 1 to 7.

Images