Cable fault detection method based on distributed nodes

Abstract

The invention discloses a cable fault detection method based on distributed nodes, relates to the field of power faults, solves the problem that the existing cable fault detection method is poor in detection effect, and comprises the following steps: S1, obtaining a target area topological graph, and carrying out modularity gain analysis on the distributed nodes in the target area topological graph to obtain a target area topological graph; s2, according to the distributed node subnet division data, carrying out node connection analysis on the node topology subnets according to the distributed node subnet division data, and carrying out node connection analysis on the node topology subnets according to the distributed node subnet division data, s3, according to the node subnet division data, respectively carrying out cable fault detection on the first type of topology subnet and the second type of topology subnet, and carrying out cable fault detection on the second type of topology subnet and the first type of topology subnet. The accuracy and pertinence of the cable fault detection method are improved.

Description

TECHNICAL FIELD

[0001] The application belongs to the field of power failure, and relates to technology, in particular to a cable failure detection method based on distributed nodes. BACKGROUND

[0002] The existing cable failure detection method has the following defects when detecting cable failure of distributed nodes: 1. When the existing cable failure detection method divides the topology subnets of the distributed nodes in the topology graph through the node gain degree traversal method, the number of nodes and the number of connection edges of the distributed nodes in the topology graph are analyzed, and the node load, cable impedance and cable length are not included in the composition parameters of the node gain degree, so that the node gain calculation process lacks comprehensiveness, which easily leads to the lack of rationality of the node topology subnet division result, and further affects the cable failure detection result. 2. When the existing cable failure detection method detects cable failure, a fixed cable failure detection method is usually adopted, the effective neighbor nodes and the invalid neighbor nodes in the node topology subnet are not marked respectively, the cable connection density of the effective neighbor nodes is analyzed while the cable connection distance of the invalid neighbor nodes is analyzed, the node topology subnet cannot be classified according to the analysis result, and the targeted cable failure detection method cannot be matched, so that the cable failure detection efficiency is easily low.

[0003] Therefore, the cable failure detection method based on distributed nodes is proposed. SUMMARY

[0004] In view of the defects of the prior art, the application aims to provide a cable failure detection method based on distributed nodes, and aims to improve the pertinence and detection efficiency of the cable failure detection method.

[0005] In order to achieve the above-mentioned purpose, the application adopts the following technical scheme: the cable failure detection method based on distributed nodes comprises the following steps: Step S1: acquiring a target area topology graph, and performing module gain analysis on the distributed nodes in the target area topology graph, and according to the analysis result, the distributed nodes are divided into a plurality of node topology subnets, and distributed node subnet division data is obtained; Step S2: according to the distributed node subnet division data, performing node connection analysis on each node topology subnet, and according to the analysis result, the node topology subnet is divided into a first type node subnet and a second type node subnet, and node subnet division data is obtained; Step S3: according to the node subnet division data, performing cable failure detection on the first type topology subnet and the second type topology subnet respectively.

[0006] Furthermore, step S1 also includes the following steps: Step S11: Obtain the distributed nodes that need to be detected for cable faults, obtain multiple distributed nodes, and obtain the node topology map corresponding to the multiple distributed nodes to obtain the target area topology map. Step S12: Mark the distributed nodes existing in the topology graph of the target area to obtain distributed nodes from v1 to vn. Step S13: Mark any two distributed node connection cables from distributed node V1 to distributed node Vn to obtain the connection cable from e1 to em. Step S14: Obtain the circuit impedance from the e1 connecting cable to the em connecting cable respectively, and obtain the circuit impedance values ​​from e1 to em. Obtain the cable connection length values ​​from the e1 connecting cable to the em connecting cable respectively, and obtain the circuit length values ​​from e1 to em. Step S15: Select distributed nodes Vi and Vj from distributed nodes V1 to Vn, and select the topology subnet nodes to which distributed nodes Vi and Vj belong to obtain the sample topology subnet. Step S16: Randomly select two distributed nodes that are not in the sample topology subnet in the target area topology map to obtain the first sample distributed node and the second sample distributed node. Perform Vnew distributed node traversal on the area topology subnet formed by the first sample distributed node and the second sample distributed node. Use Vnew distributed nodes to complete the traversal of each distributed node to obtain the node topology subnet corresponding to the first sample distributed node and the second sample distributed node. Repeat the above process to divide the target area topology map into multiple node topology subnets and obtain distributed node subnet partitioning data.

[0007] Furthermore, step S15 also includes the following steps: Connect the e1 connecting cable to the em connecting cable and the V1 distributed node to the Vn distributed node. If the ei connecting cable is the connecting cable between the Vi distributed node and the Vj distributed node, set the product of the inverse of the ei circuit impedance value and the e1 circuit length value as the cable connection weight between the Vi distributed node and the Vj distributed node to obtain the Wi,j cable connection weight. The node load intensity values ​​from distributed node v1 to distributed node vn at the current moment are obtained to obtain the real-time load from node V1 to node Vn. Treat the Vi distributed nodes and Vj distributed nodes as an independent topological sub-region, and obtain the weighted modularity corresponding to the independent topological sub-region; The weighted modularity is calculated using the following formula: ; Where Qzy is the weighted modularity, m is the number of cables connecting the distributed nodes, Wi,j is the cable connection weight between distributed nodes Vi and Vj, si is the real-time load of node Vi, and sj is the real-time load of node Vj.

[0008] Furthermore, step S15 also includes the following steps: Add a new distributed node Vnew to the independent topological sub-region formed by the distributed nodes Vi and Vj, and calculate the updated weighted modularity after adding the new distributed node Vnew. The updated weighted modularity is calculated using the following formula: ; Among them, Qzy ’ To update the weighted modularity, p ’ The value represents the number of cables connecting the distributed nodes after the addition of the Vnew distributed node, Wnew,j represents the cable connection weight between the Vnew distributed node and the Vj distributed node, snew represents the real-time load of the Vnew node, and sj represents the real-time load of the Vj node. Calculate the difference between the updated weighted modularity and the weighted modularity, and take the absolute value of the difference to obtain the modularity gain after adding the Vnew distributed node. If the modularity gain is greater than 0, the Vnew distributed node is retained; if the modularity gain is less than or equal to 0, the Vnew distributed node is deleted. Repeat the above operation to complete the traversal of each distributed node using the Vnew distributed node to obtain the sample topology subnet.

[0009] Furthermore, step S2 also includes the following steps: Step S21: Obtain distributed node subnet partitioning data, obtain multiple node topology subnets based on the distributed node subnet partitioning data, and arbitrarily select a sample topology subnet from the multiple obtained node topology subnets; Step S22: Perform connection clustering analysis on the distributed nodes in the sample topology subnet, and obtain the topology type division index value corresponding to the sample topology subnet based on the analysis results; Step S23: Obtain the topology type division index value corresponding to each node's topology subnet; Step S24: Obtain the preset range of topology type division index. If the topology type division index value is within the preset range, the corresponding node topology subnet is divided into the first type of topology subnet. If the topology type division index value is not within the preset range, the corresponding node topology subnet is divided into the second type of topology subnet, thus obtaining the node subnet division data.

[0010] Furthermore, step S22 also includes the following steps: Step S221: Obtain the node topology graph corresponding to the sample topology subnet, obtain the sample subnet topology graph, mark each distributed node in the sample subnet topology graph, and arbitrarily select a sample distributed node from the multiple marked distributed nodes. Step S222: Perform neighbor node connectivity analysis on the sample distributed nodes in the sample subnet topology graph to obtain the connectivity density ratio corresponding to the sample distributed nodes; Step S223: Obtain the connectivity density ratio corresponding to each distributed node, and calculate the average value of the multiple connectivity density ratios to obtain the topology type division index value corresponding to the sample topology subnet.

[0011] Furthermore, step S222 also includes the following steps: Step S2221: Set the distributed nodes in the sample subnet topology graph other than the sample distributed nodes as candidate neighbor nodes, and mark the obtained candidate neighbor nodes as candidate nodes D1 to Dc respectively. Step S2222: In the sample subnet topology diagram, create a connection between the sample distributed node and the candidate node D1, obtaining an auxiliary connection for node D1. If the auxiliary connection for node D1 does not pass through any candidate neighbor node, then the candidate node D1 is set as a valid neighbor node. If the auxiliary connection for node D1 passes through any candidate neighbor node, then the candidate node D1 is set as an invalid neighbor node. Create a connection between the sample distributed node and the candidate node D2, obtaining an auxiliary connection for node D2. If the auxiliary connection for node D2 does not pass through any candidate neighbor node... If a neighbor node is found, then the candidate node D2 is set as a valid neighbor node. If the auxiliary connection of node D2 passes through any candidate neighbor node, then the candidate node D2 is set as an invalid neighbor node. And so on, creating connections between sample distributed nodes and candidate nodes Dc to obtain auxiliary connections for node Dc. If the auxiliary connection of node Dc does not pass through any candidate neighbor node, then the candidate node Dc is set as a valid neighbor node. If the auxiliary connection of node Dc passes through any candidate neighbor node, then the candidate node Dc is set as an invalid neighbor node. Step S2223: Perform effective neighbor node analysis on the sample distributed nodes, and obtain the neighbor node connection density based on the analysis results; Step S2224: Perform cable connection distance analysis on the sample distributed nodes for invalid neighbor nodes, and obtain the direct connection deviation of invalid nodes based on the analysis results; Step S2225: The ratio of the effective neighbor node connection density to the ineffective node direct connection deviation of the sample distributed node is used to obtain the connectivity density ratio of the sample distributed node.

[0012] Furthermore, step S2223 also includes the following steps: The actual number of cable connection edges is obtained by counting the number of cable lines that directly connect the sample distributed nodes and their effective neighbor nodes via physical cables. The number of effective neighbor nodes is obtained, and the effective neighbor node connection density corresponding to the sample distributed node is calculated by combining the actual number of cable connection edges and the number of effective neighbor nodes. The connection density of the effective neighbor nodes corresponding to the sample distributed nodes is calculated using the following formula: ; Where Lmy is the effective neighbor node connection density corresponding to the sample distributed node, Sls is the actual number of cable connection edges, and Jls is the number of effective neighbor nodes.

[0013] Furthermore, step S2224 also includes the following steps: In the sample subnet topology diagram, invalid neighbor nodes relative to the sample distributed nodes are obtained, and the obtained invalid neighbor nodes are numbered from W1 invalid neighbor node to Wd invalid neighbor node respectively. The cable connection paths between invalid neighbor nodes of W1 and sample distributed nodes in the sample subnet topology are obtained, resulting in multiple connection paths of W1 nodes. The path length values ​​of the multiple connection paths of W1 nodes are compared, and the connection path of W1 nodes with the smallest length value is set as the shortest W1 cable path. The length value of the shortest W1 cable path is obtained to get the shortest length value of W1 path. The straight-line distance between the invalid neighbor node of W1 and the sample distributed node in the sample subnet topology is obtained to obtain the direct cable connection distance of W1. The number of distributed nodes covered by the shortest path of cable W1 is obtained to get the number of nodes covered by the W1 path. Calculate the difference between the shortest path length of W1 and the direct connection distance of W1 cable, and calculate the ratio of the difference to the direct connection distance of W1 cable to obtain the deviation of the shortest path distance of W1. The shortest path distance deviation of W1 and the number of path nodes covered by W1 are used to calculate the direct connection deviation of W1. The direct connection deviation of W1 is calculated using the following formula: ; Where Zlw1 is the direct connection deviation of W1, Jfw1 is the number of path nodes covered by W1, and Lpw1 is the shortest path distance deviation of W1. The direct connection deviation from the invalid neighbor node W2 to the invalid neighbor node Wd is obtained respectively, thus obtaining the direct connection deviation from W2 to Wd. The average of the direct connection deviations from W1 to Wd is calculated to obtain the direct connection deviation of the invalid nodes corresponding to the sample distributed nodes.

[0014] Furthermore, step S3 also includes the following steps: Obtain node subnet partitioning data, and then obtain the first type of topology subnet and the second type of topology subnet respectively based on the node subnet partitioning data; Distributed fiber optic temperature measurement technology is used to detect cable faults in the first type of topology subnet, and cable fault warnings are given based on the detection results. Traveling wave detection technology is used to detect cable faults in the second type of topology subnet, and cable fault warnings are given based on the detection results.

[0015] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are: 1. When dividing the distributed nodes in the topology graph into subnets by traversing the node gain degree, this invention incorporates node load, cable impedance, and cable length into the composition parameters of node gain degree for comprehensive analysis, thereby improving the comprehensiveness of the node gain calculation process, further ensuring the rationality of the node topology subnet division results, and thus improving the accuracy of cable fault detection results. 2. This invention marks distributed nodes in a node topology subnet as valid neighbor nodes and invalid neighbor nodes respectively. While performing cable connection density analysis on valid neighbor nodes, it also performs cable connection distance analysis on invalid neighbor nodes. Based on the analysis results, the node topology subnet is classified into types, and a targeted cable fault detection method is matched for it, thereby further improving the efficiency of cable fault detection. Attached Figure Description

[0016] To facilitate understanding by those skilled in the art, the present invention will be further described below with reference to the accompanying drawings.

[0017] Figure 1 This is an overall system block diagram of the present invention; Figure 2 This is a schematic diagram of the neighbor nodes of the present invention; Figure 3This is a schematic diagram of the node connection path of the present invention. Detailed Implementation

[0018] The technical solution of the present invention will be clearly and completely described below with reference to the embodiments. 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 of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0019] Example Please see Figure 1 This invention provides a technical solution: a cable fault detection method based on distributed nodes, comprising the following steps: Step S1: Obtain the topology map of the target area, and perform modularity gain analysis on the distributed nodes in the topology map of the target area. Based on the analysis results, split the distributed nodes into multiple node topology subnets to obtain distributed node subnet partitioning data. Step S1 further includes the following steps: The distributed nodes that need to be detected for cable faults are acquired, resulting in multiple distributed nodes. The node topology map corresponding to these multiple distributed nodes is then acquired to obtain the target area topology map. The topology of the target area is partially split to obtain multiple node topology subnets; Specifically as follows: Mark the distributed nodes that exist in the topology of the target region to obtain distributed nodes from v1 to vn. It should be noted here that: In this application, v1, v2, v3...vn in the distributed nodes from v1 to vn are the numbers corresponding to the distributed nodes, and n is the quantity value corresponding to the distributed nodes, and n is an integer greater than 0.

[0020] Mark any two distributed node connection cables from distributed node V1 to distributed node Vn to obtain the connection cable from e1 to em. It should be noted here that: In this application, e1, e2, e3...em in the e1 connecting cable to the em connecting cable are the numbers corresponding to the distributed node connecting cables, and m is the quantity value corresponding to the distributed node connecting cables, and m is an integer greater than 0.

[0021] The circuit impedances from the E1 connecting cable to the EM connecting cable are obtained separately, and the cable lengths from the E1 connecting cable to the EM connecting cable are obtained separately, and the circuit lengths from the E1 connecting cable to the EM connecting cable are obtained separately. Select distributed nodes Vi and Vj from distributed nodes V1 to Vn, and select the topology subnet nodes to which distributed nodes Vi and Vj belong to obtain the sample topology subnet. Connect the e1 connecting cable to the em connecting cable and the V1 distributed node to the Vn distributed node. If the ei connecting cable is the connecting cable between the Vi distributed node and the Vj distributed node, set the product of the inverse of the ei circuit impedance value and the e1 circuit length value as the cable connection weight between the Vi distributed node and the Vj distributed node to obtain the Wi,j cable connection weight. It should be noted here that: In this application, the ei connecting cable mentioned above can be any one of the e1 connecting cable to the em connecting cable, and the Vi distributed node and Vj distributed node mentioned above can be any two distributed nodes with a cable connection relationship among the V1 distributed node to the Vn distributed node. Furthermore, the cable connection weight Wi,j represents the cable connection weight between distributed nodes Vi and Vj, and can be represented by the symbol Wi,j = -impedanceij × lengthij.

[0022] The node load intensity values ​​from distributed node v1 to distributed node vn at the current moment are obtained to obtain the real-time load from node V1 to node Vn. Treat the Vi distributed nodes and Vj distributed nodes as an independent topological sub-region, and obtain the weighted modularity corresponding to the independent topological sub-region; The weighted modularity is calculated using the following formula: ; Where Qzy is the weighted modularity, m is the number of cables connecting the distributed nodes, Wi,j is the cable connection weight between distributed nodes Vi and Vj, si is the real-time load of node Vi, and sj is the real-time load of node Vj. Add a new distributed node Vnew to the independent topological sub-region formed by the distributed nodes Vi and Vj, and calculate the updated weighted modularity after adding the new distributed node Vnew. The updated weighted modularity is calculated using the following formula: ; Among them, Qzy’ To update the weighted modularity, p ’ The value represents the number of cables connecting the distributed nodes after the addition of the Vnew distributed node, Wnew,j represents the cable connection weight between the Vnew distributed node and the Vj distributed node, snew represents the real-time load of the Vnew node, and sj represents the real-time load of the Vj node. Calculate the difference between the updated weighted modularity and the weighted modularity, and take the absolute value of the difference to obtain the modularity gain after adding the Vnew distributed node. If the modularity gain is greater than 0, the Vnew distributed node is retained; if the modularity gain is less than or equal to 0, the Vnew distributed node is deleted. Repeat the above operation to complete the traversal of each distributed node using the Vnew distributed node to obtain the sample topology subnet. Repeat the process of obtaining the sample topology subnet. In the target region topology map, arbitrarily select two distributed nodes that are not in the sample topology subnet to obtain the first sample distributed node and the second sample distributed node. Then, perform Vnew distributed node traversal on the region topology subnet formed by the first sample distributed node and the second sample distributed node. Use Vnew distributed nodes to complete the traversal of each distributed node to obtain the node topology subnet corresponding to the first sample distributed node and the second sample distributed node. Repeat the above process to divide the target region topology map into multiple node topology subnets and obtain the distributed node subnet partitioning data. It should be noted here that: In this application, the process of dividing the target area topology map into multiple node topology subnets has achieved the area division for each distributed node; The above steps have the following advantages: In this application, the above-mentioned method of dividing the distributed nodes into topology subnets by traversing the node gain degree can significantly improve the speed, accuracy and prevention capability of cable fault detection by prioritizing the processing of key nodes and limiting the scope of fault impact. At the same time, it optimizes the allocation of operation and maintenance resources and reduces the risk of power outages caused by cable detection. In existing technologies, when dividing distributed nodes in a topology graph into subnets by traversing node gain, only the number of nodes and the number of connecting edges in the topology graph are analyzed. The power indicators (node ​​load, cable impedance, and cable length) of the distributed nodes are not included in the composition parameters of node gain. This invention significantly improves the practicality of topology division and the reliability of the system by incorporating node load, cable impedance, and length into the node gain calculation.

[0023] Step S2: Perform node connection analysis on each node topology subnet according to the distributed node subnet partitioning data, and divide the node topology subnet into the first type of node subnet and the second type of node subnet according to the analysis results to obtain the node subnet partitioning data; Step S2 further includes the following steps: Obtain distributed node subnet partitioning data, obtain multiple node topology subnets based on the distributed node subnet partitioning data, and arbitrarily select a sample topology subnet from the multiple obtained node topology subnets; Perform connectivity clustering analysis on the distributed nodes in the sample topology subnet, and obtain the topology type classification index value corresponding to the sample topology subnet based on the analysis results; Specifically as follows: Obtain the node topology graph corresponding to the sample topology subnet to get the sample subnet topology graph. Mark each distributed node in the sample subnet topology graph and arbitrarily select a sample distributed node from the multiple marked distributed nodes. Neighbor node connectivity analysis is performed on the sample distributed nodes in the sample subnet topology to obtain the connectivity density ratio corresponding to the sample distributed nodes; Specifically as follows: In the sample subnet topology graph, all distributed nodes except the sample distributed nodes are set as candidate neighbor nodes, and the obtained candidate neighbor nodes are marked as candidate nodes D1 to Dc respectively. It should be noted here that: In this application, D1, D2, D3...Dc, from candidate node D1 to candidate node Dc, are the numbers corresponding to the candidate neighbor nodes, and c is the quantity value corresponding to the candidate neighbor nodes, and c is an integer greater than 0; Please see Figure 2In the sample subnet topology, a connection is created between the sample distributed node and the candidate node D1, resulting in an auxiliary connection for node D1. If the auxiliary connection for node D1 does not pass through any candidate neighbor node, then the candidate node D1 is set as a valid neighbor node; if the auxiliary connection for node D1 passes through any candidate neighbor node, then the candidate node D1 is set as an invalid neighbor node. A connection is then created between the sample distributed node and the candidate node D2, resulting in an auxiliary connection for node D2. If the auxiliary connection for node D2 does not pass through any candidate neighbor node, then the candidate node D2 is set as a valid neighbor node; if the auxiliary connection for node D2 passes through any candidate neighbor node, then the candidate node D2 is set as an invalid neighbor node. This process continues until a connection is created between the sample distributed node and the candidate node Dc, resulting in an auxiliary connection for node Dc. If the auxiliary connection for node Dc does not pass through any candidate neighbor node, then the candidate node Dc is set as a valid neighbor node; if the auxiliary connection for node Dc passes through any candidate neighbor node, then the candidate node Dc is set as an invalid neighbor node. Perform effective neighbor node analysis on the sample distributed nodes, and obtain the neighbor node connection density based on the analysis results; Specifically as follows: The actual number of cable connection edges is obtained by counting the number of cable lines that directly connect the sample distributed nodes and their effective neighbor nodes via physical cables. It should be noted here that: The actual number of cable connection edges is the number of direct physical cable connections in the subnet. Each edge corresponds to one cable and connects two nodes.

[0024] The number of effective neighbor nodes is obtained, and the effective neighbor node connection density corresponding to the sample distributed node is calculated by combining the actual number of cable connection edges and the number of effective neighbor nodes.

[0025] The connection density of the effective neighbor nodes corresponding to the sample distributed nodes is calculated using the following formula: ; Where Lmy is the effective neighbor node connection density corresponding to the sample distributed node, Sls is the actual number of cable connection edges, and Jls is the number of effective neighbor nodes. An analysis of the cable connection distance between invalid neighbor nodes is performed on the sample distributed nodes, and the direct connection deviation of invalid nodes is obtained based on the analysis results. Specifically as follows: In the sample subnet topology diagram, invalid neighbor nodes relative to the sample distributed nodes are obtained, and the obtained invalid neighbor nodes are numbered from W1 invalid neighbor node to Wd invalid neighbor node respectively. It should be noted here that: In this application, W1, W2, W3...Wd are the numbers corresponding to the invalid neighbor nodes from W1 to Wd, and d is the number of invalid neighbor nodes, which is an integer greater than 0.

[0026] Please see Figure 3 The cable connection paths between invalid neighbor nodes of W1 and sample distributed nodes in the sample subnet topology are obtained, resulting in multiple connection paths of W1 nodes. The path length values ​​of the multiple connection paths of W1 nodes are compared, and the connection path of W1 nodes with the smallest length value is set as the shortest cable path of W1. The length value of the shortest cable path of W1 is obtained to obtain the shortest length value of W1 path. The straight-line distance between the invalid neighbor node of W1 and the sample distributed node in the sample subnet topology is obtained to obtain the direct cable connection distance of W1. The number of distributed nodes covered by the shortest path of cable W1 is obtained to get the number of nodes covered by the W1 path. Calculate the difference between the shortest path length of W1 and the direct connection distance of W1 cable, and calculate the ratio of the difference to the direct connection distance of W1 cable to obtain the deviation of the shortest path distance of W1. The shortest path distance deviation of W1 and the number of path nodes covered by W1 are used to calculate the direct connection deviation of W1. The direct connection deviation of W1 is calculated using the following formula: ; Where Zlw1 is the direct connection deviation of W1, Jfw1 is the number of path nodes covered by W1, and Lpw1 is the shortest path distance deviation of W1. Repeat the process of obtaining the direct connection deviation of W1, and obtain the direct connection deviation corresponding to the invalid neighbor node of W2 to the invalid neighbor node of Wd respectively, to obtain the direct connection deviation of W2 to the direct connection deviation of Wd. The average of the direct connection deviations from W1 to Wd is calculated to obtain the direct connection deviation of the invalid nodes corresponding to the sample distributed nodes. The connectivity density ratio of the sample distributed node is obtained by comparing the effective neighbor node connection density to the ineffective node direct connection deviation. Repeat the process of obtaining the connectivity density ratio corresponding to the sample distributed nodes, obtain the connectivity density ratio corresponding to each distributed node, and calculate the average value of the obtained multiple connectivity density ratios to obtain the topology type division index value corresponding to the sample topology subnet. Repeat the process of obtaining the topology type division index value corresponding to the sample topology subnet, and obtain the topology type division index value corresponding to each node topology subnet respectively; Obtain the preset range of topology type division index. If the topology type division index value is within the preset range, the corresponding node topology subnet is divided into the first type of topology subnet. If the topology type division index value is not within the preset range, the corresponding node topology subnet is divided into the second type of topology subnet, thus obtaining the node subnet division data. It should be noted here that: In this application, the first type of topology subnet involved herein is approximately a fully connected cable topology subnet, and the second type of topology subnet involved herein is approximately a partially connected topology subnet; Obtain several historical topological sub-regions that are divided into first-type topological subnets. Obtain the topological type division index value corresponding to each historical topological sub-region. Compare the values ​​of the obtained topological type division index values. Set the topological type division index value with the largest value as the upper limit of the preset interval of the topological type division index. Set the topological type division index value with the smallest value as the lower limit of the preset interval of the topological type division index. Set the numerical interval formed by the upper limit of the preset density interval of the point connection and the lower limit of the preset interval of the topological type division index as the preset interval of the topological type division index. Step S3: Perform cable fault detection on the first type of topology subnet and the second type of topology subnet according to the node subnet division data; Step S3 further includes the following steps: Obtain node subnet partitioning data, and then obtain the first type of topology subnet and the second type of topology subnet respectively based on the node subnet partitioning data; Distributed fiber optic temperature measurement technology is used to detect cable faults in the first type of topology subnet, and cable fault warnings are given based on the detection results. Specifically as follows: Temperature-measuring optical fibers are laid for all power cables in the first type of topology subnet, and real-time optical fiber temperature values ​​are obtained for different temperature-measuring optical fibers. Multiple real-time optical fiber temperature values ​​are obtained, and a reasonable range of optical fiber temperature is determined. If the real-time optical fiber temperature value is within the reasonable range, the corresponding power cable is not abnormal. If the real-time optical fiber temperature value is not within the reasonable range, the corresponding power cable is abnormal, and a cable fault warning is issued. It should be noted here that: In this application, the reasonable range of fiber optic temperature refers to the range of values ​​consisting of the real-time temperature values ​​of the fiber optic cable in the corresponding cable area under normal operating conditions. In this application, the power cables involved did not exhibit any abnormalities, including the real-time temperature value of the optical fiber being within the reasonable temperature range of the optical fiber.

[0027] Traveling wave detection technology is used to detect cable faults in the second type of topology subnet, and cable fault early warning is given based on the detection results. Specifically as follows: In the first type of topology subnet, any power cable is selected as the sample power cable, and the distributed nodes at both ends of the sample power cable are respectively set as the first distributed node and the second distributed node. The first distributed node and the second distributed node send a traveling wave signal. The first distributed node obtains the transmission strength corresponding to the traveling wave signal strength value to obtain the first signal strength value. The second distributed node obtains the reception strength corresponding to the traveling wave signal strength value to obtain the second signal strength value. The difference between the first signal strength value and the second signal strength value is calculated, and the absolute value of the difference is taken to obtain the traveling wave signal attenuation value. The cable connection length between the first distributed node and the second distributed node is numerically obtained to obtain the propagation distance of the traveling wave signal. The ratio of the traveling wave signal attenuation value to the traveling wave signal propagation distance is calculated to obtain the traveling wave signal attenuation distance ratio. Obtain the reference range of the traveling wave signal attenuation distance ratio. If the traveling wave signal attenuation distance ratio is within the reference range, it is determined that the corresponding power cable is not abnormal. If the traveling wave signal attenuation distance ratio is not within the reference range, the corresponding power cable is abnormal, and a cable fault warning is issued. It should be noted here that: In this application, the reference range of traveling wave signal attenuation distance ratio is the numerical range composed of the traveling wave signal attenuation distance ratio of the corresponding cable area under normal operating conditions; In this application, the power cables involved do not exhibit any abnormalities, including situations where the traveling wave signal attenuation distance ratio is within the boundary of the traveling wave signal attenuation distance ratio reference range.

[0028] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to any specific implementation. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims

1. A cable fault detection method based on distributed nodes, characterized in that, Includes the following steps: Step S1: Obtain the topology map of the target area, and perform modularity gain analysis on the distributed nodes in the topology map of the target area. Based on the analysis results, split the distributed nodes into multiple node topology subnets to obtain distributed node subnet partitioning data. Step S2: Perform node connection analysis on the node topology subnet based on the distributed node subnet partitioning data. Based on the analysis results, divide the node topology subnet into a first type of node subnet and a second type of node subnet to obtain the node subnet partitioning data. Step S3: Based on the node subnet division data, perform cable fault detection on the first type of topology subnet and the second type of topology subnet respectively.

2. The cable fault detection method based on distributed nodes according to claim 1, characterized in that, Step S1 further includes the following steps: Step S11: Obtain the topology map of the distributed nodes performing cable fault detection to obtain the topology map of the target area; Step S12: Mark the distributed nodes existing in the topology graph of the target area to obtain distributed nodes from v1 to vn. Step S13: Mark the node connection cables from distributed node V1 to distributed node Vn to obtain the connection cable from e1 to em. Step S14: Obtain the circuit impedance from the e1 connecting cable to the em connecting cable, and obtain the circuit impedance values ​​from e1 to em. Obtain the cable connection length from the e1 circuit length to the em circuit length. Step S15: Select distributed nodes Vi and Vj from distributed nodes V1 to Vn, and select the topology subnet nodes to which distributed nodes Vi and Vj belong to obtain the sample topology subnet. Step S16: Randomly select two distributed nodes that are not in the sample topology subnet in the target area topology map to obtain the first sample distributed node and the second sample distributed node. Perform Vnew distributed node traversal on the area topology subnet formed by the first sample distributed node and the second sample distributed node to obtain the node topology subnet corresponding to the first sample distributed node and the second sample distributed node. Repeat this process to divide the target area topology map into multiple node topology subnets and obtain distributed node subnet partitioning data.

3. The cable fault detection method based on distributed nodes according to claim 2, characterized in that, Step S15 further includes the following steps: Connect the e1 connecting cable to the em connecting cable and the V1 distributed node to the Vn distributed node. If the ei connecting cable is the connecting cable between the Vi distributed node and the Vj distributed node, set the product of the inverse of the ei circuit impedance value and the e1 circuit length value as the cable connection weight between the Vi distributed node and the Vj distributed node to obtain the Wi,j cable connection weight. The node load intensity values ​​from distributed node v1 to distributed node vn at the current moment are obtained to obtain the real-time load from node V1 to node Vn. Treat the Vi distributed nodes and Vj distributed nodes as an independent topological sub-region, and obtain the weighted modularity corresponding to the independent topological sub-region; The weighted modularity is calculated using the following formula: ； Where Qzy is the weighted modularity, m is the number of cables connecting the distributed nodes, Wi,j is the cable connection weight between distributed nodes Vi and Vj, si is the real-time load of node Vi, and sj is the real-time load of node Vj.

4. The cable fault detection method based on distributed nodes according to claim 3, characterized in that, Step S15 further includes the following steps: Add a new distributed node Vnew to the independent topological sub-region formed by the distributed nodes Vi and Vj, and calculate the update weighted modularity after adding the distributed node Vnew. The specific formula is as follows: ； Among them, Qzy ’ To update the weighted modularity, p ’ The value represents the number of cables connecting the distributed nodes after the addition of the Vnew distributed node, Wnew,j represents the cable connection weight between the Vnew distributed node and the Vj distributed node, snew represents the real-time load of the Vnew node, and sj represents the real-time load of the Vj node. Calculate the difference between the updated weighted modularity and the weighted modularity, and take the absolute value of the difference to obtain the modularity gain after adding the Vnew distributed node. If the modularity gain is greater than 0, the Vnew distributed node is retained; if the modularity gain is less than or equal to 0, the Vnew distributed node is deleted. Repeat this process to traverse each distributed node using the Vnew distributed node to obtain the sample topology subnet.

5. The cable fault detection method based on distributed nodes according to claim 1, characterized in that, Step S2 further includes the following steps: Step S21: Obtain distributed node subnet partitioning data, obtain multiple node topology subnets based on the distributed node subnet partitioning data, and arbitrarily select a sample topology subnet from the multiple obtained node topology subnets; Step S22: Perform connection clustering analysis on the distributed nodes in the sample topology subnet, and obtain the topology type division index value corresponding to the sample topology subnet based on the analysis results; Step S23: Obtain the topology type division index value corresponding to each node's topology subnet; Step S24: Obtain the preset range of topology type division index. If the topology type division index value is within the preset range, the corresponding node topology subnet is divided into the first type of topology subnet. If the topology type division index value is not within the preset range, the corresponding node topology subnet is divided into the second type of topology subnet, thus obtaining the node subnet division data.

6. The cable fault detection method based on distributed nodes according to claim 5, characterized in that, Step S22 further includes the following steps: Step S221: Obtain the node topology graph corresponding to the sample topology subnet to obtain the sample subnet topology graph, and arbitrarily select a sample distributed node in the sample subnet topology graph; Step S222: Perform neighbor node connectivity analysis on the sample distributed nodes in the sample subnet topology graph to obtain the connectivity density ratio corresponding to the sample distributed nodes; Step S223: Obtain the connectivity density ratio corresponding to each distributed node, and calculate the average value of the multiple connectivity density ratios to obtain the topology type division index value corresponding to the sample topology subnet.

7. The cable fault detection method based on distributed nodes according to claim 5, characterized in that, Step S222 further includes the following steps: Step S2221: Set the distributed nodes in the sample subnet topology graph other than the sample distributed nodes as candidate neighbor nodes, and mark the obtained candidate neighbor nodes as candidate nodes D1 to Dc respectively. Step S2222: In the sample subnet topology diagram, create a connection between the sample distributed node and the candidate node D1 to obtain the auxiliary connection of node D1. If the auxiliary connection of node D1 does not pass through any candidate neighbor node, then the candidate node D1 is set as a valid neighbor node. If the auxiliary connection of node D1 passes through any candidate neighbor node, then the candidate node D1 is set as an invalid neighbor node. And so on, create a connection between the sample distributed node and the candidate node Dc to obtain the auxiliary connection of node Dc. If the auxiliary connection of node Dc does not pass through any candidate neighbor node, then the candidate node Dc is set as a valid neighbor node. If the auxiliary connection of node Dc passes through any candidate neighbor node, then the candidate node Dc is set as an invalid neighbor node. Step S2223: Perform effective neighbor node analysis on the sample distributed nodes, and obtain the neighbor node connection density based on the analysis results; Step S2224: Perform cable connection distance analysis on the sample distributed nodes for invalid neighbor nodes, and obtain the direct connection deviation of invalid nodes based on the analysis results; Step S2225: The ratio of the effective neighbor node connection density to the ineffective node direct connection deviation of the sample distributed node is used to obtain the connectivity density ratio of the sample distributed node.

8. The cable fault detection method based on distributed nodes according to claim 7, characterized in that, Step S2223 further includes the following steps: The actual number of cable connection edges is obtained by counting the number of cable lines that directly connect the sample distributed nodes and their effective neighbor nodes via physical cables. The number of effective neighbor nodes is obtained, and the effective neighbor node connection density corresponding to the sample distributed node is calculated by combining the actual number of cable connection edges and the number of effective neighbor nodes.

9. The cable fault detection method based on distributed nodes according to claim 7, characterized in that, Step S2224 further includes the following steps: In the sample subnet topology diagram, invalid neighbor nodes relative to the sample distributed nodes are obtained, resulting in invalid neighbor nodes W1 to Wd. The cable connection paths between invalid neighbor nodes of W1 and sample distributed nodes in the sample subnet topology are obtained, resulting in multiple connection paths of W1 nodes. The path length values ​​of the multiple connection paths of W1 nodes are compared, and the connection path of W1 nodes with the smallest length value is set as the shortest W1 cable path. The length value of the shortest W1 cable path is obtained to get the shortest length value of W1 path. The straight-line distance between the invalid neighbor node of W1 and the sample distributed node in the sample subnet topology is obtained to obtain the direct cable connection distance of W1. The number of distributed nodes covered by the shortest path of cable W1 is obtained to get the number of nodes covered by the W1 path. Calculate the difference between the shortest path length of W1 and the direct connection distance of W1 cable, and calculate the ratio of the difference to the direct connection distance of W1 cable to obtain the deviation of the shortest path distance of W1. The shortest path distance deviation of W1 and the number of path nodes covered by W1 are used to calculate the direct connection deviation of W1. Calculate the direct connection deviation from W2 to Wd; The average of the direct connection deviations from W1 to Wd is calculated to obtain the direct connection deviation of the invalid nodes corresponding to the sample distributed nodes.

10. The cable fault detection method based on distributed nodes according to claim 1, characterized in that, Step S3 further includes the following steps: Obtain node subnet partitioning data, and then obtain the first type of topology subnet and the second type of topology subnet respectively based on the node subnet partitioning data; Distributed fiber optic temperature measurement technology is used to detect cable faults in the first type of topology subnet, and cable fault warnings are given based on the detection results. Traveling wave detection technology is used to detect cable faults in the second type of topology subnet, and cable fault warnings are given based on the detection results.

Images