Power line mechanical external failure monitoring system and method based on digital twin technology

Abstract

The present application relates to the field of line external failure warning technology, and provides a power line mechanical external failure monitoring system and method based on digital twin technology, the system comprising a point cloud data processing module, a map display module, a communication module, and an early warning module. The point cloud data processing module of the present application aggregates a large range of point cloud data through a clustering algorithm, identifies the cluster set through an image recognition algorithm, obtains the precise cluster set of the corresponding cable segment and vertical equipment, and then generates a high-precision digital twin model of a large range of lines on the map, which improves the recognition accuracy of external failure warning on the basis of expanding the monitoring range.

Description

Technical Field

[0001] The present application belongs to the technical field of line external failure warning, and specifically relates to a power line mechanical external failure monitoring system and method based on digital twin technology. Background Art

[0002] With the rapid development of the power industry, the scale of the power grid is expanding, and the power lines are crisscrossed. Their safe and stable operation is crucial to ensuring social production, life and economic development. However, power lines have been erected in the wild for a long time and are easily damaged by external factors, especially special operation vehicles such as cranes that frequently accidentally touch high-voltage lines, resulting in power outages and damage to power facilities, causing significant economic losses and safety hazards to society.

[0003] At present, there are many technical solutions for monitoring external force damage to power lines on the market, among which the more common ones are systems based on video monitoring and radar detection. These systems usually monitor the surrounding environment in real time by installing cameras or radar equipment along the power lines, and issue warning signals when abnormal behavior or objects approaching are detected. However, these existing technical solutions have the following major shortcomings in practical applications: Limited monitoring range: Traditional video surveillance systems are limited by the camera's field of view and installation location, making it difficult to achieve all-round monitoring of power lines without blind spots. Especially in complex terrain and severe weather conditions, the monitoring effect is greatly reduced.

[0004] Insufficient early warning accuracy: Although radar detection technology can detect the movement of objects, in complex environments, such as tree obstruction, electromagnetic interference, etc., another method of measuring distance through monitoring images can only rely on reference objects to calculate the approximate distance, which seriously affects the accuracy and reliability of the early warning. Summary of the invention

[0005] The present application provides a power line mechanical external failure monitoring system and method based on digital twin technology to solve or partially solve the problems raised in the above background technology.

[0006] This application provides a power line mechanical external failure monitoring system based on digital twin technology, including a point cloud data processing module, a map display module, a communication module, and an early warning module; The point cloud data processing module receives the point cloud data, performs multi-layer clustering processing and image recognition on the point cloud data, and finally generates a cluster set list C_final_i corresponding to each cable segment and vertical device, where i is the set number; The communication module communicates with the vehicle-mounted RTK positioning device to obtain vehicle positioning information and vehicle contour information; The map display module obtains the cluster set C_final_i and generates a simulated map, obtains the vehicle positioning information and displays it on the simulated map; The early warning module monitors the distance between the vehicle and each cable segment and vertical equipment in real time. When the distance between the vehicle and any cable segment or vertical equipment is less than the preset threshold D, the vehicle contour and the point cloud in the corresponding cluster set C_final_i are judged as externally broken and an early warning is issued.

[0007] Preferably, the point cloud data processing module performs preprocessing on the point cloud data before performing multi-layer clustering processing and image recognition on the point cloud data. The specific method is as follows: Use statistical filtering algorithm to remove noise points and outliers; Use the RANSAC algorithm to fit the plane and remove ground points; Remove points outside the preset height threshold range.

[0008] Preferably, the point cloud data processing module uses the DBSCAN clustering algorithm to perform clustering processing on the point cloud data. The specific method is as follows: S301: confirming clustering parameters according to the number of identification times N, and initializing parameters of the DBSCAN algorithm, wherein the parameters at least include an eps range and a MinPts value, and the initial value of N is 1; S302: performing clustering processing on the total set of point cloud data to obtain multiple intermediate cluster sets C_j, where j is the number of the element in the set; S303: Perform local feature recognition and judgment on the point cloud set elements in the intermediate cluster set C_j in turn. If the device type and posture corresponding to the corresponding element are identified, a unique device identification code corresponding to the element is generated by combining the point cloud position distribution and the preset device information library; S304: According to the identified device type and posture, confirm the device contour range, add the point cloud data within the contour range to the point cloud set of the corresponding element, delete the corresponding point cloud data from the total point cloud data set, and add the unique device identification code and the corresponding element in C_First_j to the cluster set list C_final_i; S305: Count the unique device identification codes and determine whether there are any unidentified unique device identification codes in combination with the device information database. If yes, clear the intermediate cluster set C_j, increase the value of N by 1, and jump to step S301; If not, jump to step S306; S306: End the clustering and identification process, and send the cluster set list C_final_i to the map display module.

[0009] Preferably, in step S303, the specific method for performing device local feature recognition and judgment on the point cloud set elements is as follows: S331: Obtain a top view P1 of a point cloud set and generate a minimum circular airspace Cir covering all points in P1; S332: Rotate the point cloud based on the central axis of the circular airspace Cir, and obtain M plan views P2_m at uniform intervals, where m ranges from 1 to M; S333: Compare the preset device image set, perform image recognition on the top view P1 and the horizontal view P2_m, and identify the device type and posture.

[0010] Preferably, in step S333, if it is identified that the corresponding point cloud set corresponds to multiple possible device types or postures, multiple device contour ranges are generated based on the combination of device type and posture, and other point cloud set elements that are within or partially within the corresponding device contour range are combined with the identified elements into a new point cloud set, and secondary identification is performed according to steps S331 to S333.

[0011] Preferably, the specific method of the early warning module for external damage judgment and early warning is as follows: S601: Generate a warning contour based on the actual contour point cloud of each cable segment and vertical equipment; S602: Receive the vehicle positioning information and vehicle contour information in real time. When the distance between the vehicle and any cable segment or vertical device is less than the preset threshold D, determine whether the vehicle contour and the warning contour of the corresponding cable segment or vertical device overlap according to the vehicle's travel direction. If so, generate warning information and send it to the vehicle through the communication module.

[0012] Preferably, the method for generating the warning profile is as follows: S611: Obtaining a set of actual contour point clouds of a corresponding cable segment or vertical device; S612: Calculate the maximum and minimum values ​​of the X, Y, and Z coordinates of all elements in the actual contour point cloud set, and generate a contour cube a based on the maximum and minimum values ​​of the three-axis coordinates; S613: Expand each face of the contour cube a by a preset offset size d to form a warning cube A, where the contour of A is the warning contour.

[0013] Preferably, the warning module issues warning information in a graded manner based on the expected minimum distance between the vehicle contour and the warning contour.

[0014] Preferably, the map display module integrates particle animation effects. When the warning module issues a warning message, the screen automatically flies to the warning coordinates and simulates the external damage effect. The present application also provides a method for monitoring mechanical external failure of power lines based on digital twin technology, comprising the following steps: S100: Acquire point cloud data; S200: Perform multi-layer clustering processing and image recognition on the point cloud data based on the DBSCAN algorithm and the image recognition algorithm, and generate a cluster set list C_final_i corresponding to each cable segment and vertical device; S300: Monitor the distance between the vehicle and each cable segment and vertical equipment in real time. When the distance between the vehicle and any cable segment or vertical equipment is less than the preset threshold D, make an external breach judgment and issue an early warning for the vehicle outline and the point cloud in the corresponding cluster set C_final_i.

[0015] Compared with the prior art, the present invention has the following beneficial effects: This application aggregates a large range of point cloud data through a clustering algorithm, identifies the cluster set through an image recognition algorithm, obtains the accurate cluster set of the corresponding cable segments and vertical equipment, and then generates a high-precision digital twin model of a large range of lines on the map, which improves the recognition accuracy of external damage warning on the basis of expanding the monitoring range. This application uses the DBSCAN clustering algorithm in conjunction with the image recognition algorithm to cluster and identify the total set of point cloud data through multi-layer polling, completes the point cloud of the identified point cloud data based on the original contour of the corresponding equipment type, and deletes the completed point cloud from the total set of point cloud data, while ensuring the clustering speed and clustering accuracy. The recognition speed is improved at the same time, effectively avoiding the loss of point clouds or the addition of noise points during clustering, which is conducive to the rapid and accurate generation of line twin models on the map module. This application increases the redundant space for external damage by setting the warning contour, avoiding collision accidents in extreme cases where the distance between vehicles and line equipment is extremely small. BRIEF DESCRIPTION OF THE DRAWINGS The present application is further described below in conjunction with the accompanying drawings and embodiments.

[0016] Figure 1 This is a schematic diagram of the system composition of this application. Figure 2 This is a schematic diagram of the method flow of this application, Figure 3 This is a schematic diagram of the implementation of local feature recognition of the device of this application. DETAILED DESCRIPTION

[0017] For example, certain words are used in the specification and claims to refer to specific components. Those skilled in the art should understand that hardware manufacturers may use different terms to refer to the same component. This specification and claims do not use differences in names as a way to distinguish components, but use differences in the functions of components as the criteria for distinction. For example, "including" mentioned throughout the specification and claims is an open term, so it should be interpreted as "including but not limited to". "Approximately" means that within an acceptable error range, those skilled in the art can solve the technical problem within a certain error range and basically achieve the technical effect.

[0018] In the description of the present application, it should be understood that the terms "upper", "lower", "front", "rear", "left", "right", "horizontal", etc., indicating directions or positional relationships, are based on the directions or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present application and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific direction, be constructed and operated in a specific direction, and therefore should not be understood as limiting the present application.

[0019] In this application, unless otherwise clearly specified and limited, the terms "installed", "connected", "connected", "fixed" and the like should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be an indirect connection through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in this application can be understood according to specific circumstances.

[0020] Example 1 like Figures 1 to 3 As shown, the present application provides a power line mechanical external failure monitoring system based on digital twin technology, including a point cloud data processing module, a map display module, a communication module, and an early warning module. The functions of each module are as follows: The point cloud data processing module receives the point cloud data, performs multi-layer clustering processing and image recognition on the point cloud data, and finally generates a cluster set list C_final_i corresponding to each cable segment and vertical device, where i is the set number; The communication module communicates with the vehicle-mounted RTK positioning device to obtain vehicle positioning information and vehicle contour information; The map display module obtains the cluster set C_final_i and generates a simulated map, obtains the vehicle positioning information and displays it on the simulated map; The early warning module monitors the distance between the vehicle and each cable segment and vertical equipment in real time. When the distance between the vehicle and any cable segment or vertical equipment is less than the preset threshold D, the vehicle contour and the point cloud in the corresponding cluster set C_final_i are judged as externally broken and an early warning is issued.

[0021] The point cloud data is acquired through drone scanning, the cable segments are divided based on the intervals of cable connections and jumper devices such as cable towers, and the vertical equipment includes cable support equipment such as cable towers or other ground equipment that may affect the operation of the vehicle.

[0022] The RTK (Real Time Kinematic) positioning device, i.e., a real-time dynamic carrier phase differential positioning device, is a high-precision satellite positioning and measurement device. The vehicle profile information is obtained by scanning with a high-definition scanning camera before the vehicle runs.

[0023] Preferably, the point cloud data processing module performs preprocessing on the point cloud data before performing multi-layer clustering processing and image recognition on the point cloud data. The specific method is as follows: Use statistical filtering algorithm to remove noise points and outliers; Use the RANSAC algorithm to fit the plane and remove ground points; Remove points outside the preset height threshold range.

[0024] The statistical filtering algorithm is a filtering technology used to process point cloud data. Its basic principle is to perform statistical analysis on the neighborhood of each point in the point cloud and remove points that do not meet certain standards based on the statistical results. This application uses the statistical filtering algorithm to remove obvious noise points.

[0025] The RANSAC (Random Sample Consensus) algorithm is an algorithm used to fit a model from a set of data and identify inliers and outliers. In the field of plane fitting, the RANSAC algorithm has significant advantages, especially when processing data containing a lot of noise and outliers. This application uses the RANSAC algorithm to remove ground points to reduce the noise of point cloud data, which is beneficial to the subsequent clustering speed and clustering accuracy.

[0026] The height threshold interval is set based on the installation height of the cable and the extension height of the vertical surface of the relevant vertical equipment, and the purpose is also to reduce the noise of the point cloud data.

[0027] Specifically, the point cloud data processing module uses the DBSCAN clustering algorithm to perform clustering processing on the point cloud data. The specific method is as follows: S301: confirming clustering parameters according to the number of identification times N, and initializing parameters of the DBSCAN algorithm, wherein the parameters at least include an eps range and a MinPts value, and the initial value of N is 1; S302: performing clustering processing on the total set of point cloud data to obtain multiple intermediate cluster sets C_j, where j is the number of the element in the set; S303: Perform local feature recognition and judgment on the point cloud set elements in the intermediate cluster set C_j in turn. If the device type and posture corresponding to the corresponding element are identified, a unique device identification code corresponding to the element is generated by combining the point cloud position distribution and the preset device information library; S304: According to the identified device type and posture, confirm the device contour range, add the point cloud data within the contour range to the point cloud set of the corresponding element, delete the corresponding point cloud data from the total point cloud data set, and add the unique device identification code and the corresponding element in C_First_j to the cluster set list C_final_i; S305: Count the unique device identification codes and determine whether there are any unidentified unique device identification codes in combination with the device information database. If yes, clear the intermediate cluster set C_j, increase the value of N by 1, and jump to step S301; If not, jump to step S306; S306: End the clustering and identification process, and send the cluster set list C_final_i to the map display module.

[0028] This application uses the DBSCAN clustering algorithm in conjunction with the image recognition algorithm to perform multi-layer polling clustering and identification on the point cloud data set, completes the identified point cloud data based on the original contour of the corresponding device type, and deletes the completed point cloud from the point cloud data set. It improves the recognition speed while ensuring the clustering speed and clustering accuracy, effectively avoids the loss of point clouds or the addition of noise points during clustering, and is conducive to the rapid and accurate generation of route twin models on the map module.

[0029] In step S301, the eps range is the neighborhood radius, and the MinPts value is the minimum number of samples.

[0030] In step S302, the clustering execution method in the DBSCAN algorithm is a conventional setting in the art and will not be described in detail in this application.

[0031] In step S303, the parameters of the elements in the device information library include a unique device identification code, a device type, and basic longitude and latitude coordinates of the device. The basic longitude and latitude coordinates of the device are the longitude and latitude coordinates of any point of the corresponding device during construction or planning; the device posture refers to the posture of the device in a fixed coordinate system.

[0032] Specifically, in step S303, the specific method for performing device local feature recognition and judgment on the point cloud set elements is as follows: S331: Obtain a top view P1 of a point cloud set and generate a minimum circular airspace Cir covering all points in P1; S332: Rotate the point cloud based on the central axis of the circular airspace Cir, and obtain M plan views P2_m at uniform intervals, where m ranges from 1 to M; S333: Compare the preset device image set, perform image recognition on the top view P1 and the horizontal view P2_m, and identify the device type and posture.

[0033] like Figure 3 As shown, Figure 3 The upper middle image is a top view, the lower image is a plan view, and the value of M is 4.

[0034] In the steps S331 to S333, image recognition is performed using the top view and multiple plan views of the point cloud image, which is beneficial to the rapid identification of the device type and device posture.

[0035] Preferably, in step S333, if it is identified that the corresponding point cloud set corresponds to multiple possible device types or postures, multiple device contour ranges are generated based on the combination of device type and posture, and other point cloud set elements that are within or partially within the corresponding device contour range are combined with the identified elements into a new point cloud set, and secondary identification is performed according to steps S331 to S333.

[0036] For example: if the initial recognition shows that the device type may be S1 or S2, if the device type is S1, its posture may be P1 or P2, if the device type is S2, its posture can only be P3, then the possible contour ranges are {S1, P1}, {S1, P2}, {S2, P3}, three contour ranges are generated respectively, and new point cloud sets are combined, and the three combinations are judged and identified using the recognition methods of steps S331 to S333.

[0037] Specifically, the specific method of the early warning module to judge and warn of external damage is as follows: S601: Generate a warning contour based on the actual contour point cloud of each cable segment and vertical equipment; S602: Receive the vehicle positioning information and vehicle contour information in real time. When the distance between the vehicle and any cable segment or vertical device is less than the preset threshold D, determine whether the vehicle contour and the warning contour of the corresponding cable segment or vertical device overlap according to the vehicle's travel direction. If so, generate warning information and send it to the vehicle through the communication module.

[0038] Specifically, the method for generating the warning profile is as follows: S611: Obtaining a set of actual contour point clouds of a corresponding cable segment or vertical device; S612: Calculate the maximum and minimum values ​​of the X, Y, and Z coordinates of all elements in the actual contour point cloud set, and generate a contour cube a based on the maximum and minimum values ​​of the three-axis coordinates; S613: Expand each face of the contour cube a by a preset offset size d to form a warning cube A, where the contour of A is the warning contour.

[0039] This application increases the external damage redundant space by setting up the early warning contour, avoiding collision accidents in extreme cases where the distance between vehicles and line equipment is extremely small.

[0040] Preferably, the warning module issues warning information in a graded manner based on the expected minimum distance between the vehicle contour and the warning contour.

[0041] Preferably, the map display module integrates particle animation effects, and when the warning module issues a warning message, the screen automatically flies to the warning coordinates and simulates an external damage effect.

[0042] The map display module of this application uses the CesiumLab platform to slice the point cloud processed by the point cloud data processing module, and uses the Cesium.Cesium3DTileset method to render the sliced ​​point cloud data in the form of a 3DTiles model, thereby optimizing loading efficiency and ensuring fast loading and smooth display of large-scale point cloud data; combined with the Cesium.js map framework and BingMap map, the flyTo method and the Cesium particle system animation effect are used to achieve seamless integration and display of map information and point cloud slice data. When an external breach warning occurs, the system automatically flies to the warning coordinate point and simulates the external breach effect.

[0043] Based on the above system, the present application also provides a method for monitoring mechanical external failure of power lines based on digital twin technology, comprising the following steps: S100: Acquire point cloud data; S200: Perform multi-layer clustering processing and image recognition on the point cloud data based on the DBSCAN algorithm and the image recognition algorithm, and generate a cluster set list C_final_i corresponding to each cable segment and vertical device; S300: Monitor the distance between the vehicle and each cable segment and vertical equipment in real time. When the distance between the vehicle and any cable segment or vertical equipment is less than the preset threshold D, make an external breach judgment and issue an early warning for the vehicle outline and the point cloud in the corresponding cluster set C_final_i.

[0044] The implementation methods of the present application are described in detail above in conjunction with the accompanying drawings, but the present application is not limited to the above implementation methods, and various changes can be made within the knowledge scope of ordinary technicians in the relevant technical field without departing from the purpose of the present application.

Claims

1. The power line mechanical external failure monitoring system based on digital twin technology is characterized by: include: Point cloud data processing module, map display module, communication module, and early warning module; The point cloud data processing module receives the point cloud data, performs multi-layer clustering processing and image recognition on the point cloud data, and finally generates a cluster set list C_final_i corresponding to each cable segment and vertical device, where i is the set number; The communication module communicates with the vehicle-mounted RTK positioning device to obtain vehicle positioning information and vehicle contour information; The map display module obtains the cluster set C_final_i and generates a simulated map, obtains the vehicle positioning information and displays it on the simulated map; The early warning module monitors the distance between the vehicle and each cable segment and vertical equipment in real time. When the distance between the vehicle and any cable segment or vertical equipment is less than the preset threshold D, the vehicle contour and the point cloud in the corresponding cluster set C_final_i are judged as externally broken and an early warning is issued.

2. According to the digital twin technology-based power line mechanical external failure monitoring system of claim 1, it is characterized by: The point cloud data processing module performs preprocessing on the point cloud data before performing multi-layer clustering processing and image recognition on the point cloud data. The specific method is as follows: Use statistical filtering algorithm to remove noise points and outliers; Use the RANSAC algorithm to fit the plane and remove ground points; Remove points outside the preset height threshold range.

3. The power line mechanical external failure monitoring system based on digital twin technology according to any one of claims 1 or 2 is characterized in that: The point cloud data processing module uses the DBSCAN clustering algorithm to perform clustering processing on the point cloud data. The specific method is as follows: S301: confirm clustering parameters according to the number of identification times N, and initialize the parameters of the DBSCAN algorithm, wherein the parameters at least include the eps range and the MinPts value, and the initial value of N is 1; S302: performing clustering processing on the total set of point cloud data to obtain multiple intermediate cluster sets C_j, where j is the number of the element in the set; S303: Perform local feature recognition and judgment on the point cloud set elements in the intermediate cluster set C_j in turn. If the device type and posture corresponding to the corresponding element are identified, a unique device identification code corresponding to the element is generated by combining the point cloud position distribution and the preset device information library; S304: According to the identified device type and posture, confirm the device contour range, add the point cloud data within the contour range to the point cloud set of the corresponding element, delete the corresponding point cloud data from the total point cloud data set, and add the unique device identification code and the corresponding element in C_First_j to the cluster set list C_final_i; S305: Count the unique device identification codes and determine whether there are any unrecognized unique device identification codes in combination with the device information database. If yes, clear the intermediate cluster set C_j, increase the value of N by 1, and jump to step S301; If not, jump to step S306; S306: End the clustering and identification process, and send the cluster set list C_final_i to the map display module.

4. The power line mechanical external failure monitoring system based on digital twin technology according to claim 3 is characterized by: In step S303, the specific method for identifying and judging the local features of the device on the point cloud set elements is as follows: S331: Obtain a top view P1 of a point cloud set and generate a minimum circular airspace Cir covering all points in P1; S332: Rotate the point cloud based on the central axis of the circular airspace Cir, and obtain M plan views P2_m at uniform intervals, where m ranges from 1 to M; S333: Compare the preset device image set, perform image recognition on the top view P1 and the horizontal view P2_m, and identify the device type and posture.

5. The power line mechanical external failure monitoring system based on digital twin technology according to claim 4 is characterized by: In the step S333, if it is identified that the corresponding point cloud set corresponds to multiple possible device types or postures, multiple device contour ranges are generated based on the combination of device type and posture, and other point cloud set elements that are within or partially within the corresponding device contour range are combined with the identified elements into a new point cloud set, and secondary identification is performed according to steps S331 to S333.

6. The power line mechanical external failure monitoring system based on digital twin technology according to claim 1 is characterized by: The specific method of the early warning module to judge and warn of external damage is as follows: S601: Generate a warning contour based on the actual contour point cloud of each cable segment and vertical equipment; S602: Receive the vehicle positioning information and vehicle contour information in real time. When the distance between the vehicle and any cable segment or vertical device is less than the preset threshold D, determine whether the vehicle contour and the warning contour of the corresponding cable segment or vertical device overlap according to the vehicle's travel direction. If so, generate warning information and send it to the vehicle through the communication module.

7. The power line mechanical external failure monitoring system based on digital twin technology according to claim 6 is characterized by: The method for generating the early warning profile is specifically as follows: S611: Obtaining a set of actual contour point clouds of a corresponding cable segment or vertical device; S612: Calculate the maximum and minimum values ​​of the X, Y, and Z coordinates of all elements in the actual contour point cloud set, and generate a contour cube a based on the maximum and minimum values ​​of the three-axis coordinates; S613: Expand each face of the contour cube a by a preset offset size d to form a warning cube A, where the contour of A is the warning contour.

8. The power line mechanical external failure monitoring system based on digital twin technology according to any one of claims 6 or 7 is characterized in that: The warning module issues warning information in a graded manner based on the expected minimum distance between the vehicle contour and the warning contour.

9. The power line mechanical external failure monitoring system based on digital twin technology according to claim 1 is characterized by: The map display module integrates particle animation effects. When the early warning module issues an early warning message, the picture automatically flies to the early warning coordinates and simulates the external damage effect.

10. A method for monitoring mechanical failure of power lines based on digital twin technology, characterized in that: The steps include: S100: Acquire point cloud data; S200: Perform multi-layer clustering processing and image recognition on the point cloud data based on the DBSCAN algorithm and the image recognition algorithm, and generate a cluster set list C_final_i corresponding to each cable segment and vertical device; S300: Monitor the distance between the vehicle and each cable segment and vertical equipment in real time. When the distance between the vehicle and any cable segment or vertical equipment is less than the preset threshold D, make an external breach judgment and issue an early warning for the vehicle outline and the point cloud in the corresponding cluster set C_final_i.

Images