Ground wire broken strand automatic analysis method and device based on point cloud data

Abstract

The invention provides a ground wire broken strand automatic analysis method and device based on point cloud data, and the method and device achieve the automatic and intelligent detection of ground wire broken strands, overcome the defects that in the prior art, the subjectivity is high, and the detection is liable to be affected by the environment through the use of the point cloud data, reduce the misjudgment and missing detection of manual inspection, and improve the detection efficiency. Thousands of tower data can be processed through single analysis, the detection speed is increased by more than 10 times compared with manual detection, and the method is suitable for intelligent operation and maintenance of large-scale power transmission lines. Meanwhile, according to the scheme disclosed by the invention, through dynamic slicing and baseline fitting, the broken strand identification accuracy can reach 90% or above; in addition, the anomaly detection parameters in the invention are open to adapt to line scenes of different voltage levels and ground wire types; in addition, the detection result is directly connected with an operation and maintenance system, and the ground wire broken strand fault can be quickly positioned and the defect can be quickly repaired.

Description

TECHNICAL FIELD

[0001] The present disclosure relates to the field of electric power operation and maintenance, and in particular to a ground wire broken strand automatic analysis method and device based on point cloud data. BACKGROUND

[0002] Ground wire broken strand detection is an important part of power system maintenance, mainly used to ensure the integrity and safety of overhead transmission line ground wires (lightning wires). The current schemes commonly used for ground wire broken strand detection include visual inspection, infrared detection, ultrasonic detection, etc. Although the current schemes are relatively mature, they all have defects. For example, the visual inspection scheme has strong subjectivity and poor stability; the infrared detection scheme is easily affected by sunlight and temperature difference, and has a high false positive rate; the ultrasonic detection has a reduced signal-to-noise ratio in a strong wind or vibration environment. In summary, the current ground wire broken strand detection has the defects of insufficient automation and intelligence, and is easily affected by the environment; and also has the defect of low efficiency from detection to operation implementation, which causes the ground wire broken strand fault to be unable to be eliminated at the fastest speed. SUMMARY

[0003] The present disclosure at least provides a ground wire broken strand automatic analysis method and device based on point cloud data to solve at least one of the above technical defects.

[0004] According to an aspect of the present disclosure, a ground wire broken strand automatic analysis method based on point cloud data is provided, comprising:

[0005] Obtaining original point cloud data of a transmission line, and performing classification processing on the original point cloud data to obtain ground wire point cloud data;

[0006] Obtaining an abnormality detection parameter; wherein the abnormality detection parameter includes a point cloud slice thickness, a height difference threshold, a detection direction, and a starting distance;

[0007] Marking a tower pole, and performing section segmentation on the ground wire point cloud data based on the starting distance and the point cloud slice thickness to obtain a plurality of point cloud slices;

[0008] For each point cloud slice, performing reference line fitting using the point cloud data in the point cloud slice to obtain a ground wire baseline; extracting point cloud data within a preset distance range according to the detection direction with the ground wire baseline as a reference, and calculating the perpendicular distance of each point cloud to the ground wire baseline; regarding the point cloud with a perpendicular distance greater than the height difference threshold as a broken strand abnormal point; clustering the broken strand abnormal points, removing noise points based on the clustering result, and determining a broken strand position based on the remaining broken strand abnormal points.

[0009] In a possible implementation, the ground wire broken strand automatic analysis method based on point cloud data further comprises:

[0010] generate and display a three-dimensional point cloud image with the broken strand position points based on the broken strand position and the ground wire point cloud data; wherein the three-dimensional point cloud image further includes a broken strand position point list;

[0011] in response to a click operation on a target broken strand position point in the broken strand position point list, highlight the corresponding broken strand position point in the three-dimensional point cloud image and display the feature information of the broken strand position point; the feature information includes the ground wire type and the broken strand position;

[0012] in response to a click operation on a review button in the three-dimensional point cloud image, display a revision page for revising the broken strand position points; the revision page includes a revision component for revising each broken strand position point and an adding component for adding a new broken strand position point; the revision component includes a modification component and a deletion component;

[0013] in response to an operation on the revision component, delete the feature information of the broken strand position point or obtain new feature information of the broken strand position point, and synchronously update the three-dimensional point cloud image;

[0014] in response to an operation on the adding component, obtain the feature information of the newly added broken strand position point, and synchronously update the three-dimensional point cloud image based on the feature information of the newly added broken strand position point.

[0015] In a possible implementation, the ground wire broken strand automatic analysis method based on point cloud data further includes: exporting the feature information of each broken strand position point in a predetermined format.

[0016] In a possible implementation, the ground wire broken strand automatic analysis method based on point cloud data further includes:

[0017] input the three-dimensional point cloud image and the feature information of each broken strand position point into the trained point cloud operation and maintenance model;

[0018] the point cloud operation and maintenance processing submodel in the point cloud operation and maintenance model processes the three-dimensional point cloud image to obtain a first operation and maintenance feature map;

[0019] the feature operation and maintenance submodel in the point cloud operation and maintenance model processes the feature information of each broken strand position point to obtain a second operation and maintenance feature map;

[0020] the operation and maintenance comprehensive processing submodel in the point cloud operation and maintenance model performs feature extraction and fusion processing on the first operation and maintenance feature map and the second operation and maintenance feature map to obtain operation and maintenance decision data;

[0021] transmit the operation and maintenance decision data to an operation and maintenance system.

[0022] In a possible implementation, the ground wire broken strand automatic analysis method based on point cloud data further includes:

[0023] obtaining operation and maintenance execution data fed back by the operation and maintenance system;

[0024] extracting completion information in the operation and maintenance execution data, and in a case where the completion information indicates that the repair is completed, re-obtaining three-dimensional point cloud data in a point cloud slice range corresponding to the broken strand position point, and classifying the three-dimensional point cloud data to obtain a repaired ground wire point cloud, performing baseline fitting on the repaired ground wire point cloud to obtain a repaired ground wire baseline; extracting point cloud data within a preset distance range according to a detection direction with the repaired ground wire baseline as a reference, and calculating a perpendicular distance of each point cloud extracted from the repaired ground wire baseline; taking point cloud with a perpendicular distance greater than the height difference threshold as a repaired broken strand abnormal point; clustering the repaired broken strand abnormal points, eliminating noise points based on a clustering result, and determining a repaired broken strand position based on the remaining repaired broken strand abnormal points; in a case where the repaired broken strand position does not include the broken strand position point, extracting operation and maintenance operation information from the operation and maintenance execution data;

[0025] matching the operation and maintenance operation information with the operation and maintenance decision data, and if the matching is unsuccessful, re-training the point cloud operation and maintenance model by using the operation and maintenance operation information.

[0026] In a possible implementation, the point cloud operation and maintenance processing sub-model includes a fault classification sub-model and an operation and maintenance feature processing sub-model.

[0027] processing the three-dimensional point cloud image by using the fault classification sub-model to obtain a fault classification feature map; and processing the fault classification feature map by using the operation and maintenance feature processing sub-model to obtain a first operation and maintenance feature map.

[0028] In a possible implementation, the fault classification sub-model includes a lightweight YOLOv10 model.

[0029] In a possible implementation, the tower pole is marked, and the ground wire point cloud data is segmented into multiple point cloud slices based on the starting distance and the point cloud slice thickness, including:

[0030] The ground wire point cloud data between two tower poles and having a distance greater than the starting distance from the tower pole is segmented into multiple point cloud slices according to the point cloud slice thickness.

[0031] In a possible implementation, the segmenting the ground wire point cloud data includes:

[0032] The ground wire point cloud data is segmented along the line direction.

[0033] According to another aspect of the present disclosure, a ground wire broken strand automatic analysis device based on point cloud data is provided, including:

[0034] a point cloud processing module, configured to acquire original point cloud data of a power transmission line, and perform classification processing on the original point cloud data to obtain ground wire point cloud data;

[0035] a parameter acquisition module, configured to acquire abnormality detection parameters; wherein the abnormality detection parameters include point cloud slice thickness, height difference threshold, detection direction, and starting distance;

[0036] a point cloud slice module, configured to mark a tower pole, and perform section segmentation on the ground wire point cloud data based on the starting distance and the point cloud slice thickness to obtain a plurality of point cloud slices;

[0037] an abnormality detection module, configured to, for each point cloud slice, perform baseline fitting on the point cloud data in the point cloud slice to obtain a ground wire baseline; extract point cloud data within a preset distance range according to the detection direction with the ground wire baseline as a reference, and calculate the perpendicular distance between each point cloud and the ground wire baseline; regard point cloud with a perpendicular distance greater than the height difference threshold as a broken strand abnormal point; perform clustering on the broken strand abnormal points, remove noise points based on the clustering result, and determine a broken strand position based on the remaining broken strand abnormal points.

[0038] The ground wire breakage automatic analysis method and device based on point cloud data of the present disclosure first acquires the original point cloud data of the power transmission line, and classifies and processes the original point cloud data to obtain ground wire point cloud data; and acquires abnormality detection parameters; wherein the abnormality detection parameters include point cloud slice thickness, height difference threshold, detection direction, and starting distance; then marks the tower pole, and based on the starting distance and the point cloud slice thickness, performs section segmentation on the ground wire point cloud data to obtain a plurality of point cloud slices; finally, for each point cloud slice, performs baseline fitting on the point cloud data in the point cloud slice to obtain a ground wire baseline; taking the ground wire baseline as a reference, extracts the point cloud data within a preset distance range according to the detection direction, and calculates the perpendicular distance of each point cloud extracted from the ground wire baseline; the point cloud with a perpendicular distance greater than the height difference threshold is regarded as a breakage abnormal point; the breakage abnormal points are clustered, and based on the clustering result, noise points are removed, and the breakage position is determined based on the remaining breakage abnormal points. The present disclosure realizes the automatic and intelligent detection of ground wire breakage by using the above-mentioned scheme, overcomes the defects of strong subjectivity and easy environmental influence in the prior art by using point cloud data, reduces the misjudgment and missed detection of manual inspection, can process thousands of base towers per analysis, and the detection speed is more than 10 times faster than manual inspection, which is suitable for intelligent operation and maintenance of large-scale power transmission lines. At the same time, the scheme of the present disclosure realizes a breakage identification accuracy of more than 90% through dynamic slicing and baseline fitting; and the abnormality detection parameters in the present disclosure are open to adapt to different voltage levels and ground wire types of line scenes; in addition, the detection result of the present disclosure directly interfaces with the operation and maintenance system to realize rapid positioning of ground wire breakage faults and rapid repair of defects.

[0039] It should be understood that the content described in this part is not intended to identify key or important features of the embodiments of the present disclosure, nor to limit the scope of the present disclosure. Other features of the present disclosure will become apparent through the following description. BRIEF DESCRIPTION OF DRAWINGS

[0040] The accompanying drawings are used to better understand the present scheme and do not constitute a limitation on the present disclosure. Among them:

[0041] Figure 1 is a flowchart of the ground wire breakage automatic analysis method based on point cloud data according to the present disclosure;

[0042] Figure 2 is an operation schematic diagram of starting ground wire breakage analysis according to an embodiment of the present disclosure;

[0043] Figure 3 is an abnormality detection parameter setting page schematic diagram according to an embodiment of the present disclosure;

[0044] Figure 4 is a detection result viewing page schematic diagram according to an embodiment of the present disclosure;

[0045] Figure 5 This is a schematic diagram of the exported data page according to an embodiment of this disclosure;

[0046] Figure 6 This is a schematic diagram of data derived according to embodiments of this disclosure;

[0047] Figure 7 This is a schematic diagram illustrating the addition of the breakage location point according to an embodiment of this disclosure;

[0048] Figure 8 This is a schematic diagram after adding the break point according to the embodiments of this disclosure;

[0049] Figure 9 This is a schematic diagram of the structure of the automatic ground wire strand breakage analysis device based on point cloud data according to this disclosure. Detailed Implementation

[0050] The exemplary embodiments of this disclosure are described below with reference to the accompanying drawings, including various details of the embodiments to aid understanding, and should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of this disclosure. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0051] This disclosure addresses the shortcomings of current ground wire strand breakage detection methods, such as insufficient automation and intelligence, susceptibility to environmental influences, and low efficiency in implementation from detection to maintenance. It proposes an automatic ground wire strand breakage analysis method and device based on point cloud data. The technical solution of this disclosure achieves automated and intelligent detection of ground wire strand breaks. By utilizing point cloud data, it overcomes the inherent subjectivity and susceptibility to environmental influences in existing technologies, reducing misjudgments and missed detections caused by manual inspections. A single analysis can process data from thousands of towers, increasing detection speed by more than 10 times compared to manual methods, making it suitable for intelligent maintenance of large-scale transmission lines. Furthermore, the solution of this disclosure achieves a strand breakage identification accuracy of over 90% through dynamic slicing and baseline fitting. The anomaly detection parameters in this disclosure are adaptable to different voltage levels and ground wire types. Additionally, the detection results of this disclosure are directly integrated with the maintenance system, enabling rapid location of ground wire strand breakage faults and rapid defect repair.

[0052] The technical solution of this disclosure will be described below through specific embodiments.

[0053] like Figure 1 The diagram shown is a flowchart of the automatic ground wire breakage analysis method based on point cloud data in this embodiment. The execution subject of this embodiment is a computing device or component with data processing capabilities. Specifically, the method of this embodiment may include the following steps:

[0054] S110, obtain original point cloud data of the power transmission line, and perform classification processing on the original point cloud data to obtain ground wire point cloud data.

[0055] In the present disclosure, broken strand recognition is performed for the ground wire, and therefore the "ground wire" category in the original point cloud data is selected by default; in addition, the technical concept of the present disclosure supports detection of other types of lines, and therefore manual adjustment is supported to adapt to different detection scenarios.

[0056] The scheme of the present embodiment starts to execute in response to a start operation of a user, for example, as shown in Figure 2 The user clicks "ground wire broken strand analysis" to start execution of the method of the present embodiment.

[0057] The present step can also include checking the classification quality of the ground wire category, and manually modifying the misclassified points.

[0058] S120, obtain abnormality detection parameters; wherein the abnormality detection parameters include point cloud slice thickness, height difference threshold, detection direction, and starting distance.

[0059] The above abnormality detection parameters support user adjustment, as shown in Figure 3 The user can adjust the following key parameters to adapt to different line environments and detection requirements:

[0060] Starting distance (default 15m): exempted range before and after the tower, to avoid false positives due to interference of the tower structure. Slice thickness (default 3.5m): horizontal slicing along the ground wire direction, the thinner the slice, the higher the detection accuracy, but the calculation amount increases. Height difference threshold (default 0.5m): height deviation threshold of the broken strand position point and the ground wire baseline, exceeding the threshold is determined as an abnormality. Detection direction: by default, only the ground wire points below are detected (the broken strand is usually downward), the input of this parameter can be checked by the user "detect ground wire below" to check special abnormalities. Of course, the detection direction can also include the ground wire below and the bottom line above.

[0061] In addition, as shown in Figure 3 The default is not to check the ground wire above, and if you want to detect the ground wire above, you can check it.

[0062] S130, mark the tower and pole, and perform section segmentation on the ground wire point cloud data based on the starting distance and the point cloud slice thickness to obtain a plurality of point cloud slices.

[0063] Here specifically can be between two tower poles, with the distance greater than the starting distance of the ground wire point cloud data, according to the point cloud slice thickness along the line direction, the ground wire point cloud data is segmented to obtain a plurality of point cloud slices. For example, the ground wire point cloud data is segmented into a plurality of horizontal point cloud slices along the line direction according to the point cloud slice thickness (such as 3.5m).

[0064] The present disclosure marks the tower and segment of the ground wire point cloud data, and ensures that the ground wire point cloud data is correctly classified according to the line direction.

[0065] S140, for each point cloud slice, the baseline fitting is carried out by using the point cloud data in the point cloud slice to obtain the ground wire baseline; according to the detection direction, the point cloud data in the preset distance range is extracted with the ground wire baseline as the reference, and the vertical distance of each point cloud extracted from the ground wire baseline is calculated; the point cloud with a vertical distance greater than the height difference threshold is regarded as a broken strand abnormal point; the broken strand abnormal points are clustered, the noise points are removed based on the clustering result, and the broken strand position is determined based on the remaining broken strand abnormal points.

[0066] In this step, the ground wire point cloud data in each point cloud slice is linearly fitted to generate a ground wire baseline (representing the normal form of the ground wire), and then all point cloud data within a preset distance range (1.5m) above and below the ground wire baseline is extracted. The vertical distance of each point cloud from the ground wire baseline is calculated, and if it exceeds the height difference threshold (such as 0.5m), it is marked as a broken strand abnormal point. Then the result is aggregated: the adjacent broken strand abnormal points are clustered, the isolated noise points are excluded, and then the broken strand position (horizontal distance from the small tower) is output.

[0067] The linear fitting method is specifically: the least square method is used to linearly fit the point cloud of each point cloud slice to determine the ground wire baseline, such as y = kx + b, as the basis for judging the broken strand abnormal point.

[0068] The height difference threshold determination is specifically: the vertical distance d of each point P(x, y) in the extracted point cloud to the ground wire baseline is calculated:

[0069]

[0070] If d> height difference threshold, it is determined as a broken strand abnormal point.

[0071] In addition, the present embodiment can also include the following misjudgment optimization strategy:

[0072] Direction filtering: the direction of the abnormal broken strand point is counted, and the upward point cloud (mostly noise or auxiliary equipment) is removed.

[0073] Noise suppression: isolated points are removed by spatial clustering (such as DBSCAN), and only continuous abnormal areas are retained.

[0074] Quality check: combined with point cloud density and local smoothness, to exclude false positives of low quality data segments.

[0075] The above scheme realizes the function of automatic analysis of ground wire breakage, which is based on laser point cloud data and automatically detects whether the ground wire (overhead ground wire or OPGW optical cable) of a power transmission line has breakage defects through intelligent algorithms. This function can greatly improve the detection efficiency, reduce false positives and missed detections of manual inspection, and is suitable for intelligent operation and maintenance of large-scale power transmission lines.

[0076] In some embodiments, after detecting the breakage abnormal point, the ground wire breakage automatic analysis method based on point cloud data of the present disclosure further includes the following steps:

[0077] Step one, generate and display a three-dimensional point cloud image with a breakage position point based on the breakage position and the ground wire point cloud data; wherein the three-dimensional point cloud image further includes a breakage position point list. The breakage position point list includes information of the breakage position point.

[0078] Step two, in response to a click operation on a target breakage position point in the breakage position list, highlight the corresponding breakage position point in the three-dimensional point cloud image and display the characteristic information of the breakage position point; the characteristic information includes the ground wire type and the breakage position.

[0079] The above click operation can be a double-click operation, as shown in Figure 4 After the double-click operation, the distance between the corresponding breakage position point and the small tower is displayed, and in addition, the ground wire name, the name of the section where the breakage position point is located, etc. can also be displayed.

[0080] Step three, in response to a click operation on the review button in the three-dimensional point cloud image, display a revision page for revising the breakage position point; the revision page includes a revision component for revising each breakage position point and an addition component for adding a new breakage position point; the revision component includes a modification component and a deletion component.

[0081] Step four, in response to the operation of the revision component, delete the characteristic information of the breakage position point or obtain new characteristic information of the breakage position point, and synchronously update the three-dimensional point cloud image.

[0082] Step five, in response to the operation of the addition component, obtain the characteristic information of the newly added breakage position point, and synchronously update the three-dimensional point cloud image based on the characteristic information of the newly added breakage position point.

[0083] As shown in Figure 7 , 8As shown, the user can manually add the broken strand position point mark. If some broken strand position points need to be manually marked, click the "Add" button, select the section name, and click the broken strand position point.

[0084] In some embodiments, the following steps can also be implemented: exporting the feature information of each broken strand position point in a predetermined format. For example, as shown in Figure 5 、 6 the user clicks the export button to export the CSV table report.

[0085] The above embodiments implement the following functions:

[0086] Visual interaction: double-click the list to locate the broken strand position point, and the 3D point cloud view highlights the abnormal position.

[0087] Report generation: support exporting CSV format report, including broken strand position, severity and other key information.

[0088] Manual review: provides manual marking function to allow operation and maintenance personnel to supplement suspected broken strand points.

[0089] In some embodiments, in order to overcome the defects of low detection and operation efficiency in the prior art, based on the detection result, the following steps can also be performed to quickly perform the ground wire broken strand operation and maintenance processing:

[0090] Step one, input the three-dimensional point cloud image and the feature information of each broken strand position point into the trained point cloud operation and maintenance model.

[0091] Step two, the point cloud operation and maintenance processing sub-model in the point cloud operation and maintenance model processes the three-dimensional point cloud image to obtain a first operation and maintenance feature map.

[0092] Step three, the feature operation and maintenance sub-model in the point cloud operation and maintenance model processes the feature information of each broken strand position point to obtain a second operation and maintenance feature map.

[0093] Step four, the operation and maintenance comprehensive processing sub-model in the point cloud operation and maintenance model performs feature extraction and fusion processing on the first operation and maintenance feature map and the second operation and maintenance feature map to obtain operation and maintenance decision data.

[0094] Step five, transmit the operation and maintenance decision data to the operation and maintenance system.

[0095] The above point cloud operation and maintenance model is obtained by using the three-dimensional point cloud image and the feature information of each broken strand position point as samples, and then training the operation and maintenance decision data.

[0096] In order to improve the accuracy of operation and maintenance and adapt to changes in the scene, the point cloud operation and maintenance model will be continuously supervised and trained, and the steps are as follows:

[0097] Step one, obtain the operation and maintenance execution data fed back by the operation and maintenance system.

[0098] Step two, extract the completion information in the operation and maintenance execution data, and in the case that the completion information indicates that the repair is completed, reacquire the three-dimensional point cloud data in the point cloud slice range corresponding to the broken strand position point, and classify the three-dimensional point cloud data to obtain a repaired ground wire point cloud. The repaired ground wire point cloud is used for baseline fitting to obtain a repaired ground wire baseline. According to the detection direction, the point cloud data in a preset distance range is extracted with the repaired ground wire baseline as the reference, and the perpendicular distance of each point cloud extracted to the repaired ground wire baseline is calculated. The point cloud with a perpendicular distance greater than the height difference threshold is taken as a repaired broken strand abnormal point. The repaired broken strand abnormal points are clustered, and the noise points are removed based on the clustering result. The remaining repaired broken strand abnormal points are used to determine the repaired broken strand position. In the case that the repaired broken strand position does not include the broken strand position point, the operation and maintenance operation information is extracted from the operation and maintenance execution data.

[0099] Step three, match the operation and maintenance operation information with the operation and maintenance decision data. If the matching is unsuccessful, the operation and maintenance operation information is used to retrain the point cloud operation and maintenance model.

[0100] In the actual operation and maintenance process, the operation and maintenance system may directly perform operation and maintenance repair based on the operation and maintenance decision data, may perform operation and maintenance repair using a manually set repair decision, or may perform operation and maintenance repair based on the operation and maintenance decision data and the manually set repair decision. These are all fed back by the operation and maintenance system in the form of operation and maintenance execution data. Therefore, the operation and maintenance execution data fed back is processed to improve the operation and maintenance accuracy of the point cloud operation and maintenance model.

[0101] The above point cloud operation and maintenance processing sub-model includes a fault classification sub-model and an operation and maintenance feature processing sub-model. The three-dimensional point cloud image is processed using the fault classification sub-model to obtain a fault classification feature map. The fault classification feature map is processed using the operation and maintenance feature processing sub-model to obtain a first operation and maintenance feature map. The fault classification sub-model includes a lightweight YOLOv10 model.

[0102] The feature operation and maintenance sub-model, the operation and maintenance comprehensive processing sub-model, and the operation and maintenance feature processing sub-model can include a fully connected layer, a convolutional layer, and an attention mechanism. The present disclosure does not limit this.

[0103] The above embodiments of the present disclosure achieve the following technical effects:

[0104] 1. Efficient automation: a single analysis can process thousands of foundation tower data, and the detection speed is more than 10 times higher than manual detection.

[0105] 2. High-precision detection: through dynamic slicing and fitting algorithms, the broken strand recognition accuracy can reach more than 90% (depending on the quality of the point cloud).

[0106] 3. Flexible adjustment: parameter opening adapts to different voltage levels, ground wire types, and line scenes.

[0107] 4. Lightweight output: the results directly interface with the operation and maintenance system, supporting rapid positioning and defect repair.

[0108] Embodiments of the present disclosure are applicable to scenarios such as regular inspection of power transmission lines, rapid investigation after disasters such as lightning strikes and icing, and special detection of hidden dangers in old lines.

[0109] Based on the same inventive concept, the present disclosure provides a ground wire broken strand automatic analysis device based on point cloud data. The components of the device perform the same or similar steps as the above method, and therefore similar parts will not be repeated. As shown in Figure 9 The ground wire broken strand automatic analysis device based on point cloud data of the present embodiment includes:

[0110] The point cloud processing module 910 is configured to obtain original point cloud data of the power transmission line and perform classification processing on the original point cloud data to obtain ground wire point cloud data.

[0111] The parameter acquisition module 920 is configured to acquire abnormality detection parameters. The abnormality detection parameters include point cloud slice thickness, height difference threshold, detection direction, and starting distance.

[0112] The point cloud slicing module 930 is configured to mark the tower pole and perform section segmentation on the ground wire point cloud data based on the starting distance and the point cloud slice thickness to obtain a plurality of point cloud slices.

[0113] The abnormality detection module 940 is configured to, for each point cloud slice, perform baseline fitting on the point cloud data in the point cloud slice to obtain a ground wire baseline; extract point cloud data within a preset distance range according to the detection direction with the ground wire baseline as a reference, and calculate the perpendicular distance between each point cloud and the ground wire baseline; regard the point cloud with a perpendicular distance greater than the height difference threshold as a broken strand abnormal point; cluster the broken strand abnormal points, remove noise points based on the clustering result, and determine the broken strand position based on the remaining broken strand abnormal points.

[0114] Various implementations of the techniques described above can be realized in digital electronic circuitry, integrated circuitry, a field programmable gate array (FPGA), an application specific integrated circuit (ASIC), a system on a chip (SOC), a programmable logic device (CPLD), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include implementation in one or more computer programs that are executable and / or interpretable on a programmable system including at least one programmable processor, which can be special or general purpose, coupled to receive data and instructions from, and to transmit data and instructions to, a storage system, at least one input device, and at least one output device.

[0115] Program code for carrying out methods of the present disclosure can be written in any combination of one or more programming languages. The program code can be provided to a processor or controller of a general purpose computer, special purpose computer, or other programmable data processing apparatus to produce a machine, such that the program code, when executed by the processor or controller, produces the functions / operations specified in the flowcharts and / or the block diagrams. The program code can be executed entirely on a machine, partially on a machine, partially on a machine as a stand-alone software package, partially on a machine and partially on a remote machine or entirely on a remote machine or server.

[0116] In the context of the present disclosure, a machine-readable medium can be a tangible medium that contains or stores a program for use by or in connection with an instruction execution system, apparatus, or device. The machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can include but is not limited to an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples of the machine-readable storage medium will include one or more lines of electrical connections, portable computer disks, hard disk drives, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or Flash memory), optical fibers, portable compact disc read-only memories (CD-ROMs), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0117] To provide for interaction with a user, the systems and techniques described here can be implemented on a computer having a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user and a keyboard and a pointing device (e.g., a mouse or a trackball) by which the user can provide input to the computer. Other kinds of devices can be used to provide for interaction with a user as well; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form, including acoustic, speech, or tactile input.

[0118] The systems and techniques described here can be implemented in a computing system that includes a back end component (e.g., as a data server), or that includes a middleware component (e.g., an application server), or that includes a front end component (e.g., a user computer having a graphical user interface or a Web browser through which a user can interact with an implementation of the systems and techniques described here), or any combination of such back end, middleware, or front end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area network (LAN), a wide area network (WAN), and the Internet.

[0119] The computer system can include clients and servers. A client and server are generally remote from each other and typically interact through a communication network. The relationship of client and server is generally established by computer programs running on the respective computers and having a client-server relationship to each other. The servers can be cloud servers, servers of a distributed system, or servers combined with a blockchain.

[0120] It should be understood that various forms of flow shown above can be used, with steps reordered, added, or removed. For example, the steps recited in the present disclosure can be performed in parallel, in series, or in a different order, without limitation, as long as the desired results of the technology disclosed in the present disclosure are achieved.

[0121] The specific embodiments described above are not intended to limit the scope of the present disclosure. Those skilled in the art will understand that various modifications, combinations, sub-combinations, and alternatives can be made to the specific embodiments without departing from the spirit and principles of the present disclosure. Any further modifications, equivalent substitutions, improvements, and the like, either alone or in some combination or sub-combination, are intended to be included within the scope of the present disclosure.

Claims

1. A ground strand breakage automatic analysis method based on point cloud data, characterized in that, The application comprises: Obtaining original point cloud data of a power transmission line and classifying the original point cloud data to obtain ground wire point cloud data; Obtaining abnormality detection parameters; wherein the abnormality detection parameters include point cloud slice thickness, height difference threshold, detection direction, and starting distance; Marking tower poles and segmenting the ground wire point cloud data based on the starting distance and the point cloud slice thickness to obtain a plurality of point cloud slices; For each point cloud slice, fitting a reference line using point cloud data in the point cloud slice to obtain a ground wire reference line; extracting point cloud data within a preset distance range according to the detection direction with the ground wire reference line as a reference, and calculating the perpendicular distance of each point cloud to the ground wire reference line; taking point cloud with a perpendicular distance greater than the height difference threshold as a broken strand abnormal point; clustering the broken strand abnormal points, removing noise points based on the clustering results, and determining the broken strand position based on the remaining broken strand abnormal points.

2. The method of claim 1, wherein, Further comprising: Generating and displaying a three-dimensional point cloud image with a broken strand position point based on the broken strand position, the ground wire point cloud data; wherein the three-dimensional point cloud image further includes a broken strand position point list; In response to a click operation on a target broken strand position point in the broken strand position point list, highlighting the corresponding broken strand position point in the three-dimensional point cloud image and displaying the feature information of the broken strand position point; the feature information includes the ground wire type and the broken strand position; In response to a click operation on a review button in the three-dimensional point cloud image, displaying a revision page for revising the broken strand position point; the revision page includes a revision component for revising each broken strand position point and an addition component for adding a new broken strand position point; the revision component includes a modification component and a deletion component; In response to an operation on the revision component, deleting the feature information of the broken strand position point or obtaining new feature information of the broken strand position point, and synchronously updating the three-dimensional point cloud image; In response to an operation on the addition component, obtaining the feature information of the newly added broken strand position point, and synchronously updating the three-dimensional point cloud image based on the feature information of the newly added broken strand position point.

3. The method of claim 2, wherein, Further comprising: Exporting the feature information of each broken strand position point in a predetermined format.

4. The method of claim 2, wherein, Further comprising: Inputting the three-dimensional point cloud image and the feature information of each broken strand position point into a trained point cloud operation and maintenance model; A point cloud operation and maintenance processing submodel in the point cloud operation and maintenance model processes the three-dimensional point cloud image to obtain a first operation and maintenance feature map; A feature operation and maintenance submodel in the point cloud operation and maintenance model processes the feature information of each broken strand position point to obtain a second operation and maintenance feature map; An operation and maintenance comprehensive processing submodel in the point cloud operation and maintenance model performs feature extraction and fusion processing on the first operation and maintenance feature map and the second operation and maintenance feature map to obtain operation and maintenance decision data; Transmitting the operation and maintenance decision data to an operation and maintenance system.

5. The method of claim 4, wherein, Further comprising: Obtaining operation and maintenance execution data fed back by the operation and maintenance system; extracting completion information in the operation and maintenance execution data, and in the case that the completion information indicates that the repair is completed, re-acquiring three-dimensional point cloud data in a point cloud slice range corresponding to the broken strand position point, and classifying the three-dimensional point cloud data to obtain a repaired ground wire point cloud, and performing baseline fitting on the repaired ground wire point cloud to obtain a repaired ground wire baseline; taking the repaired ground wire baseline as a reference, extracting point cloud data within a preset distance range according to a detection direction, and calculating the vertical distance of each point cloud extracted from the repaired ground wire baseline; taking the point cloud with a vertical distance greater than the height difference threshold as a repaired broken strand abnormal point; performing clustering on the repaired broken strand abnormal point, removing noise points based on the clustering result, and determining a repaired broken strand position based on the remaining repaired broken strand abnormal points; in the case that the repaired broken strand position does not include the broken strand position point, extracting operation and maintenance operation information from the operation and maintenance execution data; matching the operation and maintenance operation information with the operation and maintenance decision data, and if the matching is unsuccessful, re-training the point cloud operation and maintenance model by using the operation and maintenance operation information.

6. The method of claim 4, wherein, The point cloud operation and maintenance processing sub-model includes a fault classification sub-model and an operation and maintenance feature processing sub-model. The fault classification sub-model includes a lightweight YOLOv10 model.

7. The method of claim 6, wherein, The tower pole is marked, and based on the starting distance and the point cloud slice thickness, the ground wire point cloud data is segmented into multiple point cloud slices, including:

8. The method of claim 1, wherein, The ground wire point cloud data between two tower poles and with a distance greater than the starting distance from the tower pole is segmented into multiple point cloud slices according to the point cloud slice thickness. The ground wire point cloud data is segmented into multiple point cloud slices.

9. The method of claim 1, wherein, The ground wire point cloud data is segmented into multiple point cloud slices. The ground wire point cloud data is segmented into multiple point cloud slices.

10. A ground strand breakage automatic analysis device based on point cloud data, characterized by, including: The point cloud processing module is configured to acquire original point cloud data of the power transmission line, and perform classification processing on the original point cloud data to obtain ground wire point cloud data. The parameter acquisition module is configured to acquire abnormality detection parameters; wherein the abnormality detection parameters include a point cloud slice thickness, a height difference threshold, a detection direction, and a starting distance. The point cloud slice module is configured to mark a tower pole, and based on the starting distance and the point cloud slice thickness, segment the ground wire point cloud data into multiple point cloud slices. The abnormality detection module is configured to, for each point cloud slice, perform baseline fitting on the point cloud data in the point cloud slice to obtain a ground wire baseline; take the ground wire baseline as a reference, extract point cloud data within a preset distance range according to a detection direction, and calculate the vertical distance of each point cloud extracted from the ground wire baseline; take the point cloud with a vertical distance greater than the height difference threshold as a broken strand abnormal point; perform clustering on the broken strand abnormal point, remove noise points based on the clustering result, and determine a broken strand position based on the remaining broken strand abnormal points.

Images