A power transmission corridor regional construction supervision method and system based on unmanned aerial vehicle swarm cooperative operation

Abstract

The present application relates to the technical field of power construction supervision, and more particularly to a power transmission corridor regional construction supervision method and system based on unmanned aerial vehicle group cooperative operation. The method linearly divides the supervision area according to the geographical trend of the power transmission line corridor, divides the line segment according to the supervision area, generates a sub-area, uses an unmanned aerial vehicle to collect images of the construction area, forms a difference comparison with the historical image, identifies changes through pixel-level comparison, outputs a change layer containing position and type information, constructs a power transmission corridor regional construction knowledge graph, makes an abnormality judgment on the changed area, triggers a rule reasoning mechanism to determine whether there is a violation, automatically generates a rectification notice and pushes it, reviews and patrols after rectification to realize a closed loop, dynamically adjusts the flight task priority and frequency according to the violation identification density and rectification progress, realizes differentiated high-frequency patrol and iterative monitoring of high-risk areas, and significantly improves the intelligent level and monitoring efficiency of the strip corridor regional construction supervision.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of power construction supervision, and in particular to a power transmission corridor regional construction supervision method and system based on unmanned aerial vehicle group cooperative operation. BACKGROUND

[0002] At present, the supervision of the construction area mainly relies on manual patrol or phased auditing means based on remote sensing images, which has problems such as large personnel investment, low monitoring accuracy, and delayed response. Especially in power engineering, the power transmission line and its tower site usually present a "linear strip" distribution, passing through various terrain environments, with complex terrain and wide distribution. Manual monitoring has great difficulty in covering and is low in efficiency, and traditional remote sensing means cannot effectively identify and track in real time.

[0003] In recent years, high-resolution image patrol methods based on unmanned aerial vehicles have been widely used in engineering supervision. Although it can better improve the dynamic supervision capability, most of the existing schemes rely on single machine flight and lack of cluster cooperation mechanism, which has problems such as low flight efficiency, uneven task scheduling, and overlapping patrol. At the same time, a large number of unmanned aerial vehicle images need to be uploaded to the cloud for processing, which has problems of bandwidth pressure and response delay, and it is difficult to realize real-time violation detection.

[0004] In addition, the existing image-based violation identification technology mostly uses rule templates or simple image difference methods, which can only detect "changes" and cannot understand whether the change behavior is compliant, and lacks the ability to integrate and judge the semantic information of land use, approval history, and construction phase. As a structured semantic modeling and reasoning technology, knowledge graph technology has shown great ability in intelligent question answering, risk control compliance, urban planning, and other fields in recent years. By constructing a triple relationship graph between land and use, approval, planning control, and combining image recognition labels, cross-modal decision assistance from image "change detection" to "behavior compliance reasoning" can be realized, which significantly improves the accuracy and intelligence level of construction violation identification. However, there is no mature method to introduce knowledge graph into the unmanned aerial vehicle construction supervision system, especially in the application of power corridor type strip area. SUMMARY

[0005] To solve the above technical problems, the present application provides a power transmission corridor regional construction supervision method based on unmanned aerial vehicle group cooperative operation, comprising the following steps:

[0006] S1. Obtain power transmission line GIS data and tower site coordinate data, construct a power transmission corridor construction monitoring buffer zone based on the center line of the power transmission line, and dynamically adjust the power transmission corridor construction monitoring buffer zone by combining the power transmission line terrain data with the digital elevation model to obtain the power transmission corridor construction monitoring area;

[0007] S2. Based on the tower coordinate data, the power transmission corridor construction monitoring area is divided into regions, and a plurality of construction monitoring sub-regions are obtained, the unmanned aerial vehicle monitoring path optimization model of each sub-region is constructed with the unmanned aerial vehicle operation condition parameter as the constraint, and the unmanned aerial vehicle monitoring path of each sub-region is calculated with the unmanned aerial vehicle minimum flight time and the unmanned aerial vehicle maximum monitoring coverage rate as the target;

[0008] S3. According to the unmanned aerial vehicle monitoring path of each sub-region, based on the unmanned aerial vehicle group cooperative patrol mechanism, image data of each construction monitoring sub-region is collected by the unmanned aerial vehicle group, and a power transmission corridor regional monitoring image data set is constructed;

[0009] S4. According to the power transmission corridor regional monitoring image data set, the semantic feature labeling and feature recognition are carried out through the lightweight deep learning algorithm, and the feature comparison is carried out with the power transmission corridor regional historical image data, and the abnormal feature data set of the power transmission corridor region is obtained;

[0010] S5. The preset power transmission corridor regional land data and standard working condition data are obtained, the power transmission corridor regional construction knowledge graph is constructed, the feature comparison is carried out based on the rule reasoning algorithm and the abnormal feature data set, the abnormal construction data set of the power transmission corridor region is obtained, the power transmission corridor regional construction rectification report is generated and pushed to the construction unit for rectification;

[0011] S6. The rectification frequency data, rectification progress data and construction area sensitivity data of each construction unit are obtained, a construction violation risk value calculation model is constructed, the construction violation risk level of each construction monitoring sub-region is calculated, and the unmanned aerial vehicle group cooperative monitoring of each construction monitoring sub-region is carried out based on the construction violation risk level calculation until the construction rectification is completed.

[0012] The application also provides a power transmission corridor regional construction monitoring system based on unmanned aerial vehicle group cooperative operation, which is realized based on the power transmission corridor regional construction monitoring method based on unmanned aerial vehicle group cooperative operation described above, and comprises a multi-source data acquisition and fusion module for acquiring power transmission line GIS data and tower coordinate data, constructing a power transmission corridor construction monitoring buffer zone based on the center line of the power transmission line, and dynamically adjusting the power transmission corridor construction monitoring buffer zone by combining the power transmission line terrain data with the digital elevation model to obtain the power transmission corridor construction monitoring region.

[0013] The monitoring region division and path generation module is used for dividing the power transmission corridor construction monitoring region based on the tower coordinate data, obtaining a plurality of construction monitoring sub-regions, constructing the unmanned aerial vehicle monitoring path optimization model of each sub-region with the unmanned aerial vehicle operation condition parameter as the constraint, and calculating the unmanned aerial vehicle monitoring path of each sub-region with the unmanned aerial vehicle minimum flight time and the unmanned aerial vehicle maximum monitoring coverage rate as the target;

[0014] The unmanned aerial vehicle group control and monitoring module is used for image data collection of each construction monitoring sub-region by the unmanned aerial vehicle group based on a distributed cooperative control algorithm according to the unmanned aerial vehicle monitoring path of each sub-region, and a power transmission corridor region monitoring image dataset is constructed.

[0015] The image abnormal data comparison module performs semantic feature labeling and feature identification through a lightweight deep learning algorithm according to the power transmission corridor region monitoring image dataset, and performs feature comparison with historical image data of the power transmission corridor region, so as to obtain an abnormal feature dataset of the power transmission corridor region.

[0016] The knowledge graph construction and construction abnormality monitoring module is used for obtaining preset power transmission corridor region plot data and standard working condition data, constructing a power transmission corridor region construction knowledge graph, and performing feature comparison based on a rule reasoning algorithm and the abnormal feature dataset, so as to obtain an abnormal construction dataset of the power transmission corridor region, generate a power transmission corridor region construction rectification report, and push the report to a construction unit for rectification.

[0017] The construction rectification monitoring module is used for obtaining rectification frequency data, rectification progress data and construction region sensitivity data of each construction unit, constructing a construction violation risk value calculation model, calculating a construction violation risk level of each construction monitoring sub-region, and performing unmanned aerial vehicle group cooperative monitoring on each construction monitoring sub-region until the construction rectification is completed.

[0018] The method and system fully exert the advantages of the unmanned aerial vehicle group and the reasoning ability of the knowledge graph, realize the whole-process closed-loop management of the illegal construction from discovery, pushing, rectification to review, and significantly improve the intelligent level and response efficiency of the strip corridor region construction supervision. BRIEF DESCRIPTION OF DRAWINGS

[0019] Figure 1 is a power transmission corridor region construction supervision method flowchart based on unmanned aerial vehicle group cooperative operation.

[0020] Figure 2 is a structural schematic diagram of a power transmission corridor regional construction supervision system based on unmanned aerial vehicle swarm cooperative operation of the present application.

[0021] Figure 3 is a regulatory area determination flowchart of an embodiment of the present application.

[0022] Figure 4 is a UAV swarm cooperative patrol flowchart of an embodiment of the present application.

[0023] Figure 5 is a real-time processing flowchart of aerial survey data of an embodiment of the present application.

[0024] Figure 6 is a violation identification and closed-loop management flowchart of an embodiment of the present application.

[0025] Figure 7 is a dynamic patrol strategy optimization flowchart of an embodiment of the present application. DETAILED DESCRIPTION

[0026] In order to make the objects, technical solutions and advantages of the present application clearer, the present application will be further described in detail with reference to the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and do not limit the present application, i.e., the described embodiments are only a part of the embodiments of the present application, but not all the embodiments. The components of the embodiments of the present application generally described and shown in the drawings can be arranged and designed in various different configurations.

[0027] Therefore, all other embodiments obtained by those skilled in the art based on the embodiments of the present application without creative labor fall within the scope of protection of the present application. It should be noted that the relationship terms such as "first" and "second" and the like are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between the entities or operations.

[0028] Moreover, the term "comprising" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed or inherent to such a process, method, article or device.

[0029] Embodiment 1:

[0030] The present application provides a power transmission corridor regional construction supervision method based on unmanned aerial vehicle swarm cooperative operation, comprising the following steps:

[0031] S1. Obtain the GIS data of the power transmission line and the tower site coordinate data, construct a power transmission corridor construction monitoring buffer zone based on the center line of the power transmission line, and dynamically adjust the power transmission corridor construction monitoring buffer zone by combining the digital elevation model with the power transmission line terrain data to obtain the power transmission corridor construction monitoring area;

[0032] S2. Divide the power transmission corridor construction monitoring area into multiple construction monitoring sub-areas based on the tower site coordinate data, construct a UAV monitoring path optimization model for each sub-area based on the UAV operation condition parameters, and calculate the UAV monitoring path of each sub-area based on the minimum flight time of the UAV and the maximum monitoring coverage rate of the UAV as the target;

[0033] S3. According to the UAV monitoring path of each sub-area, based on the UAV group cooperative patrol mechanism, the image data of each construction monitoring sub-area is collected by the UAV group to construct a power transmission corridor area monitoring image data set;

[0034] S4. According to the power transmission corridor area monitoring image data set, the semantic feature labeling and feature recognition are performed through the lightweight deep learning algorithm, and the feature comparison is performed with the historical image data of the power transmission corridor area to obtain the abnormal feature data set of the power transmission corridor area;

[0035] S5. Obtain the preset power transmission corridor area plot data and standard working condition data, construct a power transmission corridor area construction knowledge graph, and perform feature comparison based on the rule reasoning algorithm and the abnormal feature data set to obtain the abnormal construction data set of the power transmission corridor area, generate a power transmission corridor area construction rectification report and push it to the construction unit for rectification;

[0036] S6. Obtain the rectification frequency data, rectification progress data and construction area sensitivity data of each construction unit, construct a construction risk value calculation model, calculate the construction risk level of each construction monitoring sub-area, and perform UAV group cooperative monitoring on each construction monitoring sub-area based on the construction risk level calculation until the construction rectification is completed.

[0037] Further, step S1 includes the following sub-steps:

[0038] S101. Obtain the GIS data of the power transmission line and the tower site coordinate data, and perform three-dimensional space registration in the GIS platform to construct a three-dimensional space scene containing the geometric features of the power transmission line, the terrain features of the power transmission line and the object features of the power transmission line;

[0039] S102. According to the geometric feature data of the power transmission line, take the center line of the power transmission line as the reference, calculate the basic buffer width of the monitoring area according to the voltage grade data of the power transmission line, and obtain the first monitoring corridor boundary;

[0040] S103. According to the power transmission line terrain feature data, based on the pre-set power transmission line terrain safety data, the first monitoring corridor boundary is dynamically adjusted through the digital elevation model to obtain a second monitoring corridor boundary;

[0041] S104. According to the power transmission line terrain feature data, based on the pre-set power transmission line terrain safety distance data, the second monitoring corridor boundary is corrected in safety distance avoidance combined with the terrain contour, and the range of the corrected corridor boundary is fused and spliced with the first monitoring corridor boundary through a spatial Boolean algorithm to obtain a power transmission corridor construction monitoring area.

[0042] Specifically, the implementation principle flow of each sub-step in the above embodiment is as follows:

[0043] As Figure 3 , first, input the power transmission line GIS data, including power transmission line coordinates, tower position, voltage level and design safety distance parameters, and combine DEM terrain data and terrain vector data to complete spatial registration of multi-source data and build a three-dimensional space scene containing line geometric features, terrain undulations and surrounding terrain; Then, taking the line centerline as the reference, generate a basic buffer zone according to the voltage level setting initial buffer width, and correct the complex terrain based on the slope and slope calculated by DEM, when the slope is greater than 25°, the upslope side is expanded by 5-10 meters, the downslope side is reduced by 2-5 meters, and when crossing a river or canyon, the two sides are extended by 1.5 times the tower height distance along the crossing point; Next, buildings, trees and other structures within the safety distance of the line are processed for terrain avoidance according to the safety distance, and Boolean operation is performed with the basic buffer zone to eliminate the monitoring blind area; Finally, output the monitoring corridor vector polygon data with attributes such as line section ID, voltage level and monitoring accuracy requirements, etc., to provide accurate spatial range basis for subsequent patrol tasks.

[0044] Specifically, Geographic Information System (Geographic Information System or Geo-Informationsystem, GIS) is a specific and very important spatial information system. It is a technical system that collects, stores, manages, calculates, analyzes, displays and describes the spatial information of the entire or part of the earth's surface (including the atmosphere) under the support of computer hardware and software system. Digital Elevation Model (Digital Elevation Model, DEM) is a digital simulation of the terrain (i.e. digital expression of the terrain surface form) realized by limited terrain elevation data. It is a solid ground model that represents the ground elevation in the form of an ordered numerical array.

[0045] Further, step S2 includes the following sub-steps:

[0046] S201. dividing the power transmission corridor construction monitoring area into a plurality of construction monitoring sub-areas based on the tower location coordinate data;

[0047] S202. obtaining drone endurance time data, flight speed data, camera resolution data, and visual range data, and establishing a drone performance parameter matrix;

[0048] S203. taking the boundary of the power transmission corridor construction monitoring area as a reference, and establishing a drone obstacle avoidance parameter matrix according to the drone performance parameter matrix;

[0049] S204. respectively dispersing each construction monitoring sub-area into a node network based on a graph model construction algorithm, and respectively constructing a drone monitoring path optimization model corresponding to each construction monitoring sub-area node network by taking the drone performance parameter matrix as the edge weight value and the drone obstacle avoidance parameter matrix as the node value;

[0050] S205. according to the drone monitoring path optimization model corresponding to each node network, taking the minimum flight time of the drone and the highest monitoring coverage rate of the drone as the target, and through Dijkstra algorithm or A * algorithm to obtain the drone monitoring path of each construction monitoring sub-area.

[0051] Specifically, the implementation principle flow of each sub-step in the above embodiment is as follows:

[0052] As Figure 4 , first, the line is divided into N tower-to-tower sections according to the location of the power transmission tower, and the coordinate sequence of the center line of each section is extracted as the basic path skeleton; then, a path planning constraint condition model is established, combined with the performance parameter constraints of the maximum endurance time of the drone, the maximum flight speed of the drone, and the camera resolution and maximum visual range of the drone camera, the optimal flight height is calculated through the formula H = focal length x ground distance / sensor size to meet the monitoring accuracy requirement according to the minimum shooting resolution GSD≤10cm, and a safety flight channel is formed by shrinking 10% within the monitoring corridor boundary, and the obstacles such as trees and buildings are converted into polygonal no-fly zones; then, the tower-to-tower sections are dispersed into a node network containing coordinate and flight height constraints, and the flight time or energy consumption considering the influence of wind speed is taken as the edge weight value to construct a path optimization model, and Dijkstra algorithm is used to generate an optimal inspection path containing reciprocating routes and hovering shooting points with the target of "full coverage monitoring + shortest flight time"; finally, the three-dimensional space path coordinate sequence (X, Y, Z) of each tower-to-tower section and the corresponding control parameters such as flight speed and shooting trigger point are output, providing precise execution instructions for drone flight control.

[0053] Further, the unmanned aerial vehicle group in step S3 comprises a plurality of unmanned aerial vehicle subgroups, the unmanned aerial vehicle subgroups comprise master unmanned aerial vehicles and slave unmanned aerial vehicles, and correspond to respective construction monitoring sub-regions one by one, the master unmanned aerial vehicles and the slave unmanned aerial vehicles perform real-time transmission of unmanned aerial vehicle operation condition parameters through Mesh ad hoc networks.

[0054] Further, the unmanned aerial vehicle group cooperative patrol mechanism comprises a multi-unmanned aerial vehicle cluster scheduling sub-mechanism and an adaptive patrol and obstacle avoidance sub-mechanism; the multi-unmanned aerial vehicle cluster scheduling sub-mechanism divides each construction monitoring sub-region into a plurality of task groups according to an unmanned aerial vehicle performance parameter matrix through a block parallel and dynamic load balancing algorithm, takes a master unmanned aerial vehicle as a center node and a slave unmanned aerial vehicle as a neighboring node, and performs real-time transmission of unmanned aerial vehicle operation condition parameters of the slave unmanned aerial vehicle to the master unmanned aerial vehicle through a Mesh ad hoc network.

[0055] Further, the adaptive patrol and obstacle avoidance sub-mechanism performs flight attitude control of the master unmanned aerial vehicle and the master-slave unmanned aerial vehicles according to an unmanned aerial vehicle obstacle avoidance parameter matrix and based on a digital elevation model through a PID algorithm, performs terrain obstacle avoidance control of the master unmanned aerial vehicle and the master-slave unmanned aerial vehicles through a V-Graph algorithm to generate a local obstacle avoidance path, performs real-time monitoring of relative positions of the master unmanned aerial vehicle and the slave unmanned aerial vehicle through GPS and ultra-wideband positioning technology, and publishes a unified timestamp by the master unmanned aerial vehicle, and the slave unmanned aerial vehicle performs master-slave unmanned aerial vehicle obstacle avoidance control according to the timestamp.

[0056] Specifically, the sub-steps of step S3 in the above embodiment and the implementation principle flow are as follows:

[0057] As Figure 4Firstly, according to the monitoring corridor section division task unit, combined with the number of unmanned aerial vehicles and endurance, the "block parallel + dynamic load balancing" strategy is adopted to allocate N inter-tower sections to M task groups (5-10 sections per group, time difference ≤15%), and the main unmanned aerial vehicle is responsible for global scheduling, and the slave unmanned aerial vehicle synchronizes the state data in real time through Mesh ad hoc network to realize cluster cooperation. Meanwhile, for the case where the length of the straight section exceeds 2 km, a virtual node is set at the midpoint of the tower to facilitate shooting from both sides; Secondly, in the flight process, the parameters are adjusted adaptively based on weather conditions and terrain features. When the wind speed is greater than 10 m / s, the flight speed is reduced to Vmax×80% and the height is increased by 5%, and in thin fog weather, the shooting interval is shortened to improve the image overlap rate. When the short distance elevation exceeds 10 meters, the DEM data and PID algorithm are used to climb in advance to maintain a relative height difference of ≤2 meters; Thirdly, the three-dimensional grid map and Voronoi diagram algorithm are used to generate a local obstacle avoidance path that maintains a safety distance of ≥2 meters from obstacles, and in a multi-machine environment, obstacle coordinates are broadcast and priority sorting is used to realize cooperative obstacle avoidance; Finally, by setting the multi-machine horizontal distance ≥50 meters and the vertical height difference ≥10 meters, and combining GPS and ultra-wideband positioning technology to monitor in real time, the speed compensation algorithm is triggered to ensure safety, and in the cross-section task, the "main path synchronization + sub-path independent" mode is adopted, and the unified timestamp and synchronization point are used to calibrate the position to ensure the consistency of the image acquisition angle, thereby improving the multi-machine cooperation accuracy.

[0058] Specifically, Mesh ad hoc network is a special Mesh networking method that emphasizes the autonomous cooperation and self-organizing ability between nodes. In Mesh ad hoc network, nodes communicate with each other through wireless connections and automatically adjust routing paths according to network topology and communication requirements. This self-organizing feature makes Mesh ad hoc network more adaptable and robust in complex environments. Voronoi diagram, also known as Thiessen polygon, is a set of continuous polygons composed of vertical bisectors connecting two adjacent points.

[0059] Further, step S4 includes the following sub-steps:

[0060] S401. According to the image data set of the power transmission corridor area, the construction target feature recognition result is obtained by using the improved YOLOv8n model for construction target feature recognition.

[0061] S402. According to the construction target feature recognition result, the historical image data of the power transmission corridor area is obtained, the pyramid level index is established, and the abnormal feature data set of the power transmission corridor area is obtained through SIFT feature point matching.

[0062] Specifically, the implementation principle flow of each sub-step in the above embodiment is as follows:

[0063] As Figure 5First, the fisheye lens is corrected by Zhang Zhengyou calibration method to ensure the residual distortion <0.5%. Combined with non-local mean filtering, the Gaussian noise is removed to improve the signal-to-noise ratio by ≥15 dB and improve the image quality. The visual SLAM positioning adopts the ORB-SLAM3 algorithm to construct a sparse point cloud map in real time, achieving high-precision positioning with an accuracy of ≤0.3 m on plains and ≤0.5 m in mountainous areas. Multi-source data fusion fuses GPS / IMU data through extended Kalman filter (EKF) to correct positioning errors and output georeferenced images (resolution 0.1 m / pixel) with UTM coordinates, providing a geographic coordinate reference for subsequent analysis.A lightweight model (parameters <10MB) based on YOLOv8n is improved for power construction, which can detect 9 types of targets such as construction machinery and temporary buildings in real time from transmission corridor monitoring images, and generate five-tuple [category, confidence, geographic coordinates, pixel coordinates, size] for each detection target. The Protocol Buffers format is used to compress the data to improve transmission efficiency by 40% and reduce data transmission pressure. Then, a historical image library containing RGB / multispectral images (interval 1 day / time) of the same region in the previous 3 months is constructed and indexed by pyramid level to support fast multi-scale matching and provide historical data support for change detection. Sub-pixel level alignment is achieved through SIFT (Scale-invariant feature transform) feature point matching (combined with RANSAC (Random sample consensus) to remove false matches, and the number of matching pairs is greater than or equal to 20). SSIM (Structure Similarity Index Measure) and MSE (Mean Square Error) are used for change detection, and NDVI (Normalized Difference Vegetation Index) is used to exclude seasonal vegetation changes. Only the changed area ROI with a compression ratio of 10:1 or more and an accurate timestamp is uploaded. Then, Thin Plate Spline (TPS) is used for geometric correction of the image, and ground targets are used for periodic control point calibration to ensure an error of less than 0.2 pixels. At the same time, histogram matching (R / G / B band correlation coefficient >0.95) is performed on images of different time periods to realize radiation normalization and ensure image registration accuracy. Subsequently, multi-level change detection is carried out. First, the brightness mutation area (ΔL>20 and area >5m²) is extracted based on the difference image for rough detection. Then, the texture of the rough detection area is analyzed by LBP (Local Binary Pattern) to identify the characteristics of artificial structures (straight edge density >10 / m²), and combined with land use classification data to shield natural changes, only the changed area of construction land and unused land is retained. Finally, the GeoTIFF format change layer and attribute table are generated to record the change type, geographic coordinates, pixel coordinates, area, change confidence, associated historical image ID, and recommended verification level, etc. to ensure that the area detection error of the plain area is less than or equal to 5%, the position deviation is less than or equal to 0.5m, and the position deviation in the mountainous area is less than or equal to 1m to meet the accuracy requirements of transmission corridor change detection.

[0064] Further, step S5 comprises the following sub-steps:

[0065] S501. Obtain preset power transmission corridor regional land data and construction condition data, and establish a multi-entity association network of land data entities and construction condition data entities to construct a power transmission corridor regional construction knowledge graph;

[0066] S502. Based on a SWRL (Semantic Web Rule Language) rule reasoning algorithm, according to the power transmission corridor regional construction knowledge graph, perform feature compliance comparison on the abnormal feature data set, and take abnormal features that do not meet the requirements of the power transmission corridor regional construction knowledge graph as abnormal construction data to obtain an abnormal construction data set;

[0067] S503. Based on a spatial superposition analysis algorithm, compare the abnormal construction data set with each construction monitoring sub-region, and when the intersection area is greater than or equal to 50%, associate the current abnormal construction data with the current construction monitoring sub-region, and generate a power transmission corridor regional construction rectification report.

[0068] S504. Push the obtained power transmission corridor regional construction rectification report to the corresponding construction unit for rectification.

[0069] Specifically, the implementation principle process of each sub-step in the above embodiment is as follows:

[0070] As Figure 6 , first, obtain preset power transmission corridor regional land data and construction condition data, the power transmission corridor regional land data including land number data, land boundary coordinate data and land use category data, and the construction condition data including construction permit data, basic farmland protection line data, ecological protection line data, construction type data and construction duration data.

[0071] Define the land number data, land boundary coordinate data, land use category data, construction permit data, basic farmland protection line data, ecological protection line data, construction type data and construction duration data as entity types in the graph structure, define the land and use relationship type, land and construction permit relationship type, land and protection area relationship type, land and construction type relationship type, and land and construction duration relationship type as relationship types in the graph structure network, establish a multi-entity association network as a power transmission corridor regional construction knowledge graph.

[0072] Then, based on the SWRL (Semantic Web Rule Language) rule inference algorithm as an inference language to define rules, according to the power transmission corridor regional construction knowledge graph, the feature compliance comparison of the abnormal feature data set is performed, the abnormal feature data that does not meet the requirements of the power transmission corridor regional construction knowledge graph is regarded as abnormal construction data, and the abnormal construction data set is obtained, such as: when the local entity is basic farmland and the construction type entity is concrete hardening, the current land is defined as a land and use relationship violation; when the construction permit entity is an expired construction permit, and the land boundary coordinate entity and the basic farmland protection line entity or the ecological protection line entity exist overlap, the current land is defined as a land and protection area relationship violation. Specifically, during reasoning, the matching time of a single rule inference is less than 100 milliseconds, and at least 1000 pieces of abnormal feature image data can be processed in a single rule inference, which meets the real-time reasoning requirements.

[0073] Then, based on the spatial overlay analysis algorithm, the abnormal construction data set and each construction monitoring sub-region are subjected to spatial overlay analysis, when the abnormal image data in the abnormal construction data set and the image data of each construction monitoring sub-region are greater than or equal to 50%, the current abnormal construction data is associated with the current construction monitoring sub-region, and is positioned as a violation land, and a power transmission corridor regional construction rectification report is generated, which includes the construction violation position, the construction before and after comparison graph data, the construction violation relationship type, the construction rectification period, the construction responsibility unit, and realizes the accurate push of the violation information.

[0074] According to the construction rectification report, the construction unit rectifies the violation construction behavior according to the rectification report, and after the task triggering responsibility unit uploads the rectification certificate, the unmanned aerial vehicle focuses on the patrol of the violation region (the flight height is reduced to 50m, and the resolution is improved to 0.05m / pixel), and obtains high-definition review images.

[0075] The rectification effect verification adopts the DeepLabv3+ model for image semantic segmentation, the judgment standard is that the artificial feature area is reduced by more than 90% and no new construction equipment is added, the closed loop is marked through the review, otherwise it is automatically upgraded to manual checking, and the rectification is ensured to be in place.

[0076] Further, step S6 includes the following sub-steps:

[0077] S601. Obtain the rectification frequency data and rectification progress data of each construction unit, and construct a construction risk value calculation model, which is represented as:

[0078] ;

[0079] wherein R represents a construction risk value, D represents a current rectification frequency, L represents a current rectification progress, S represents a construction area sensitivity, and respectively represent weight parameters;

[0080] S602. According to the construction risk value calculation model, the construction risk grade of each construction monitoring sub-area is calculated. When the construction risk value is greater than 0.8, it is a high-risk construction area; when the construction risk value is greater than 0.5 and less than or equal to 0.8, it is a medium-risk construction area; and when the construction risk value is less than or equal to 0.5, it is a low-risk construction area.

[0081] S603. According to the construction risk grade calculation result, the repeated monitoring of each construction monitoring sub-area is performed based on the unmanned aerial vehicle monitoring path of each sub-area, and the feature of the power transmission corridor area construction knowledge graph is updated according to the repeated monitoring result.

[0082] S604. According to the repeated monitoring result, the unmanned aerial vehicle group monitoring path based on the Archimedes spiral is constructed for the construction area with at least two same abnormal construction feature data through the reinforcement learning algorithm, and the unmanned aerial vehicle group monitoring is performed until the construction rectification is completed.

[0083] Specifically, the implementation principle process of each sub-step in the above embodiment is as follows:

[0084] For example, Figure 7 ,

[0085] First, the risk value calculation model is constructed wherein R represents a construction risk value, D represents a current rectification frequency, L represents a current rectification progress, S represents a construction area sensitivity, and respectively represent weight parameters; the risk assessment model is established by comprehensively considering the current illegal density, rectification progress delay rate, and number of adjacent sensitive targets, and the regulatory risk of different areas is quantitatively calculated. When the construction risk value is greater than 0.8, it is a high-risk construction area; when the construction risk value is greater than 0.5 and less than or equal to 0.8, it is a medium-risk construction area; and when the construction risk value is less than or equal to 0.5, it is a low-risk construction area.

[0086] Then, the differentiated scheduling strategy is formulated according to the risk value division level, wherein the high-risk area is patrolled 3 times a day and the flight height is reduced to 80 m, the camera frame rate is increased to 30 fps, the medium-risk area is patrolled 2 times a day and the multi-spectral detection is enabled, and the low-risk area maintains the regular patrol once a day and dynamically allocates resources to the newly accessed high-risk area, thereby realizing intelligent patrol scheduling.

[0087] Next, the illegal identification result is fed back to the target detection model regularly, the training data set is updated every quarter and typical illegal samples are added, so as to improve the identification ability of the model to hidden construction such as underground foundation excavation, and realize the iterative optimization of the identification model.

[0088] Subsequently, the repeated violation area is generated by a reinforcement learning algorithm (Proximal Policy Optimization, PPO) to generate a spiral encryption unmanned aerial vehicle group monitoring path based on an Archimedes spiral, and the route is dynamically optimized to improve the patrol coverage and accuracy.

[0089] Finally, the recurrence rate of each area violation and the timely rate of rectification and other indicators are regularly counted, the results are input into the intelligent scheduling system, and the risk assessment model parameters are adjusted by dynamic weight to make the patrol strategy continuously adapt to the actual supervision demand until the construction rectification is completed.

[0090] Embodiment 2

[0091] As shown in Figure 2 Embodiment 2 of the present application also provides a power transmission corridor regional construction supervision system based on unmanned aerial vehicle group cooperative operation, which is realized based on the power transmission corridor regional construction supervision method based on unmanned aerial vehicle group cooperative operation in Embodiment 1, comprising a multi-source data acquisition and fusion module for acquiring power transmission line GIS data and tower site coordinate data, constructing a power transmission corridor construction monitoring buffer zone based on the center line of the power transmission line, and dynamically adjusting the power transmission corridor construction monitoring buffer zone by combining the power transmission line terrain data with the digital elevation model to obtain the power transmission corridor construction monitoring region.

[0092] A monitoring region division and path generation module is used to divide the power transmission corridor construction monitoring region based on the tower site coordinate data to obtain a plurality of construction monitoring sub-regions, to construct an unmanned aerial vehicle monitoring path optimization model for each sub-region based on the unmanned aerial vehicle operation condition parameters, and to calculate the unmanned aerial vehicle monitoring path of each sub-region based on the minimum flight time of the unmanned aerial vehicle and the highest monitoring coverage rate of the unmanned aerial vehicle.

[0093] An unmanned aerial vehicle group control and monitoring module is used to collect image data of each construction monitoring sub-region by the unmanned aerial vehicle group based on the distributed cooperative control algorithm according to the unmanned aerial vehicle monitoring path of each sub-region, and to construct a power transmission corridor regional monitoring image data set.

[0094] An image abnormal data comparison module is used to perform semantic feature labeling and feature recognition by a lightweight deep learning algorithm based on the power transmission corridor regional monitoring image data set, and to compare the features with the historical image data of the power transmission corridor region to obtain an abnormal feature data set of the power transmission corridor region.

[0095] The knowledge graph construction and construction anomaly monitoring module is configured to obtain preset power transmission corridor regional plot data and standard working condition data, construct a power transmission corridor regional construction knowledge graph, and perform feature comparison based on a rule-based reasoning algorithm and an abnormal feature data set to obtain an abnormal construction data set of the power transmission corridor region, generate a power transmission corridor regional construction rectification report, and push the report to a construction unit for rectification.

[0096] The construction rectification monitoring module is configured to obtain rectification frequency data, rectification progress data, and construction region sensitivity data of each construction unit, construct a construction risk value calculation model, calculate the construction risk level of each construction monitoring sub-region, and perform unmanned aerial vehicle group cooperative monitoring on each construction monitoring sub-region based on the construction risk level calculation until the construction rectification is completed.

[0097] Specifically, the specific workflow is as follows:

[0098] First, the system obtains power line GIS data and tower coordinate data through the multi-source data acquisition and fusion module, and constructs a power transmission corridor construction monitoring buffer area based on the power line center line. In combination with a digital elevation model (DEM) and power line terrain data, the monitoring buffer area is dynamically adjusted, and finally a power transmission corridor construction monitoring region covering the actual terrain conditions is generated.

[0099] Then, the monitoring region division and path generation module divides the power transmission corridor construction monitoring region into multiple construction monitoring sub-regions according to the tower coordinate data. In combination with unmanned aerial vehicle operation condition parameters (such as maximum flight time, maximum monitoring range, flight speed, etc.), an unmanned aerial vehicle monitoring path optimization model is established, and the optimal unmanned aerial vehicle monitoring path of each construction monitoring sub-region is calculated with “minimum flight time + highest monitoring coverage” as the objective function.

[0100] Next, the unmanned aerial vehicle group control and monitoring module uses a distributed cooperative control algorithm to dispatch the unmanned aerial vehicle group according to the optimal monitoring path of each sub-region to perform efficient patrol and image acquisition tasks. The unmanned aerial vehicle collects construction image data in real time during covering all sub-regions, and generates a power transmission corridor regional monitoring image data set.

[0101] Subsequently, the image abnormal data comparison module performs lightweight deep learning processing on the collected monitoring image data set to realize semantic feature labeling and target feature recognition. The extracted monitoring features are compared with historical image data of the power transmission corridor region to automatically generate an abnormal feature data set of the power transmission corridor region, providing a basis for subsequent construction compliance judgment.

[0102] On this basis, the knowledge graph construction and construction anomaly monitoring module obtains preset plot data and standard working condition data, and establishes a power transmission corridor construction knowledge graph. Based on the rule reasoning algorithm, the system compares the abnormal feature data set obtained in the previous stage with the knowledge graph, identifies abnormal construction behavior, generates a power transmission corridor regional construction anomaly data set, and automatically generates a construction rectification report, which is pushed to the responsible construction unit to start the rectification process.

[0103] Finally, the construction rectification monitoring module dynamically tracks the rectification of each construction unit, collects rectification frequency data, rectification progress data and construction area sensitivity data, establishes a construction risk value calculation model, and evaluates the risk level of each construction monitoring sub-region. According to the evaluation results, dynamically adjust the monitoring frequency and flight path planning of the UAV group, and intensify the patrol of the key risk areas until the construction rectification is completed, so as to realize the whole-process closed-loop management.

[0104] The application is described from the viewpoints of use purpose, efficiency, progress and novelty, and has practical progressiveness, which meets the function improvement and use requirements emphasized by the patent law. The above description and drawings are only preferred embodiments of the application, and are not limited to the application. Therefore, all similar, identical, equivalent, modified and improved applications within the scope of the application are included in the protection scope of the application.

[0105] The above specific embodiments further detail the purpose, technical solutions and beneficial effects of the application. It should be understood that the above description is only a specific embodiment of the application and does not limit the protection scope of the application. Any modification, equivalent replacement, improvement, etc. within the spirit and principles of the application should be included in the protection scope of the application.

Claims

1. A power transmission corridor regional construction supervision method based on cooperation of a UAV swarm, characterized in that, Comprise the following steps: S1. Obtain the transmission line GIS data and tower site coordinate data, construct a transmission corridor construction monitoring buffer zone based on the transmission line center line, and dynamically adjust the transmission corridor construction monitoring buffer zone through a digital elevation model combined with the transmission line terrain data to obtain a transmission corridor construction monitoring area; S2. Based on the tower site coordinate data, the transmission corridor construction monitoring area is divided into multiple construction monitoring sub-areas, and the UAV monitoring path optimization model of each sub-area is constructed based on the UAV operation condition parameters as the constraint, and the minimum flight time of the UAV and the maximum monitoring coverage rate of the UAV are taken as the target to calculate the UAV monitoring path of each sub-area; S3. According to the UAV monitoring path of each sub-area, based on the UAV group cooperative patrol mechanism, the image data of each construction monitoring sub-area is collected by the UAV group to construct a transmission corridor area monitoring image data set; S4. According to the transmission corridor area monitoring image data set, the semantic feature labeling and feature recognition are carried out through the lightweight deep learning algorithm, and the feature comparison is carried out with the historical image data of the transmission corridor area to obtain the abnormal feature data set of the transmission corridor area; S5. Obtain the preset transmission corridor area plot data and standard working condition data, construct a transmission corridor area construction knowledge graph, and compare the features based on the rule reasoning algorithm and the abnormal feature data set to obtain the abnormal construction data set of the transmission corridor area, generate a transmission corridor area construction rectification report and push it to the construction unit for rectification; S6. Obtain the rectification frequency data, rectification progress data and construction area sensitivity data of each construction unit, construct a construction violation risk value calculation model, calculate the construction violation risk level of each construction monitoring sub-area, and based on the construction violation risk level calculation, the UAV group cooperative monitoring is carried out for each construction monitoring sub-area until the construction rectification is completed; Step S2 comprises the following steps: S201. Based on the tower site coordinate data, the transmission corridor construction monitoring area is divided into multiple construction monitoring sub-areas; S202. Obtain the UAV endurance time data, flight speed data, camera resolution data and visual range data, and establish a UAV performance parameter matrix; S203. Taking the boundary of the transmission corridor construction monitoring area as the reference, the UAV obstacle avoidance parameter matrix is established according to the UAV performance parameter matrix; S204. Based on the graph model construction algorithm, each construction monitoring sub-area is discretized into a node network, and the UAV performance parameter matrix is taken as the edge weight value and the UAV obstacle avoidance parameter matrix is taken as the node value. The UAV monitoring path optimization model corresponding to each construction monitoring sub-area node network is constructed; S205. According to the UAV monitoring path optimization model corresponding to each node network, taking the minimum flight time of the UAV and the maximum monitoring coverage rate of the UAV as the target, the UAV monitoring path of each construction monitoring sub-area is calculated through the Dijkstra algorithm. 2.The power transmission corridor regional construction monitoring method based on the cooperation of the UAV group according to claim 1, characterized in that, Step S1 comprises the following steps: S101. Obtain the GIS data of the power transmission line and the tower coordinate data, and perform three-dimensional space registration in the GIS platform to construct a three-dimensional space scene containing the geometric features of the power transmission line, the topographic features of the power transmission line, and the object features of the power transmission line; S102. According to the geometric feature data of the power transmission line, taking the center line of the power transmission line as the reference, the basic buffer width of the monitoring area is calculated according to the voltage grade data of the power transmission line, and the first monitoring corridor boundary is obtained; S103. According to the topographic feature data of the power transmission line, based on the pre-set topographic safety data of the power transmission line, the first monitoring corridor boundary is dynamically adjusted through the digital elevation model, and the second monitoring corridor boundary is obtained; S104. According to the object feature data of the power transmission line, based on the pre-set object safety distance data of the power transmission line, the second monitoring corridor boundary is corrected by avoiding the safety distance combined with the object contour, and the corrected corridor boundary range is fused and spliced with the first monitoring corridor boundary through the space Boolean algorithm, and the power transmission corridor construction monitoring area is obtained. 3.The power transmission corridor regional construction monitoring method based on the cooperation of the UAV group according to claim 1, characterized in that, The UAV group in step S3 includes multiple UAV subgroups; The UAV subgroups include master UAVs and slave UAVs, and each corresponds to a respective construction monitoring sub-area; The master UAVs and slave UAVs transmit real-time UAV operating condition parameters through Mesh ad hoc networking.

4. The power transmission corridor regional construction monitoring method based on the cooperation of the UAV group according to claim 3, characterized in that, The UAV group cooperative patrol mechanism in step S3 includes a multi-UAV cluster scheduling sub-mechanism, an adaptive patrol and obstacle avoidance sub-mechanism; The multi-UAV cluster scheduling sub-mechanism divides each construction monitoring sub-area into multiple task groups according to the UAV performance parameter matrix through block parallel and dynamic load balancing algorithms, takes the master UAV as the center node and the slave UAV as the adjacent node, and transmits the UAV operating condition parameters of the slave UAV to the master UAV in real time through Mesh ad hoc networking; The adaptive patrol and obstacle avoidance sub-mechanism controls the flight attitude of the master UAV and the master-slave UAV through PID algorithm based on digital elevation model, generates local obstacle avoidance path through V-NO graph algorithm to control the object obstacle avoidance of the master UAV and the master-slave UAV, and monitors the relative position of the master UAV and the slave UAV in real time through GPS and ultra-bandwidth positioning technology, and the master UAV issues a unified timestamp, and the slave UAV controls the master-slave UAV obstacle avoidance according to the timestamp. 5.The power transmission corridor regional construction monitoring method based on the cooperation of the UAV group according to claim 1, characterized in that, Step S4 includes the following steps: S401. According to the power transmission corridor area monitoring image data set, the construction target feature recognition result is obtained by performing construction target feature recognition through the power construction special model based on the improved YOLOv8n; S402. According to the construction target feature recognition result, the power transmission corridor area historical image data is obtained, the pyramid level index is established, and the abnormal feature data set of the power transmission corridor area is obtained through SIFT feature point matching. 6.The power transmission corridor regional construction monitoring method based on the cooperation of the UAV group according to claim 1, characterized in that, Step S5 includes the following steps: S501. Obtain the pre-set power transmission corridor area plot data and construction condition data, and establish a multi-entity association network of plot data entities and construction condition data entities to construct a power transmission corridor area construction knowledge graph; S502. Based on the SWRL rule inference algorithm, the feature compliance comparison is performed on the abnormal feature data set according to the power transmission corridor regional construction knowledge graph, the abnormal feature data that does not meet the requirements of the power transmission corridor regional construction knowledge graph is taken as the abnormal construction data, and an abnormal construction data set is obtained; S503. Based on the spatial overlay analysis algorithm, the abnormal construction data set is compared with each construction monitoring sub-region, and when the intersection area is greater than or equal to 50%, the current abnormal construction data is associated with the current construction monitoring sub-region, and a power transmission corridor regional construction rectification report is generated; S504. The obtained power transmission corridor regional construction rectification report is pushed to the corresponding construction unit for rectification.

7. The power transmission corridor regional construction monitoring method based on the cooperation of the UAV group according to claim 1, characterized in that, Step S6 includes the following steps: S601. Obtain the rectification frequency data and rectification progress data of each construction unit, and construct a construction violation risk value calculation model, which is represented as: ； Wherein, R represents the construction violation risk value, D represents the current rectification frequency, L represents the current rectification progress, and S represents the construction area sensitivity; S602. According to the construction violation risk value calculation model, the construction violation risk level of each construction monitoring sub-region is calculated, when the construction violation risk value is greater than 0.8, it is a high violation risk construction region; when the construction violation risk value is greater than 0.5 and less than or equal to 0.8, it is a medium violation risk construction region; when the construction violation risk value is less than or equal to 0.5, it is a low violation risk construction region; S603. According to the construction violation risk level calculation result, based on the unmanned aerial vehicle monitoring path of each sub-region, each construction monitoring sub-region is repeatedly monitored, and the feature of the power transmission corridor regional construction knowledge graph is updated according to the repeated monitoring result; S604. According to the repeated monitoring result, through the reinforcement learning algorithm, for the construction region with at least two same abnormal construction feature data, the unmanned aerial vehicle group monitoring path based on the Archimedes spiral is constructed and the unmanned aerial vehicle group monitoring is carried out until the construction rectification is completed.

8. A power transmission corridor area construction monitoring system based on cooperative operation of a UAV group, the system being implemented based on the power transmission corridor area construction monitoring method based on cooperative operation of a UAV group according to any one of claims 1-7, characterized in that, It includes: A multi-source data acquisition and fusion module is used to acquire power line GIS data and tower coordinate data, to construct a power transmission corridor construction monitoring buffer zone based on the center line of the power transmission line, and to dynamically adjust the power transmission corridor construction monitoring buffer zone by combining the power transmission line terrain data with the digital elevation model, to obtain the power transmission corridor construction monitoring region; A monitoring region division and path generation module is used to divide the power transmission corridor construction monitoring region into multiple construction monitoring sub-regions based on the tower coordinate data, to construct an unmanned aerial vehicle monitoring path optimization model for each sub-region with unmanned aerial vehicle operation condition parameters as constraints, and to calculate the unmanned aerial vehicle monitoring path of each sub-region with the minimum flight time of the unmanned aerial vehicle and the maximum monitoring coverage rate of the unmanned aerial vehicle as targets; An unmanned aerial vehicle group control and monitoring module is used to collect image data of each construction monitoring sub-region by the unmanned aerial vehicle group based on the distributed collaborative control algorithm according to the unmanned aerial vehicle monitoring path of each sub-region, and to construct a power transmission corridor regional monitoring image data set. The image anomaly data comparison module is configured to perform semantic feature labeling and feature recognition on the power transmission corridor area monitoring image dataset by using a lightweight deep learning algorithm, and compare the features with historical image data of the power transmission corridor area to obtain an anomaly feature dataset of the power transmission corridor area. The knowledge graph construction and construction anomaly monitoring module is configured to obtain preset power transmission corridor area plot data and standard working condition data, construct a construction knowledge graph of the power transmission corridor area, and compare features based on a rule reasoning algorithm and the anomaly feature dataset to obtain an anomaly construction dataset of the power transmission corridor area, generate a construction rectification report for the power transmission corridor area, and push the report to a construction unit for rectification. The construction rectification monitoring module is configured to obtain rectification frequency data, rectification progress data, and construction area sensitivity data of each construction unit, construct a construction violation risk value calculation model, calculate the construction risk level of each construction monitoring sub-area, and perform coordinated monitoring of each construction monitoring sub-area by using a UAV group based on the construction risk level calculation until the construction rectification is completed.

Images