An Unmanned Aerial Vehicle Group Collaboration Method and Device for Power Line Inspection

Abstract

The present invention discloses a method and device for power line inspection by cooperation of unmanned aerial vehicle groups, belonging to the field of power line inspection. The method includes: establishing a three-dimensional equivalent model of the power line inspection area, planning multiple power line inspection routes, and configuring power line inspection tasks; configuring unmanned aerial vehicle groups to execute power line inspection tasks, obtaining multi-modal detection data of power lines, and locking the abnormal points of the lines when line abnormalities are detected; determining the three-dimensional model of the abnormal area, conducting abnormal evolution, obtaining the prediction result of abnormal development, and performing power line maintenance on the abnormal points of the lines. This application solves the technical problem in the prior art that power line maintenance is only carried out when obvious abnormalities occur in power line inspection, lacking timely maintenance of early abnormalities and affecting the operation reliability of power lines, and achieves the technical effect of timely discovering early abnormalities of power lines, predicting future development trends through abnormal evolution, conducting early maintenance of power lines, and improving the operation reliability of power lines.

Description

Technical Field

[0001] The present invention relates to the field of power line inspection, and particularly to a method and device for power line inspection by cooperation of unmanned aerial vehicle groups. Background Art

[0002] As an important part of the power transmission network, the safe operation of power lines directly affects the stability and reliability of the power grid. In order to ensure the safe operation of power lines, regular inspection and maintenance of power lines are required. In the prior art, the inspection of power lines realizes the detection of obvious abnormalities of power lines by setting various detection thresholds, lacks the timely discovery and treatment of early abnormalities, and is prone to the gradual deterioration of abnormal situations, thus triggering serious power accidents and affecting the operation reliability of power lines. At the same time, due to the lack of prediction and evolution analysis of abnormal situations, the development trend of abnormalities cannot be predicted, resulting in passive response to maintenance work, missing the best treatment opportunity, and increasing maintenance costs. Therefore, in the prior art, power line maintenance is only carried out when obvious abnormalities occur in power line inspection, lacking timely maintenance of early abnormalities and affecting the operation reliability of power lines. Summary of the Invention

[0003] The present application provides a method and device for power line inspection by cooperation of unmanned aerial vehicle groups, aiming to solve the technical problem that in the prior art, power line maintenance is only carried out when obvious abnormalities occur in power line inspection, lacking timely maintenance of early abnormalities and affecting the operation reliability of power lines.

[0004] In view of the above problems, the present application provides a method and device for power line inspection by cooperation of unmanned aerial vehicle groups.

[0005] In the first aspect disclosed in this application, a method for power line inspection by collaborative unmanned aerial vehicle (UAV) groups is provided. The method includes: establishing a three-dimensional equivalent model of the power inspection area, planning multiple power inspection routes according to the three-dimensional equivalent model, and configuring multiple power inspection tasks for each power inspection route; configuring a UAV group according to the multiple power inspection tasks of each power inspection route. The UAV group includes a main control UAV and multiple inspection UAVs. Among them, the main control UAV is used to collaboratively control the multiple inspection UAVs to complete the power inspection tasks and summarize and process the collected data of the multiple inspection UAVs. The multiple inspection UAVs respectively execute different types of inspection tasks; controlling the UAV group to execute the power inspection tasks along the corresponding power inspection routes, and obtaining multi-modal inspection data of the power lines in real time. When the main control UAV detects a line anomaly, locking the line anomaly point, and sending the line anomaly point, multi-modal inspection data and anomaly detection results to the central processing platform; the central processing platform determines the three-dimensional model of the anomaly area in the three-dimensional equivalent model according to the line anomaly point, and performs anomaly evolution on the line anomaly point by combining the anomaly detection results and multi-modal inspection data to obtain an anomaly development prediction result; performing power line maintenance on the line anomaly point according to the anomaly development prediction result.

[0006] In another aspect disclosed in this application, a device for power line inspection by collaborative UAV groups is provided. The device includes: a task planning unit for establishing a three-dimensional equivalent model of the power inspection area, planning multiple power inspection routes according to the three-dimensional equivalent model, and configuring multiple power inspection tasks for each power inspection route; a UAV configuration unit for configuring a UAV group according to the multiple power inspection tasks of each power inspection route. The UAV group includes a main control UAV and multiple inspection UAVs. Among them, the main control UAV is used to collaboratively control the multiple inspection UAVs to complete the power inspection tasks and summarize and process the collected data of the multiple inspection UAVs. The multiple inspection UAVs respectively execute different types of inspection tasks; a task execution unit for controlling the UAV group to execute the power inspection tasks along the corresponding power inspection routes, obtaining multi-modal inspection data of the power lines in real time. When the main control UAV detects a line anomaly, locking the line anomaly point, and sending the line anomaly point, multi-modal inspection data and anomaly detection results to the central processing platform; an anomaly evolution unit for the central processing platform to determine the three-dimensional model of the anomaly area in the three-dimensional equivalent model according to the line anomaly point, and perform anomaly evolution on the line anomaly point by combining the anomaly detection results and multi-modal inspection data to obtain an anomaly development prediction result; a line maintenance unit for performing power line maintenance on the line anomaly point according to the anomaly development prediction result.

[0007] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0008] Due to the adoption of a three-dimensional equivalent model for establishing a power inspection area, multiple power inspection routes are planned based on the three-dimensional equivalent model, and multiple power inspection tasks are configured for each power inspection route, providing a basis for the subsequent inspection work of the unmanned aerial vehicle (UAV) group. According to the multiple power inspection tasks of each power inspection route, a UAV group is configured. The UAV group includes a main control UAV and multiple inspection UAVs, realizing the division of labor and cooperation within the UAV group. The main control UAV is responsible for task allocation and data aggregation, and the inspection UAVs are responsible for executing different types of inspection tasks, improving the inspection efficiency and the diversity of data collection. Control the UAV group to execute the power inspection tasks along the corresponding power inspection routes, and obtain multi-modal inspection data of the power lines in real time. When the main control UAV detects a line anomaly, lock the line anomaly point, and send the line anomaly point, multi-modal inspection data and anomaly detection results to the central processing platform, realizing the real-time detection and anomaly discovery of the power lines by the UAV group. The central processing platform determines the three-dimensional model of the anomaly area in the three-dimensional equivalent model according to the line anomaly point, combines the anomaly detection results and multi-modal inspection data to perform anomaly evolution on the line anomaly point, obtains the anomaly development prediction result, and realizes the development prediction of the line anomaly. According to the anomaly development prediction result, perform power line maintenance on the line anomaly point, realizing the technical solution of early maintenance of the line, solving the technical problem in the prior art that power line inspection only performs power line maintenance when obvious anomalies occur, lacking timely maintenance of early anomalies and affecting the operation reliability of power lines, and achieving the technical effect of timely discovering early anomalies of power lines, predicting future development trends through anomaly evolution, performing early maintenance of power lines, and improving the operation reliability of power lines.

[0009] The above description is only an overview of the technical solution of the present application. In order to be able to understand the technical means of the present application more clearly, it can be implemented according to the content of the specification. And in order to make the above and other purposes, features and advantages of the present application more obvious and understandable, the specific embodiments of the present application are specifically given below. Brief Description of the Drawings

[0010] Figure 1 FIG. is a schematic flow chart of a method for power line inspection by cooperation of a UAV group provided by an embodiment of the present application;

[0011] Figure 2 FIG. is a schematic structural diagram of a device for power line inspection by cooperation of a UAV group provided by an embodiment of the present application.

[0012] Description of the reference numerals: Task planning unit 11, UAV configuration unit 12, Task execution unit 13, Anomaly evolution unit 14, Line maintenance unit 15. Detailed Description of the Embodiments

[0013] The overall idea of the technical solution provided by the present application is as follows:

[0014] The embodiment of the present application provides a method and device for power line inspection by cooperative unmanned aerial vehicle (UAV) groups. First, a three-dimensional digital model of the power inspection area is established. On this basis, inspection routes and tasks are planned to provide guidance for the cooperative inspection of UAV groups. Second, a master-slave cooperative UAV group scheme is adopted. The master UAV coordinates and schedules the inspection UAVs to carry out multimodal data collection and anomaly detection, realizing efficient cooperation within the UAV group. Third, during the real-time inspection of the UAV group, once an anomaly is found, the anomaly point is locked in time and data is reported to the central processing platform. The central processing platform determines the three-dimensional model of the anomaly area according to the location of the anomaly point, and comprehensively analyzes the multi-source detection data to carry out anomaly evolution analysis and predict the development trend of the anomaly. Then, according to the prediction result of the anomaly development, the maintenance work of the line anomaly point is carried out in time to realize the early maintenance of the line.

[0015] After introducing the basic principle of the present application, the various non-limiting implementation manners of the present application will be specifically introduced below with reference to the accompanying drawings of the specification.

[0016] Embodiment 1, as Figure 1 shown, the embodiment of the present application provides a method for power line inspection by cooperative UAV groups, and the method includes:

[0017] S1: Establish a three-dimensional equivalent model of the power inspection area, plan multiple power inspection routes according to the three-dimensional equivalent model, and configure multiple power inspection tasks for each power inspection route.

[0018] Specifically, first, collect the geographical information data and power line information of the power inspection area, such as terrain and landform, vegetation distribution, line orientation, tower position, etc. Second, through three-dimensional modeling software, integrate the collected geographical information data and power line information into a three-dimensional equivalent model, which not only includes natural geographical elements but also accurately presents the spatial distribution of power lines. Then, based on the established three-dimensional equivalent model, further plan multiple power inspection routes. During the planning process, comprehensively consider the terrain complexity, line distribution density, flight characteristics of the UAVs (such as endurance, wind resistance, etc.) and inspection requirements (such as inspection accuracy, coverage, etc.), and combine the decision-making of the expert group to generate multiple power inspection routes to cover all power lines that need to be inspected and avoid obstacles that may affect flight safety at the same time. After completing the planning of the power inspection routes, configure multiple specific power inspection tasks for each power inspection route, including but not limited to visual inspection tasks, infrared inspection tasks, laser scanning tasks, etc. Among them, the configuration of the power inspection tasks needs to be adjusted according to the line characteristics and actual inspection requirements.

[0019] By establishing a three-dimensional equivalent model and planning power inspection routes and power inspection tasks based on it, a solid foundation is provided for the subsequent collaborative inspection of unmanned aerial vehicle (UAV) groups, improving the efficiency, quality, and safety of power line inspections.

[0020] S2: Configure UAV groups according to multiple power inspection tasks of each power inspection route. The UAV groups include one master UAV and multiple inspection UAVs. Among them, the master UAV is used to collaboratively control multiple inspection UAVs to complete power inspection tasks and aggregate and process the acquisition data of multiple inspection UAVs. Multiple inspection UAVs respectively perform different types of inspection tasks.

[0021] Specifically, first, based on the planned power inspection routes and power inspection tasks, configure the UAV groups required for each power inspection route. The configured UAV groups include one master UAV and multiple inspection UAVs. Among them, the master UAV has strong computing and communication capabilities and is used to coordinate the actions of the entire UAV group. Multiple inspection UAVs are configured according to the types of power inspection tasks. Each inspection UAV specifically performs one type of inspection task, corresponding one-to-one with the power inspection task. The master UAV is responsible for planning and coordinating the actions of the inspection UAVs and receiving and processing the data transmitted back by the inspection UAVs in real time. Each inspection UAV is equipped with corresponding inspection equipment, such as high-definition cameras, infrared thermal imagers, laser scanners, etc., according to the task type of the power inspection task it specifically performs, for performing specific types of inspection tasks.

[0022] Through the configuration of UAV groups, the collaborative application of multiple inspection means can be realized, improving the efficiency and scalability of power line inspections. Each inspection UAV focuses on a specific type of inspection task, can better play its advantages, and improve the inspection quality. At the same time, the unified coordination of the master UAV ensures the orderly progress of the entire inspection process, while the collaborative work of multiple inspection UAVs improves the inspection coverage and data acquisition capabilities, providing hardware support for the subsequent execution of power inspection tasks.

[0023] S3: Control the UAV groups to perform power inspection tasks along the corresponding power inspection routes, and obtain multi-modal inspection data of power lines in real time. When the master UAV detects a line anomaly, lock the line anomaly point and send the line anomaly point, the multi-modal inspection data, and the anomaly detection result to the central processing platform.

[0024] Specifically, first, according to the configured unmanned aerial vehicle (UAV) group, control the UAV group to start executing the power inspection task according to the planned power inspection route. During the inspection process, multiple inspection UAVs, according to their respective power inspection tasks, use the equipped inspection equipment to collect the inspection data of the power line in real time. Multiple inspection UAVs form multi-modal inspection data, including but not limited to high-definition images, infrared thermal images, laser scanning data, etc., providing multi-dimensional information for comprehensively evaluating the status of the power line. Secondly, while coordinating the inspection UAVs to execute tasks, the master UAV receives and processes the multi-modal inspection data transmitted back by multiple inspection UAVs in real time, and performs real-time analysis on the received multi-modal inspection data. For example, for different types of power inspection tasks, corresponding normal thresholds are set based on the design parameters and historical operation data of the power line to determine whether the multi-modal inspection data is within the normal range; compare the multi-modal inspection data received in real time with the preset normal threshold; according to the comparison result, determine whether there is an abnormal situation; when one or more inspection data exceed their corresponding normal thresholds, it is determined as an abnormal situation.

[0025] When it is detected that the multi-modal inspection data indicates an abnormal situation of the power line, the master UAV immediately locks the position information of the power line and determines the abnormal point of the line. Then, the master UAV packages the detected abnormal point information of the line, the relevant multi-modal inspection data, and the abnormal detection result, and sends them to the central processing platform through the wireless communication network, ensuring that the central processing platform can quickly obtain the on-site situation and providing the necessary information support for subsequent in-depth analysis and decision-making of the abnormality.

[0026] Through real-time monitoring and data transmission, abnormal situations of the power line can be quickly discovered, and relevant information can be timely transmitted to the central processing platform, improving the timeliness and accuracy of power line inspection, providing detailed data support for subsequent abnormality handling, and thus enhancing the operation reliability and safety of the power line.

[0027] S4: The central processing platform determines the three-dimensional model of the abnormal area in the three-dimensional equivalent model according to the abnormal point of the line, and performs abnormal evolution on the abnormal point of the line by combining the abnormal detection result and the multi-modal inspection data to obtain the prediction result of abnormal development.

[0028] Specifically, first, the central processing platform receives the line anomaly point information, multi-modal detection data, and anomaly detection results sent by the master unmanned aerial vehicle (UAV). Among them, the anomaly detection result is the information obtained after the master UAV conducts anomaly detection. For example, an anomaly detection result is as follows: Anomaly type: abnormal conductor temperature; Anomaly degree: temperature exceeding the standard by 8°C (measured 75°C, normal threshold 67°C); Anomaly location: conductor between tower No. 10 and tower No. 11; Related parameters: current load current 580A, ambient temperature 28°C, wind speed 1.5 m / s, etc. Based on the received line anomaly points, the central processing platform locates the exact positions of the line anomaly points in the three-dimensional equivalent model. Secondly, centered on the located line anomaly points, a certain range of area is delimited in the three-dimensional equivalent model, and the three-dimensional model information of this area is extracted to form a three-dimensional model of the anomaly area, including topographic and geomorphic features, distribution of power facilities, etc. around the line anomaly points, providing a spatial reference for subsequent anomaly evolution analysis.

[0029] Then, the central processing platform combines the anomaly detection results and multi-modal detection data to conduct anomaly evolution analysis on the line anomaly points. Specifically, the central processing platform first determines the anomaly type according to the anomaly detection results, such as abnormal conductor temperature, abnormal line insulation layer, etc. Then, it calls the anomaly evolution model corresponding to this anomaly type, inputs the multi-modal detection data into the anomaly evolution model, and simulates the development trend of the abnormal state over time. During the anomaly evolution process, various factors are comprehensively considered, such as meteorological conditions, load changes, material aging, etc., to improve the accuracy of prediction. By running the anomaly evolution model, the changes in the abnormal state within a certain period in the future are simulated, including the acceleration rate of the anomaly degree, the expansion of the influence range, etc. After that, based on the results of the anomaly evolution, an anomaly development prediction result is generated, including information such as the time axis of the anomaly development, the change curve of the influence degree, the potential risk level, etc., providing a reference for subsequent maintenance decisions.

[0030] Through the anomaly evolution analysis based on the three-dimensional model and multi-modal data, the abnormal conditions of the power line can be comprehensively evaluated and predicted, effectively improving the preventive maintenance ability of the power line, reducing the risk of sudden failures, and thus enhancing the reliability and security of power supply.

[0031] S5: Perform power line maintenance on the line anomaly points according to the anomaly development prediction result.

[0032] Specifically, first, based on the obtained abnormal development prediction results, the central processing platform comprehensively considers factors such as the type, severity, development trend, and potential risks of the abnormality, and formulates corresponding power line maintenance strategies. Second, according to the formulated power line maintenance strategies, a specific maintenance task list is generated, including the maintenance priority, required professional technicians, necessary repair tools and spare parts, estimated maintenance time, etc. Then, the central processing platform issues the abnormal development prediction results and the maintenance task list to the relevant maintenance teams. After receiving them, the maintenance teams carry the necessary equipment and materials and go to the line abnormality point for on-site processing, such as replacing damaged components, cleaning dirt, adjusting line tension, and strengthening the support structure.

[0033] Through the maintenance based on the abnormal development prediction results, accurate maintenance and preventive maintenance of power lines can be achieved, which not only improves the maintenance efficiency, reduces the maintenance cost, but also effectively prevents the occurrence of potential faults, thereby enhancing the overall reliability and safety of power lines.

[0034] Furthermore, the embodiment of the present application further includes:

[0035] For the first power inspection route, traverse the multiple power inspection tasks to determine the first power inspection task; extract the first task execution point in the first power inspection task, and control the unmanned aircraft group to fly to the first task execution point, where the first task execution point corresponds to the first power line to be inspected; according to the first power inspection task, use the multiple inspection unmanned aircraft to synchronously perform power inspection on the first power line to be inspected to obtain multi-modal inspection data; transmit the multi-modal inspection data to the master unmanned aircraft, perform abnormal detection on the multi-modal inspection data, and when a line abnormality is detected, use the first task execution point as the line abnormality point.

[0036] In a feasible implementation manner, the first power inspection route refers to any one of multiple power inspection routes, and the inspection methods of other power inspection routes are the same. For the first power inspection route, the master unmanned aircraft sequentially traverses the multiple power inspection tasks configured on the first power inspection route according to the inspection time sequence, and determines one power inspection task each time as the first power inspection task. Then, extract the first task execution point in the first power inspection task. The first task execution point refers to the specific position where the inspection needs to be performed in the first power inspection task, and is represented by three-dimensional space coordinates. The first task execution point corresponds to the first power line to be inspected, and the first power line to be inspected refers to the power line segment that needs to be inspected corresponding to the first task execution point. The master unmanned aircraft extracts the coordinate information of the first task execution point and controls the entire unmanned aircraft group to fly to this position accordingly.

[0037] Then, multiple inspection drones perform synchronous inspection on the first power line to be inspected according to their respective tasks, so as to obtain comprehensive multi-modal inspection data, including various forms of inspection data, such as visible light images, infrared thermal images, laser scanning data, etc. Subsequently, the multi-modal inspection data is transmitted to the master drone, and the master drone receives and analyzes the multi-modal inspection data. When an anomaly is detected, the current first task execution point is marked as a line anomaly point, providing precise positioning for subsequent processing.

[0038] Through the division of labor and cooperation of multiple drones, efficient real-time monitoring and precise anomaly detection of power lines are achieved, improving the efficiency and comprehensiveness of inspection, and providing a strong guarantee for the safe operation of power lines.

[0039] Furthermore, the embodiment of the present application further includes:

[0040] According to the first power line to be inspected, replicate the line model of the three-dimensional equivalent model to obtain a first three-dimensional power line, which represents the normal state of the first power line to be inspected; before the power line inspection task, pre-store the first three-dimensional power line in the master drone; after the master drone receives the multi-modal inspection data, retrieve the first three-dimensional power line and perform anomaly detection on the multi-modal inspection data.

[0041] In a preferred implementation manner, first, identify the precise position and scope of the first power line to be inspected in the power line inspection area, and extract the corresponding local area from the constructed three-dimensional equivalent model to achieve targeted replication of the three-dimensional equivalent model and obtain the first three-dimensional power line, which not only includes the geometric features of the line itself (such as conductor sag, tower structure, etc.), but also includes the surrounding environmental information (such as terrain undulation, vegetation distribution, etc.). The first three-dimensional power line represents the parameters and characteristics of the first power line to be inspected under ideal or normal operating conditions, providing a standard reference for subsequent anomaly detection.

[0042] Execute before the actual inspection task starts. Process the generated first three-dimensional power line through data compression and optimization to adapt to the storage and processing capabilities of the master drone. Subsequently, transmit and store the first three-dimensional power line in the storage device of the master drone, such as a solid-state drive or high-performance flash memory, thereby reducing the need for real-time data transmission and avoiding possible network latency or bandwidth shortage problems during the inspection process. At the same time, it provides a standard reference that can be called at any time for the master drone, improving the response speed and reliability of anomaly detection.

[0043] After the master UAV receives the multi-modal detection data from each detection UAV, it immediately starts the anomaly detection process. First, the master UAV quickly retrieves the pre-stored first 3D power line from its storage device; then, it conducts multi-dimensional comparative analysis on the real-time obtained multi-modal detection data and the first 3D power line, including but not limited to image registration, point cloud matching, thermal imaging analysis, etc. Through precise comparison, the deviations between the actual line state and the normal state are identified, such as problems like abnormal conductor sag and conductor damage, so as to timely discover potential safety hazards.

[0044] Through the anomaly detection of the power line, by using the pre-constructed 3D model and combining with the real-time multi-modal data, while ensuring the detection accuracy, the detection efficiency is significantly improved, providing strong technical support for the preventive maintenance and safe operation of the power system.

[0045] Furthermore, the embodiments of the present application further include:

[0046] Based on the first 3D power line, extract the multi-modal feature templates of the first power line to be inspected in the normal state, and construct multiple modal normal feature libraries; compare the multi-modal detection data with the multiple modal normal feature libraries respectively, calculate the feature deviation degrees of different types of detections to obtain multiple feature deviation degrees; extract the deviation threshold values of different types of detections, and determine whether the multiple feature deviation degrees meet the corresponding deviation threshold values; if the multiple feature deviation degrees all meet the corresponding deviation threshold values, determine that the state of the first power line to be inspected is normal; if any one of the multiple feature deviation degrees does not meet the corresponding deviation threshold value, determine that the state of the first power line to be inspected is abnormal.

[0047] In a preferred embodiment, first, a multi-dimensional analysis is performed on a pre-stored first three-dimensional power line. By means of image processing algorithms, point cloud analysis techniques, thermal imaging feature extraction, etc., normal state features in multiple modalities are extracted from the first three-dimensional power line as multi-modal feature templates, including but not limited to conductor geometric parameters (such as sag, inter-phase distance), conductor surface texture features, normal operating temperature distribution of the conductor, etc. Subsequently, these extracted multi-modal feature templates are classified and sorted to construct multiple independent modal normal feature libraries. Each modal normal feature library corresponds to a specific detection modality, such as a visible light feature library, an infrared feature library, a laser point cloud feature library, etc., as a reference for subsequent anomaly detection. Subsequently, the real-time multi-modal detection data obtained is accurately compared with the constructed modal normal feature libraries. For each detection modality, the deviation degree between the actual detection data and the normal features is calculated to obtain a quantified feature deviation degree. For example, for visible light images, the structural similarity index is calculated; for thermal imaging data, the root mean square error of the temperature distribution is calculated; for laser point cloud data, the average distance deviation after point cloud registration is calculated, etc., so as to obtain multiple feature deviation degrees corresponding to different detection modalities.

[0048] After that, the deviation thresholds corresponding to various detection types are extracted from the preset parameter library. Among them, the deviation thresholds are preset based on a large amount of historical data and expert experience, representing the maximum acceptable deviation degree of various features. Subsequently, the calculated multiple feature deviation degrees are compared with the corresponding deviation thresholds, and each detection modality has its own independent judgment criterion. If the feature deviation degrees of all detection modalities do not exceed their respective deviation thresholds, from multiple dimensions, the indicators of the first power line to be inspected are within the acceptable range, and there are no obvious anomalies or potential risks, it is considered that the first power line to be inspected is in a normal state. On the contrary, if the feature deviation degree of any detection modality exceeds the corresponding deviation threshold, it is determined that the first power line to be inspected is in an abnormal state, ensuring that any potential anomalies can be discovered in time.

[0049] Through comprehensive power line anomaly detection, comprehensively utilizing multi-modal data, various types of line anomalies can be effectively identified, improving the reliability and safety of power line inspections, and providing a strong guarantee for the stable operation of the power system.

[0050] Furthermore, the embodiments of the present application further include:

[0051] In the three-dimensional equivalent model, obtain the three-dimensional spatial coordinates of the line anomaly point; with the three-dimensional spatial coordinates as the center, establish a three-dimensional buffer zone with a preset radius, extract the three-dimensional equivalent model within the three-dimensional buffer zone, and generate a three-dimensional model of the abnormal area; according to the abnormal detection result, obtain the abnormal type of the line anomaly point, and according to the abnormal type, match the corresponding abnormal evolution model in the abnormal evolution knowledge base; based on the abnormal detection result and the multi-modal detection data, simulate the dynamic change process of the abnormal state over time through the abnormal evolution model, and obtain the time-series spatial distribution of the abnormal state; generate an abnormal development prediction result according to the time-series spatial distribution.

[0052] In a feasible implementation manner, after the central processing platform receives the line anomaly point reported by the master unmanned aerial vehicle, it locates the line anomaly point in the pre-constructed three-dimensional equivalent model, extracts its three-dimensional spatial coordinates, lays a foundation for subsequent abnormal area modeling and analysis, and ensures the precise positioning of abnormal handling. Then, with the obtained three-dimensional coordinates of the anomaly point as the center, a three-dimensional buffer zone with a preset radius is constructed. Among them, the size of the preset radius is dynamically adjusted according to factors such as the abnormal type and line characteristics. Subsequently, all information within the three-dimensional buffer zone is extracted from the complete three-dimensional equivalent model, including power facilities, terrain and landforms, surrounding environments, etc., to form a local three-dimensional model of the abnormal area, providing a comprehensive spatial background for subsequent abnormal evolution analysis.

[0053] Then, analyze the abnormal detection result to determine the specific abnormal type of the line anomaly point, such as wire overheating, line breakage, etc. Access the pre-established abnormal evolution knowledge base, which contains abnormal evolution models for various types of power line anomalies. The abnormal evolution model is a simulation prediction model constructed based on historical abnormal development data of different types of anomalies, and is used to simulate and predict the development process of specific types of anomalies. Subsequently, input the abnormal detection result and multi-modal detection data into the selected abnormal evolution model, and through simulation prediction, simulate the dynamic change process of the abnormal state over time, such as environmental conditions, load changes, etc. Through repeated iteration, a series of spatial distribution data representing the abnormal state at different time points are generated, forming the time-series spatial distribution of the abnormal state. Then, based on the obtained time-series spatial distribution, extract key features, identify the development trend, and generate a comprehensive abnormal development prediction result.

[0054] Through the precise positioning, comprehensive analysis and prediction of power line anomalies, not only the current abnormal state is considered, but also its future development trend is simulated, providing an important decision-making basis for the preventive maintenance and handling of the power system, and effectively improving the safety and reliability of the power system.

[0055] Furthermore, the embodiments of the present application further include:

[0056] Based on the time series spatial distribution, calculate the rate of change of the abnormal points on the line over time to obtain the abnormal development speed; statistically analyze the main direction of the change of the abnormal points on the line with respect to the spatial distribution to obtain the abnormal development direction; along the abnormal development direction, extract the abnormal influence range in the three-dimensional model of the abnormal area; generate an abnormal development prediction result according to the abnormal development speed, the abnormal development direction, and the abnormal influence range.

[0057] In a preferred embodiment, first, analyze the obtained time series spatial distribution. By comparing the abnormal states at different time points, calculate the change rate of the abnormal degree to obtain the abnormal development speed. Among them, the abnormal development speed includes multiple parameters, such as the temperature rise speed, the deformation rate of the insulating layer, etc. At the same time, through spatial statistical analysis of the time series spatial distribution, analyze the diffusion trend of the abnormal state in space, identify the main direction of the abnormal development to obtain the abnormal development direction, and form a spatial pattern of the abnormal extending along the line. Subsequently, along the predicted abnormal development direction, combined with the characteristics of the abnormal type and surrounding environmental factors, calculate and extract the possibly affected area in the simulation environment to generate the abnormal influence range. After that, generate a curve of the change of the abnormal degree over time, a three-dimensional visualization model of the abnormal influence range, prediction of key time points, potential risk assessment, etc. according to the obtained abnormal development speed, abnormal development direction, and abnormal influence range as the abnormal development prediction result to provide intuitive and comprehensive information support.

[0058] Through accurate simulation analysis and prediction of the power line abnormality, considering the development speed, direction of the abnormality, and the specific spatial environment, provide personalized and contextualized prediction results to support the formulation of targeted maintenance strategies.

[0059] Furthermore, the embodiment of the present application further includes:

[0060] Calculate the correlation confidence level between the abnormal points on the line and the abnormal development prediction result to obtain the abnormal prediction confidence level; if the abnormal prediction confidence level is less than or equal to the preset confidence threshold, send a data supplement instruction to the master unmanned aerial vehicle; after receiving the data supplement instruction, the master unmanned aerial vehicle cooperates to guide the multiple detection unmanned aerial vehicles to fly to re-collect the multi-modal supplementary data of the abnormal points on the line; send the multi-modal supplementary data to the master unmanned aerial vehicle, summarize and then transmit it back to the central processing platform, re-perform abnormal evolution on the abnormal points on the line, and obtain an updated abnormal development prediction result.

[0061] In a preferred embodiment, after obtaining the abnormal development prediction result, the central processing platform conducts a reliability assessment on it. First, it analyzes the development of similar abnormalities in the historical database, compares the similarity between the current abnormal development prediction result and historical cases, and takes it as the quantified abnormal prediction confidence level, that is, the correlation confidence level, which reflects the reliability of the current abnormal development prediction result. Subsequently, the calculated abnormal prediction confidence level is compared with a pre-set confidence threshold. When the confidence level is lower than or equal to the confidence threshold, it indicates that there is uncertainty in the current abnormal development prediction result or a significant difference from historical experience. In this case, the central processing platform generates and sends a data supplement instruction to the master unmanned aerial vehicle (UAV) to initiate an additional data collection process.

[0062] After receiving the data supplement instruction, the master UAV coordinates multiple detection UAVs to the location of the line abnormality. Based on the previous abnormality type and the abnormal development prediction result, a targeted supplementary data collection strategy is formulated, including adjusting sensor parameters, changing the collection angle, or increasing the sampling frequency, etc. Under the command of the master UAV, multiple detection UAVs synchronously execute this optimized data collection task to obtain more comprehensive and accurate multi-modal supplementary data. Subsequently, after the multi-modal supplementary data collected by multiple detection UAVs is transmitted to the master UAV for preliminary aggregation and processing, the master UAV transmits the processed multi-modal supplementary data back to the central processing platform. After receiving the new multi-modal supplementary data, the central processing platform integrates it with the previous data, re-performs the abnormal evolution analysis, and generates an updated abnormal development prediction result with higher accuracy and reliability.

[0063] Through the evaluation based on historical experience, data supplement, and re-analysis, the accuracy and reliability of abnormal prediction are improved, which can effectively cope with the complex and changeable power line abnormalities and provide more reliable support for the safety operation and maintenance decision-making of the power system.

[0064] In summary, the unmanned aerial vehicle group collaborative power line inspection method provided by the embodiments of the present application has the following technical effects:

[0065] Establish a three-dimensional equivalent model of the power inspection area, plan multiple power inspection routes according to the three-dimensional equivalent model, and configure multiple power inspection tasks for each power inspection route, providing an action guide for the collaborative inspection of the unmanned aerial vehicle (UAV) group to ensure full coverage and high efficiency of the inspection. According to the multiple power inspection tasks of each power inspection route, configure the UAV group, which includes a main control UAV and multiple inspection UAVs. Among them, the main control UAV is used to collaboratively control multiple inspection UAVs to complete the power inspection tasks and summarize and process the collected data of multiple inspection UAVs. Multiple inspection UAVs respectively execute different types of inspection tasks. Through task division and collaboration, the overall efficiency of the UAV group is exerted to improve the inspection efficiency. Control the UAV group to execute the power inspection tasks along the corresponding power inspection routes, and obtain multi-modal inspection data of the power lines in real time. When the main control UAV detects a line anomaly, lock the line anomaly point, and send the line anomaly point, multi-modal inspection data and anomaly detection results to the central processing platform to achieve rapid discovery and positioning of the line anomaly and provide data for subsequent anomaly analysis. The central processing platform determines the three-dimensional model of the anomaly area in the three-dimensional equivalent model according to the line anomaly point, and performs anomaly evolution on the line anomaly point by combining the anomaly detection results and multi-modal inspection data to obtain the prediction result of the anomaly development, realizing from anomaly detection to anomaly warning, mastering the law of anomaly development, and providing a decision-making basis for formulating maintenance strategies. Perform power line maintenance on the line anomaly point according to the prediction result of the anomaly development, adopt a predictive maintenance method, reduce the maintenance cost, reduce the impact on the line operation, and improve the reliability of the line operation.

[0066] Embodiment 2, based on the same inventive concept as the method for power line inspection collaborative with a UAV group in the foregoing embodiment, as Figure 2 shown, the embodiment of the present application provides a device for power line inspection collaborative with a UAV group, and the device includes:

[0067] A task planning unit 11, configured to establish a three-dimensional equivalent model of the power inspection area, plan multiple power inspection routes according to the three-dimensional equivalent model, and configure multiple power inspection tasks for each power inspection route.

[0068] A UAV configuration unit 12, configured to configure a UAV group according to the multiple power inspection tasks of each power inspection route. The UAV group includes a main control UAV and multiple inspection UAVs. Among them, the main control UAV is used to collaboratively control multiple inspection UAVs to complete the power inspection tasks and summarize and process the collected data of multiple inspection UAVs. Multiple inspection UAVs respectively execute different types of inspection tasks.

[0069] The task execution unit 13 is used to control the drone group to perform power inspection tasks along the corresponding power inspection routes, and to obtain multi-modal detection data of the power lines in real time. When the master drone detects a line anomaly, it locks the line anomaly point and sends the line anomaly point, the multi-modal detection data, and the anomaly detection result to the central processing platform.

[0070] The anomaly evolution unit 14 is used for the central processing platform to determine the three-dimensional model of the anomaly area in the three-dimensional equivalent model according to the line anomaly point, and to perform anomaly evolution on the line anomaly point by combining the anomaly detection result and the multi-modal detection data to obtain the anomaly development prediction result.

[0071] The line maintenance unit 15 is used to perform power line maintenance on the line anomaly point according to the anomaly development prediction result.

[0072] Furthermore, the task execution unit 13 includes the following execution steps:

[0073] For the first power inspection route, traverse the multiple power inspection tasks to determine the first power inspection task; extract the first task execution point in the first power inspection task, and control the drone group to fly to the first task execution point, where the first task execution point corresponds to the first power line to be inspected; according to the first power inspection task, use the multiple detection drones to synchronously perform power inspection on the first power line to be inspected to obtain multi-modal detection data; transmit the multi-modal detection data to the master drone, perform anomaly detection on the multi-modal detection data, and when a line anomaly is detected, use the first task execution point as the line anomaly point.

[0074] Furthermore, the task execution unit 13 also includes the following execution steps:

[0075] According to the first power line to be inspected, copy the line model of the three-dimensional equivalent model to obtain the first three-dimensional power line, which represents the normal state of the first power line to be inspected; before the power inspection task, pre-store the first three-dimensional power line in the master drone; after the master drone receives the multi-modal detection data, retrieve the first three-dimensional power line and perform anomaly detection on the multi-modal detection data.

[0076] Furthermore, the task execution unit 13 also includes the following execution steps:

[0077] Based on the first three-dimensional power line, extract the multi-modal feature templates of the first power line to be inspected under normal conditions, and construct multiple modal normal feature libraries; compare the multi-modal detection data with the multiple modal normal feature libraries respectively, calculate the feature deviation degrees of different types of detections, and obtain multiple feature deviation degrees; extract the deviation threshold values of different types of detections, and determine whether the multiple feature deviation degrees meet the corresponding deviation threshold values; if the multiple feature deviation degrees all meet the corresponding deviation threshold values, determine that the state of the first power line to be inspected is normal; if any one of the multiple feature deviation degrees does not meet the corresponding deviation threshold value, determine that the state of the first power line to be inspected is abnormal.

[0078] Further, the abnormal evolution unit 14 includes the following execution steps:

[0079] In the three-dimensional equivalent model, obtain the three-dimensional spatial coordinates of the line abnormal point; take the three-dimensional spatial coordinates as the center, establish a three-dimensional buffer zone with a preset radius, extract the three-dimensional equivalent model in the three-dimensional buffer zone, and generate an abnormal area three-dimensional model; according to the abnormal detection result, obtain the abnormal type of the line abnormal point, and match the corresponding abnormal evolution model in the abnormal evolution knowledge base according to the abnormal type; based on the abnormal detection result and the multi-modal detection data, simulate the dynamic change process of the abnormal state over time through the abnormal evolution model, and obtain the time series spatial distribution of the abnormal state; according to the time series spatial distribution, generate an abnormal development prediction result.

[0080] Further, the abnormal evolution unit 14 includes the following execution steps:

[0081] Based on the time series spatial distribution, calculate the rate of change of the line abnormal point over time to obtain the abnormal development speed; count the main direction of the change of the line abnormal point with the spatial distribution to obtain the abnormal development direction; along the abnormal development direction, extract the abnormal influence range in the abnormal area three-dimensional model; according to the abnormal development speed, the abnormal development direction and the abnormal influence range, generate an abnormal development prediction result.

[0082] Further, the embodiment of the present application further includes a supplementary evolution unit, and this unit includes the following execution steps:

[0083] Calculate the correlation confidence between the abnormal point of the line and the abnormal development prediction result to obtain the abnormal prediction confidence. If the abnormal prediction confidence is less than or equal to the preset confidence threshold, send a data supplement instruction to the master UAV. After receiving the data supplement instruction, the master UAV collaboratively guides the multiple detection UAVs to fly and re-collect the multimodal supplement data of the abnormal point of the line. Send the multimodal supplement data to the master UAV, summarize and then send it back to the central processing platform to re-evolve the abnormality of the abnormal point of the line and obtain an updated abnormal development prediction result.

[0084] Any step of the method described above can be stored as computer instructions or programs in an unrestricted computer memory and can be called and recognized by an unrestricted computer processor to implement any one of the methods in the embodiments of the present application, and no redundant restrictions are made here.

[0085] Furthermore, the first or second described above may not only represent an order relationship, but may also represent a specific concept, and / or refer to the selection of multiple elements individually or in whole. Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalent technologies, the present application is intended to include these changes and deformations.

Claims

1. A method for power line inspection by collaborative operation of unmanned aerial vehicles, characterized in that, Including: Establish a three-dimensional equivalent model of the power inspection area, plan multiple power inspection routes according to the three-dimensional equivalent model, and configure multiple power inspection tasks for each power inspection route; Configure a drone group according to the multiple power inspection tasks of each power inspection route. The drone group includes a main control drone and multiple inspection drones. Among them, the main control drone is used to cooperate with the multiple inspection drones to complete the power inspection tasks and summarize and process the collected data of the multiple inspection drones. The multiple inspection drones perform different types of inspection tasks respectively; Control the drone group to perform power inspection tasks along the corresponding power inspection routes, and obtain multi-modal inspection data of the power lines in real time. When the main control drone detects a line anomaly, lock the line anomaly point, and send the line anomaly point, the multi-modal inspection data and the anomaly detection result to the central processing platform; The central processing platform determines the three-dimensional model of the anomaly area in the three-dimensional equivalent model according to the line anomaly point, and performs anomaly evolution on the line anomaly point by combining the anomaly detection result and the multi-modal inspection data to obtain the anomaly development prediction result; Perform power line maintenance on the line anomaly point according to the anomaly development prediction result; The central processing platform determines the three-dimensional model of the anomaly area in the three-dimensional equivalent model according to the line anomaly point, and performs anomaly evolution on the line anomaly point by combining the anomaly detection result and the multi-modal inspection data to obtain the anomaly development prediction result, including: In the three-dimensional equivalent model, obtain the three-dimensional spatial coordinates of the line anomaly point; Taking the three-dimensional spatial coordinates as the center, establish a three-dimensional buffer zone with a preset radius, extract the three-dimensional equivalent model in the three-dimensional buffer zone, and generate the three-dimensional model of the anomaly area; According to the anomaly detection result, obtain the anomaly type of the line anomaly point, and match the corresponding anomaly evolution model in the anomaly evolution knowledge base according to the anomaly type; Based on the anomaly detection result and the multi-modal inspection data, simulate the dynamic change process of the anomaly state over time through the anomaly evolution model to obtain the time series spatial distribution of the anomaly state; Generate an anomaly development prediction result according to the time series spatial distribution.

2. The method for power line inspection by cooperation of unmanned aerial vehicle groups according to claim 1, wherein, Obtain multi-modal inspection data of the power lines in real time. When the main control drone detects a line anomaly, lock the line anomaly point, including: For the first power inspection route, traverse the multiple power inspection tasks to determine the first power inspection task; Extract the first task execution point in the first power inspection task, and control the drone group to fly to the first task execution point, where the first task execution point corresponds to the first power line to be inspected; According to the first power inspection task, use the multiple inspection drones to synchronously perform power inspection on the first power line to be inspected to obtain multi-modal inspection data; Transmit the multi-modal inspection data to the main control drone, perform anomaly detection on the multi-modal inspection data, and when a line anomaly is detected, use the first task execution point as the line anomaly point.

3. A method for power line inspection by cooperation of unmanned aerial vehicle groups according to claim 2, characterized in that Transmit the multi-modal detection data to the master unmanned aerial vehicle (UAV) for anomaly detection of the multi-modal detection data, including: According to the first power line to be inspected, replicate the line model of the three-dimensional equivalent model to obtain a first three-dimensional power line, where the first three-dimensional power line represents the normal state of the first power line to be inspected; Before the power line inspection task, pre-store the first three-dimensional power line in the master UAV; After receiving the multi-modal detection data, the master UAV retrieves the first three-dimensional power line to perform anomaly detection on the multi-modal detection data.

4. A method for power line inspection by cooperation of unmanned aerial vehicle groups according to claim 3, characterized in that, After receiving the multi-modal detection data, the master UAV retrieves the first three-dimensional power line to perform anomaly detection on the multi-modal detection data, including: Based on the first three-dimensional power line, extract the multi-modal feature templates of the first power line to be inspected in the normal state, and construct multiple modal normal feature libraries; Compare the multi-modal detection data with the multiple modal normal feature libraries respectively, calculate the feature deviation degrees of different types of detections, and obtain multiple feature deviation degrees; Extract the deviation threshold values of different types of detections, and determine whether the multiple feature deviation degrees meet the corresponding deviation threshold values; If all the multiple feature deviation degrees meet the corresponding deviation threshold values, determine that the state of the first power line to be inspected is normal; If any one of the multiple feature deviation degrees does not meet the corresponding deviation threshold value, determine that the state of the first power line to be inspected is abnormal.

5. A method for power line inspection by cooperation of unmanned aerial vehicle groups according to claim 1, characterized in that, Generate an abnormal development prediction result according to the time series spatial distribution, including: Based on the time series spatial distribution, calculate the rate of change of the abnormal points on the line over time to obtain the abnormal development speed; Statistically analyze the main direction of the change of the abnormal points on the line with respect to the spatial distribution to obtain the abnormal development direction; Along the abnormal development direction, extract the abnormal influence range in the three-dimensional model of the abnormal area; Generate an abnormal development prediction result according to the abnormal development speed, the abnormal development direction, and the abnormal influence range.

6. A method for power line inspection by collaborative unmanned aerial vehicles according to claim 1, characterized in that The method further includes: Calculate the correlation confidence degree between the abnormal points on the line and the abnormal development prediction result to obtain the abnormal prediction confidence degree; If the abnormal prediction confidence degree is less than or equal to the preset confidence threshold value, send a data supplement instruction to the master UAV; After receiving the data supplement instruction, the master UAV cooperatively guides the multi-rotor detection UAVs to fly to re-collect the multi-modal supplementary data of the abnormal points on the line; Send the multi-modal supplementary data to the master UAV, summarize and transmit it back to the central processing platform, and re-perform abnormal evolution on the abnormal points on the line to obtain an updated abnormal development prediction result.

7. An unmanned aerial vehicle group collaborative power line inspection device, characterized in that A power line inspection method for collaborative operation of an unmanned aerial vehicle group according to any one of claims 1-6, including: A task planning unit, which is used to establish a three-dimensional equivalent model of the power line inspection area, plan multiple power line inspection routes according to the three-dimensional equivalent model, and configure multiple power line inspection tasks for each power line inspection route; UAV Configuration Unit, which is used to configure a UAV group according to multiple power inspection tasks of each power inspection route. The UAV group includes a master UAV and multiple inspection UAVs. Among them, the master UAV is used to cooperate with and control multiple inspection UAVs to complete power inspection tasks, and aggregate and process the collected data of multiple inspection UAVs. Multiple inspection UAVs respectively perform different types of inspection tasks; Task Execution Unit, which is used to control the UAV group to perform power inspection tasks along the corresponding power inspection route, and obtain multi-modal detection data of the power line in real time. When the master UAV detects a line anomaly, lock the line anomaly point, and send the line anomaly point, the multi-modal detection data and the anomaly detection result to the central processing platform; Anomaly Evolution Unit, which is used for the central processing platform to determine the 3D model of the anomaly area in the 3D equivalent model according to the line anomaly point, and combine the anomaly detection result and the multi-modal detection data to perform anomaly evolution on the line anomaly point to obtain the anomaly development prediction result; Line Maintenance Unit, which is used to perform power line maintenance on the line anomaly point according to the anomaly development prediction result; The Anomaly Evolution Unit includes the following execution steps: In the 3D equivalent model, obtain the 3D spatial coordinates of the line anomaly point; with the 3D spatial coordinates as the center, establish a 3D buffer with a preset radius, extract the 3D equivalent model in the 3D buffer, and generate the 3D model of the anomaly area; according to the anomaly detection result, obtain the anomaly type of the line anomaly point, and according to the anomaly type, match the corresponding anomaly evolution model in the anomaly evolution knowledge base; based on the anomaly detection result and the multi-modal detection data, simulate the dynamic change process of the anomaly state over time through the anomaly evolution model to obtain the time series spatial distribution of the anomaly state; according to the time series spatial distribution, generate the anomaly development prediction result.

Images