Power transmission line intelligent inspection system based on air, space and ground

Abstract

The invention relates to power transmission line inspection, in particular to a space-air-ground-based power transmission line intelligent inspection system, which comprises a satellite remote sensing module, a satellite remote sensing module, a satellite remote sensing image acquisition module, a satellite remote sensing image preprocessing module, a satellite remote sensing image segmentation module, a satellite remote sensing image target detection module and a satellite remote sensing image target detection module, risk information is generated and sent to a task scheduling center; the ground monitoring module collects state index data and meteorological index data of the power transmission line in real time, generates corresponding early warning information when any index triggers alarm, and sends the early warning information to the task scheduling center; the task scheduling center generates an inspection task and an optimal inspection path according to the risk information and the early warning information, allocates the inspection task to the ground monitoring module and the unmanned aerial vehicle inspection module, and sends the optimal inspection path to the unmanned aerial vehicle; according to the technical scheme provided by the invention, the defect that accurate, efficient and intelligent inspection of the power transmission line is difficult in the prior art can be effectively overcome.

Description

Technical Field

[0001] This invention relates to power transmission line inspection, specifically to an intelligent power transmission line inspection system based on air-space-ground communication. Background Technology

[0002] To ensure power supply reliability, transmission line inspections typically employ methods such as manual ground inspections, vehicle / helicopter inspections, fixed online monitoring (e.g., video, micro-meteorology, tension / tilt, conductor temperature, icing sensors, partial discharge detection, etc.), drone close-range inspections, and satellite remote sensing monitoring.

[0003] However, the aforementioned methods of power transmission line inspection have the following main problems: manual / vehicle inspections are time-consuming and limited by terrain; helicopters are costly and easily affected by weather conditions and airspace; drones have limited endurance / payload and are difficult to complete long-distance missions in one go; satellite remote sensing monitoring has the advantage of wide area coverage, but it is easily blocked by clouds, rain, and snow, and thermal infrared is affected by surface radiation; SAR also has speckle noise and geometric distortion; fixed online monitoring provides point-like information, which is difficult to reflect the status of the entire line. Data from multiple sources, including satellites, drones, and ground monitoring, differ significantly in spatial resolution, temporal resolution, coordinate and projection references, timestamp accuracy, and image / point cloud / electrical parameter data formats. There is a lack of cross-source spatiotemporal registration and quality evaluation mechanisms, making it difficult to form a continuous, comparable, and consistent view.

[0004] Single-modal image recognition or threshold alarms are susceptible to factors such as weather, season, terrain, and equipment calibration drift, leading to missed detections or false alarms. Furthermore, the lack of constraints and verification mechanisms integrated with conductor thermal-mechanical-electrical mechanisms (such as sag-temperature-current carrying capacity, icing load-wind vibration, corona / partial discharge thresholds, etc.) results in inconsistent state assessment standards and weak interpretability. The link from "anomaly detection" to "organization for review / handling" relies heavily on human experience. Task allocation and path planning struggle to integrate multiple factors such as weather, no-fly zones, charging / landing points, obstruction geometry, and operational safety for optimization. The lack of uncertainty-based re-inspection triggering and priority management leads to slow response times and high costs.

[0005] Furthermore, the lack of a unified health index and fault probability quantification framework makes it difficult to synthesize global risks from different hidden dangers; channel hazards (such as excessive vegetation height, external damage and intrusion) and electrical / structural hazards (such as abnormal sag, partial discharge, etc.) are often presented on separate platforms, lacking integrated visualization and work order linkage. Dispatch communication and timing are easily restricted in complex terrain / weak coverage areas; engineering issues such as link interruption between edge devices and the cloud, data consistency and encrypted transmission, and compliance with no-fly / height restrictions urgently need to be systematically resolved; "information silos" still exist between SCADA / PMU / online monitoring and remote sensing / UAV data, making linkage difficult.

[0006] Based on the above situation, there is an urgent need for an integrated "air-space-ground" intelligent inspection system for power transmission lines that can quickly detect suspicious points over a wide area, accurately confirm them with high resolution in the near field, and conduct continuous ground monitoring and cross-verification. This system should provide a unified spatiotemporal benchmark and quality assessment, a fusion mechanism for multimodal data, state quantification and uncertainty output based on physical mechanisms, and a closed-loop operation capability that spans "discovery—task issuance—collection—fusion—assessment—re-inspection / handling—knowledge regression" to improve detection rate and interpretability, reduce false alarms, shorten handling time, and balance operating costs with safety and compliance. Summary of the Invention

[0007] (a) Technical problems to be solved In view of the above-mentioned shortcomings of the existing technology, the present invention provides an intelligent inspection system for power transmission lines based on air and ground, which can effectively overcome the shortcomings of the existing technology in that it is difficult to carry out accurate, efficient and intelligent inspection of power transmission lines.

[0008] (II) Technical Solution To achieve the above objectives, the present invention provides the following technical solution: The intelligent inspection system for power transmission lines based on air-space-ground communication includes the following functional modules: The satellite remote sensing module acquires and preprocesses satellite remote sensing images, performs image segmentation and target detection on the preprocessed satellite remote sensing images, extracts image features for risk classification, and generates risk information to be sent to the mission scheduling center. The ground monitoring module collects real-time status index data and meteorological index data of the transmission line, and generates corresponding early warning information and sends it to the task scheduling center when any index triggers an alarm. The task scheduling center generates inspection tasks and optimal inspection paths based on risk information and early warning information, assigns inspection tasks to the ground monitoring module and the UAV inspection module, and sends the optimal inspection path to the UAV. The UAV inspection module performs near-field precision inspections of suspected risk areas identified by the satellite remote sensing module and the ground monitoring module according to the optimal inspection path, and sends the collected real-time monitoring data to the data fusion and evaluation module. The data fusion and evaluation module integrates multi-source data collected by the satellite remote sensing module, ground monitoring module, and UAV inspection module. It performs health assessment and fault prediction by calculating health index and failure probability. At the same time, it performs risk assessment, change monitoring, and report generation based on the verification results of UAVs. The generated evaluation report is sent to engineers to facilitate subsequent maintenance and upkeep.

[0009] Preferably, the satellite remote sensing module acquires satellite remote sensing images and performs preprocessing, including: Geometric correction eliminates geometric distortions in images, ensuring they align with the actual geographic coordinate system and guaranteeing the accuracy of subsequent data analysis. ; Where C represents the geometrically corrected coordinates, and O represents the original coordinates. This is due to correction deviations caused by satellite orbit and sensor perspective; Eliminating atmospheric effects and sensor errors through radiation correction: ; Among them, I corr I represents the intensity of the radiometrically corrected image. meas K represents the original image intensity, and K is the radiometric correction factor used to adjust the sensor response to ensure that the image's radiometric values ​​match the actual reflectivity of the ground objects. Use Gaussian filtering or median filtering to remove noise from the image: ; Among them, I filtered (x,y) represents the filtered pixel value, I(x+i,y+j) represents the neighboring pixel values ​​in the original image, and k is the window size of the filter.

[0010] Preferably, the satellite remote sensing module performs image segmentation and target detection on the preprocessed satellite remote sensing image, extracts image features for risk classification, and generates risk information to be sent to the task scheduling center, including: Convolutional Neural Networks (CNNs) are used to perform image segmentation and object detection on the preprocessed images, and risk classification is performed by extracting image features. ; Where y is the risk classification result, x is the image feature vector, W1 is the classification weight matrix, b1 is the bias term, and softmax is the activation function; Risk information D is generated based on the risk classification results of the preprocessed image. sat And generate risk information D sat Send to the task scheduling center: ; Where t is the timestamp, L is the risk level, and class is the risk category. Here are the latitude and longitude coordinates, s is the confidence level, and r is the radius of influence.

[0011] Preferably, the satellite remote sensing images include optical imaging images, multispectral imaging images, thermal infrared imaging images, and synthetic aperture radar data. The optical imaging images are used to detect vegetation invasion and external damage, the thermal infrared imaging images are used to monitor abnormal temperature problems of power lines, the thermal infrared imaging images help identify different types of objects or materials, and the synthetic aperture radar data can provide clear monitoring in poor weather or at night.

[0012] Preferably, the ground monitoring module collects real-time status indicator data and meteorological indicator data of the transmission line, and generates corresponding early warning information and sends it to the task scheduling center when any indicator triggers an alarm, including: The condition indicators of transmission lines include conductor temperature, conductor tension, and partial discharge current: The conductor temperature is measured by a sensor as T(x,y). If the conductor temperature T(x,y) exceeds a preset temperature threshold T... thresh This will trigger the wire temperature fault alarm F: ; Where (x,y) are the position coordinates; Conductor tension T and conductor sag Relatedly, by calculating the conductor sag Determine if the conductor is within a safe tension range to avoid conductor breakage or displacement due to excessive tension: ; Where w is the weight of the conductor per unit length, and L is the conductor span; Partial discharge monitoring is used to detect electrical faults in conductors; the partial discharge current I... PD Proportional to voltage U, measured by monitoring partial discharge current I PD Able to detect potential electrical faults: ; Where k is a proportionality constant; Meteorological data includes wind speed V wind Temperature and humidity, if wind speed V wind Exceeding the preset wind speed threshold V thresh If this occurs, a wind speed risk alarm R will be triggered. ; When any indicator triggers an alarm, the ground monitoring module will generate a corresponding early warning message D. ground And the generated warning information D ground Send to the task scheduling center: .

[0013] Preferably, the task scheduling center generates inspection tasks and optimal inspection paths based on risk information and early warning information, assigns inspection tasks to the ground monitoring module and the UAV inspection module, and simultaneously sends the optimal inspection path to the UAV, including: According to risk information D sat and early warning information D ground Summary of risk points N is the number of risk points, and the i-th risk point r i Represented as: ; Among them, t i Risk point r i timestamp, L i Risk point r i The risk level, based on the early warning information D ground The generated risk points all correspond to a risk level of 1, c. i Risk point r i Risk categories, Risk point r i latitude and longitude coordinates; Project each risk point from its geographic coordinates to the planned plane coordinates to construct a distance matrix: ; Where, d i,j Risk point r i With risk point r j The distance between them Risk point r j The coordinates are latitude and longitude, and dist is a distance function; Set priority weights: ; in, Risk point r i Priority weight; Starting from the base station The inspection sequence is represented as follows The optimization goal is to shorten the path and prioritize higher-level items: ; in, To optimize the objective function, Starting point With the first inspection point The distance between them For the kth inspection point With the (k+1)th inspection point The distance between them For the kth inspection point The priority weight of the level, k is the inspection position number, , All are weighting coefficients; Determine flight time / battery constraints: ; Where v is the flight speed of the drone, T max This refers to the maximum permitted flight time for the drone. Under the conditions of satisfying the flight time / power constraints and airspace constraints, the optimal inspection path is obtained by solving. And send it to the drone: .

[0014] Preferably, the UAV inspection module performs near-field precision inspection of suspected risk areas identified by the satellite remote sensing module and the ground monitoring module according to the optimal inspection path, and sends the collected real-time monitoring data to the data fusion and evaluation module, including: The drone follows the optimal inspection path Upon reaching the vicinity of each inspection point, high-definition cameras are used to collect high-precision images, thermal imagers are used to collect infrared images, and lidar is used to collect 3D point cloud data. Through the collaboration of vision and lidar, near-field precision inspections are carried out on suspected risk areas identified by satellite remote sensing modules and ground monitoring modules to verify the risks. High-definition cameras capture data at the i-th inspection point. High-precision image I i After feature extraction, the classifier outputs: ; ; Among them, y i This is the probability vector output by the classifier. Here, W1 is the feature extraction function, W2 is the classification weight matrix, b2 is the bias term, and p... img,i For high-precision image I i The probability of belonging to the most likely category; After verifying the risks, verify information D. drone Send to the data fusion and evaluation module: ; Among them, t i 'For inspection points' timestamp, L i 'For inspection points' The risk level, based on the early warning information D ground The generated inspection points all correspond to a risk level of 1, c. i 'For inspection points' Risk categories, Inspection point Latitude and longitude coordinates.

[0015] Preferably, the data fusion and evaluation module fuses multi-source data collected by the satellite remote sensing module, the ground monitoring module, and the UAV inspection module, and performs health assessment and fault prediction by calculating health indices and fault probabilities, including: The Kalman filter algorithm is used to fuse multi-source data collected by satellite remote sensing module, ground monitoring module and UAV inspection module. An adaptive confidence weighting mechanism is adopted to dynamically adjust the weight of each data source according to the reliability and quality of different data sources, so as to perform health assessment and fault prediction by calculating health index and fault probability.

[0016] Preferably, the data fusion and evaluation module performs risk assessment, change monitoring, and report generation based on the UAV's verification results, and sends the generated evaluation report to engineers to facilitate subsequent maintenance and upkeep work, including: Receive verification information D sent by the drone inspection module drone For verification information within the same spatial-temporal neighborhood, perform association and deduplication. If two pieces of verification information satisfy the following: ; The two verification messages will then be grouped into the same event cluster; in, , These are the latitude and longitude coordinates of the inspection points corresponding to the two verified information entries. , These represent the influence radii of the inspection points corresponding to the two verified information items, t. a '、t b 'These are the timestamps of the inspection points corresponding to the two verification messages,' For time windows; All verified information is aggregated into multiple event clusters, each representing a potential risk occurring in a corresponding space and time. Risk assessment, change monitoring, and report generation are performed on all verified information belonging to the same event cluster. The generated assessment reports are sorted by priority and sent to engineers to facilitate subsequent maintenance and upkeep.

[0017] (III) Beneficial Effects Compared with existing technologies, the intelligent inspection system for power transmission lines based on air and space provided by this invention has the following advantages: 1) This invention combines the wide-area monitoring capabilities of satellite remote sensing, the high-resolution data support of UAVs, and real-time ground monitoring, enabling comprehensive monitoring of the health status of power transmission lines. By calculating health indices and fault probabilities, the system can perform health assessments and fault predictions, accurately predict potential faults, and provide early warnings of high-risk areas, thereby reducing the occurrence of sudden faults. 2) This invention adopts a change detection mechanism, which can automatically identify and trigger an alarm when there are significant changes in the state of equipment or environment. By comparing with historical data, the system can capture changes in potential problems of the line in real time, ensuring timely detection and handling of anomalies, and further improving the system's response speed and operation and maintenance efficiency. 3) This invention employs advanced data fusion algorithms such as Kalman filtering, which can integrate multi-source data from satellite remote sensing, UAV inspection, and ground monitoring in real time to provide accurate and comprehensive health assessments. This data fusion method ensures complementary advantages between different functional modules, improving the accuracy of health assessments and the reliability of fault prediction. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without any creative effort.

[0019] Figure 1 This is a schematic diagram of the system of the present invention. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0021] The following describes the specific functional modules of the intelligent transmission line inspection system based on air-space-ground communication provided by this invention, using concrete examples (such as...). Figure 1 As shown), the system functional modules include: The satellite remote sensing module acquires and preprocesses satellite remote sensing images, performs image segmentation and target detection on the preprocessed satellite remote sensing images, extracts image features for risk classification, and generates risk information to be sent to the mission scheduling center. The ground monitoring module collects real-time status index data and meteorological index data of the transmission line, and generates corresponding early warning information and sends it to the task scheduling center when any index triggers an alarm. The task scheduling center generates inspection tasks and optimal inspection paths based on risk information and early warning information, assigns inspection tasks to the ground monitoring module and the UAV inspection module, and sends the optimal inspection path to the UAV. The UAV inspection module performs near-field precision inspections of suspected risk areas identified by the satellite remote sensing module and the ground monitoring module according to the optimal inspection path, and sends the collected real-time monitoring data to the data fusion and evaluation module. The data fusion and evaluation module integrates multi-source data collected by the satellite remote sensing module, ground monitoring module, and UAV inspection module. It performs health assessment and fault prediction by calculating health index and failure probability. At the same time, it performs risk assessment, change monitoring, and report generation based on the verification results of UAVs. The generated evaluation report is sent to engineers to facilitate subsequent maintenance and upkeep.

[0022] I. Satellite Remote Sensing Module 1) The satellite remote sensing module acquires satellite remote sensing images and performs preprocessing, including: Geometric correction eliminates geometric distortions in images, ensuring they align with the actual geographic coordinate system and guaranteeing the accuracy of subsequent data analysis. ; Where C represents the geometrically corrected coordinates, and O represents the original coordinates. This is due to correction deviations caused by satellite orbit and sensor perspective; Eliminating atmospheric effects and sensor errors through radiation correction: ; Among them, I corr I represents the intensity of the radiometrically corrected image. meas K represents the original image intensity, and K is the radiometric correction factor used to adjust the sensor response to ensure that the image's radiometric values ​​match the actual reflectivity of the ground objects. Use Gaussian filtering or median filtering to remove noise from the image: ; Among them, I filtered (x,y) represents the filtered pixel value, I(x+i,y+j) represents the neighboring pixel values ​​in the original image, and k is the window size of the filter.

[0023] 2) The satellite remote sensing module performs image segmentation and target detection on the preprocessed satellite remote sensing images, extracts image features for risk classification, and generates risk information which is sent to the mission scheduling center, including: Convolutional Neural Networks (CNNs) are used to perform image segmentation and object detection on the preprocessed images, and risk classification is performed by extracting image features. ; Where y is the risk classification result, x is the image feature vector, W1 is the classification weight matrix, b1 is the bias term, and softmax is the activation function; Risk information D is generated based on the risk classification results of the preprocessed image. sat And the generated risk information D sat Send to the task scheduling center: ; Where t is the timestamp, L is the risk level, and class is the risk category. Here are the latitude and longitude coordinates, s is the confidence level, and r is the radius of influence.

[0024] 3) Satellite remote sensing images include optical imaging images, multispectral imaging images, thermal infrared imaging images, and synthetic aperture radar data. Optical imaging images are used to detect vegetation invasion and external damage, thermal infrared imaging images are used to monitor abnormal temperature problems of power lines, thermal infrared imaging images help identify different types of objects or materials, and synthetic aperture radar data can provide clear monitoring in bad weather or at night.

[0025] II. Ground Monitoring Module The ground monitoring module collects real-time status and meteorological data of the transmission lines and generates corresponding early warning information when any indicator triggers an alarm, sending it to the task dispatch center. This includes: The condition indicators of transmission lines include conductor temperature, conductor tension, and partial discharge current: The conductor temperature is measured by a sensor as T(x,y). If the conductor temperature T(x,y) exceeds a preset temperature threshold T... thresh This will trigger the wire temperature fault alarm F: ; Where (x,y) are the position coordinates; Conductor tension T and conductor sag Relatedly, by calculating the conductor sag Determine if the conductor is within a safe tension range to avoid conductor breakage or displacement due to excessive tension: ; Where w is the weight of the conductor per unit length, and L is the conductor span; Partial discharge monitoring is used to detect electrical faults in conductors; the partial discharge current I... PD Proportional to voltage U, measured by monitoring partial discharge current IPD Able to detect potential electrical faults: ; Where k is a proportionality constant; Meteorological data includes wind speed V wind Temperature and humidity, if wind speed V wind Exceeding the preset wind speed threshold V thresh If this occurs, a wind speed risk alarm R will be triggered. ; When any indicator triggers an alarm, the ground monitoring module will generate a corresponding early warning message D. ground And the generated warning information D ground Send to the task scheduling center: .

[0026] III. Task Scheduling Center The task scheduling center generates inspection tasks and optimal inspection paths based on risk and early warning information, assigns inspection tasks to the ground monitoring module and the UAV inspection module, and simultaneously sends the optimal inspection path to the UAV, including: According to risk information D sat and early warning information D ground Summary of risk points N is the number of risk points, and the i-th risk point r i Represented as: ; Among them, t i Risk point r i timestamp, L i Risk point r i The risk level, based on the early warning information D ground The generated risk points all correspond to a risk level of 1, c. i Risk point r i Risk categories, Risk point r i latitude and longitude coordinates; Project each risk point from its geographic coordinates to the planned plane coordinates to construct a distance matrix: ; Where, d i,j Risk point r i With risk point r j The distance between them Risk point r j The coordinates are latitude and longitude, and dist is a distance function; Set priority weights: ; in, Risk point r i Priority weight; Starting from the base station The inspection sequence is represented as follows The optimization goal is to shorten the path and prioritize higher-level items: ; in, To optimize the objective function, Starting point With the first inspection point The distance between them For the kth inspection point With the (k+1)th inspection point The distance between them For the kth inspection point The priority weight of the level, k is the inspection position number, , All are weighting coefficients; Determine flight time / battery constraints: ; Where v is the flight speed of the drone, T max This refers to the maximum permitted flight time for the drone. Under the conditions of satisfying the flight time / power constraints and airspace constraints, the optimal inspection path is obtained by solving. And send it to the drone: .

[0027] In the technical solution of this application, the task scheduling center can also trigger composite / dynamic replanning based on feedback from the UAV.

[0028] IV. Unmanned Aerial Vehicle Inspection Module The UAV inspection module performs near-field precision inspections of suspected risk areas identified by the satellite remote sensing module and the ground monitoring module according to the optimal inspection path, and sends the collected real-time monitoring data to the data fusion and evaluation module, including: The drone follows the optimal inspection path Upon reaching the vicinity of each inspection point, high-definition cameras are used to collect high-precision images, thermal imagers are used to collect infrared images, and lidar is used to collect 3D point cloud data. Through the collaboration of vision and lidar, near-field precision inspections are carried out on suspected risk areas identified by satellite remote sensing modules and ground monitoring modules to verify the risks. High-definition cameras capture data at the i-th inspection point. High-precision image I iAfter feature extraction, the classifier outputs: ; ; Among them, y i This is the probability vector output by the classifier. Here, W1 is the feature extraction function, W2 is the classification weight matrix, b2 is the bias term, and p... img,i For high-precision image I i The probability of belonging to the most likely category; After verifying the risks, verify information D. drone Send to the data fusion and evaluation module: ; Among them, t i 'For inspection points' timestamp, L i 'For inspection points' The risk level, based on the early warning information D ground The generated inspection points all correspond to a risk level of 1, c. i 'For inspection points' Risk categories, Inspection point Latitude and longitude coordinates.

[0029] V. Data Fusion and Evaluation Module 1) The data fusion and evaluation module fuses multi-source data collected by the satellite remote sensing module, ground monitoring module, and UAV inspection module. It performs health assessments and fault predictions by calculating health indices and fault probabilities, including: The Kalman filter algorithm is used to fuse multi-source data collected by satellite remote sensing module, ground monitoring module and UAV inspection module. An adaptive confidence weighting mechanism is adopted to dynamically adjust the weight of each data source according to the reliability and quality of different data sources, so as to perform health assessment and fault prediction by calculating health index and fault probability.

[0030] 2) The data fusion and evaluation module performs risk assessment, change monitoring, and report generation based on the UAV's verification results. The generated evaluation report is then sent to engineers to facilitate subsequent maintenance and upkeep work, including: Receive verification information D sent by the drone inspection module drone For verification information within the same spatial-temporal neighborhood, perform association and deduplication. If two pieces of verification information satisfy the following: ; The two verification messages will then be grouped into the same event cluster; in, , These are the latitude and longitude coordinates of the inspection points corresponding to the two verified information entries. , These represent the influence radii of the inspection points corresponding to the two verified information items, t. a '、t b 'These are the timestamps of the inspection points corresponding to the two verification messages,' For time windows;

[0031] All verified information is aggregated into multiple event clusters, each representing a potential risk occurring in a corresponding space and time. Risk assessment, change monitoring, and report generation are performed on all verified information belonging to the same event cluster. The generated assessment reports are sorted by priority and sent to engineers to facilitate subsequent maintenance and upkeep.

[0032] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions will not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A space-air-ground based intelligent inspection system for power transmission lines, characterized in that: Includes the following functional modules: The satellite remote sensing module acquires and preprocesses satellite remote sensing images, performs image segmentation and target detection on the preprocessed satellite remote sensing images, extracts image features for risk classification, and generates risk information to be sent to the mission scheduling center. The ground monitoring module collects real-time status index data and meteorological index data of the transmission line, and generates corresponding early warning information and sends it to the task scheduling center when any index triggers an alarm. The task scheduling center generates inspection tasks and optimal inspection paths based on risk information and early warning information, assigns inspection tasks to the ground monitoring module and the UAV inspection module, and sends the optimal inspection path to the UAV. The UAV inspection module performs near-field precision inspections of suspected risk areas identified by the satellite remote sensing module and the ground monitoring module according to the optimal inspection path, and sends the collected real-time monitoring data to the data fusion and evaluation module. The data fusion and evaluation module integrates multi-source data collected by the satellite remote sensing module, ground monitoring module, and UAV inspection module. It performs health assessment and fault prediction by calculating health index and failure probability. At the same time, it performs risk assessment, change monitoring, and report generation based on the verification results of UAVs. The generated evaluation report is sent to engineers to facilitate subsequent maintenance and upkeep.

2. The intelligent inspection system for power transmission lines based on air-space-ground communication as described in claim 1, characterized in that: The satellite remote sensing module acquires satellite remote sensing images and performs preprocessing, including: Geometric correction eliminates geometric distortions in images, ensuring they align with the actual geographic coordinate system and guaranteeing the accuracy of subsequent data analysis. ； Where C represents the geometrically corrected coordinates, and O represents the original coordinates. This is due to correction deviations caused by satellite orbit and sensor perspective; Eliminating atmospheric effects and sensor errors through radiation correction: ； Among them, I corr I represents the intensity of the radiometrically corrected image. meas K represents the original image intensity, and K is the radiometric correction factor used to adjust the sensor response to ensure that the image's radiometric values ​​match the actual reflectivity of the ground objects. Use Gaussian filtering or median filtering to remove noise from the image: ； Among them, I filtered (x,y) represents the filtered pixel value, I(x+i,y+j) represents the neighboring pixel values ​​in the original image, and k is the window size of the filter.

3. The intelligent inspection system for power transmission lines based on air-space-ground communication as described in claim 2, characterized in that: The satellite remote sensing module performs image segmentation and target detection on the preprocessed satellite remote sensing images, extracts image features for risk classification, and generates risk information which is sent to the task scheduling center, including: Convolutional Neural Networks (CNNs) are used to perform image segmentation and object detection on the preprocessed images, and risk classification is performed by extracting image features. ； Where y is the risk classification result, x is the image feature vector, W1 is the classification weight matrix, b1 is the bias term, and softmax is the activation function; Risk information D is generated based on the risk classification results of the preprocessed image. sat And generate risk information D sat Send to the task scheduling center: ； Where t is the timestamp, L is the risk level, and class is the risk category. Here are the latitude and longitude coordinates, s is the confidence level, and r is the radius of influence.

4. The intelligent inspection system for transmission lines based on air-space-ground communication as described in claim 3, characterized in that: The satellite remote sensing images include optical imaging images, multispectral imaging images, thermal infrared imaging images, and synthetic aperture radar data. The optical imaging images are used to detect vegetation invasion and external damage, the thermal infrared imaging images are used to monitor abnormal temperature problems of power lines, the thermal infrared imaging images help identify different types of objects or materials, and the synthetic aperture radar data can provide clear monitoring in poor weather or at night.

5. The intelligent inspection system for transmission lines based on air-space-ground communication as described in claim 4, characterized in that: The ground monitoring module collects real-time status indicator data and meteorological indicator data of the transmission lines, and generates corresponding early warning information when any indicator triggers an alarm, sending it to the task scheduling center, including: The condition indicators of transmission lines include conductor temperature, conductor tension, and partial discharge current: The conductor temperature is measured by a sensor as T(x,y). If the conductor temperature T(x,y) exceeds a preset temperature threshold T... thresh This will trigger the wire temperature fault alarm F: ； Where (x,y) are the position coordinates; Conductor tension T and conductor sag Relatedly, by calculating the conductor sag Determine if the conductor is within a safe tension range to avoid conductor breakage or displacement due to excessive tension: ； Where w is the weight of the conductor per unit length, and L is the conductor span; Partial discharge monitoring is used to detect electrical faults in conductors; the partial discharge current I... PD Proportional to voltage U, measured by monitoring partial discharge current I PD Able to detect potential electrical faults: ； Where k is a proportionality constant; Meteorological data includes wind speed V wind Temperature and humidity, if wind speed V wind Exceeding the preset wind speed threshold V thresh If this occurs, a wind speed risk alarm R will be triggered. ； When any indicator triggers an alarm, the ground monitoring module will generate a corresponding early warning message D. ground And the generated warning information D ground Send to the task scheduling center: 。 6. The intelligent inspection system for transmission lines based on air-space-ground communication as described in claim 5, characterized in that: The task scheduling center generates inspection tasks and optimal inspection paths based on risk and early warning information, assigns inspection tasks to the ground monitoring module and the UAV inspection module, and simultaneously sends the optimal inspection path to the UAV, including: According to risk information D sat and early warning information D ground Summary of risk points N is the number of risk points, and the i-th risk point r i Represented as: ； Among them, t i Risk point r i timestamp, L i Risk point r i The risk level, based on the early warning information D ground The generated risk points all correspond to a risk level of 1, c. i Risk point r i Risk categories, Risk point r i latitude and longitude coordinates; Project each risk point from its geographic coordinates to the planned plane coordinates to construct a distance matrix: ； Where, d i,j Risk point r i Risk point r j The distance between them Risk point r j The coordinates are latitude and longitude, and dist is a distance function; Set priority weights: ； in, Risk point r i Priority weight; Starting from the base station The inspection sequence is represented as follows The optimization goal is to shorten the path and prioritize higher-level items: ； in, To optimize the objective function, Starting point With the first inspection point The distance between them For the kth inspection point With the (k+1)th inspection point The distance between them For the kth inspection point The priority weight of the level, k is the inspection position number, , All are weighting coefficients; Determine flight time / battery constraints: ； Where v is the flight speed of the drone, T max This refers to the maximum permitted flight time for the drone. Under the conditions of satisfying the flight time / power constraints and airspace constraints, the optimal inspection path is obtained by solving. And send it to the drone: 。 7. The intelligent inspection system for transmission lines based on air-space-ground communication as described in claim 6, characterized in that: The UAV inspection module performs near-field precision inspection of suspected risk areas identified by the satellite remote sensing module and the ground monitoring module according to the optimal inspection path, and sends the collected real-time monitoring data to the data fusion and evaluation module, including: The drone follows the optimal inspection path Upon reaching the vicinity of each inspection point, high-definition cameras are used to collect high-precision images, thermal imagers are used to collect infrared images, and lidar is used to collect 3D point cloud data. Through the collaboration of vision and lidar, near-field precision inspections are carried out on suspected risk areas identified by satellite remote sensing modules and ground monitoring modules to verify the risks. High-definition cameras capture data at the i-th inspection point. High-precision image I i After feature extraction, the classifier outputs: ； ； Among them, y i This is the probability vector output by the classifier. Here, W1 is the feature extraction function, W2 is the classification weight matrix, b2 is the bias term, and p... img,i For high-precision image I i The probability of belonging to the most likely category; After verifying the risks, verify information D. drone Send to the data fusion and evaluation module: ； Among them, t i 'For inspection points' timestamp, L i 'For inspection points' The risk level, based on the early warning information D ground The generated inspection points all correspond to a risk level of 1, c. i 'For inspection points' Risk categories, Inspection point Latitude and longitude coordinates.

8. The intelligent inspection system for transmission lines based on air-space-ground communication as described in claim 7, characterized in that: The data fusion and evaluation module fuses multi-source data collected by the satellite remote sensing module, ground monitoring module, and UAV inspection module, and performs health assessment and fault prediction by calculating health index and fault probability, including: The Kalman filter algorithm is used to fuse multi-source data collected by satellite remote sensing module, ground monitoring module and UAV inspection module. An adaptive confidence weighting mechanism is adopted to dynamically adjust the weight of each data source according to the reliability and quality of different data sources, so as to perform health assessment and fault prediction by calculating health index and fault probability.

9. The intelligent inspection system for transmission lines based on air-space-ground communication as described in claim 8, characterized in that: The data fusion and evaluation module performs risk assessment, change monitoring, and report generation based on the UAV's verification results. The generated evaluation report is then sent to engineers to facilitate subsequent maintenance and upkeep work, including: Receive verification information D sent by the drone inspection module drone For verification information within the same spatial-temporal neighborhood, perform association and deduplication. If two pieces of verification information satisfy the following: ； The two verification messages will then be grouped into the same event cluster; in, , These are the latitude and longitude coordinates of the inspection points corresponding to the two verified information entries. , These represent the influence radii of the inspection points corresponding to the two verified information items, t. a '、t b 'These are the timestamps of the inspection points corresponding to the two verification messages,' For time windows; All verified information is aggregated into multiple event clusters, each representing a potential risk occurring in a corresponding space and time. Risk assessment, change monitoring, and report generation are performed on all verified information belonging to the same event cluster. The generated assessment reports are sorted by priority and sent to engineers to facilitate subsequent maintenance and upkeep.

Images