Power distribution network cable health assessment system and method based on unmanned aerial vehicle

Abstract

The invention discloses a power distribution network cable health assessment system and method based on an unmanned aerial vehicle. The system comprises a data acquisition module, a data processing module and a health early warning module. Relates to the technical field of power distribution network cable health assessment, and solves the technical problem that a health assessment result of a power distribution network cable by an unmanned aerial vehicle is not accurate enough in the prior art. The method comprises the following steps: performing regional division on a power distribution network cable based on line information and environment information to obtain a plurality of inspection regions; acquiring network cable original images of each inspection area in a plurality of continuous periods according to the unmanned aerial vehicle; calculating a network cable health score based on the abnormal feature data; and performing health early warning on the power distribution network cable based on the network cable health score. Regional division is carried out based on the line information and the environment information of the power distribution network cable, the unmanned aerial vehicle is used for collecting the image, the abnormal features are extracted in combination with the image processing technology, then the network cable health score is calculated, the inspection efficiency and accuracy are improved, and safe and stable operation of a power grid is guaranteed.

Description

Technical Field

[0001] The present invention belongs to the technical field of distribution network line health assessment, and specifically provides a distribution network line health assessment system and method based on drones. Background Art

[0002] Traditional distribution network health assessments rely on manual inspections, which are time-consuming and labor-intensive, and lack comprehensive coverage, especially in complex or hazardous environments. With the rapid development of drone technology, it offers advantages in high-altitude viewing angles and flexible maneuverability for power inspections. However, existing drone inspection technology is still limited to single-source data collection and lacks an intelligent health assessment system. Therefore, the present invention proposes a distribution network health assessment system and method based on drones. By using drones equipped with high-precision equipment to collect multidimensional data and combining it with intelligent analysis technology, a comprehensive, efficient, and accurate health assessment of distribution network lines can be achieved.

[0003] Existing technology uses drones to collect line images of distribution lines, extracts features from the line images, obtains abnormal characteristics of the network lines, and then conducts health assessments on the distribution lines based on the abnormal characteristics of the network lines. However, in actual environments, due to the different amounts of adverse factors such as bird activity and weather in different areas, existing technology makes it difficult to control drones to conduct targeted data collection on distribution lines based on the number of adverse factors in each area, resulting in inaccurate health assessment results for distribution lines.

[0004] The present invention provides a distribution network line health assessment system and method based on drones to solve the above technical problems. Summary of the Invention

[0005] The present invention aims to solve at least one of the technical problems existing in the prior art. To this end, the present invention proposes a distribution line health assessment system and method based on drones, which is used to solve the problem that the prior art makes it difficult to control drones to conduct targeted data collection on distribution lines based on the number of adverse factors in each area, resulting in inaccurate health assessment results of distribution lines. In addition, the drones in the prior art lack a line collision warning mechanism, resulting in low safety of drone image collection of distribution lines.

[0006] To achieve the above-mentioned object, a first aspect of the present invention provides a distribution network line health assessment system based on a drone, comprising: a data processing module, and a data acquisition module and a health warning module connected thereto; The data acquisition module is used to obtain line information and environmental information of the distribution network; divide the distribution network into several inspection areas based on the line information and environmental information; and collect original images of the network cables in each inspection area within several consecutive cycles using the drone; The data processing module is used to extract features from the original image of the network cable based on image processing technology to obtain abnormal feature data; The health warning module calculates the health score of the network line based on the abnormal characteristic data; and issues health warnings to the distribution network line based on the health score of the network line.

[0007] Preferably, the regional division of the distribution network lines based on line information and environmental information includes: A1: Extract line information and environmental information of the distribution network. Line information includes power line length, total number of towers, transmission voltage level, and tower location information. Environmental information includes the average daily number of birds landing on power lines and the frequency of lightning strikes. A2: Divide the distribution network into several sub-areas according to the location of the towers, and mark the sub-areas as i; where i = 1, 2, ..., n-1, and n is the total number of towers; A3: Determine whether the daily average number of birds landing on power lines and the frequency of lightning strikes in each sub-area are both greater than the corresponding preset thresholds. If so, mark the inspection level of the corresponding sub-area as level one; if not, jump to A4. A4: Determine whether the daily average number of bird landings on power lines and the frequency of lightning strikes in each sub-area are both less than the corresponding preset thresholds. If so, mark the inspection level of the corresponding sub-area as level three; if not, mark the inspection level of the corresponding sub-area as level two. A5: Merge adjacent sub-areas of the same inspection level into one inspection area.

[0008] Preferably, the collecting of original network images of each inspection area in a plurality of consecutive cycles by the drone includes: Obtain the discharge distance and wind speed of each inspection area; the discharge distance is determined according to the voltage-distance mapping table; The safe collection distance for drones is calculated based on the discharge distance of the distribution grid, wind speed, and the drone's response delay. The image collection strategy for drones in each inspection area is determined based on the safe collection distance and the inspection level of each inspection area. The drone collects original images of the distribution network lines in several consecutive cycles according to the image acquisition strategy of each inspection area.

[0009] It should be noted that the flight direction of the UAV in the present invention is parallel to the power distribution network line.

[0010] Preferably, the calculation of the safe collection distance of the drone based on the discharge distance of the distribution network line, wind speed, humidity and the response delay of the drone includes: Extract the discharge distance of the distribution network line, wind speed, and response delay of the drone; through the formula Calculate the safe collection distance ACL of the drone; where FDL is the discharge distance, XYC is the response delay of the drone, and FS is the wind speed; a1 and a2 are proportional coefficients, and a1 ≧ 1 and a2 > 0; ln() is a logarithmic function with a natural constant as the base; a3 is the wind speed influence coefficient, which is determined based on the positional relationship between the wind speed, drone, and distribution line.

[0011] Preferably, the wind speed influence coefficient is determined according to the positional relationship between the wind speed, the UAV, and the distribution grid line, including: Obtaining wind speed and decomposing it into a first wind speed and a second wind speed; wherein the first wind speed is a wind speed component parallel to the direction of the distribution grid line, and the second wind speed is a wind speed component perpendicular to the direction of the distribution grid line; Obtain the position and flight direction of the UAV, and mark the flight direction of the UAV as direction vector 1; generate direction vector 2 perpendicular to the power distribution line with the position of the UAV as the starting point; Determining whether the direction of the first wind speed is the same as the direction of the direction vector 1; If yes, mark the first tag as 0; if no, mark the first tag as 1; Determine whether the direction of the second wind speed is the same as the direction of the second direction vector; If yes, mark the second tag as 0; if no, mark the second tag as 1; The wind speed influence coefficient a3 is set according to the values ​​of the first tag and the second tag, wherein the values ​​of the first tag and the second tag include {[0,0], [0,1], [1,0], [1,1]}.

[0012] Preferably, the determining of the image acquisition strategy of the drone in each inspection area based on the safe acquisition distance and the inspection level of each inspection area includes: B1: Extract the safe collection distance of the drone and the inspection level of each inspection area; B2: Collect the relative distance between the drone and the power distribution line in the current inspection area in real time, and mark the relative distance as the real-time collection distance; B3: Determine whether the real-time acquisition distance is less than the safe acquisition distance. If so, generate a line collision prompt command and send it to the drone operator's smart terminal. If not, the drone continues to collect original images of the distribution network lines in the current inspection area. Smart terminals include mobile phones, computers, and tablets. B4: Determine whether the inspection level of the inspection area is level one; If so, during the process of the UAV flying from the starting point to the ending point of the corresponding inspection area, control the UAV to perform the first acquisition of the original image of the power distribution network lines within the corresponding inspection area; after the UAV flies to the ending point of the corresponding inspection area, the UAV operator sends a command to adjust the position of the UAV to the opposite side of the power distribution network lines, and the distance between the adjusted UAV and the power distribution network lines within the corresponding inspection area should be greater than the safe acquisition distance; during the process of the UAV flying from the ending point to the starting point of the corresponding inspection area, control the UAV to perform the second acquisition of the original image of the power distribution network lines within the corresponding inspection area; If not, during the process of the UAV flying from the starting point to the ending point of the corresponding inspection area, control the UAV to acquire several original images of the power distribution network lines within the corresponding inspection area.

[0013] Preferably, the feature extraction of the original image of the network line based on the image processing technology includes: Extract the original images of the network lines in each inspection area; perform gray-scale processing on the original images of the network lines to obtain gray-scale images of the network lines; perform contour extraction on the gray-scale images of the network lines based on the edge detection algorithm to obtain several target contour features of the power distribution network lines; compare each of the several target contour features with the standard contour features respectively, and mark the target contour features with different comparison results as abnormal contour features; among them, the standard contour features are obtained by image processing of the images of the power distribution network lines in the normal state; Mark the proportion of the pixel area of the abnormal contour features in the original image of the network line as the proportion of potential fault points; mark the number of occurrences of the abnormal contour features in the original image of the network line as the number of potential fault points; Integrate the proportion of potential fault points and the number of potential fault points into abnormal feature data.

[0014] Preferably, the calculation of the health score of the power distribution network lines based on the abnormal feature data includes: Extract the abnormal feature data; calculate the health score JKP of the network line through the formula where QGB is the proportion of potential fault points, and 0 < QGB < 1, QGS is the number of potential fault points; b1 and b2 are both proportionality coefficients greater than 0.

[0015] Preferably, the health warning of the power distribution network lines based on the health score of the network line includes: Y1: Extract the health score of the network line; Y2: Set the health range threshold [FY1, FY2] of the power distribution network lines, and 0 < FY1 < FY2; Y3: Judge whether the health score of the network line is less than the minimum value FY1 of the health range threshold; if so, mark the health warning level as level one, generate a maintenance command and send it to the smart terminal of the maintenance personnel; if not, jump to Y4; Y4: Determines whether the network cable health score is within the healthy range threshold [FY1, FY2]. If yes, the health warning level is marked as level 2; if not, the health warning level is marked as level 3.

[0016] A second aspect of the present invention provides a distribution network line health assessment method based on a drone, comprising: S1: Obtain line information and environmental information of the distribution network; S2: Divide the distribution network into several inspection areas based on line information and environmental information; S3: The drone collects the original images of the network cables in each inspection area within several consecutive cycles; S4: Extract features from the original image of the network cable based on image processing technology to obtain abnormal feature data; S5: Calculate the network cable health score based on abnormal feature data; S6: Provide health warnings for distribution network lines based on the network line health score.

[0017] Compared with the prior art, the present invention has the following beneficial effects: 1. This invention integrates distribution network line information with environmental information to achieve precise regional demarcation. Using drones to continuously capture images, combined with image processing technology, it extracts abnormal features and calculates a network health score, enabling health warnings. This solution improves inspection efficiency and accuracy, allowing for the timely identification of potential problems, thereby ensuring the safe and stable operation of the power grid.

[0018] 2. The present invention calculates the safe collection distance of the drone by substituting the discharge information of the distribution network line, wind speed and response delay of the drone into a specific formula; comprehensively considering the safety impact of the discharge distance, wind speed and response delay of the drone on the process of drone collecting distribution network lines, so that the calculated safe collection distance is more accurate, which facilitates the planning of the path for the drone to collect distribution network line information based on the safe collection distance, thereby improving the safety of the drone's health assessment of the distribution network line.

[0019] 3. This invention monitors the distance between the drone and the distribution lines within the inspection area in real time. Once the distance falls below a safety threshold, a collision warning is triggered, ensuring that operators can promptly adjust the drone to a safe position, significantly improving the safety of drone inspections. Furthermore, for areas with frequent bird activity and high lightning strike rates, this invention utilizes a dual-sided drone image acquisition strategy to increase the amount of image data in key areas, effectively reducing false detections and missed detections, thereby improving the comprehensiveness and accuracy of drone health assessments of distribution lines. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0021] Figure 1 This is an overall flow chart of the UAV-based distribution network health assessment method of the present invention; Figure 2 Schematic diagram of the principle of the UAV-based distribution network health assessment system of the present invention; Figure 3 This is a flow chart of dividing the distribution network lines into regions in the present invention. DETAILED DESCRIPTION

[0022] The technical solutions of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the embodiments described are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0023] See also Figure 1-Figure 3 , a first embodiment of the present invention provides a distribution network line health assessment system based on a drone, comprising: a data processing module, and a data acquisition module and a health warning module connected thereto; Data acquisition module: used to obtain line information and environmental information of the distribution network; based on the line information and environmental information, the distribution network is divided into several inspection areas; the original images of the network cables in each inspection area within several consecutive cycles are collected by drones; Data processing module: used to extract features from the original image of the network cable based on image processing technology to obtain abnormal feature data; Health warning module: Calculates network cable health scores based on abnormal feature data; issues health warnings for distribution network lines based on network cable health scores.

[0024] In this embodiment, the distribution network lines are divided into regions based on line information and environmental information, including: A1: Extract line information and environmental information of the distribution network. Line information includes power line length, total number of towers, transmission voltage level, and tower location information. Environmental information includes the average daily number of birds landing on power lines and the frequency of lightning strikes. A2: Divide the distribution network into several sub-areas according to the location of the towers, and mark the sub-areas as i; where i = 1, 2, ..., n-1, and n is the total number of towers; A3: Determine whether the daily average number of birds landing on power lines and the frequency of lightning strikes in each sub-area are both greater than the corresponding preset thresholds. If so, mark the inspection level of the corresponding sub-area as level one; if not, jump to A4. A4: Determine whether the daily average number of bird landings on power lines and the frequency of lightning strikes in each sub-area are both less than the corresponding preset thresholds. If so, mark the inspection level of the corresponding sub-area as level three; if not, mark the inspection level of the corresponding sub-area as level two. A5: Merge adjacent sub-areas of the same inspection level into one inspection area.

[0025] In this embodiment, the drone collects original network images of each inspection area within several consecutive cycles, including: Obtain the discharge distance and wind speed of each inspection area; wherein the discharge distance is determined according to the voltage-distance mapping table; the voltage-distance mapping table is used by those skilled in the art to determine the specific discharge distance according to the specific voltage value of the distribution network line; the specific details are shown in the following table:

[0026] The safe collection distance for drones is calculated based on the discharge distance of the distribution grid, wind speed, and the drone's response delay. The image collection strategy for drones in each inspection area is determined based on the safe collection distance and the inspection level of each inspection area. The drone collects original images of the distribution network lines in several consecutive cycles according to the image acquisition strategy of each inspection area.

[0027] It should be noted that the flight direction of the UAV in the present invention is parallel to the power distribution network line.

[0028] In this embodiment, the safe collection distance of the drone is calculated based on the discharge distance of the distribution line, wind speed, humidity, and the response delay of the drone, including: Extract the discharge distance of the distribution network line, wind speed, and response delay of the drone; through the formula Calculate the safe collection distance ACL of the drone; where FDL is the discharge distance, XYC is the response delay of the drone, and FS is the wind speed; a1 and a2 are proportional coefficients, and a1 ≧ 1 and a2 > 0; ln() is a logarithmic function with a natural constant as the base; a3 is the wind speed influence coefficient, which is determined based on the positional relationship between the wind speed, drone, and distribution line.

[0029] For example, the proportional coefficients a1=1.2, a2=0.3, and the wind speed influence coefficient a3=0.3 are set; the discharge distance FDL=1m, the drone's response delay XYC=50ms, and the wind speed FS=4m / s; the safe collection distance ACL of the drone is calculated to be 3.58m by the formula.

[0030] The present invention calculates the safe collection distance of the drone by substituting the discharge information of the distribution network line, wind speed and response delay of the drone into a specific formula; comprehensively considering the safety impact of the discharge distance, wind speed and response delay of the drone on the process of drone collecting distribution network line information, so that the calculated safe collection distance is more accurate, which facilitates the planning of the path for the drone to collect distribution network line information based on the safe collection distance, thereby improving the safety of the drone's health assessment of the distribution network line.

[0031] In this embodiment, the wind speed influence coefficient is determined based on the positional relationship between the wind speed, the drone, and the distribution grid line, including: Obtaining wind speed and decomposing it into a first wind speed and a second wind speed; wherein the first wind speed is a wind speed component parallel to the direction of the distribution grid line, and the second wind speed is a wind speed component perpendicular to the direction of the distribution grid line; Obtain the position and flight direction of the UAV, and mark the flight direction of the UAV as direction vector 1; generate direction vector 2 perpendicular to the power distribution line with the position of the UAV as the starting point; Determining whether the direction of the first wind speed is the same as the direction of the direction vector 1; If yes, mark the first tag as 0; if no, mark the first tag as 1; Determine whether the direction of the second wind speed is the same as the direction of the second direction vector; If yes, mark the second tag as 0; if no, mark the second tag as 1; The wind speed influence coefficient a3 is set according to the values ​​of the first tag and the second tag, wherein the values ​​of the first tag and the second tag include {[0,0], [0,1], [1,0], [1,1]}.

[0032] For example, the specific value of the wind speed influence coefficient a3 is set by relevant experts based on experience. The mapping relationship between the value of the wind speed influence coefficient and the first label and the second label is shown in the following table:

[0033] The present invention comprehensively considers the influence of different positional relationships among wind speed, UAV and distribution network lines on the calculation of safe collection distance of UAV, and decomposes the wind speed parallel to the distribution network line to obtain a first wind speed, decomposes the wind speed perpendicular to the distribution network line to obtain a second wind speed, and obtains direction vector 1 and direction vector 2 according to the position and flight direction of the UAV; according to the direction consistency judgment result of the first wind speed and direction vector 1, and the direction consistency judgment result of the second wind speed and direction vector 2, the values ​​of the first label and the second label are set, and then the values ​​of the wind speed influence coefficient in various situations are set according to relevant experts; so that the calculated safe collection distance is more accurate, and the shooting clarity of the original image of the network cable is guaranteed while ensuring that the UAV is in a safe collection position, which is conducive to improving the safety of the UAV performing health assessment on the distribution network line.

[0034] In this embodiment, the image acquisition strategy of the drone in each inspection area is determined based on the safe acquisition distance and the inspection level of each inspection area, including: B1: Extract the safe collection distance of the drone and the inspection level of each inspection area; B2: Collect the relative distance between the drone and the power distribution line in the current inspection area in real time, and mark the relative distance as the real-time collection distance; B3: Determine whether the real-time acquisition distance is less than the safe acquisition distance. If so, generate a line collision prompt command and send it to the drone operator's smart terminal. If not, the drone continues to collect original images of the distribution network lines in the current inspection area. Smart terminals include mobile phones, computers, and tablets. B4: Determine whether the inspection level of the inspection area is level one; If yes, then during the process of the drone flying from the starting point to the end point of the corresponding inspection area, the drone is controlled to collect the original image of the network cables of the distribution network lines in the corresponding inspection area for the first time; after the drone flies to the end point of the corresponding inspection area, the drone operator sends a command to adjust the position of the drone to the opposite side of the distribution network lines, and after the adjustment, the distance between the drone and the distribution network lines in the corresponding inspection area is greater than the safe collection distance; during the process of the drone flying from the end point to the starting point of the corresponding inspection area, the drone is controlled to collect the original image of the network cables of the distribution network lines in the corresponding inspection area for the second time; If not, the UAV is controlled to collect a number of original images of the power distribution network cables in the corresponding inspection area during the flight from the starting point to the end point of the corresponding inspection area.

[0035] The present invention collects the real-time distance between the drone and the power distribution lines in the current inspection area in real time. When the real-time collected distance is less than the safe collection distance, a line collision prompt command is generated to timely remind the drone operator to adjust the drone to a safe area, which is conducive to improving the safety of the drone for health assessment of the power distribution lines. In addition, for the inspection area with the first-level inspection level (that is, the area with frequent bird activities and lightning strikes), based on the conventional image collection method, the drone collects images on both sides of the power distribution lines, so that the drone collects more original images of the power lines in the area where the power distribution lines are prone to failure, reducing the probability of false detection and missed detection, which is conducive to improving the accuracy of the drone for health assessment of the power distribution lines.

[0036] In this embodiment, feature extraction is performed on the original wire images based on image processing technology, including: Extracting the original wire images of each inspection area; performing gray processing on the original wire images to obtain wire gray images; performing contour extraction on the wire gray images based on an edge detection algorithm to obtain several target contour features of the power distribution lines; comparing each of the several target contour features with the standard contour features respectively, and marking the target contour features with different comparison results as abnormal contour features; wherein, the standard contour features are obtained by image processing of the power distribution lines in the normal state. Marking the proportion of the pixel area of the abnormal contour features in the original wire image as the proportion of potential fault points; marking the number of occurrences of the abnormal contour features in the original wire image as the number of potential fault points. Integrating the proportion of potential fault points and the number of potential fault points into abnormal feature data.

[0037] In this embodiment, calculating the wire health score based on the abnormal feature data includes: Extracting the abnormal feature data; calculating the wire health score JKP through the formula where QGB is the proportion of potential fault points, and 0 < QGB < 1, QGS is the number of potential fault points; b1 and b2 are both proportionality coefficients greater than 0, and the specific values of b1 and b2 are set by experts in relevant fields according to experience.

[0038] Exemplarily, set the proportionality coefficients b1 = 2, b2 = 30; set the proportion of potential fault points QGB = 5%, and the number of potential fault points QGS = 1.2; calculate the wire health score JKP = 66.61 through the formula.

[0039] The present invention calculates the network cable health score by substituting the proportion of potential fault points and the number of potential fault points in the abnormal characteristic data into a specific formula. The proportion of potential fault points and the number of potential fault points can well reflect the health status of the distribution network line, making the calculated network cable health score more accurate, which is conducive to the subsequent maintenance of the distribution network line based on the network cable health score.

[0040] In this embodiment, health warnings for distribution network lines are performed based on the network line health scores, including: Y1: Extract network cable health score; Y2: Set the health range threshold of the distribution network line [FY1, FY2], and 0 <FY1<FY2; Y3: Determine whether the network cable health score is less than the minimum value FY1 of the health range threshold. If yes, mark the health warning level as level 1, generate a maintenance command and send it to the maintenance personnel's smart terminal. If no, jump to Y4. Y4: Determines whether the network cable health score is within the healthy range threshold [FY1, FY2]. If yes, the health warning level is marked as level 2; if not, the health warning level is marked as level 3.

[0041] For example, the network cable health score JKP is set to 66.61, and the health range threshold of the distribution network line [FY1, FY2] is set to [70, 90]. Since the network cable health score is less than the minimum value FY1 of the health range threshold, the health warning level is marked as level one, and a maintenance command is generated and sent to the maintenance personnel's smart terminal.

[0042] A second embodiment of the present invention provides a method for evaluating the health of a distribution network line based on a drone, comprising: S1: Obtain line information and environmental information of the distribution network; S2: Divide the distribution network into several inspection areas based on line information and environmental information; S3: The drone collects the original images of the network cables in each inspection area within several consecutive cycles; S4: Extract features from the original image of the network cable based on image processing technology to obtain abnormal feature data; S5: Calculate the network cable health score based on abnormal feature data; S6: Provide health warnings for distribution network lines based on the network line health score.

[0043] Some of the data in the above formula are calculated by removing the dimensions and taking their numerical values. The formula is a formula that is closest to the actual situation obtained by software simulation of a large amount of collected data; the preset parameters and preset thresholds in the formula are set by technical personnel in this field according to actual conditions or obtained through simulation of a large amount of data.

[0044] Working principle of the present invention: The present invention obtains line information and environmental information of the distribution network line; divides the distribution network line into several inspection areas based on the line information and environmental information; collects original images of the network lines in each inspection area within several consecutive cycles using a drone; extracts features from the original images of the network lines based on image processing technology to obtain abnormal feature data; calculates the network line health score based on the abnormal feature data; and issues health warnings for the distribution network line based on the network line health score.

[0045] The above embodiments are only used to illustrate the technical method of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. A UAV-based distribution network health assessment system, comprising: The data processing module, and the data acquisition module and health warning module connected thereto are characterized in that: The data acquisition module is used to obtain line information and environmental information of the distribution network; divide the distribution network into several inspection areas based on the line information and environmental information; and collect original images of the network cables in each inspection area within several consecutive cycles using the drone; The data processing module is used to extract features from the original image of the network cable based on image processing technology to obtain abnormal feature data; The health warning module calculates the health score of the network line based on the abnormal characteristic data; and issues health warnings to the distribution network line based on the health score of the network line.

2. The UAV-based distribution network health assessment system according to claim 1 is characterized in that: The regional division of the distribution network based on line information and environmental information includes: A1: Extract line information and environmental information of the distribution network. Line information includes power line length, total number of towers, transmission voltage level, and tower location information. Environmental information includes the average daily number of birds landing on power lines and the frequency of lightning strikes. A2: Divide the distribution network into several sub-areas according to the location of the towers, and mark the sub-areas as i; where i = 1, 2, ..., n-1, and n is the total number of towers; A3: Determine whether the daily average number of birds landing on power lines and the frequency of lightning strikes in each sub-area are both greater than the corresponding preset thresholds. If so, mark the inspection level of the corresponding sub-area as level one; if not, jump to A4. A4: Determine whether the daily average number of bird landings on power lines and the frequency of lightning strikes in each sub-area are both less than the corresponding preset thresholds. If so, mark the inspection level of the corresponding sub-area as level three; if not, mark the inspection level of the corresponding sub-area as level two. A5: Merge adjacent sub-areas of the same inspection level into one inspection area.

3. The UAV-based distribution network health assessment system according to claim 2 is characterized in that: The method of collecting original network images of each inspection area in a number of consecutive cycles based on the drone includes: Obtain the discharge distance and wind speed of each inspection area; the discharge distance is determined according to the voltage-distance mapping table; The safe collection distance for drones is calculated based on the discharge distance of the distribution grid, wind speed, and the drone's response delay. The image collection strategy for drones in each inspection area is determined based on the safe collection distance and the inspection level of each inspection area. The drone collects original images of the distribution network lines in several consecutive cycles according to the image acquisition strategy of each inspection area.

4. The UAV-based distribution network health assessment system according to claim 3 is characterized in that: The method of calculating the safe collection distance of the drone based on the discharge distance of the distribution grid, the wind speed, and the response delay of the drone includes: Extract the discharge distance of the distribution network line, wind speed, and response delay of the drone; through the formula Calculate the safe collection distance ACL of the drone; where FDL is the discharge distance, XYC is the response delay of the drone, and FS is the wind speed; a1 and a2 are proportional coefficients, and a1 ≧ 1 and a2 > 0; ln() is a logarithmic function with a natural constant as the base; a3 is the wind speed influence coefficient, which is determined based on the positional relationship between the wind speed, drone, and distribution line.

5. The UAV-based distribution network health assessment system according to claim 4 is characterized in that: The wind speed influence coefficient is determined based on the positional relationship between the wind speed, the UAV, and the distribution grid line, including: Obtaining wind speed and decomposing it into a first wind speed and a second wind speed; wherein the first wind speed is a wind speed component parallel to the direction of the distribution grid line, and the second wind speed is a wind speed component perpendicular to the direction of the distribution grid line; Obtain the position and flight direction of the UAV, and mark the flight direction of the UAV as direction vector 1; generate direction vector 2 perpendicular to the power distribution line with the position of the UAV as the starting point; Determining whether the direction of the first wind speed is the same as the direction of the direction vector 1; If yes, mark the first tag as 0; if no, mark the first tag as 1; Determine whether the direction of the second wind speed is the same as the direction of the second direction vector; If yes, mark the second tag as 0; if no, mark the second tag as 1; The wind speed influence coefficient a3 is set according to the values ​​of the first tag and the second tag.

6. The UAV-based distribution network health assessment system according to claim 3 is characterized in that: The image acquisition strategy of the drone in each inspection area is determined based on the safe acquisition distance and the inspection level of each inspection area, including: B1: Extract the safe collection distance of the drone and the inspection level of each inspection area; B2: Collect the relative distance between the drone and the power distribution line in the current inspection area in real time, and mark the relative distance as the real-time collection distance; B3: Determine whether the real-time acquisition distance is less than the safe acquisition distance; if so, generate a line collision prompt command and send it to the drone operator's smart terminal; if not, the drone continues to collect original images of the distribution network cables in the current inspection area; B4: Determine whether the inspection level of the inspection area is level one; If yes, the drone is controlled to collect original images of the distribution network cables within the corresponding inspection area. After the drone completes the inspection, the drone is adjusted to the opposite side of the distribution network cables, and the drone is controlled to collect original images of the distribution network cables within the corresponding inspection area again. After the adjustment, the distance between the drone and the distribution network cables within the corresponding inspection area is greater than the safe collection distance. If not, the drone is controlled to collect several original images of the distribution network cables in the corresponding inspection area.

7. The UAV-based distribution network health assessment system according to claim 1, characterized in that: The feature extraction of the original image of the network cable based on the image processing technology includes: Extract the original image of the network cable in each inspection area; grayscale process the original image of the network cable to obtain a grayscale image of the network cable; perform contour extraction on the grayscale image of the network cable based on the edge detection algorithm to obtain several target contour features of the distribution network line; compare the several target contour features with the standard contour features respectively, and mark the target contour features with different comparison results as abnormal contour features; among which, the standard contour features are obtained by image processing from the image of the distribution network line in normal state; The proportion of the pixel area of ​​the abnormal contour feature in the original image of the network cable is marked as the proportion of potential fault points; the number of abnormal contour features appearing in the original image of the network cable is marked as the number of potential fault points; The proportion of potential fault points and the number of potential fault points are integrated into abnormal feature data.

8. The UAV-based distribution network health assessment system according to claim 1, characterized in that: The calculation of the network cable health score based on the abnormal feature data includes: Extract abnormal feature data; through the formula Calculate the network cable health score JKP; where QGB is the proportion of potential fault points, and 0 < QGB < 1, QGS is the number of potential fault points; b1 and b2 are both proportionality coefficients greater than 0.

9. The UAV-based distribution network health assessment system according to claim 1, characterized in that: The health warning of the distribution network line based on the network line health score includes: Y1: Extract network cable health score; Y2: Set the health range threshold of the distribution network line [FY1, FY2], and 0 <FY1<FY2; Y3: Determine whether the network cable health score is less than the minimum value FY1 of the health range threshold. If yes, mark the health warning level as level 1, generate a maintenance command and send it to the maintenance personnel's smart terminal. If no, jump to Y4. Y4: Determines whether the network cable health score is within the healthy range threshold [FY1, FY2]. If yes, the health warning level is marked as level 2; if not, the health warning level is marked as level 3.

10. A method for evaluating the health of a distribution network line based on a drone, based on the operation of a distribution network line health assessment system based on a drone according to any one of claims 1 to 9, characterized in that: include: S1: Obtain line information and environmental information of the distribution network; S2: Divide the distribution network into several inspection areas based on line information and environmental information; S3: The drone collects the original images of the network cables in each inspection area within several consecutive cycles; S4: Extract features from the original image of the network cable based on image processing technology to obtain abnormal feature data; S5: Calculate the network cable health score based on abnormal feature data; S6: Provide health warnings for distribution network lines based on the network line health score.

Images