Intelligent fault studying and judging method for distribution line of unmanned aerial vehicle

Abstract

The invention discloses an intelligent fault studying and judging method for a distribution line of an unmanned aerial vehicle, and relates to the technical field of line detection, and the method comprises the steps: classifying each segment according to the environment and fault record of each segment in the distribution line, and determining a detection equipment set of each segment according to a classification result; meanwhile, according to the environment, the detection difficulty and the distance in each segment, the sequence of each segment is determined, then routing inspection is performed, after routing inspection is finished, whether data abnormity exists in each segment is analyzed by using routing inspection data, then associated data corresponding to the data abnormity is analyzed, then secondary routing inspection is performed, and targeted detection is performed on a line. The line inspection comprehensiveness and efficiency are improved, a more reliable basis is provided for fault analysis and diagnosis, the fault recognition accuracy is improved, and the unmanned aerial vehicle inspection effect is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of line detection, and in particular to a method for intelligent fault analysis of a power distribution line of an unmanned aerial vehicle. Background Art

[0002] Using drones to inspect distribution lines can improve the efficiency and accuracy of fault detection and reduce the cost and risk of manual inspections. In traditional technology, when using drones to inspect distribution lines, the inspection content and inspection routes basically need to be formulated by staff, and then the data collected after the inspection is compared with the threshold to identify the faults in the line. Obviously, this inspection has at least the following shortcomings: 1. When inspecting, drones need to carry a lot of collection equipment, but all collection equipment detects the line, which greatly increases the amount of collected data and reduces the efficiency of data collection. At the same time, it also increases the subsequent data storage pressure and the processing pressure of fault identification. The existing technology lacks the ability to classify each line according to the environment and fault records of each line, and then select the collection equipment required for each line according to the classification results, so that each line cannot be targeted for detection, the efficiency of line data collection cannot be improved, and the comprehensiveness of the detection data cannot be guaranteed while accurately detecting each line, which reduces the effect of line inspection.

[0003] 2. In traditional technologies, the inspection routes of drones are mostly confirmed and controlled manually based on experience, without confirming the inspection order of each route according to the complexity of the environment in each route and the difficulty of detecting each route. As a result, it is impossible to complete difficult area inspections when the drone is in the best condition, and it is impossible to ensure the quality of the collected data, and it is impossible to reduce problems such as data ambiguity and missing data caused by equipment performance degradation or environmental factors. It is impossible to provide a more reliable basis for subsequent analysis and diagnosis.

[0004] 3. When a line fails, a chain reaction may occur, resulting in multiple data anomalies in the line. Therefore, comprehensive detection of related chain data based on the abnormal data in the line can ensure the comprehensiveness of line data collection. However, traditional technologies lack secondary detection of the line based on the abnormal data in the line, and cannot obtain the status of related chain data, thereby failing to ensure the comprehensiveness of line inspections and providing an effective reference for subsequent fault identification, reducing the accuracy of fault identification and affecting line operation. Summary of the invention

[0005] In view of the above-mentioned technical deficiencies, the purpose of the present invention is to provide a method for intelligent fault analysis of UAV power distribution lines.

[0006] In order to solve the above technical problems, the present invention adopts the following technical solution: The present invention provides a method for intelligent fault analysis of UAV distribution lines, comprising the following steps: S1, dividing the distribution lines into sections, obtaining the detection records, terrain data and meteorological data of each historical detection of each section, and classifying each section.

[0007] S2. According to the classification results of each segment, a set of detection equipment corresponding to the drone when detecting each segment is selected, and then the drone is used to detect each segment to obtain the detection data of each segment.

[0008] S3. Utilize the detection data of each segment to analyze whether the detection data of each segment is sufficient, record each segment with sufficient detection data as a Class I segment, and analyze the safety of each Class I segment. If the safety is insufficient, identify the fault data of each Class I segment with insufficient safety. Record each segment with insufficient detection data as a Class II segment, and analyze the detection equipment set corresponding to the secondary detection of each Class II segment based on the detection data of each segment.

[0009] S4. Start the UAV for secondary inspection to obtain secondary inspection data corresponding to each Class II segment, and use the inspection data and secondary inspection data corresponding to each Class II segment to analyze the safety of each Class II segment. If the safety is insufficient, identify the fault data of each Class II segment with insufficient safety.

[0010] S5. Display the fault data of each Category I segment and each Category II segment with insufficient safety.

[0011] The beneficial effects of the present invention are: 1. The present invention provides an intelligent fault analysis method for drone distribution lines. According to the environment of each segment in the distribution line and the fault records, each segment is classified, and the detection equipment set of each segment is confirmed according to the classification result. At the same time, according to the environment, detection difficulty and spacing in each segment, the order of each segment is confirmed, and then an inspection is carried out. After the inspection, the inspection data is used to analyze whether there are data anomalies in each segment, and then the related data corresponding to the data anomalies are analyzed, and then a secondary inspection is carried out to perform targeted detection on the line, thereby improving the comprehensiveness and efficiency of the line inspection, providing a more reliable basis for fault analysis and diagnosis, increasing the accuracy of fault identification, and improving the effect of drone inspection.

[0012] 2. The present invention classifies each section in the distribution line according to the environment and fault records of each section, and confirms the detection equipment set of each section based on the classification results, thereby realizing targeted detection of each section of the line, greatly improving the efficiency of line data collection, and ensuring the comprehensiveness of the detection data while accurately detecting each section of the line, ensuring the effectiveness of line inspection.

[0013] 3. The present invention analyzes the environmental complexity and line detection difficulty in each segment, and then confirms the inspection order of each segment according to the distance between the segments, to ensure that the drone completes the inspection of difficult areas when it is in the best state, to ensure the quality of the collected data, to reduce problems such as data ambiguity and missing data caused by equipment performance degradation or environmental interference, to provide a more reliable basis for subsequent analysis and diagnosis, and at the same time, combined with the distance between the segments, to reduce the flight time and energy consumption of the drone, extend the flight time of the drone, and save power resources.

[0014] 4. After the first inspection, the present invention screens out abnormal data based on the detection data of each segment, and then confirms the associated data corresponding to the abnormal data based on the historical detection records, and then selects the corresponding acquisition equipment for secondary detection to ensure the comprehensiveness of the line inspection, provide an effective reference for the identification of subsequent faults, and improve the accuracy of fault identification. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. 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 creative work.

[0016] Figure 1 The present invention is a schematic flow chart of the steps for implementing the method. DETAILED DESCRIPTION

[0017] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0018] See also Figure 1 As shown, a method for intelligent fault analysis of UAV distribution lines includes the following steps: S1, dividing the distribution lines into sections, obtaining the detection records, terrain data and meteorological data of each historical detection of each section, and classifying each section.

[0019] In the above, the inspection records, terrain data and meteorological data of each historical inspection of each section are obtained from the historical inspection records of the line.

[0020] The terrain data includes but is not limited to the slope, horizontal curvature and vertical curvature of the ground surface. The meteorological data includes but is not limited to the temperature and wind speed.

[0021] The terrain changes due to artificial cultivation or natural disasters, so the terrain data needs to be detected and updated every time it is detected.

[0022] In a specific embodiment, the classification process of each segment is as follows: obtain reference smooth terrain data and reference safe meteorological data from the line control center, compare the terrain data and meteorological data of each historical detection of each segment with the reference smooth terrain data and reference safe meteorological data, and analyze the environmental characteristic values ​​of each segment.

[0023] It should be noted that the reference smooth terrain data and the reference safe meteorological data represent the corresponding critical values ​​of terrain data and meteorological data when there is no impact on the line. They are jointly formulated by multiple experts based on historical data and their own experience.

[0024] The environmental characteristic value includes values ​​of 1 and 0. When the environmental characteristic value is 1, it indicates that the environment is normal; when the environmental characteristic value is 0, it indicates that the environment is dangerous.

[0025] In the above, the analysis process of the environmental characteristic values ​​of each segment is as follows: the reference stable terrain data and the reference safe meteorological data are recorded as D and Q respectively, and the terrain data and meteorological data Q of each historical detection of each segment are recorded as f The average terrain data and meteorological data of each segment are obtained by calculating the mean value, which are denoted as D f and, f represents the number of each segment, and f is a positive integer.

[0026] The environmental characteristic value expression is: D f Represents the environmental characteristic value of the fth segment.

[0027] Based on the detection records of each historical detection of each segment, the location area and fault type of each fault in each historical detection of each segment are obtained, thereby counting the location areas corresponding to each fault type in each segment, and analyzing the dangerous characteristic values ​​of each fault type in each segment.

[0028] The hazard characteristic value includes values ​​of 1 and 0. When the hazard characteristic value of a certain fault type in a segment is 1, it indicates that the possibility of the fault type existing in the segment is low. When the hazard characteristic value of a certain fault type in a segment is 0, it indicates that the possibility of the fault type existing in the segment is high.

[0029] In the above, the analysis process of the dangerous characteristic value of each fault type in each segment is as follows: the three-dimensional model of the distribution line is obtained from the line control center, and each position area corresponding to each fault type in each segment is imported into the three-dimensional model of the distribution line. In the three-dimensional model, the set color corresponding to each fault type is used to mark each position area corresponding to each fault type in each segment; thereby, the total area and maximum saturation of the color area corresponding to each fault type in each segment are obtained from the three-dimensional model of the distribution line, which are respectively recorded as S fn and B fn , f represents the number of each segment, n represents the number of each fault type, and both f and n are positive integers.

[0030] It should be noted that when two or more position areas overlap, the color will become darker and the saturation will be higher.

[0031] Dangerous characteristic value analysis model expression: In the formula, α fn represents the dangerous characteristic value of the nth fault type in the fth segment, S f represents the total area of ​​the line in the fth segment, B n It indicates the saturation of the set color corresponding to the nth fault type, ε1 and ε2 respectively indicate the proportional coefficient of the set color area corresponding to the total area and the proportional coefficient of the maximum saturation, e indicates the natural constant, and β is the preset hazard assessment coefficient threshold.

[0032] It should be noted that the Pearson correlation coefficient between the total area and maximum saturation of the color area corresponding to the fault type and the dangerous characteristic value of the fault type is analyzed by using the detection records and the calculation method of the Pearson correlation coefficient, which are recorded as a1 and a2 respectively. The calculation method of the Pearson correlation coefficient is common knowledge and will not be described in detail here.

[0033] The hazard assessment coefficient threshold is a reference value or critical value for judging the probability of occurrence of a fault type, and is jointly formulated by multiple experts based on historical data and their own experience. When the calculated data is greater than the hazard assessment coefficient threshold, it indicates that the probability of occurrence is high, otherwise it indicates that the probability of occurrence is low; for example, if the hazard assessment coefficient threshold is 5, The calculated value of is 4, 4<5, indicating that the probability of the nth fault type in the fth segment is low.

[0034] If the environmental characteristic value in a segment is 0, it indicates that all fault types in the segment need precise inspection, and the segment is classified as a comprehensive precise inspection type; when the environmental characteristic value in a segment is 1, then all fault types in the segment with a dangerous characteristic value of 0 need precise inspection, and the segment is classified as a partial precise inspection type; if the environmental characteristic value in a segment is 1, and the dangerous characteristic values ​​of all fault types in the segment are 1, then all fault types in the segment do not need precise inspection, and the segment is classified as a basic detection type, thereby completing the classification of each segment.

[0035] S2. According to the classification results of each segment, a set of detection equipment corresponding to the drone when detecting each segment is selected, and then the drone is used to detect each segment to obtain the detection data of each segment.

[0036] In a specific embodiment, the use of drones to inspect each segment is as follows: based on the classification results of each segment, statistics are collected on the segments corresponding to the comprehensive precision inspection type, partial precision inspection type and basic inspection type, and at the same time, the set of inspection equipment corresponding to the basic line inspection and the inspection equipment corresponding to the precision inspection of each fault type are obtained from the database, and then the set of inspection equipment for each segment is confirmed.

[0037] Preferably, the process of confirming the detection equipment set of each segment is as follows: the detection equipment set corresponding to the basic detection type for each segment is the detection equipment set corresponding to the basic detection of the line; the detection equipment set corresponding to the comprehensive fine inspection type for each segment is the sum of the detection equipment set corresponding to the basic detection of the line and the detection equipment corresponding to the fine inspection of each fault type; the fault types in each segment corresponding to the partial fine inspection type with a dangerous characteristic value of 0 are obtained as the fine inspection fault types, and the detection equipment set corresponding to each segment of the partial fine inspection type is the sum of the detection equipment set corresponding to the basic detection of the line and the detection equipment corresponding to the fine inspection of each fine inspection fault type.

[0038] Preferably, the basic detection types include terrain detection and basic image detection, and the detection equipment set used includes lidar, high-definition camera and GPD positioning system. When other detection equipment collects data, the GPD positioning system collects the precise location of the collected data and associates it with the data collected by other detection equipment to obtain accurate data collection location information.

[0039] Fault types include temperature anomaly and discharge fault, etc. The corresponding detection equipment for temperature anomaly and discharge fault precision inspection are infrared monitoring equipment and ultraviolet monitoring equipment respectively;

[0040] Assuming that the precise inspection fault type in some precise inspection types is a discharge fault, the detection equipment set corresponding to some precise inspection types includes lidar, high-definition camera, GPD positioning system and ultraviolet monitoring equipment.

[0041] The set of testing equipment corresponding to the comprehensive and precise inspection type includes lidar, high-definition camera, GPD positioning system and testing equipment corresponding to all fault types.

[0042] The present invention classifies each section in the distribution line according to the environment and fault records of each section, and confirms the detection equipment set of each section according to the classification result, thereby realizing targeted detection of each section of the line, greatly improving the efficiency of line data collection, and ensuring the comprehensiveness of the detection data while accurately detecting each section of the line, ensuring the effect of line inspection.

[0043] The segments corresponding to the comprehensive and precise inspection type are recorded as marked segments. The real-time meteorological data corresponding to each marked segment, the terrain data corresponding to the last historical inspection, and the three-dimensional model of the distribution line are extracted. The priority value of each marked segment is analyzed, the starting inspection segment of the UAV is confirmed, and then the inspection order of each segment of the UAV is confirmed.

[0044] In the above, the specific process of confirming the detection order of each segment by the drone is as follows: S2-11, obtaining the real-time meteorological data corresponding to each marked segment and the terrain data corresponding to the last historical detection, and using the three-dimensional model of the distribution line to calculate the priority value of each marked segment, and select the marked segment corresponding to the maximum priority value as the starting detection segment.

[0045] Preferably, the priority value of each marked segment is calculated by the following specific process: obtaining the real-time meteorological data corresponding to each marked segment from the meteorological center, recording the real-time meteorological data corresponding to each marked segment and the terrain data corresponding to the last historical detection as Q j and D j , substitute into the calculation formula Get the environmental difficulty coefficient of the drone when detecting the jth marked segment j represents the number of each marked segment, and j is a positive integer.

[0046] The line density, maximum line height and maximum line height difference of each marked segment are obtained from the three-dimensional model of the distribution line, and the line density, maximum line height and maximum line height difference of each marked segment are normalized and marked as b1 and b2. j 、b2 j and b3 j , In the formula, They are the proportional coefficient of the set line density, the proportional coefficient of the maximum line height, and the proportional coefficient of the maximum line height difference. It indicates the difficulty coefficient of the line detection when the UAV detects the jth marked segment; the priority value calculation formula is used: Get the priority value of the jth marked segment

[0047] In the above, line density is the total length of railways or roads per unit area.

[0048] It should be noted that the detection records were used to calculate the Pearson correlation coefficient to analyze the line density, the maximum line height, and the Pearson correlation coefficient between the maximum line height and the line detection difficulty coefficient, which were recorded as a3, a4, and a5 respectively.

[0049] The greater the line density, the denser the lines, the more complex the line structure, and the more difficult it is to collect data. Similarly, the greater the line height, the higher the flying altitude of the drone during collection, which increases the difficulty of collection. The greater the line height difference, the greater the line fluctuation, and the more complex the flight control of the drone during inspection, resulting in greater difficulty in collection. Therefore, the difficulty of line collection is evaluated by line density, maximum line height, and maximum line height, and the confirmation of subsequent inspection order is used to ensure priority collection in difficult areas, thereby improving the collection effect.

[0050] S2-12. Calculate the priority value of each segment according to the calculation method of the priority value of each marked segment, and record it as At the same time, obtain the corresponding position of each segment and confirm the distance between each segment and the actual detection segment, which is recorded as L f , using the calculation formula Get the first priority value corresponding to each segment Where L max The maximum deployment distance of the set drone; select the segment corresponding to the maximum first priority value as the second detection segment.

[0051] S2-13. Obtain the distance between each remaining segment and the second detection segment, calculate the second priority value corresponding to each remaining segment according to the analysis method of S2-12, and select the segment corresponding to the maximum second priority value as the third detection segment. In this way, confirm the detection order of each segment by the drone.

[0052] The drone is equipped with all the collection equipment to perform inspections in the order of the inspection of each segment. When inspecting each segment, the set of inspection equipment corresponding to each segment is turned on for inspection, and the collection data of each inspection equipment in each segment is obtained as the inspection data of each segment.

[0053] The present invention analyzes the environmental complexity and line detection difficulty in each segment, and then confirms the inspection order of each segment according to the distance between the segments, ensuring that the drone completes the difficult area inspection when it is in the best state, ensuring the quality of the collected data, and reducing problems such as data ambiguity and missing due to equipment performance degradation or environmental interference, providing a more reliable basis for subsequent analysis and diagnosis. At the same time, combined with the distance between the segments, the flight time and energy consumption of the drone are reduced, the flight time of the drone is extended, and power resources are saved.

[0054] S3. Utilize the detection data of each segment to analyze whether the detection data of each segment is sufficient, record each segment with sufficient detection data as a Class I segment, and analyze the safety of each Class I segment. If the safety is insufficient, identify the fault data of each Class I segment with insufficient safety. Record each segment with insufficient detection data as a Class II segment, and analyze the detection equipment set corresponding to the secondary detection of each Class II segment based on the detection data of each segment.

[0055] In a specific embodiment, the analysis of whether the detection data of each segment is sufficient is carried out as follows: the collected data corresponding to each detection device is obtained from the detection data of each segment, and the safety interval of the collected data corresponding to each detection device is obtained from the line control center, and the collected data corresponding to each detection device in each segment is compared with the safety interval. When there is at least one detection device in a segment whose collected data is not within the safety interval, the detection devices whose collected data is not within the safety interval are recorded as marked devices; the marked devices in each segment are counted, and the fault type corresponding to the abnormal collected data of each marked device is obtained from the line control center, the fault type corresponding to each marked device in each segment is obtained, and then the associated fault types of the fault types corresponding to the fault types of each marked device in each segment are screened out.

[0056] In the above, the screening process of each associated fault type corresponding to each marked device in each segment is as follows: obtain each fault type corresponding to each historical detection in each segment from the historical detection record, compare the fault type corresponding to each marked device in each segment with the fault type corresponding to each historical detection in each segment, if the fault type corresponding to a marked device in a segment is the same as the fault type corresponding to a historical detection in a segment, then use the historical detection in the segment as the associated historical detection corresponding to the marked device in the segment, so as to obtain each associated historical detection corresponding to each marked device in each segment.

[0057] Extract each fault type corresponding to each associated historical detection corresponding to each marked device in each segment, count the number of each fault type in the associated historical detection corresponding to each marked device in each segment, and then divide it by the number of associated historical detections to obtain the association rate of each fault type in the associated historical detection corresponding to each marked device in each segment, and take each fault type with a correlation rate greater than a preset correlation rate threshold as each associated fault type corresponding to the fault type of each marked device in each segment.

[0058] It should be noted that the correlation rate threshold represents the critical value of the correlation strength. When the correlation rate is greater than the correlation rate threshold, it indicates that the correlation is strong, otherwise it indicates that the correlation is weak. The correlation rate threshold is set by multiple experts based on experience and historical data. Assuming that the correlation rate threshold is 60%, the correlation between the two fault types is 58%, and 58% < 60%, it indicates that the correlation between the two fault types is weak.

[0059] The detection equipment corresponding to each fault type precision inspection is obtained from the line control center, thereby obtaining the detection equipment for precision inspection of each associated fault type corresponding to each marking device in each section as the detection equipment for precision inspection of each associated fault type corresponding to each marking device in each section.

[0060] Compare the detection devices of the associated fine inspection corresponding to each marking device in each segment with the detection device set of each segment. If the detection devices of the associated fine inspection corresponding to each marking device in a segment are all in the detection device set of the segment, then the detection data of the segment is sufficient, and the segment is recorded as a first-class segment.

[0061] If a detection device of an associated fine inspection corresponding to a marking device in a segment is not in the detection device set of the segment, the detection data of the segment is insufficient, and the segment is recorded as a second-category segment. At the same time, the detection device of the associated fine inspection is used as the detection device corresponding to the secondary inspection of the segment. This is used to analyze whether the detection data of each segment is sufficient, and the detection device set corresponding to the secondary inspection of each second-category segment is counted.

[0062] In another specific embodiment, the specific process of analyzing the safety of each Class I segment is as follows: extracting the collected data of each detection device in each Class I segment, and comparing it with the safety interval of the collected data corresponding to each detection device; when there is at least one detection device in a certain Class I segment whose collected data is not within the safety interval, it is determined that the safety of the Class I segment is insufficient and there is a fault, and the fault type of the detection device corresponding to the collected data not within the safety interval is extracted, and each fault type of each Class I segment with insufficient safety is obtained; at the same time, the location area of ​​each fault type is obtained from the detection data of each Class I segment with insufficient safety as the fault data of each Class I segment with insufficient safety.

[0063] S4. Start the UAV for secondary inspection to obtain secondary inspection data corresponding to each Class II segment, and use the inspection data and secondary inspection data corresponding to each Class II segment to analyze the safety of each Class II segment. If the safety is insufficient, identify the fault data of each Class II segment with insufficient safety.

[0064] In a specific embodiment, the drone is started for secondary inspection, and the specific process is as follows: extract the real-time meteorological data corresponding to each second category segment and the terrain data corresponding to the last historical inspection, and use the three-dimensional model of the distribution line to confirm the starting inspection segment corresponding to the drone secondary inspection.

[0065] It should be noted that the confirmation method of the starting detection segment corresponding to the drone secondary detection is the same as the confirmation method of the starting detection segment corresponding to the drone detection, which will not be repeated here.

[0066] Then obtain the distance between other second-category segments and the starting detection segment, calculate the first priority value of other second-category segments, confirm the second detection segment corresponding to the drone secondary detection, and thereby confirm the detection order of each second-category segment corresponding to the drone secondary detection.

[0067] It should be noted that the detection order of each second-category segment corresponding to the secondary detection of the drone is the same as the detection order confirmation method of each segment of the drone detection, which will not be repeated here.

[0068] The UAV is equipped with a set of detection equipment corresponding to the secondary detection of each Class II segment, and performs detection in sequence according to the detection order of each Class II segment. When detecting each Class II segment, the set of detection equipment corresponding to the secondary detection of each Class II segment is started to collect secondary detection data corresponding to each Class II segment.

[0069] After the first inspection, the present invention screens out abnormal data based on the detection data of each segment, then confirms the associated data corresponding to the abnormal data based on the historical detection records, and then selects the corresponding acquisition equipment for secondary detection, thereby ensuring the comprehensiveness of the line inspection, providing an effective reference for the identification of subsequent faults, and improving the accuracy of fault identification.

[0070] In another specific embodiment, the specific process of analyzing the safety of each Class II segment is as follows: acquiring the collection data corresponding to each detection device from the detection data and secondary detection data corresponding to each Class II segment, and analyzing the safety of each Class II segment according to the safety analysis process of each Class I segment, and simultaneously obtaining the fault data of each Class II segment with insufficient safety.

[0071] S5. Display the fault data of each Category I segment and each Category II segment with insufficient safety.

[0072] The embodiment of the present invention classifies each segment according to the environment and fault records of each segment in the distribution line, and confirms the detection equipment set of each segment according to the classification result, and confirms the order of each segment according to the environment, detection difficulty and spacing in each segment, and then conducts inspection. After the inspection, the inspection data is used to analyze whether there is data anomaly in each segment, and then the related data corresponding to the data anomaly is analyzed, and then a secondary inspection is performed to perform targeted detection on the line, thereby improving the comprehensiveness and efficiency of the line inspection, providing a more reliable basis for fault analysis and diagnosis, increasing the accuracy of fault identification, and improving the effect of drone inspection.

[0073] The above contents are merely examples and explanations of the concept of the present invention. Those skilled in the art may make various modifications or additions to the specific embodiments described or replace them in a similar manner. As long as they do not deviate from the concept of the invention or exceed the scope defined in this specification, they shall all fall within the protection scope of the present invention.

Claims

1. A method for intelligent fault analysis of UAV power distribution lines, characterized in that: The steps include: S1. Divide the distribution line into sections, obtain the inspection records, terrain data and meteorological data of each historical inspection of each section, and classify each section; S2. According to the classification results of each segment, a set of detection equipment corresponding to the drone when detecting each segment is selected, and then the drone is used to detect each segment to obtain detection data of each segment; S3. Analyze whether the detection data of each segment is sufficient by using the detection data of each segment, record each segment with sufficient detection data as a first-class segment, and analyze the safety of each first-class segment. If the safety is insufficient, identify the fault data of each first-class segment with insufficient safety; Each segment with insufficient test data is recorded as a second-class segment, and based on the test data of each segment, the set of test equipment corresponding to the secondary test of each second-class segment is analyzed; S4, starting the drone to perform secondary inspection, obtaining secondary inspection data corresponding to each second-category segment, using the inspection data and secondary inspection data corresponding to each second-category segment, analyzing the safety of each second-category segment, and if the safety is insufficient, identifying the fault data of each second-category segment with insufficient safety; S5. Display the fault data of each Category I segment and each Category II segment with insufficient safety.

2. According to claim 1, a method for intelligent fault analysis of UAV power distribution lines is characterized in that: The classification process of each segment is as follows: Obtain reference stable terrain data and reference safe meteorological data from the line control center, compare the terrain data and meteorological data of each historical test of each segment with the reference stable terrain data and reference safe meteorological data, and analyze the environmental characteristic values ​​of each segment; The environmental characteristic value includes the values ​​of sum and 0. When the environmental characteristic value is 1, it indicates that the environment is normal; When the environmental characteristic value is 0, it indicates that the environment is dangerous; Based on the detection records of each historical detection of each segment, the location area and fault type of each fault detected in each segment are obtained, thereby counting the location areas corresponding to each fault type in each segment, and analyzing the dangerous characteristic values ​​of each fault type in each segment; The hazard characteristic value includes values ​​of 1 and 0. When the hazard characteristic value of a certain fault type in a certain segment is 1, it indicates that the possibility of the existence of the fault type in the segment is low. When the hazard characteristic value of a certain fault type in a certain segment is 0, it indicates that the possibility of the existence of the fault type in the segment is high. If the environmental characteristic value in a segment is 0, it means that all fault types in the segment need to be carefully inspected, and the segment is classified as a comprehensive and careful inspection type; When the environmental characteristic value in a segment is 1, each fault type with a dangerous characteristic value of 0 in the segment needs to be fine-checked, and the segment is classified as a partial fine-check type; If the environmental characteristic value in a segment is 1, and the dangerous characteristic values ​​of each fault type in the segment are all 1, then all fault types in the segment do not need to be fine-checked, and the segment is classified as a basic detection type, thereby completing the classification of each segment.

3. The method for intelligent fault analysis of UAV power distribution lines according to claim 2 is characterized in that: The analysis process of the dangerous characteristic value of each fault type in each segment is as follows: The three-dimensional model of the distribution line is obtained from the line control center, and each location area corresponding to each fault type in each segment is imported into the three-dimensional model of the distribution line. In the three-dimensional model, each location area corresponding to each fault type in each segment is marked with a set color corresponding to each fault type; thereby, the total area and maximum saturation of the color area corresponding to each fault type in each segment are obtained from the three-dimensional model of the distribution line, which are recorded as S fn and B fn , f represents the number of each segment, n represents the number of each fault type, and both f and n are positive integers; Dangerous characteristic value analysis model expression: In the formula, α fn represents the dangerous characteristic value of the nth fault type in the fth segment, S f represents the total area of ​​the line in the fth segment, B n It indicates the saturation of the set color corresponding to the nth fault type, ε1 and ε2 respectively indicate the proportional coefficient of the set color area corresponding to the total area and the proportional coefficient of the maximum saturation, e indicates the natural constant, and β is the preset hazard assessment coefficient threshold.

4. The method for intelligent fault analysis of UAV power distribution lines according to claim 3 is characterized in that: The specific process of using a drone to detect each segment is as follows: According to the classification results of each section, the sections corresponding to the comprehensive precision inspection type, partial precision inspection type and basic inspection type are counted, and at the same time, the detection equipment set corresponding to the basic line inspection and the detection equipment corresponding to the precision inspection of each fault type are obtained from the database, and then the detection equipment set of each section is confirmed; Each segment corresponding to the comprehensive and precise inspection type is recorded as a marked segment, and the real-time meteorological data corresponding to each marked segment, the terrain data corresponding to the last historical inspection, and the three-dimensional model of the distribution line are extracted. The priority value of each marked segment is analyzed, and the starting inspection segment of the drone is confirmed, and then the inspection order of each segment of the drone is confirmed; The drone is equipped with all the collection equipment to perform inspections in the order of the inspection of each segment. When inspecting each segment, the set of inspection equipment corresponding to each segment is turned on for inspection, and the collection data of each inspection equipment in each segment is obtained as the inspection data of each segment.

5. The method for intelligent fault analysis of UAV power distribution lines according to claim 4 is characterized in that: The specific process of confirming the detection order of each segment of the drone detection is as follows: S2-11, obtaining the real-time meteorological data corresponding to each marked segment and the terrain data corresponding to the last historical detection, and using the three-dimensional model of the distribution line to calculate the priority value of each marked segment, and selecting the marked segment corresponding to the maximum priority value as the starting detection segment; S2-12. Calculate the priority value of each segment according to the calculation method of the priority value of each marked segment, and record it as At the same time, obtain the corresponding position of each segment and confirm the distance between each segment and the actual detection segment, which is recorded as L f , using the calculation formula Get the first priority value corresponding to each segment Where L max The maximum deployment distance of the drone is set; the segment corresponding to the maximum first priority value is selected as the second detection segment; S2-13. Obtain the distance between each remaining segment and the second detection segment, calculate the second priority value corresponding to each remaining segment according to the analysis method of S2-12, and select the segment corresponding to the maximum second priority value as the third detection segment. In this way, confirm the detection order of each segment by the drone.

6. The method for intelligent fault analysis of UAV power distribution lines according to claim 1 is characterized in that: The specific process of analyzing whether the detection data of each segment is sufficient is as follows: Acquire the collected data corresponding to each detection device from the detection data of each segment, and obtain the safety interval of the collected data corresponding to each detection device from the line control center, compare the collected data corresponding to each detection device in each segment with the safety interval, and when there is at least one detection device in a segment whose collected data is not within the safety interval, record the detection devices whose collected data is not within the safety interval as marked devices; Count the marked devices in each section, obtain the fault type corresponding to the abnormal data collected by each marked device from the line control center, obtain the fault type corresponding to each marked device in each section, and then filter out the associated fault types corresponding to the fault type of each marked device in each section; Acquire the detection equipment corresponding to each fault type precision inspection from the line control center, thereby obtaining the detection equipment of each associated fault type precision inspection corresponding to each marking device in each section as each detection equipment of each associated fault type precision inspection corresponding to each marking device in each section; Compare the detection devices of the associated fine inspection corresponding to each marking device in each segment with the detection device set of each segment. If the detection devices of the associated fine inspection corresponding to each marking device in a segment are all in the detection device set of the segment, then the detection data of the segment is sufficient, and the segment is recorded as a first-class segment; If a detection device of an associated fine inspection corresponding to a marking device in a segment is not in the detection device set of the segment, the detection data of the segment is insufficient, and the segment is recorded as a second-category segment. At the same time, the detection device of the associated fine inspection is used as the detection device corresponding to the secondary inspection of the segment. This is used to analyze whether the detection data of each segment is sufficient, and the detection device set corresponding to the secondary inspection of each second-category segment is counted.

7. The method for intelligent fault analysis of UAV power distribution lines according to claim 6 is characterized in that: The screening process of each associated fault type corresponding to each marked device in each segment is as follows: Obtain each fault type corresponding to each historical detection in each segment from the historical detection record, compare the fault type corresponding to each marked device in each segment with the fault type corresponding to each historical detection in each segment, if the fault type corresponding to a marked device in a segment is the same as the fault type corresponding to a historical detection in a segment, then use the historical detection in the segment as the associated historical detection corresponding to the marked device in the segment, so as to obtain the associated historical detections corresponding to each marked device in each segment; Extract each fault type corresponding to each associated historical detection corresponding to each marked device in each segment, count the number of each fault type in the associated historical detection corresponding to each marked device in each segment, and then divide it by the number of associated historical detections to obtain the association rate of each fault type in the associated historical detection corresponding to each marked device in each segment, and take each fault type with a correlation rate greater than a preset correlation rate threshold as each associated fault type corresponding to the fault type of each marked device in each segment.

8. The method for intelligent fault analysis of UAV power distribution lines according to claim 1 is characterized in that: The specific process of analyzing the security of each type of segment is as follows: The collected data of each detection device in each first-class segment is extracted, and compared with the safety interval of the corresponding collected data of each detection device. When the collected data of at least one detection device in a certain class of segment is not within the safety interval, it is determined that the safety of the first-class segment is insufficient and a fault exists. The fault type of the corresponding detection device whose collected data is not within the safety interval is extracted, and each fault type of each first-class segment with insufficient safety is obtained. At the same time, the location area of ​​each fault type is obtained from the detection data of each first-class segment with insufficient safety as the fault data of each first-class segment with insufficient safety.

9. The method for intelligent fault analysis of UAV power distribution lines according to claim 7 is characterized in that: The specific process of starting the drone for secondary detection is as follows: Extract the real-time meteorological data corresponding to each second-class segment and the terrain data corresponding to the last historical detection, and use the three-dimensional model of the distribution line to confirm the starting detection segment corresponding to the drone secondary detection; Then obtain the distance between the other second-class segments and the starting detection segment, calculate the first priority value of the other second-class segments, confirm the second detection segment corresponding to the secondary detection of the drone, and thereby confirm the detection order of the second-class segments corresponding to the secondary detection of the drone; The UAV is equipped with a set of detection equipment corresponding to the secondary detection of each Class II segment, and performs detection in sequence according to the detection order of each Class II segment. When detecting each Class II segment, the set of detection equipment corresponding to the secondary detection of each Class II segment is started to collect secondary detection data corresponding to each Class II segment.

10. The method for intelligent fault analysis of UAV power distribution lines according to claim 1 is characterized in that: The specific process of analyzing the security of each of the two types of segments is as follows: The collected data corresponding to each detection device is obtained from the detection data and secondary detection data corresponding to each Class II segment, and the safety of each Class II segment is analyzed according to the safety analysis process of each Class I segment. At the same time, the fault data of each Class II segment with insufficient safety is obtained.

Images