Unmanned aerial vehicle high-altitude non-contact electricity testing method

Abstract

The invention belongs to the technical field of non-contact electricity testing, and particularly relates to an unmanned aerial vehicle high-altitude non-contact electricity testing method. According to the method, electric field intensity data and image information of a power line are acquired through an electric field sensor and optical imaging equipment carried by an unmanned aerial vehicle, feature point detection and matching are performed by using an AKAZE algorithm, a corona discharge area is analyzed, and then a data fusion model is established by combining an electric field intensity feature vector and a corona feature vector. And calculating the high-voltage probability of the electric power line, and judging whether the electric power line has high voltage according to a set decision threshold value, thereby improving the safety and efficiency of electric power overhaul work.

Description

Technical Field

[0001] The invention belongs to the technical field of non-contact electrical testing, and in particular relates to a high-altitude non-contact electrical testing method for an unmanned aerial vehicle. Background Art

[0002] In the operation and maintenance of modern power systems, power testing plays a pivotal role. Traditional power testing methods mainly include manual pole testing and ground non-contact power testing. Manual pole testing requires maintenance personnel to carry power testing equipment to climb up the power pole tower and get in close contact with the power line for testing. This method is not only labor-intensive and inefficient, but also exposes maintenance personnel directly to high-voltage environments, facing many serious safety risks such as electric shock and falling. According to statistics, in the past power maintenance operations, safety accidents caused by manual pole testing occurred frequently, posing a huge threat to the life safety of power workers.

[0003] Although ground non-contact electrical testing avoids the need for personnel to work at heights, it is greatly affected by environmental factors. For example, in severe weather conditions such as heavy rain and dense fog, the detection accuracy will drop significantly. At the same time, ground testing is easily restricted by topography and it is difficult to obtain accurate and effective test results for power lines in some special terrains such as mountainous areas and canyons.

[0004] With the rapid development of drone technology, its application in the power field has gradually attracted attention. Drones have the advantages of strong maneuverability, flexible operation, and remote control. They can reach areas that are difficult for humans to reach, providing new possibilities for high-altitude non-contact electrical testing. However, the existing drone electrical testing technology still needs to be improved, and how to improve the accuracy and reliability of detection has become a key issue that needs to be solved urgently. Summary of the invention

[0005] In view of the shortcomings of the existing technology, the present invention proposes a non-contact high-altitude electricity testing method for UAVs. By integrating the electric field strength and optical imaging information, through data fusion and probability calculation, the high-voltage state of the power line can be accurately and efficiently judged, thereby improving the safety and efficiency of power maintenance.

[0006] The technical solution of the present invention is as follows: A high-altitude non-contact electrical testing method for an unmanned aerial vehicle, comprising: The drone is equipped with electric field sensors and optical imaging equipment. The electric field sensors collect electric field strength data around the power lines, and the optical imaging equipment is used to photograph the power lines to obtain image information of the power lines. The AKAZE algorithm is used to identify the feature points of the power lines in the image information. The feature points are matched with the feature descriptors generated by the AKAZE algorithm in multiple sets of images. The brightness changes of the feature points matched in multiple sets of images are analyzed to determine the corona discharge area. The electric field intensity feature vector is obtained by extracting the features of the electric field intensity data, and the corona feature vector is obtained by extracting the features of the corona discharge area; based on the physical relationship between the electric field intensity feature vector and the corona feature vector, a data fusion model is established; The electric field strength characteristic vector and corona characteristic vector with the same coordinates are input to calculate the probability of high voltage in the power line, and a decision threshold is set at the same time. When the calculated probability of high voltage in the power line is greater than the decision threshold, it is determined that high voltage in the power line exists, otherwise it is determined that it does not exist.

[0007] Among them, the established data fusion model is: ; in, is the probability value of high voltage in power lines, is an exponential function, is the eigenvalue after fusion; In combination with the first aspect, in a possible implementation manner, a collection frequency of the electric field sensor is the same as a shooting frequency of the optical imaging device.

[0008] In combination with the first aspect, in a possible implementation manner, identifying feature points of power lines in image information using the AKAZE algorithm includes: Constructing the scale space of the image using the AKAZE algorithm: Defining the Gaussian kernel function ,in is the Gaussian kernel function, is the image pixel coordinate, is the scale parameter, is an exponential function; for the original grayscale image Perform Gaussian filtering to obtain images at different scales ,in represents the convolution operation, Images at different scales.

[0009] Initially detect feature points in scale space: calculate the difference between adjacent scale images ,in is the scale factor, is the difference between adjacent scale images; in the difference image, each pixel is compared with its surroundings The pixel value in the neighborhood, if the pixel point is a local maximum or minimum value, is the initially detected feature point; Screen and optimize the initially detected feature points: Calculate the Hessian matrix at the initially detected feature points ,in, , , The images are Directional second-order derivative, and Directional mixed second-order derivatives, Directional second derivative, is the Hessian matrix at the feature points initially detected; according to the determinant value of the Hessian matrix He Trace To determine whether the feature point is an edge point; set the threshold ,like , then the feature point is considered to be an edge point and is removed; after screening and optimization, stable power line feature points are obtained.

[0010] In combination with the first aspect, in a possible implementation manner, the matching method of the identified feature points in the continuous plurality of groups of images includes: For each identified feature point, the AKAZE algorithm is used to generate its corresponding feature descriptor; Calculate the feature points in the previous set of images Feature descriptor And the feature points in the latter set of images Feature descriptor The similarity between them is calculated as: ; in, for and The similarity between and Feature Descriptors and No. elements, is the dimension of the feature descriptor; Set matching threshold , if the feature descriptor and The similarity between , it is considered to be a matching feature point.

[0011] In combination with the first aspect, in a possible implementation manner, the corona discharge area is determined as follows: For the matched feature points, the brightness inter-frame difference calculation is performed: set up In the image The gray value in ,set up In the image The gray value in , then the inter-frame difference of brightness for: ; The inter-frame difference of brightness With brightness threshold For comparison, Exceeding the threshold When , it means that the brightness of the area where the feature point is located has changed; The area where brightness changes occur more than three times at the same position in multiple sets of consecutive images is determined as the corona discharge area.

[0012] In combination with the first aspect, in a possible implementation manner, the electric field strength characteristic vector is: ; in, represents the electric field strength eigenvector, is the mean value of the electric field strength, is the standard deviation of the electric field strength, and are the maximum and minimum values ​​of the electric field strength, is the mean of the first-order differences of the electric field intensity, is the standard deviation of the first-order difference of the electric field intensity; The corona eigenvector is: ; in, represents the corona eigenvector, is the corona discharge area, is the circumference of the corona discharge area, is the corona discharge area shape factor, is the average brightness of the corona discharge area, is the standard deviation of the brightness of the corona discharge area.

[0013] In combination with the first aspect, in a possible implementation manner, the fused feature value The calculation methods include: Process the electric field strength feature vector and calculate its weighted sum : ; in, is each electric field strength feature in the electric field strength feature vector, For each electric field strength characteristic The corresponding weight coefficient is, ; Process the corona eigenvector and calculate its weighted sum : ; in, For each corona feature in the corona feature vector, Each corona feature The corresponding weight coefficient is, ; Weighted sum of electric field strength features Assign weight coefficient , weighted sum of corona features Assign weight coefficient ,but: ; in, .

[0014] In conjunction with the first aspect, in a possible implementation manner, calculating the power line high voltage probability value includes: Sure , , and The value of The electric field intensity eigenvector of the same coordinate and the corona eigenvector Input into the data fusion model to calculate the probability value of high voltage on the power line ; The calculated probability value of high voltage on the power line With decision threshold For comparison. , it is determined that there is high voltage in the circuit; otherwise, it is determined that there is no high voltage.

[0015] In combination with the first aspect, in a possible implementation manner, the , , and The value of is solved using the least squares method.

[0016] The beneficial effects brought about by the technical solution provided in the embodiments of the present application include at least the following beneficial effects: The present invention uses drones to conduct high-altitude non-contact electrical testing, completely avoiding direct contact between maintenance personnel and high-voltage equipment, and fundamentally eliminating the safety risks brought by manual pole testing. Even in complex power environments, drones can complete testing tasks at a safe distance, ensuring the personal safety of maintenance personnel.

[0017] By integrating the electric field strength and optical imaging information, using advanced algorithms for feature extraction, tracking and analysis, and building a data fusion model based on physical relationships to calculate the probability of line electrification, the limitations of a single detection method have been overcome. The electric field sensor can directly measure the electric field strength around the line and has a direct physical reflection of the line electrification situation; the optical imaging device can capture visual features such as corona discharge to assist in judgment from another perspective. The combination of the two and through data fusion and probability calculation can more accurately judge the high-voltage state of the line, effectively reduce misjudgments and missed judgments, and improve the reliability of detection.

[0018] Drones have flexible and fast flight capabilities, can quickly reach the inspection location, and conduct rapid and continuous inspections of multiple power lines. Compared with traditional manual power inspection methods, it greatly shortens the inspection time and improves work efficiency. At the same time, drones can inspect the lines at different heights and angles, and can fully cover all parts of the lines to ensure the integrity of the inspection. In large-scale power line inspections, it can significantly reduce labor costs and time costs and improve the maintenance efficiency of the power system.

[0019] Drones can work in various terrains and environmental conditions, and are not restricted by ground obstacles and harsh terrain. Whether in mountainous areas, above rivers, or in areas with tall buildings in cities, as long as drones can reach them, power line inspections can be carried out. This adaptability enables the present invention to be widely used in various power scenarios, providing a strong guarantee for the comprehensive maintenance of the power system.

[0020] The rich data collected during the detection process is not only used to determine the high-voltage status of the line in real time, but can also be stored for subsequent in-depth analysis. By statistics and analysis of a large amount of detection data, potential problems and trends in power lines can be discovered, providing data basis for preventive maintenance and optimization of the power system, which helps to improve the overall operational stability and reliability of the power system. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 This is a schematic diagram of the overall process of a non-contact electrical testing method for a UAV at high altitude; Figure 2 A detailed flow chart of the steps of a non-contact electrical testing method for a drone at high altitude S100; Figure 3 A detailed flow chart of the steps S200 of a high-altitude non-contact electrical testing method for a UAV; Figure 4 A detailed flow chart of the steps S300 of a high-altitude non-contact electrical testing method for a UAV; Figure 5 The following is a detailed flow chart of the steps S400 of a high-altitude non-contact electrical testing method for a UAV. DETAILED DESCRIPTION

[0022] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0023] With the continuous development and expansion of the power system, the maintenance and inspection of power lines have become increasingly important. Accurately judging whether the power lines carry high voltage is a key link to ensure the safety of power maintenance personnel and the stable operation of the power system. Traditional power testing methods have many limitations. For example, manual pole testing not only consumes manpower and time, but also exposes maintenance personnel to high-voltage environments, facing great safety risks; although ground non-contact power testing avoids direct contact between people, it is greatly affected by the detection distance and environmental factors (such as weather, terrain, etc.), and the detection accuracy is difficult to guarantee.

[0024] To solve the above problems, please refer to Figure 1 , which shows a non-contact high-altitude electrical testing method for a drone provided by an embodiment of the present invention, the method comprising: S100: The drone is equipped with electric field sensors and optical imaging equipment. The electric field sensors collect electric field strength data around the power lines, and the optical imaging equipment is used to photograph the power lines to obtain image information of the power lines.

[0025] S200: Using the AKAZE algorithm to identify feature points of the power line in the image information, the feature points are matched with the feature descriptors generated by the AKAZE algorithm in multiple sets of images, and the brightness changes of the feature points matched in the multiple sets of images are analyzed to determine the corona discharge area.

[0026] S300: obtaining an electric field strength feature vector by extracting features of the electric field strength data, and obtaining a corona feature vector by extracting features of the corona discharge region; and establishing a data fusion model based on the physical relationship between the electric field strength feature vector and the corona feature vector.

[0027] S400: Input the electric field strength characteristic vector and the corona characteristic vector with the same coordinates to calculate the probability of high voltage in the power line, and set a decision threshold at the same time. When the calculated probability of high voltage in the power line is greater than the decision threshold, it is determined that the high voltage in the power line exists, otherwise it is determined that it does not exist.

[0028] This solution uses the electric field strength data collected by electric field sensors and the image information obtained by optical imaging equipment, combined with advanced algorithms and data fusion technology, to achieve accurate and efficient judgment of the high voltage status of power lines, thereby improving the safety and efficiency of power maintenance work, reducing safety accidents caused by misjudgment of the energized status, and providing strong support for the reliable operation of the power system.

[0029] S100 is the preparation and preliminary data collection stage before the drone takes off. During this stage, the drone is equipped with electric field sensors and optical imaging equipment to conduct a comprehensive inspection to ensure that the equipment is functioning properly. The drone then flies to the inspection area according to the preset path to collect electric field strength data and image information. S100 ensures the reliability of the equipment and the accuracy of data collection before performing the mission, providing high-quality raw data for subsequent analysis.

[0030] Please refer to Figure 2 , which shows a flowchart of an exemplary high-altitude non-contact electrical testing method S100 of a drone of the present application, and its contents include: S110: Equip the drone with electric field sensors and optical imaging equipment, and inspect and prepare the equipment.

[0031] Conduct a comprehensive inspection of the drone, including appearance inspection and power inspection.

[0032] Confirm whether the connection between the electric field sensor and the optical imaging device and the drone is stable, and whether the interface is normal without looseness or poor contact.

[0033] If there are no equipment problems, the drone will perform subsequent tasks according to the pre-planned flight path of the drone.

[0034] S120: Collecting electric field strength data of the power line through an electric field sensor.

[0035] According to the predetermined flight path, the drone is controlled to fly to the detection area near the power line. When the drone reaches the predetermined detection location, the electric field sensor is activated to collect electric field strength data around the power line. The electric field sensor measures the electric field strength in real time at a certain sampling frequency, and the measured data includes the magnitude and direction of the electric field strength.

[0036] S130: Acquire image information of the power line through an optical imaging device.

[0037] While collecting the electric field strength data, start the optical imaging equipment to photograph the power lines.

[0038] The images are taken at the same frequency as the sampling frequency of the electric field sensor to obtain image information of the dynamic changes of the power line.

[0039] S200 focuses on the processing of image information, using the AKAZE algorithm to pre-process images, detect feature points, and match feature points. S200 can accurately identify and track the feature points of power lines, and maintain the consistency and stability of feature points even in a dynamically changing environment.

[0040] Please refer to Figure 3 , which shows a flowchart of an exemplary high-altitude non-contact electrical testing method S200 of a drone of the present application, and its contents include: S210: constructing a scale space of the image based on the collected image information based on the AKAZE algorithm, and detecting feature points in the scale space.

[0041] The collected image sequences are preprocessed, including image grayscale processing, converting color images into grayscale images to reduce the amount of data and highlight the edge and texture information of the image.

[0042] The AKAZE algorithm is used to construct the scale space of the image. By constructing a series of images of different scales, the visual effect of the human eye observing objects at different distances is simulated, so that stable feature points can be detected at different scales. The specific construction process is as follows: Define the Gaussian kernel function: ,in is the image pixel coordinate, is the scale parameter, is an exponential function.

[0043] For the original grayscale image Perform Gaussian filtering to obtain images at different scales ,in Represents a convolution operation.

[0044] Preliminary detection of feature points in scale space. By calculating the difference of the image at different scales, local extreme points are found as the preliminarily detected feature points. The specific calculation method is: Calculate the difference between adjacent scale images ,in is the scale factor, usually .

[0045] In the difference image, each pixel is compared with its surroundings The pixel value in the neighborhood, if the pixel point is a local maximum or minimum value, is the feature point preliminarily detected.

[0046] Screen and optimize the initially detected feature points. Remove feature points with strong edge response, because feature points at the edge have poor stability and are easily affected by image noise and perspective changes. Calculate the Hessian matrix of the initially detected feature points ,in, , , The images are Directional second-order derivative, and Directional mixed second-order derivatives, Directional second derivative.

[0047] According to the determinant value of the Hessian matrix Heji To determine whether a feature point is an edge point. Set the threshold ,like , then the feature point is considered as an edge point and is removed. After screening and optimization, stable power line feature points are obtained.

[0048] S220: In the continuous plurality of groups of images, for feature points identified in the previous image, search and match feature points identified in the current image.

[0049] In S210, the power line feature points of the continuous plurality of image groups are identified and then feature point matching is performed.

[0050] For each identified feature point, the AKAZE algorithm is used to generate its corresponding feature descriptor. The AKAZE algorithm generates a vector with a certain length based on the image information around the feature point. , which can describe the unique properties of the feature point, including its local texture, gradient and other information. The specific generation process involves a complex nonlinear diffusion process and feature extraction of local areas to ensure that the feature descriptor has a certain invariance to the rotation, scaling and illumination changes of the image.

[0051] In multiple consecutive groups of images, for the feature points in the previous group of images Feature descriptor And the feature points in the latter set of images Feature descriptor ,calculate and The similarity between them is calculated as: ; in for and The similarity between and Feature Descriptors and No. elements, is the dimension of the feature descriptor.

[0052] Set a matching threshold , if the feature descriptor and The similarity between , it is considered to be a matching feature point, that is, it represents the same physical feature point in different groups of images. Through this matching method, corresponding feature points can be found in multiple consecutive groups of images, thereby realizing the tracking of the same feature point at different times.

[0053] S230: For the matched feature points, perform brightness frame difference calculation, and determine the corona discharge area based on the calculation result.

[0054] For the matched feature points , in consecutive images and In the calculation, the inter-frame difference of its brightness is set In the image The gray value in ,set up In the image The gray value in , then the inter-frame difference of brightness for: .

[0055] The inter-frame difference of brightness With brightness threshold For comparison, Exceeding the threshold When , it means that the brightness of the area where the feature point is located has changed. The area where the brightness changes more than three times at the same position in multiple sets of consecutive images is determined as the corona discharge area.

[0056] S300 involves the extraction of electric field strength and corona discharge area characteristics, and the establishment of a data fusion model. S300 improves the accuracy and robustness of the detection model by integrating multi-dimensional information of electric field strength and corona characteristics, making the assessment of the high voltage status of power lines more accurate.

[0057] Please refer to Figure 4 , which shows a flowchart of an exemplary high-altitude non-contact electrical testing method S300 of a drone of the present application, and its contents include: S310: Preprocess the electric field strength data, extract electric field strength features, and construct an electric field strength feature vector.

[0058] First, the collected electric field strength data is subjected to abnormal value detection and processing. After the processing is completed, the electric field strength features are extracted. The electric field strength features include: The mean value of the electric field strength , reflecting the average level of electric field strength; Standard deviation of electric field strength , which reflects the discreteness of the electric field intensity data; The maximum value of electric field strength and minimum value , providing information on the range of variation of electric field strength; The mean of the first-order differences of the electric field strength , which reflects the average rate of change of the electric field strength.

[0059] Standard deviation of the first difference of the electric field strength , which reflects the discrete degree of the rate of change of the electric field intensity.

[0060] The mean value of the electric field strength , standard deviation of electric field strength , the maximum value of the electric field strength and minimum value , the mean of the first-order difference of the electric field intensity and the standard deviation of the first-order difference of the electric field strength The electric field strength data collected by the UAV equipped with electric field sensors are calculated based on the abnormal value detection and processing of the data. The mean electric field strength , standard deviation of electric field strength , the maximum value of the electric field strength and minimum value , the mean of the first-order difference of the electric field intensity and the standard deviation of the first-order difference of the electric field strength The calculation method adopts the existing conventional calculation method.

[0061] The electric field strength features extracted above are combined into an electric field strength feature vector: ; in, represents the electric field strength eigenvector.

[0062] S320: Extract features of the corona discharge area and construct a corona feature vector.

[0063] The characteristics of the corona discharge area include geometric characteristics and brightness characteristics. The geometric characteristics include area, perimeter, and shape factor, and the brightness characteristics include average brightness and brightness standard deviation.

[0064] area It is obtained by counting the pixels in the corona discharge area.

[0065] Calculate the perimeter of the corona discharge area using a boundary tracing algorithm .

[0066] Calculate the shape factor using the ratio of the square of the perimeter to the area .

[0067] The average brightness of the corona discharge area is , reflecting the average brightness of the corona discharge area.

[0068] The standard deviation of the brightness of the corona discharge area is , reflecting the uniformity of brightness in the corona discharge area.

[0069] The geometric features and brightness features of the corona discharge area are combined into a corona feature vector: ; in, represents the corona eigenvector.

[0070] S330: Establishing a data fusion model based on the electric field intensity feature vector and the corona feature vector.

[0071] First, the electric field intensity feature vector is processed and its weighted sum is calculated. .

[0072] For each electric field strength feature in the electric field strength feature vector Assign a weight coefficient ,but: ; in, , that is, the electric field intensity feature vector The sum of the weight coefficients is 1; Secondly, the corona eigenvector is processed and its weighted sum is calculated. ; For each corona feature in the corona feature vector Assign a weight coefficient ,but: ; in, , that is, the corona feature in the corona feature vector The sum of the weight coefficients is 1.

[0073] Then calculate the fused eigenvalues : Weighted sum of electric field strength features Assign weight coefficient , weighted sum of corona features Assign weight coefficient ,but: ; in, .

[0074] Finally, a data fusion model is established to obtain the probability value of high voltage in power lines. : ; By weightedly summing the electric field intensity eigenvector and the corona eigenvector respectively, fusing and normalizing the two, the data fusion model formula for evaluating the probability of high voltage in power lines is obtained.

[0075] S400 includes the determination of weight coefficients, the calculation of high voltage probability values, and the determination of energized states. S400 determines the optimal weight coefficients through experimental data training, so that the model can more accurately predict the energized state of power lines. In addition, by setting a reasonable decision threshold, the misjudgment rate and missed judgment rate can be balanced, improving the practicality and safety of the power inspection method. Finally, by outputting the judgment results, it provides strong data support for the maintenance and management of power lines.

[0076] Please refer to Figure 5 , which shows a flowchart of an exemplary high-altitude non-contact electrical testing method S400 of a drone of the present application, and its contents include: S410: OK , , and The value of .

[0077] Through a large amount of experimental data for training, various weight coefficients are determined. The electric field strength data and corona discharge area characteristic data of power lines under different charged states (with and without high voltage) are collected and divided into training sets and test sets. Using the training set data, the least squares method is used to solve the weight coefficients so that the error between the line charged probability predicted by the model and the actual charged state is minimized.

[0078] S420: Power line high voltage probability value calculate.

[0079] The electric field sensor equipped by the drone records the corresponding coordinates when collecting electric field strength data, and the optical imaging device records the corresponding coordinates when synchronously taking the image information of the power line. Finally, the two are matched according to the coordinates to form the electric field strength data and image information corresponding to the same coordinates. The electric field strength data and image information of the same coordinates are processed to obtain the electric field strength feature vector of the same coordinates and the corona eigenvector , the electric field intensity eigenvector of the same coordinate and the corona eigenvector Input into the trained data fusion model to calculate the probability value of high voltage on the power line .

[0080] S430: Determination of the charged state Based on a large amount of experimental data and practical application experience, a reasonable decision threshold is set by comprehensively considering factors such as the misjudgment rate and missed judgment rate. .

[0081] Exemplarily, after conducting detection experiments on a large number of lines with known high voltage and no high voltage, the false positive rate and missed positive rate under different thresholds are counted, and a false positive rate-missed positive rate curve, i.e., ROC curve, is drawn. The threshold corresponding to the point close to the upper left corner of the ROC curve (where both the false positive rate and missed positive rate are low) is selected as the decision threshold. Usually, the value range of the decision threshold is between 0.5 and 0.9. If the safety requirements are extremely high and you want to minimize the misjudgment of the presence of high voltage as the absence of high voltage, you can increase the decision threshold; conversely, if you want to reduce unnecessary warnings, you can lower the decision threshold.

[0082] The calculated probability value of high voltage on the power line With decision threshold For comparison. , it is determined that there is high voltage in the circuit; otherwise, it is determined that there is no high voltage.

[0083] The judgment result is outputted in the form of high-voltage status information displayed on the display screen of the ground control station, and the maintenance personnel can be prompted by sound alarms. For live lines, alarms are sounded to remind maintenance personnel to pay attention to safety. In addition, the test result data, including electric field strength data, image information, calculated feature vectors, power line high voltage probability, etc., can be stored in the database for subsequent query and analysis, providing data support for the long-term maintenance and management of power lines.

[0084] The basic principles of the present application are described above in conjunction with specific embodiments. However, it should be noted that the advantages, strengths, effects, etc. mentioned in the present application are only examples and not limitations, and it cannot be considered that these advantages, strengths, effects, etc. are required by each embodiment of the present application. In addition, the specific details disclosed above are only for the purpose of illustration and ease of understanding, not for limitation, and the above details do not limit the present application to being implemented by adopting the above specific details.

[0085] The block diagrams of the devices, apparatuses, equipment, and systems involved in this application are only illustrative examples and are not intended to require or imply that they must be connected, arranged, and configured in the manner shown in the block diagram. As will be appreciated by those skilled in the art, these devices, apparatuses, equipment, and systems can be connected, arranged, and configured in any manner. Words such as "including", "comprising", "having", etc. are open words, referring to "including but not limited to", and can be used interchangeably with them. The words "or" and "and" used here refer to the words "and / or" and can be used interchangeably with them, unless the context clearly indicates otherwise. The words "such as" used here refer to the phrase "such as but not limited to", and can be used interchangeably with them.

[0086] It should also be noted that in the apparatus, device and method of the present application, each component or each step can be decomposed and / or recombined. Such decomposition and / or recombination should be regarded as equivalent solutions of the present application.

[0087] The above description of the disclosed aspects is provided to enable any person skilled in the art to make or use the present application. Various modifications to these aspects will be readily apparent to those skilled in the art, and the general principles defined herein may be applied to other aspects without departing from the scope of the present application. Therefore, the present application is not intended to be limited to the aspects shown herein, but rather to the widest scope consistent with the principles and novel features of the present invention.

[0088] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. A non-contact electrical testing method for a drone at high altitude, characterized in that: include: The drone is equipped with electric field sensors and optical imaging equipment. The electric field sensors collect electric field strength data around the power lines, and the optical imaging equipment is used to photograph the power lines to obtain image information of the power lines. The AKAZE algorithm is used to identify the feature points of the power lines in the image information. The feature points are matched with the feature descriptors generated by the AKAZE algorithm in multiple sets of images. The brightness changes of the feature points matched in multiple sets of images are analyzed to determine the corona discharge area. The electric field intensity feature vector is obtained by extracting the features of the electric field intensity data, and the corona feature vector is obtained by extracting the features of the corona discharge area; based on the physical relationship between the electric field intensity feature vector and the corona feature vector, a data fusion model is established; The electric field intensity characteristic vector and the corona characteristic vector with the same coordinates are input to calculate the probability of high voltage in the power line, and a decision threshold is set at the same time. When the calculated probability of high voltage in the power line is greater than the decision threshold, it is determined that the high voltage in the power line exists, otherwise it is determined that it does not exist; Among them, the established data fusion model is: ； in, is the probability value of high voltage in power lines, is an exponential function, is the fused feature value.

2. A non-contact high-altitude electrical testing method for unmanned aerial vehicles according to claim 1, characterized in that: include: The acquisition frequency of the electric field sensor is the same as the shooting frequency of the optical imaging device.

3. A non-contact high-altitude electrical testing method for unmanned aerial vehicles according to claim 1, characterized in that: The method of using the AKAZE algorithm to identify the characteristic points of the power line in the image information includes: Constructing the scale space of the image using the AKAZE algorithm: Defining the Gaussian kernel function ,in is the Gaussian kernel function, is the image pixel coordinate, is the scale parameter, is an exponential function; for the original grayscale image Perform Gaussian filtering to obtain images at different scales ,in represents the convolution operation, For images at different scales; Initially detect feature points in scale space: calculate the difference between adjacent scale images ,in is the scale factor, is the difference between adjacent scale images; in the difference image, each pixel is compared with its surroundings The pixel value in the neighborhood, if the pixel point is a local maximum or minimum value, is the initially detected feature point; Screen and optimize the initially detected feature points: Calculate the Hessian matrix at the initially detected feature points ,in, , , The images are Directional second-order derivative, and Directional mixed second-order derivatives, Directional second derivative, is the Hessian matrix at the feature points initially detected; according to the determinant value of the Hessian matrix He Trace To determine whether the feature point is an edge point; set the threshold ,like , then the feature point is considered to be an edge point and is removed; after screening and optimization, stable power line feature points are obtained.

4. A non-contact high-altitude electrical testing method for unmanned aerial vehicles according to claim 3, characterized in that: The matching method of the identified feature points in the continuous plurality of groups of images includes: For each identified feature point, the AKAZE algorithm is used to generate its corresponding feature descriptor; Calculate the feature points in the previous set of images Feature descriptor And the feature points in the latter set of images Feature descriptor The similarity between them is calculated as: ； in, for and The similarity between and Feature Descriptors and No. elements, is the dimension of the feature descriptor; Set matching threshold , if the feature descriptor and Similarity between , it is considered to be a matching feature point.

5. A non-contact high-altitude electrical testing method for unmanned aerial vehicles according to claim 4, characterized in that: The corona discharge area is determined as follows: For the matched feature points, the brightness inter-frame difference calculation is performed: set up In the image The gray value in ,set up In the image The gray value in , then the inter-frame difference of brightness for: ; The inter-frame difference of brightness With brightness threshold For comparison, Exceeding the threshold When , it means that the brightness of the area where the feature point is located has changed; The area where brightness changes occur more than three times at the same position in multiple sets of consecutive images is determined as the corona discharge area.

6. A non-contact high-altitude electrical testing method for unmanned aerial vehicles according to claim 1, characterized in that: The electric field strength characteristic vector is: ； in, represents the electric field strength eigenvector, is the mean value of the electric field strength, is the standard deviation of the electric field strength, and are the maximum and minimum values ​​of the electric field strength, is the mean of the first-order differences of the electric field intensity, is the standard deviation of the first-order difference of the electric field intensity; The corona eigenvector is: ； in, represents the corona eigenvector, is the corona discharge area, is the circumference of the corona discharge area, is the corona discharge area shape factor, is the average brightness of the corona discharge area, is the standard deviation of the brightness of the corona discharge area.

7. The high-altitude non-contact electrical testing method for unmanned aerial vehicles according to claim 1 is characterized in that: The fused feature value The calculation methods include: Process the electric field strength feature vector and calculate its weighted sum : ； in, is each electric field strength feature in the electric field strength feature vector, For each electric field strength characteristic The corresponding weight coefficient is, ; Process the corona eigenvector and calculate its weighted sum : ； in, For each corona feature in the corona feature vector, Each corona feature The corresponding weight coefficient is, ; Weighted sum of electric field strength features Assign weight coefficient , weighted sum of corona features Assign weight coefficient ,but: ； in, .

8. A non-contact high-altitude electrical testing method for unmanned aerial vehicles according to claim 7, characterized in that: Calculating the high voltage probability value of the power line includes: Sure , , and The value of The electric field intensity eigenvector of the same coordinate and the corona eigenvector Input into the data fusion model to calculate the probability value of high voltage on the power line ; The calculated probability value of high voltage on the power line With decision threshold For comparison; if , it is determined that there is high voltage in the circuit; otherwise, it is determined that there is no high voltage.

9. A non-contact high-altitude electrical testing method for unmanned aerial vehicles according to claim 8, characterized in that: Said , , and The value of is solved using the least squares method.

Images