Flashover active feature identification method and system of power transmission line image

Abstract

The invention discloses a flashover active feature recognition method for a power transmission line image, and the method comprises the steps: obtaining a historical picture, carrying out the preprocessing of the historical picture, and obtaining an image matrix, and the preprocessing comprises the first preprocessing and the second preprocessing; a sliding window technology is applied to the image matrix, the value of each window is calculated, the maximum value of the mean value of the windows is obtained, first judgment is conducted on the maximum value of the mean value of the windows, and a lightning flashover track is obtained; according to the method, the average brightness value and distribution in the lightning flashover track are extracted, the lightning flashover track is subjected to second judgment, lightning flashover is obtained, operation and maintenance personnel can understand the monitoring result more easily, and therefore operation and maintenance work can be guided more effectively.

Description

Technical Field

[0001] The present invention relates to the technical field of power transmission lines, and in particular to a flashover active feature recognition method and system for power transmission line images. Background Art

[0002] Transmission line online monitoring equipment has been widely installed on lines or towers in various regions and at various voltage levels to collect real-time data, including current, voltage, meteorological, image and other data. Among them, image-based online monitoring equipment has more intuitive output results and can clearly reflect the actual situation on site. For example, line damage caused by large mechanical equipment, or abnormal activities near towers, can all be better presented, so that line operation and maintenance personnel can be notified as soon as possible to handle the situation and ensure the reliability of power supply to the power grid.

[0003] Among transmission line faults, tripping faults caused by flashover account for the vast majority. Therefore, capturing pictures or videos of the moment when flashover hits the line can greatly accelerate fault diagnosis and line maintenance, thereby improving the stability of line operation. However, there are a large number of such devices, which generate massive amounts of image data all the time, and maintenance personnel need to identify them one by one. Only by finding out the very few that actually have flashover moments can further diagnosis and maintenance be meaningful. Therefore, it is necessary to design an algorithm to filter out pictures that do not contain flashover features in the on-site pictures of transmission lines, and only hand over pictures with flashover features to relevant personnel for further identification, which can greatly reduce labor costs. Summary of the invention

[0004] In view of the above-mentioned problems, the present invention is proposed.

[0005] Therefore, the technical problem solved by the present invention is: the present invention solves the problem that a large amount of manual labeling work is required in the early stage, the amount of calculation is too large to be suitable for deployment on terminal devices, the deep learning lacks interpretability, and the correlation between input and output is not intuitive enough.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0007] In a first aspect, the present invention provides a method for identifying active flashover features of a transmission line image, comprising:

[0008] Acquire a historical image, and preprocess the historical image to obtain an image matrix, wherein the preprocessing includes a first preprocessing and a second preprocessing;

[0009] Applying a sliding window technique to the image matrix, calculating the value of each window, obtaining the maximum value of the window mean, performing a first judgment on the maximum value of the window mean, and obtaining a lightning flashover trajectory;

[0010] The brightness mean and distribution in the lightning flashover trajectory are extracted, and a second judgment is performed on the lightning flashover trajectory to obtain the lightning flashover.

[0011] As a preferred solution of the flashover active feature recognition method of the power transmission line image of the present invention, the preprocessing of the historical image includes:

[0012] A brightness matrix is ​​obtained by the first preprocessing, and a second preprocessing is performed on the brightness matrix;

[0013] Traverse each pixel in the brightness matrix;

[0014] In the horizontal axis gradient matrix of the brightness matrix, if the horizontal axis gradient value of the current pixel is greater than the gradient value of the adjacent point in the horizontal direction or the brightness value exceeds the maximum brightness value of the first threshold, the horizontal axis gradient value of the pixel is retained;

[0015] Otherwise set to 0.

[0016] As a preferred solution of the flashover active feature recognition method of the power transmission line image described in the present invention, the second preprocessing includes:

[0017] In the vertical axis gradient matrix of the brightness matrix, if the vertical axis gradient value of the current pixel is greater than the gradient value of the adjacent point in the vertical direction or the brightness value exceeds the maximum brightness value of the first threshold, the vertical axis gradient value of the pixel is retained;

[0018] Otherwise set to 0;

[0019] The horizontal gradient matrix and the vertical gradient matrix are combined, and only the pixels with values ​​in the horizontal gradient matrix or the vertical gradient matrix are saved in the brightness matrix.

[0020] As a preferred solution of the flashover active feature recognition method of the power transmission line image described in the present invention, the second preprocessing further includes:

[0021] Iterate over each pixel in the brightness matrix;

[0022] If the horizontal axis gradient value or the vertical axis gradient value of the pixel is greater than the high threshold, it is a strong edge point and the pixel value is retained;

[0023] If the horizontal axis gradient value or the vertical axis gradient value of the pixel is greater than the low threshold and less than the high threshold, it is a weak edge point. If there are strong edge points around the pixel, the pixel value is retained, otherwise it is set to 0;

[0024] If the horizontal axis gradient value or the vertical axis gradient value of the pixel is less than the low threshold, they are both set to 0.

[0025] As a preferred solution of the flashover active feature recognition method of the transmission line image of the present invention, wherein: applying the sliding window technology to the image matrix includes:

[0026] Set the window size to slide on the image matrix;

[0027] Set the horizontal step size and vertical step size;

[0028] The average brightness value of the pixel set in the lightning flash trajectory is the average of all pixel brightness values ​​in the sliding window;

[0029] The discrete degree of pixel distribution of lightning flashover trajectory in the horizontal direction is the standard deviation of all pixels in the sliding window.

[0030] As a preferred solution of the flashover active feature recognition method of the power transmission line image of the present invention, the first judgment on the maximum value of the window mean value includes:

[0031] Based on the maximum value of the window mean, search for brightness values ​​around;

[0032] If the brightness value exceeds the maximum value of the average value of the second threshold, the corresponding pixel position is recorded to obtain the lightning flashover trajectory.

[0033] As a preferred solution of the method for identifying active lightning flashover features of a power transmission line image according to the present invention, the second determination of the lightning flashover trajectory includes:

[0034] If the mean value of the pixel set in the lightning flashover trajectory is greater than the maximum value of the brightness matrix of the second threshold, and the horizontal coordinate standard deviation of the lightning flashover trajectory is less than the set threshold, the output is flashover, otherwise no flashover.

[0035] In a second aspect, the present invention provides a system for active flashover feature recognition of a transmission line image, comprising:

[0036] A preprocessing module, used for acquiring historical images, and preprocessing the historical images to obtain an image matrix, wherein the preprocessing includes a first preprocessing and a second preprocessing;

[0037] A sliding window calculation module, used for applying a sliding window technique to the image matrix, calculating the value of each window, obtaining the maximum value of the window mean, performing a first judgment on the maximum value of the window mean, and obtaining a lightning flashover trajectory;

[0038] The feature recognition module is used to extract the brightness mean and distribution in the lightning flashover trajectory, perform a second judgment on the lightning flashover trajectory, and obtain the lightning flashover.

[0039] In a third aspect, the present invention provides a computing device, comprising:

[0040] Memory and processor;

[0041] The memory is used to store computer executable instructions, and the processor is used to execute the computer executable instructions. When the computer executable instructions are executed by the processor, the steps of the method for identifying active flashover features of transmission line images are implemented.

[0042] In a fourth aspect, the present invention provides a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, implement the steps of the method for identifying active flashover features of transmission line images.

[0043] The beneficial effects of the present invention are as follows: no data pre-labeling or model training is required, thus saving a lot of initial investment and time costs; the algorithm is simple and can be deployed on terminal devices, thus reducing the complexity of the system and communication costs, making the monitoring system more portable and practical, and the input-output association is intuitive, which means that operation and maintenance personnel can more easily understand the monitoring results, thereby more effectively guiding the operation and maintenance work. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. 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 creative labor. Among them:

[0045] Figure 1 The present invention provides an overall flow chart of a flashover active feature recognition method for a power transmission line image. DETAILED DESCRIPTION

[0046] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without creative work should fall within the scope of protection of the present invention.

[0047] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0048] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive with other embodiments.

[0049] The present invention is described in detail with reference to schematic diagrams. When describing the embodiments of the present invention, for the sake of convenience, the cross-sectional diagrams showing the device structure will not be partially enlarged according to the general scale, and the schematic diagrams are only examples, which should not limit the scope of protection of the present invention. In addition, in actual production, the three-dimensional dimensions of length, width and depth should be included.

[0050] At the same time, in the description of the present invention, it should be noted that the directions or positional relationships indicated by the terms "upper, lower, inner and outer" are based on the directions or positional relationships shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific direction, be constructed and operated in a specific direction, and therefore cannot be understood as limiting the present invention. In addition, the terms "first, second or third" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance.

[0051] In the present invention, unless otherwise clearly specified and limited, the terms "install, connect, connect" should be understood in a broad sense, for example: it can be a fixed connection, a detachable connection or an integral connection; it can also be a mechanical connection, an electrical connection or a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0052] Example 1

[0053] Reference Figure 1 , as an embodiment of the present invention, provides a flashover active feature recognition method for a transmission line image, comprising:

[0054] S100: Acquire historical images, and preprocess the historical images to obtain an image matrix, where the preprocessing includes a first preprocessing and a second preprocessing;

[0055] In the embodiment of the present application, a brightness matrix is ​​obtained by the first preprocessing, and the brightness matrix is ​​subjected to the second preprocessing;

[0056] Traverse each pixel in the brightness matrix;

[0057] In the horizontal axis gradient matrix of the brightness matrix, if the horizontal axis gradient value of the current pixel is greater than the gradient value of the adjacent point in the horizontal direction or the brightness value exceeds the maximum brightness value of the first threshold, the horizontal axis gradient value of the pixel is retained;

[0058] Otherwise set to 0.

[0059] In the vertical axis gradient matrix of the brightness matrix, if the vertical axis gradient value of the current pixel is greater than the gradient value of the adjacent point in the vertical direction or the brightness value exceeds the maximum brightness value of the first threshold, the vertical axis gradient value of the pixel is retained;

[0060] Otherwise set to 0;

[0061] The horizontal gradient matrix and the vertical gradient matrix are combined, and only the pixels with values ​​in the horizontal gradient matrix or the vertical gradient matrix are saved in the brightness matrix.

[0062] In a preferred embodiment, the first preprocessing may be a matrix transformation;

[0063] In another optional implementation, the first preprocessing may also utilize an average value method, a maximum value method or a minimum value method and a user-defined weight method.

[0064] In an embodiment of the present application, the first preprocessing option is to convert the historical image from an RGB (red, green, blue) matrix to a YUV (brightness, chroma, saturation) matrix, and extract the brightness component therefrom to achieve grayscale.

[0065] Converting the RGB matrix to a YUV matrix can be done through different methods and standards, including ITU-R BT.601, BT.709, BT.2020, etc., and will have different weight coefficients for different application scenarios (such as television, film production, etc.).

[0066] Exemplarily, the grayscale conversion can be expressed as converting the image from an RGB matrix to a YUV matrix by matrix multiplication and addition:

[0067]

[0068] It should be noted that converting RGB images to grayscale images can reduce data dimensions and computational complexity. For most edge detection algorithms, grayscale images are sufficient to capture important edge information. Grayscale focuses on brightness information and can effectively remove the interference of color information, making edge extraction more accurate.

[0069] In a preferred embodiment, the second preprocessing may be edge extraction;

[0070] In another optional embodiment, the second preprocessing may also be morphological processing, threshold segmentation and noise reduction;

[0071] In an embodiment of the present application, the second preprocessing selects edge extraction, and applies Gaussian filtering to denoise the image to make it smoother and reduce the interference of noise on edge feature extraction. The gradient strength and direction of the image are calculated by the Sobel operator to determine the strongest position of the edge change. Non-maximum suppression is performed, and each point is compared with the adjacent pixels in the corresponding gradient direction, retaining the maximum value and refining the edge. Through dual threshold detection, each point in the matrix is ​​traversed to ensure the continuity of the extracted edge. If a point value is less than the threshold, its adjacent pixels are checked to determine whether it is set as a strong edge, further improving the accuracy of edge extraction. After completing the edge extraction, the Candy algorithm is used to enhance the details of the image to improve the visual effect and highlight important features.

[0072] For example, a 3*3 Gaussian kernel is used to perform convolution operation on the brightness matrix. The Gaussian kernel is expressed as:

[0073]

[0074] Calculate the horizontal and vertical gradients of each pixel in the brightness matrix using a 3*3 window;

[0075] The horizontal axis gradient is expressed as:

[0076] Gx=sum(-x(i-1,j-1)+x(i+1,j-1)-2*x(i-1,j)+2*x(i+1,j)

[0077] -x(i-1,j+1)+x(i+1,j+1))

[0078] The vertical gradient is expressed as:

[0079] Gy=sum(x(i-1,j-1)+2*x(i,j-1)+x(i+1,j-1)-x(i-1,j+1)

[0080] -2*x(i,j+1)-x(i+1,j+1))

[0081] Among them, x(i,j) is the current pixel brightness matrix value, i, j are the horizontal and vertical coordinate positions of the pixel.

[0082] In a preferred embodiment, the first threshold value may be 0.8. Then, in the horizontal axis gradient matrix, if the horizontal axis gradient value of the current pixel is greater than the gradient value of its horizontally adjacent point or its brightness value exceeds 0.8 times the maximum brightness value, the horizontal axis gradient value of the pixel is retained, otherwise it is set to 0;

[0083] In the vertical axis gradient matrix, if the vertical axis gradient value of the current pixel is greater than the gradient value of its vertical adjacent point or its brightness value exceeds 0.8 times the maximum brightness value, the vertical axis gradient value of the pixel is retained, otherwise it is set to 0

[0084] In the embodiment of the present application, each pixel in the brightness matrix is ​​traversed;

[0085] If the horizontal axis gradient value or the vertical axis gradient value of the pixel is greater than the high threshold, it is a strong edge point and the pixel value is retained;

[0086] If the horizontal axis gradient value or the vertical axis gradient value of the pixel is greater than the low threshold and less than the high threshold, it is a weak edge point. If there are strong edge points around the pixel, the pixel value is retained, otherwise it is set to 0;

[0087] If the horizontal axis gradient value or the vertical axis gradient value of the pixel is less than the low threshold, they are both set to 0.

[0088] In a preferred embodiment, the high threshold is based on the gray value distribution of the image, generally 70% to 90% of the maximum gradient value., to ensure the selection of strong edge points;

[0089] The low threshold is set to a fraction of the high threshold, 20% to 30% of the high threshold, ensuring that pixels near the high threshold are retained, while weaker edges are judged based on the presence or absence of strong edges.

[0090] The threshold can also be selected by using an automated threshold selection method such as the Otsu method to calculate a suitable threshold based on the histogram of the image, or by conducting experiments based on actual results to adjust the threshold to obtain the best edge detection result.

[0091] It should be noted that edge is one of the most important features in an image and can effectively describe the shape and structure of an object. Edge extraction can help with subsequent image analysis and processing tasks. Gaussian filtering can smooth images and reduce the impact of noise on edge detection, thereby improving the quality and accuracy of edge extraction.

[0092] Combining grayscale and edge extraction can effectively highlight the key features of the image, especially in complex backgrounds, making it easier to identify objects. By reducing the amount of calculation through grayscale, edge extraction can significantly improve the overall processing speed and is suitable for real-time applications. After grayscale processing, unnecessary color information in the image is removed, reducing redundant information in the edge extraction process and improving the accuracy of the algorithm. Dual threshold detection can further ensure the coherence of the extracted edges, ensuring that the edges are more obvious in the image, thereby improving the quality of subsequent processing.

[0093] S200: applying a sliding window technique to the image matrix, calculating the value of each window, obtaining the maximum value of the window mean, performing a first judgment on the maximum value of the window mean, and obtaining a lightning flashover trajectory;

[0094] In an embodiment of the present application, the window size is set to slide on the image matrix;

[0095] Set the horizontal step size and vertical step size;

[0096] The average brightness value of the pixel set in the lightning flash trajectory is the average of all pixel brightness values ​​in the sliding window;

[0097] The discrete degree of pixel distribution of lightning flashover trajectory in the horizontal direction is the standard deviation of all pixels in the sliding window.

[0098] For example, a 11*11 window is set to slide on the image matrix, with a horizontal step size of 1 and a vertical step size of 11, and the mean u and the horizontal standard deviation s in each window are calculated;

[0099] The mean u represents the average brightness value of the pixel set in the lightning flash trajectory. Assume that the number of all pixels in the sliding window is N,I(x i ,y i ) is the brightness value of each pixel, expressed as:

[0100]

[0101] The standard deviation s of the horizontal axis represents the degree of discreteness of the pixel distribution in the horizontal direction of the lightning flashover trajectory. Its calculation is based on the horizontal coordinates of the pixels whose brightness values ​​exceed a certain threshold in the flashover trajectory. Assume that the horizontal coordinate x of the pixel is 1 , x 2 , …, x N , the mean of the horizontal axis is It is expressed as:

[0102]

[0103] In the embodiment of the present application, the maximum value of the window mean is used as a reference to search for brightness values ​​in the surrounding area;

[0104] If the brightness value exceeds the maximum value of the average value of the second threshold, the corresponding pixel position is recorded to obtain the lightning flashover trajectory.

[0105] In a preferred embodiment, the second threshold can be set to 0.5, then pixels with brightness values ​​exceeding 0.5 times the maximum average are searched around, and the pixel positions with brightness values ​​meeting the conditions are recorded, which is the lightning flashover trajectory.

[0106] It should be noted that by calculating the brightness mean and horizontal coordinate standard deviation of the pixels in the window, the lightning flashover trajectory can be identified more accurately. The use of sliding window and step size strategy can reduce the interference of image noise and improve the robustness of the algorithm.

[0107] S300: extracting the brightness mean and distribution in the lightning flashover trajectory, performing a second judgment on the lightning flashover trajectory, and obtaining the lightning flashover;

[0108] In an embodiment of the present application, if the mean value of the pixel set in the lightning flashover trajectory is greater than the maximum value of the brightness matrix of the second threshold, and the standard deviation of the abscissa of the lightning flashover trajectory is less than the set threshold, the output is "there is a flashover", otherwise there is no flashover;

[0109] Specifically, because the lightning flashover trajectory has the characteristics of high brightness and vertical downward, it is possible to judge whether it is a lightning flashover according to the brightness mean value and its abscissa distribution in the trajectory extracted by the above sliding window;

[0110] In a preferred embodiment, the set threshold can be 1.5. Then, if the mean value u of the pixel set in the lightning flashover trajectory > 0.5d, and the standard deviation s of the abscissa < p, the output is "there is a flashover", otherwise the output is "no flashover", where d is the maximum value of the brightness matrix.

[0111] It should be noted that by combining the two conditions of the brightness mean value and the abscissa standard deviation, lightning flashovers can be identified more accurately, reducing misjudgments. The set threshold can be adjusted according to the actual application scenario to adapt to different environments and conditions and improve the robustness of the algorithm.

[0112] The above is a schematic solution of a method for actively identifying flashover characteristics of a transmission line image in this embodiment. It should be noted that the technical solution of the device for actively identifying flashover characteristics of the transmission line image belongs to the same concept as the technical solution of the above method for actively identifying flashover characteristics of the transmission line image. For the details not described in detail in the technical solution of the device for actively identifying flashover characteristics of the transmission line image in this embodiment, reference can be made to the description of the technical solution of the method for actively identifying flashover characteristics of the transmission line image.

[0113] The device for actively identifying flashover characteristics of the transmission line image in this embodiment includes:

[0114] A preprocessing module, configured to obtain historical pictures, preprocess the historical pictures to obtain an image matrix, and the preprocessing includes a first preprocessing and a second preprocessing;

[0115] A sliding window calculation module, configured to apply a sliding window technique to the image matrix, calculate the value of each window, obtain the maximum value of the window mean, and perform a first judgment on the maximum value of the window mean to obtain a lightning flashover trajectory;

[0116] An identification feature module, configured to extract the brightness mean value and distribution in the lightning flashover trajectory, and perform a second judgment on the lightning flashover trajectory to obtain a lightning flashover.

[0117] This embodiment also provides a computing device, applicable to the situation of actively identifying flashover characteristics of a transmission line image, including:

[0118] A memory and a processor; the memory is used to store computer executable instructions, and the processor is used to execute computer executable instructions to implement the flashover active feature recognition method for transmission line images as proposed in the above embodiment.

[0119] This embodiment further provides a storage medium on which a computer program is stored. When the program is executed by a processor, the method for realizing active flashover feature recognition of transmission line images as proposed in the above embodiment is implemented.

[0120] The storage medium proposed in this embodiment and the method for realizing active flashover feature recognition of transmission line images proposed in the above embodiment belong to the same inventive concept. The technical details not fully described in this embodiment can be referred to the above embodiment, and this embodiment has the same beneficial effects as the above embodiment.

[0121] Through the above description of the implementation methods, the technicians in the relevant field can clearly understand that the present invention can be implemented by means of software and necessary general hardware, and of course can also be implemented by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as a computer's floppy disk, read-only memory (ROM), random access memory (RAM), flash memory (FLASH), hard disk or optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform the methods of various embodiments of the present invention.

[0122] Example 2

[0123] Referring to Table 1, this embodiment is different from the first embodiment in that it provides a verification test of a flashover active feature recognition method for a transmission line image, in order to verify and illustrate the technical effects adopted in this method.

[0124] Divide the data of each category into 150 groups, 90 groups of data for training and 60 groups of data for testing. The lightning strike and no lightning strike labels are defined as 0 and 1 respectively. Specify a bidirectional long short-term memory network (LSTM) layer with 95 hidden units and output the last element of the sequence.

[0125] By including a fully connected layer of size 4, followed by a softmax layer and a classification layer, 2 classes are specified. The LSTM solver is specified as adam, the gradient threshold is 11, and the maximum number of rounds is 450. A total of 2800 iterations are performed. As the training progresses, the accuracy continues to improve, and the loss function also tends to converge steadily, and the final training accuracy reaches 98.7%

[0126] In order to increase the persuasiveness, the feature vectors obtained by the traditional method and the proposed method are input into the random forest algorithm (RF) and LSTM for training and testing respectively. Table 1 shows the classification results of each method.

[0127] Table 1 Classification results of each method

[0128]

[0129] As can be seen from the above table, the combination of this method + LSTM shows the highest accuracy on both the training set and the test set, which shows that the combination of the proposed feature extraction method and the LSTM model can effectively capture the characteristics of the data and has good generalization ability on the test set.

[0130] The accuracy of the combination of traditional method + LSTM is lower than that of this method + LSTM, but still higher than that of traditional method + RF, which shows that the performance of LSTM model on time series data is generally better than that of random forest model.

[0131] The test set accuracy of the combination of our method + RF is higher than that of the traditional method + RF, which further proves the effectiveness of our method and can improve the performance even when used in combination with a traditional model such as RF.

[0132] The combination of traditional method + RF has the lowest accuracy, which may be due to the fact that the traditional feature extraction method cannot fully capture the complex characteristics of the data, resulting in limited performance of the RF model.

[0133] The test set accuracy of the LSTM and RF models under this method is higher than that of the traditional feature extraction method, which fully demonstrates the superiority of this method in extracting effective features. This method can significantly improve the performance of classification tasks when used in combination with LSTM and RF models. The LSTM model performs better than the RF model in this study, especially when used in combination with this method. This method is superior in extracting effective features and can improve the model's ability to classify situations with and without lightning strikes.

[0134] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A flashover active feature recognition method for transmission line images, characterized in that: include: Acquire a historical image, and preprocess the historical image to obtain an image matrix, wherein the preprocessing includes a first preprocessing and a second preprocessing; Applying a sliding window technique to the image matrix, calculating the value of each window, obtaining the maximum value of the window mean, performing a first judgment on the maximum value of the window mean, and obtaining a lightning flashover trajectory; The brightness mean and distribution in the lightning flashover trajectory are extracted, and a second judgment is performed on the lightning flashover trajectory to obtain the lightning flashover.

2. The method for identifying active flashover features of a power transmission line image according to claim 1, characterized in that: Preprocessing the historical images includes: A brightness matrix is ​​obtained by the first preprocessing, and a second preprocessing is performed on the brightness matrix; Traverse each pixel in the brightness matrix; In the horizontal axis gradient matrix of the brightness matrix, if the horizontal axis gradient value of the current pixel is greater than the gradient value of the adjacent point in the horizontal direction or the brightness value exceeds the maximum brightness value of the first threshold, the horizontal axis gradient value of the pixel is retained; Otherwise set to 0.

3. The method for identifying active flashover features of a power transmission line image according to claim 2, characterized in that: The second pre-processing includes: In the vertical axis gradient matrix of the brightness matrix, if the vertical axis gradient value of the current pixel is greater than the gradient value of the adjacent point in the vertical direction or the brightness value exceeds the maximum brightness value of the first threshold, the vertical axis gradient value of the pixel is retained; Otherwise set to 0; The horizontal gradient matrix and the vertical gradient matrix are combined, and only the pixels with values ​​in the horizontal gradient matrix or the vertical gradient matrix are saved in the brightness matrix.

4. The method for identifying active flashover features of a power transmission line image according to claim 1 or 3, characterized in that: The second pre-processing also includes: Iterate over each pixel in the brightness matrix; If the horizontal axis gradient value or the vertical axis gradient value of the pixel is greater than the high threshold, it is a strong edge point and the pixel value is retained; If the horizontal axis gradient value or the vertical axis gradient value of the pixel is greater than the low threshold and less than the high threshold, it is a weak edge point. If there are strong edge points around the pixel, the pixel value is retained, otherwise it is set to 0; If the horizontal axis gradient value or the vertical axis gradient value of the pixel is less than the low threshold, they are both set to 0.

5. The method for identifying active flashover features of a power transmission line image according to claim 4, characterized in that: Applying a sliding window technique to the image matrix comprises: Set the window size to slide on the image matrix; Set the horizontal step size and vertical step size; The average brightness value of the pixel set in the lightning flash trajectory is the average of all pixel brightness values ​​in the sliding window; The discrete degree of pixel distribution of lightning flashover trajectory in the horizontal direction is the standard deviation of all pixels in the sliding window.

6. The method for identifying active flashover features of a power transmission line image according to claim 5, characterized in that: The first determination of the maximum value of the window mean includes: Based on the maximum value of the window mean, search for brightness values ​​around; If the brightness value exceeds the maximum value of the average value of the second threshold, the corresponding pixel position is recorded to obtain the lightning flashover trajectory.

7. The method for identifying active flashover features of a power transmission line image according to claim 6, characterized in that: The second determination of the lightning flashover trajectory includes: If the mean value of the pixel set in the lightning flashover trajectory is greater than the maximum value of the brightness matrix of the second threshold, and the horizontal coordinate standard deviation of the lightning flashover trajectory is less than the set threshold, the output is flashover, otherwise no flashover.

8. A system for a flashover active feature recognition method of a transmission line image, characterized in that: include: A preprocessing module, used for acquiring historical images, and preprocessing the historical images to obtain an image matrix, wherein the preprocessing includes a first preprocessing and a second preprocessing; A sliding window calculation module, used for applying a sliding window technique to the image matrix, calculating the value of each window, obtaining the maximum value of the window mean, performing a first judgment on the maximum value of the window mean, and obtaining a lightning flashover trajectory; The feature recognition module is used to extract the brightness mean and distribution in the lightning flashover trajectory, perform a second judgment on the lightning flashover trajectory, and obtain the lightning flashover.

9. An electronic device, characterized in that: The device comprises: processor; a memory for storing processor-executable instructions; The processor is configured to call the instructions stored in the memory to execute the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having computer program instructions stored thereon, characterized in that: When the computer program instructions are executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images