Binocular vision-driven power transmission line windage yaw galloping monitoring device and method

Abstract

The invention discloses a binocular vision-driven power transmission line windage yaw galloping monitoring device and method, and belongs to the field of power transmission line monitoring, and the device comprises an image collection unit which is used for collecting laser contour two-dimensional image data; the feature extraction unit is used for performing feature extraction to obtain a laser contour vector matrix; the state analysis unit is used for carrying out fusion analysis to obtain a line windage yaw state vector; the track construction unit is used for constructing a line windage yaw waving track; and the alarm generation unit is used for generating a line windage yaw alarm when the line windage yaw waving track exceeds the windage yaw safety range. According to the method and the device, the technical problems of serious hysteresis and relatively high error rate caused by relatively high calculation complexity due to the fact that windage yaw galloping monitoring is realized by performing three-dimensional modeling by using laser point cloud data and then performing modeling fusion analysis at multiple moments in the prior art are solved, and the technical effects of reducing alarm hysteresis and improving monitoring accuracy are achieved.

Description

Technical Field

[0001] The present invention relates to the field of power transmission line monitoring, and in particular to a binocular vision-driven power transmission line windage fluctuation monitoring device and method. Background Art

[0002] Since transmission lines are often exposed to outdoor environments, they are easily affected by wind and cause wind deviation and fluttering. In severe cases, this can cause the conductors to collide with each other or rub against the poles and towers, affecting the safe operation of the transmission lines.

[0003] In order to monitor the wind yaw and flutter state of power transmission lines in real time, the existing technology generally adopts the method of laser point cloud scanning to obtain the three-dimensional point cloud data of the power transmission lines, and then models the three-dimensional point cloud data at multiple times, and then fuses the modeling results at multiple times to determine whether the power transmission line is in a wind yaw and flutter state. However, due to the long span of the transmission line and the huge amount of laser point cloud data, the computational complexity of modeling and the fusion analysis of multiple modeling results is very high, resulting in technical problems such as serious lag and high error rate in the judgment and alarm of wind yaw and flutter. Summary of the invention

[0004] The present application provides a binocular vision-driven transmission line wind yaw and sway monitoring device and method, aiming to solve the technical problems in the prior art that wind yaw and sway monitoring uses laser point cloud data for three-dimensional modeling, and then performs modeling fusion analysis at multiple moments, resulting in high computational complexity, severe lag and high error rate.

[0005] In view of the above problems, the present application provides a binocular vision-driven transmission line windage and fluctuation monitoring device and method.

[0006] The first aspect disclosed in the present application provides a binocular vision-driven transmission line wind deviation dance monitoring device, the device comprising: an image acquisition unit, used to acquire laser profile two-dimensional image data of a straight-line laser array through a binocular camera; a feature extraction unit, used to perform feature extraction on the laser profile two-dimensional image data through a laser profile vector analysis channel to obtain a laser profile vector matrix; a state analysis unit, used to perform fusion analysis on the laser profile vector matrix through a line wind deviation feature analysis channel to obtain a line wind deviation state vector; a trajectory construction unit, used to construct a line wind deviation dance trajectory according to the line wind deviation state vector; an alarm generation unit, used to generate a line wind deviation alarm when the line wind deviation dance trajectory exceeds a wind deviation safety range.

[0007] Another aspect disclosed in the present application provides a binocular vision-driven transmission line windage dance monitoring method, the method comprising: collecting laser profile two-dimensional image data of a straight-line laser array through a binocular camera; performing feature extraction on the laser profile two-dimensional image data through a laser profile vector analysis channel to obtain a laser profile vector matrix; performing fusion analysis on the laser profile vector matrix through a line windage feature analysis channel to obtain a line windage state vector; constructing a line windage dance trajectory based on the line windage state vector; and generating a line windage alarm when the line windage dance trajectory exceeds a windage safety range.

[0008] One or more technical solutions provided in this application have at least the following technical effects or advantages: The image acquisition unit uses a binocular camera to collect the binocular image data of the laser contour formed by the projection of a straight-line laser array on the transmission line, and binocular visual imaging is used instead of laser point cloud scanning, thereby reducing the amount of calculation for data acquisition and processing; the feature extraction unit uses the laser contour vector analysis channel to extract features from the collected laser contour binocular image data to obtain a laser contour vector matrix that can reflect the shape and position of the laser contour, thereby realizing the conversion from the laser contour image to the laser contour vector, thereby reducing the data dimension and calculation complexity of the subsequent wind deviation analysis; the state analysis unit uses the line wind deviation feature analysis channel to perform a fusion analysis on the laser contour vector matrix to obtain a line wind deviation state vector that reflects the wind deviation state of the transmission line, thereby avoiding The process of modeling three-dimensional point cloud data is simplified, and the calculation complexity is further reduced; the wind deviation dancing trajectory of the transmission line is constructed according to the line wind deviation state vector through the trajectory construction unit, and the mapping of the wind deviation state to the wind deviation dancing trajectory is realized, providing a judgment basis for the subsequent wind deviation alarm; the alarm generation unit is used to determine whether the wind deviation dancing trajectory exceeds the wind deviation safety range. If it exceeds, a line wind deviation alarm is generated, which improves the technical solution of the effectiveness of wind deviation monitoring and early warning, and solves the technical problems of high calculation complexity and serious alarm lag in the existing technology of wind deviation dancing monitoring using laser point cloud data for three-dimensional modeling, and achieves the wind deviation dancing trajectory analysis using the two-dimensional laser contour image of a straight line laser, and achieves the technical effect of reducing alarm lag and improving monitoring accuracy.

[0009] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] Figure 1A structural schematic diagram of a binocular vision-driven transmission line windage and sway monitoring device is provided for an embodiment of the present application.

[0011] Figure 2 A flow chart of a binocular vision-driven transmission line windage and sway monitoring method is provided for the embodiment of the present application; Explanation of the reference numerals: image acquisition unit 11 , feature extraction unit 12 , state analysis unit 13 , trajectory construction unit 14 , alarm generation unit 15 . DETAILED DESCRIPTION

[0012] The overall idea of ​​the technical solution provided by this application is as follows: The embodiments of the present application provide a binocular vision-driven transmission line wind deviation and dance monitoring device and method. The laser contour binocular image data formed by the projection of a straight line laser array on the transmission line is collected by a binocular camera, replacing the method of using laser point cloud scanning to obtain three-dimensional point cloud data in the prior art, thereby reducing the amount of calculation for data collection and processing. Then, by performing feature extraction on the laser contour binocular image data, a laser contour vector matrix is ​​obtained, and the conversion from the laser contour image to the laser contour vector is realized, thereby reducing the data dimension and calculation complexity of the subsequent wind deviation analysis. Next, the wind deviation state analysis and wind deviation dance trajectory construction are performed on the basis of the laser contour vector matrix, thereby avoiding the process of modeling the three-dimensional point cloud data and further reducing the calculation complexity. Afterwards, the wind deviation alarm function is realized by judging whether the wind deviation dance trajectory exceeds the safety range, thereby improving the effectiveness of wind deviation monitoring and early warning. Compared with the existing technology, the technical solution of binocular visual imaging and laser contour vector matrix analysis reduces the computational complexity while improving the real-time and effectiveness of wind deviation and fluctuation monitoring of transmission lines. It solves the technical problem that the existing technology uses laser point cloud data for three-dimensional modeling, and then performs modeling fusion analysis at multiple moments, which has high computational complexity and leads to serious lag and high error rate.

[0013] After introducing the basic principles of the present application, various non-limiting implementation methods of the present application will be specifically described below in conjunction with the drawings in the specification.

[0014] Embodiment 1 like Figure 1 As shown, the embodiment of the present application provides a binocular vision-driven transmission line windage and dancing monitoring device, the device comprising: The image acquisition unit 11 is used to acquire two-dimensional image data of the laser profile of a line laser array through a binocular camera.

[0015] Specifically, a binocular camera is used to collect images of a straight-line laser array installed on a transmission line. The straight-line laser array consists of multiple straight-line lasers, each of which emits a laser beam. When the transmission line is deflected by the wind, the laser beam emitted by the straight-line laser will deviate from its original position, forming a specific laser profile. The binocular camera simultaneously captures the laser profile of the straight-line laser array from two different angles to obtain two left and right images, namely, two-dimensional image data of the laser profile. These two-dimensional image data contain information such as the position and shape of the laser profile, providing a data basis for subsequent feature extraction and wind deviation analysis.

[0016] A binocular camera is used for image acquisition. Compared with a monocular camera, a binocular camera can obtain depth information of the scene, which helps to more accurately analyze the spatial position of the laser profile, thereby realizing accurate monitoring of wind deviation of the transmission line.

[0017] The feature extraction unit 12 is used to extract features from the laser profile two-dimensional image data through a laser profile vector analysis channel to obtain a laser profile vector matrix.

[0018] Specifically, the laser profile vector analysis channel is an algorithm module for analyzing laser profile features. It receives two-dimensional image data of the laser profile and extracts vector data that can characterize the laser profile features through a series of image processing and calculation steps. The laser profile vector includes information such as the position, shape, and direction of the laser profile. By analyzing the left and right laser profile images, a vector representation of the laser profile in three-dimensional space is obtained. The feature extraction unit 12 organizes the extracted laser profile vectors into a matrix, namely the laser profile vector matrix. Each row in the matrix corresponds to a laser profile vector and a straight-line laser, which is convenient for subsequent state analysis and dancing trajectory construction.

[0019] The laser profile vector analysis channel quickly and accurately extracts the feature information of the laser profile from the two-dimensional image data through the feature extraction unit 12. Compared with the traditional three-dimensional modeling, the efficiency and accuracy of feature extraction are improved, laying a foundation for the real-time monitoring of wind deviation of power transmission lines.

[0020] The state analysis unit 13 is used to perform fusion analysis on the laser profile vector matrix through a line windage feature analysis channel to obtain a line windage state vector.

[0021] Specifically, the line wind deviation feature analysis channel is an algorithm model used to comprehensively analyze the wind deviation characteristics of the transmission line. It receives the laser profile vector matrix, and through comprehensive analysis of each laser profile vector in the laser profile vector matrix, extracts a feature vector that can reflect the overall wind deviation state of the transmission line, that is, the line wind deviation state vector. Among them, the line wind deviation state vector includes multiple parameters such as the wind deviation angle, amplitude, and frequency of the transmission line, and describes the state characteristics of the transmission line under the action of wind deviation from different angles. The state analysis unit 13 calculates the various components of the line wind deviation state vector by integrating and analyzing the data in the laser profile vector matrix, taking into account the laser profile information of all straight-line lasers, and comprehensively and accurately evaluates the wind deviation state of the transmission line.

[0022] Among them, the line wind deviation feature analysis channel adopts a machine learning algorithm, and establishes a mapping relationship between laser profile features and wind deviation status through training of a large amount of historical data. Therefore, this channel can quickly and accurately extract wind deviation status information from the laser profile vector matrix, and realize real-time monitoring and early warning of wind deviation of transmission lines. Compared with the traditional wind deviation monitoring method through three-dimensional modeling, the state analysis using laser profile features avoids complex image processing and three-dimensional reconstruction processes, reducing computing costs and time overhead. At the same time, by comprehensively analyzing the laser profile information of multiple straight-line lasers, the reliability and accuracy of wind deviation monitoring are improved.

[0023] The trajectory construction unit 14 is used to construct a line windage dancing trajectory according to the line windage state vector.

[0024] Specifically, the line windage state vector reflects the windage state of the transmission line at a certain moment, including parameters such as windage angle, amplitude, and frequency. By analyzing and fitting the windage state vectors at multiple consecutive moments, the windage dance trajectory of the transmission line over a period of time is obtained. Specifically, the trajectory construction unit 14 uses an algorithm based on time series analysis to arrange the line windage state vectors at different moments in chronological order to form a state sequence. Then, by performing smoothing, curve fitting and other operations on the state sequence, a continuous and smooth line windage dance trajectory is generated. The line windage dance trajectory intuitively shows the movement laws and trends of the transmission line under the action of windage, and provides an important basis for the safety monitoring and maintenance of the transmission line.

[0025] By analyzing the shape, amplitude, frequency and other characteristics of the line wind deviation dancing trajectory, the wind deviation trend of the transmission line can be predicted, and abnormal wind deviation conditions can be discovered and warned in a timely manner. The trajectory construction unit 14 makes full use of the time and space information contained in the wind deviation state vector, and generates a continuous and smooth line wind deviation dancing trajectory through mathematical methods such as time series analysis and curve fitting. Compared with the traditional wind deviation analysis method based on discrete data points, the obtained wind deviation dancing trajectory is more accurate and intuitive, which is convenient for analysis and decision-making.

[0026] The alarm generating unit 15 is used to generate a line wind deviation alarm when the line wind deviation dancing trajectory exceeds the wind deviation safety range.

[0027] Specifically, the transmission line will produce different degrees of dancing under the action of wind deviation, but the amplitude and frequency of the dancing need to be controlled within a certain range to ensure the safe operation of the transmission line. The wind deviation safety range is pre-set according to the design parameters and operating experience of the transmission line, and represents the maximum wind deviation dancing amplitude and frequency that the transmission line can withstand. The alarm generation unit 15 monitors the line wind deviation dancing trajectory in real time and compares it with the wind deviation safety range. When the amplitude or frequency of the line wind deviation dancing trajectory exceeds the wind deviation safety range, the alarm generation unit 15 automatically generates a line wind deviation alarm. Among them, the line wind deviation alarm information includes key parameters such as the time, location, amplitude, and frequency of the wind deviation. This information will be transmitted to the monitoring center and maintenance personnel of the transmission line in real time through a wired or wireless communication network, so that corresponding safety measures can be taken in time, such as adjusting the tension of the transmission line, adding damping devices, etc.

[0028] By using the alarm generation unit 15, alarm judgment is made based on the wind deviation dancing trajectory of the line, so that the wind deviation status of the transmission line can be monitored in real time and continuously, and an alarm can be issued in time when the wind deviation tends to be dangerous, thereby providing reliable guarantee for the safe operation of the transmission line and improving the safety and reliability of the transmission line.

[0029] Furthermore, the image acquisition unit 11 includes the following execution steps: By means of the binocular camera, two-dimensional image data of a first laser beam of a first line laser of the line laser array on which a first electronic tag is installed is collected, wherein the first line laser is deployed at a first preset monitoring point of the transmission line, and when no wind deviation occurs, the first laser beam coincides with a preset area of ​​the first transmission line; Until the binocular camera is used to collect two-dimensional image data of the Nth laser beam of the Nth line laser of the line laser array equipped with the Nth electronic tag, wherein the Nth line laser is deployed at the Nth preset monitoring point of the transmission line, and when no wind deviation occurs, the Nth laser beam coincides with the preset area of ​​the Nth transmission line; adding the first laser beam two-dimensional image data to the Nth laser beam two-dimensional image data into the laser profile two-dimensional image data; Among them, the first transmission line preset area to the Nth transmission line preset area constitute a transmission line.

[0030] In a feasible implementation, a binocular camera is used to collect two-dimensional image data of the laser beam of each line laser in the line laser array, and these data are added to the two-dimensional image data of the laser profile to realize wind deviation monitoring of the entire transmission line.

[0031] A plurality of monitoring points are preset at different positions of the transmission line, and a line laser is installed at each monitoring point. Each line laser is identified by a unique electronic tag, such as the first electronic tag corresponds to the first line laser, and the Nth electronic tag corresponds to the Nth line laser. In the absence of wind deflection, the laser beam emitted by each line laser should coincide with the preset area of ​​the transmission line corresponding to it. For example, the first laser beam emitted by the first line laser should coincide with the preset area of ​​the first transmission line, and the Nth laser beam emitted by the Nth line laser should coincide with the preset area of ​​the Nth transmission line. These preset areas are arranged in sequence to form a complete transmission line. The binocular camera collects the two-dimensional image data of the laser beam of each line laser in sequence according to the preset order. Specifically, the binocular camera first collects the two-dimensional image data of the first laser beam of the first line laser, and then collects the two-dimensional image data of the second laser beam of the second line laser, and so on, until the two-dimensional image data of the Nth laser beam of the Nth line laser is collected. The collected first laser beam two-dimensional image data, second laser beam two-dimensional image data, ..., Nth laser beam two-dimensional image data are sequentially added to the laser contour two-dimensional image data to form a complete transmission line laser contour image.

[0032] The laser profile of the entire transmission line is collected by using a straight line laser array and a binocular camera, which improves the coverage and spatial resolution of wind deviation monitoring. By dividing the transmission line into multiple preset areas and assigning a straight line laser to each area, the laser profile is accurately matched with the position of the transmission line, which provides a basis for subsequent wind deviation status analysis and positioning. In addition, the straight line laser is uniquely identified by an electronic tag, and the correspondence between the laser beam image data and the straight line laser is established. This method is simple and reliable, and it is easy to realize automatic control and data management. Compared with traditional monitoring methods, full coverage and automation of wind deviation monitoring of transmission lines is achieved, which improves monitoring efficiency and accuracy, and reduces costs and safety risks. At the same time, the obtained laser profile two-dimensional image data provides high-quality data input for subsequent feature extraction, status analysis and other steps, ensuring the reliability and accuracy of wind deviation monitoring.

[0033] Furthermore, the feature extraction unit 12 includes the following execution steps: The laser profile two-dimensional image data includes first laser beam two-dimensional image data, wherein the first laser beam two-dimensional image data includes a first eye laser beam image and a second eye laser beam image, the first eye laser beam image and the second eye laser beam image belong to the same plane, and the first eye camera optical axis of the first eye laser beam image and the second eye camera optical axis of the second eye laser beam image are parallel to each other, the first eye laser beam image and the second eye laser beam image have a one-to-one correspondence in pixel points, the first eye camera is a left camera, and the second eye camera is a right camera; The first image input channel of the laser profile vector analysis channel receives the first eye laser beam image and the first eye camera space position to extract horizontal coordinate features, and obtains the horizontal pixel value of the first eye laser beam image; The second image input channel of the laser profile vector analysis channel receives the second laser beam image and the second camera space position to extract horizontal coordinate features, and obtains the horizontal pixel value of the second laser beam image; Performing depth analysis through the horizontal pixel values ​​of the first laser beam image and the horizontal pixel values ​​of the second laser beam image to obtain a depth map of the first laser beam image; A first laser profile vector is constructed according to the first laser beam image depth map, the first eyepiece laser beam image and the second eyepiece laser beam image, added into the matrix position associated with the first electronic tag, and the laser profile vector matrix is ​​updated.

[0034] In a preferred embodiment, the binocular vision principle is used to extract features and perform depth analysis on the two-dimensional image data of the laser beam collected by the binocular camera, thereby constructing a laser profile vector matrix to achieve accurate characterization of the windage state of the transmission line.

[0035] The binocular camera consists of two cameras on the left and right, namely the first eye camera and the second eye camera. The two cameras are located in the same plane, with their optical axes parallel to each other and imaging pixels corresponding to each other. The first eye camera acquires the first eye laser beam image, and the second eye camera acquires the second eye laser beam image. These two images together constitute the two-dimensional image data of the first laser beam, which is a component of the two-dimensional image data of the laser profile. In the feature extraction process, the laser profile vector parsing channel includes a first image input channel and a second image input channel, which receive the first eye laser beam image and the second eye laser beam image respectively. At the same time, in order to obtain the spatial position information of the laser beam, the first image input channel also receives the spatial position of the first eye camera, and the second image input channel receives the spatial position of the second eye camera.

[0036] Next, the first image input channel determines the horizontal position of the laser beam on the left camera imaging plane by combining the image and the camera position based on the first eye laser beam image and the first eye camera spatial position information, and represents it as the horizontal pixel value of the first eye laser beam image. Similarly, the second image input channel determines the horizontal position of the laser beam on the right camera imaging plane by processing the second eye laser beam image and the second eye camera spatial position information, and represents it as the horizontal pixel value of the second eye laser beam image. Although both of these horizontal pixel values ​​represent the horizontal coordinates of the laser beam, there are differences due to the different spatial positions of the two cameras. Therefore, after obtaining the horizontal pixel values ​​of the first eye laser beam image and the horizontal pixel values ​​of the second eye laser beam image, the two pixel values ​​will be subjected to depth analysis to obtain the depth map of the first laser beam image.

[0037] Since the two cameras are located at different spatial positions, there is a certain position offset in the observed laser beam image. This offset is parallax, which is related to the distance from the laser beam to the camera. The larger the parallax, the closer the laser beam is to the camera; conversely, the smaller the parallax, the farther the laser beam is from the camera. In the depth analysis process, the laser profile vector parsing channel first calculates the parallax between the horizontal pixel values ​​of the first laser beam image and the horizontal pixel values ​​of the second laser beam image; then, the parallax is converted into depth information using the camera's intrinsic parameters (such as focal length, principal point coordinates, etc.) and extrinsic parameters (such as camera spacing, rotation matrix, etc.). By performing depth calculation on each pixel in the laser beam image, the laser profile vector parsing channel obtains the depth map of the first laser beam image. The depth map is a grayscale image of the same size as the laser beam image, in which the grayscale value of each pixel represents the distance from the laser beam to the camera corresponding to the point, providing the position information of the laser beam in three-dimensional space. Through the depth analysis step, the depth map of the first laser beam image is obtained, which will be used together with the first laser beam image and the second laser beam image to construct the first laser profile vector, and then update the laser profile vector matrix.

[0038] After obtaining the first laser beam image depth map, the feature extraction unit 12 first extracts the two-dimensional coordinate information of the laser beam from the first laser beam image and the second laser beam image, and then combines these two-dimensional coordinates with the depth values ​​of the corresponding points in the first laser beam image depth map to obtain the coordinates of the laser beam in three-dimensional space, thereby obtaining the first laser profile vector. After constructing the first laser profile vector, the feature extraction unit 12 adds it to a specific position of the laser profile vector matrix. The laser profile vector matrix is ​​a two-dimensional array, in which each row corresponds to a one-line laser and each column corresponds to a time point. Each element in the matrix is ​​a laser profile vector, which represents the spatial information of the laser beam emitted by a one-line laser at a certain time point. The feature extraction unit 12 uses the first electronic tag to find the row corresponding to the first one-line laser, and then writes the first laser profile vector to the column corresponding to the current time point of the row. In this way, the laser profile vector matrix is ​​updated and the new laser beam information is recorded. Among them, the laser profile vector matrix is ​​a dynamically updated data structure. As time goes by and new laser beam images are collected, the feature extraction unit 12 continuously constructs new laser profile vectors and adds them to the matrix. By analyzing the laser profile vectors at different time points in the matrix, the motion trajectory of the laser beam can be tracked, and the windage state of the transmission line can be determined.

[0039] Compared with the traditional laser scanning ranging method, the use of binocular vision and laser profile analysis technology avoids the complex point cloud processing and 3D reconstruction process, reducing the calculation cost and time overhead. At the same time, through the matrix management of laser profile vectors, efficient storage and fast access of laser beam features are achieved, laying the foundation for subsequent wind deviation status analysis and early warning.

[0040] Furthermore, the feature extraction unit 12 further includes the following execution steps: Get the binocular camera distance, the focal length of the first eye camera and the focal length of the second eye camera; Obtain a disparity map of the first eye laser beam image by performing a one-to-one comparison of horizontal pixel values ​​of the first eye laser beam image and horizontal pixel values ​​of the second eye laser beam image; The binocular camera distance, the first binocular camera focal length, the second binocular camera focal length, and the first binocular laser beam image disparity map are fused through the depth map analysis channel of the laser profile vector analysis channel to obtain the first laser beam image depth map.

[0041] In a preferred embodiment, the feature extraction unit 12 obtains the parameters of the binocular camera, calculates the disparity map of the first laser beam image, and generates the depth map of the first laser beam image by using the depth map analysis channel of the laser profile vector analysis channel.

[0042] First, the feature extraction unit 12 obtains the key parameters of the binocular camera: binocular camera distance, first eye camera focal length and second eye camera focal length. The binocular camera distance refers to the distance between the optical centers of the left and right cameras, which determines the baseline length of the binocular vision; the first eye camera focal length and the second eye camera focal length respectively represent the focal length values ​​of the left and right cameras, which determine the imaging ratio of the camera. These parameters are obtained during the camera calibration process and are an important basis for depth estimation. Next, the feature extraction unit 12 calculates the disparity map of the first eye laser beam image using the horizontal pixel value of the first eye laser beam image and the horizontal pixel value of the second eye laser beam image. Specifically, by one-to-one correspondence and comparison of each pixel point in the left and right images, the coordinate deviation of each pixel point in the horizontal direction is calculated, thereby obtaining a disparity map. The disparity map is the same size as the original image, where the grayscale value of each pixel point represents the horizontal coordinate deviation of the point in the left and right images. Afterwards, the feature extraction unit 12 fuses the binocular camera distance, the focal length of the first binocular camera, the focal length of the second binocular camera, and the disparity map of the first binocular laser beam image through the depth map analysis channel of the laser profile vector analysis channel to generate a depth map of the first laser beam image. The depth map analysis channel adopts the principle of triangulation and uses the position relationship and disparity information of the binocular cameras to calculate the depth value corresponding to each pixel. Specifically, the disparity value and parameters such as the binocular camera distance and focal length are substituted into the triangulation formula to obtain the depth value of the pixel, and the depth values ​​of all pixels are combined into the depth map of the first laser beam image.

[0043] Furthermore, the feature extraction unit 12 further includes the following execution steps: Obtaining an acquisition timestamp of the first laser beam two-dimensional image data; When the acquisition timestamp does not belong to the preset correction time zone, the depth calculation formula is obtained: , in, Represents the depth value of the i-th pixel, Characterizes the focal length of the first eye camera, Characterizes the focal length of the second eye camera, Characterizes the disparity of the i-th pixel, Characterize the binocular camera distance; According to the depth calculation formula, a linear calculation is performed based on the binocular camera distance, the first binocular camera focal length, the second binocular camera focal length, and the first binocular laser beam image disparity map to obtain the first laser beam image depth map.

[0044] In a preferred embodiment, the feature extraction unit 12 first obtains the acquisition timestamp of the first laser beam two-dimensional image data, and then selects different depth calculation methods according to whether the timestamp belongs to a preset correction time zone to obtain the first laser beam image depth map.

[0045] The feature extraction unit 12 obtains the acquisition timestamp of the data while acquiring the two-dimensional image data of the first laser beam. The acquisition timestamp records the acquisition time of the image data and is an important basis for judging the current ambient lighting conditions. In some cases, the ambient lighting conditions will affect the depth estimation. In particular, during the period of direct sunlight during the day, the strong light will interfere with the imaging of the laser beam, resulting in nonlinear errors in the depth estimation. In order to cope with this situation, a preset correction time zone is introduced. The preset correction time zone is a time interval pre-set according to the local sunshine time and light intensity. Within this time zone, the ambient lighting conditions are usually ideal and will not have a significant impact on the depth estimation. Outside this time zone, such as at night, the impact of light is limited, affecting the imaging effect, or during the period of direct sunlight during the day, the strong light will interfere with the imaging of the laser beam, resulting in nonlinear errors in the depth estimation.

[0046] When the acquisition timestamp does not belong to the preset correction time zone, it means that the current ambient lighting conditions will interfere with the depth estimation. In order to reduce the influence of lighting, the feature extraction unit 12 uses a linear depth calculation formula to estimate the depth value of the pixel. The formula takes into account the baseline distance of the binocular camera, the focal length of the left and right cameras, and the parallax value of the pixel to calculate the depth value of the pixel. Specifically, for the i-th pixel, its depth value is The calculation formula is: ,in, Represents the depth value of the i-th pixel, and Represent the focal lengths of the first and second eyepiece cameras, respectively. represents the disparity value of the i-th pixel, Represents the binocular camera distance. The feature extraction unit 12 substitutes the binocular camera distance, the focal length of the first binocular camera, the focal length of the second binocular camera and the disparity map of the first binocular laser beam image into the above formula, and obtains the first laser beam image depth map by linear calculation. In the depth map, the gray value of each pixel corresponds to its depth value, reflecting the distance information of the point in the three-dimensional space, thereby obtaining the first laser beam image depth map.

[0047] By introducing the acquisition timestamp and the preset correction time zone, the feature extraction unit 12 adaptively selects the depth estimation method according to the changes in the ambient lighting conditions, thereby improving the accuracy and reliability of depth map generation, effectively reducing lighting interference, and improving environmental adaptability.

[0048] Furthermore, the feature extraction unit 12 further includes the following execution steps: When the acquisition timestamp belongs to the preset correction time zone, obtaining an ambient brightness monitoring value; By using the ambient brightness monitoring value and the acquisition timestamp, based on the binocular camera distance, the first binocular camera focal length, and the second binocular camera focal length, depth value sample acquisition is performed on the first binocular laser beam image disparity map to obtain a plurality of sample depth distribution maps; The mean values ​​of the depth distribution maps of the plurality of samples are counted to obtain the first laser beam image depth map.

[0049] In a feasible implementation, when the acquisition timestamp belongs to the preset correction time zone, first, the feature extraction unit 12 obtains the current ambient brightness monitoring value. The ambient brightness monitoring value is measured by the light intensity sensor and reflects the light intensity of the current scene. Combined with the acquisition timestamp, the feature extraction unit 12 can determine the current lighting conditions and provide a reference for subsequent depth estimation. Next, the feature extraction unit 12 collects depth value samples for the first eye laser beam image disparity map based on the binocular camera distance, the focal length of the first eye camera, and the focal length of the second eye camera. Specifically, a number of representative pixels are selected in the first eye laser beam image disparity map, and the depth values ​​of these pixels are calculated to obtain a number of sample depth distribution maps. Each sample depth distribution map reflects the distribution of depth values ​​of different pixels under specific lighting conditions. Afterwards, the feature extraction unit 12 performs mean statistics on a number of sample depth distribution maps to obtain a first laser beam image depth map. Specifically, for each pixel, there is a corresponding depth value in different sample depth distribution maps. The feature extraction unit 12 calculates the mean value of the depth values ​​of each pixel in all samples as the depth value of the pixel in the first laser beam image depth map. In this way, the error of a single sample is effectively reduced, and the accuracy and stability of depth estimation are improved.

[0050] In the preset correction time zone, the ambient lighting conditions may change with time and location. Collecting multiple sample depth distribution maps can better adapt to changes in lighting conditions and improve the robustness of depth estimation. At the same time, by performing mean statistics on multiple samples, the fluctuations caused by lighting changes can be smoothed to obtain more stable and reliable depth estimation results.

[0051] Furthermore, the line wind deviation feature analysis channel construction step includes: Obtaining a laser profile vector matrix record data set and a line windage state vector identification array data set; Using the line windage state vector identification data set as supervision, the laser profile vector matrix record data set is trained to generate the line windage feature analysis channel; Among them, the wind deviation state vector identification array of any line includes wind deviation angle, dancing amplitude, vertical dancing amplitude, horizontal dancing amplitude, and elliptical tilt angle data.

[0052] Specifically, in order to construct a line wind deviation feature analysis channel, two data sets are first obtained, namely the laser profile vector matrix recording data set and the line wind deviation state vector identification array data set. The laser profile vector matrix recording data set records the laser profile vector matrix of the transmission line at different time points. Each matrix represents the spatial position and direction information of the laser beams of all straight-line lasers at a certain moment, reflecting the geometric deformation characteristics of the transmission line under the action of wind deviation. The line wind deviation state vector identification array data set records the wind deviation state of the transmission line corresponding to the laser profile vector matrix recording data set. Each wind deviation state vector identification array contains several parameters describing the wind deviation state, including wind deviation angle, dancing amplitude, vertical dancing amplitude, horizontal dancing amplitude and elliptical inclination angle, which characterizes the movement characteristics of the transmission line under the action of wind deviation from different angles. Among them, the wind deflection angle indicates the angle at which the transmission line deviates from its initial position in the horizontal plane, reflecting the influence of wind direction on the transmission line; the dancing amplitude indicates the vibration amplitude of the transmission line in space, reflecting the intensity of the wind; the vertical dancing amplitude and the horizontal dancing amplitude respectively indicate the vibration amplitude of the transmission line in the vertical and horizontal directions, reflecting the components of wind in different directions; the elliptical inclination angle indicates the angle between the major axis of the ellipse of the vibration trajectory of the transmission line and the horizontal direction, reflecting the spatial distribution characteristics of the wind.

[0053] After obtaining the above two data sets, we started to train the line wind deviation feature analysis channel. The training process uses supervised learning, using the line wind deviation state vector identification array data set as a label to train the laser profile vector matrix record data set. By optimizing the model parameters, the analysis channel can accurately predict the corresponding wind deviation state vector based on the input laser profile vector matrix. The trained line wind deviation feature analysis channel establishes a mapping relationship between the laser profile feature and the wind deviation state. In practical applications, this channel can convert the laser profile vector matrix into a line wind deviation state vector in real time, realizing rapid diagnosis and early warning of wind deviation of transmission lines.

[0054] Through the line wind deviation feature analysis channel, the automatic mapping between laser profile features and wind deviation status is realized, which improves the intelligent level of wind deviation monitoring. Compared with the traditional wind deviation diagnosis method based on empirical models, it can adaptively capture the complex characteristics of transmission line wind deviation, has stronger generalization ability for unknown wind deviation conditions, and provides tools for all-weather, high-precision monitoring of transmission lines.

[0055] Furthermore, the trajectory construction unit 14 includes the following execution steps: The line windage state vector includes the line windage state vector at the first moment to the line windage state vector at the Mth moment, the first moment to the Mth moment are continuous moments, the Mth moment is the current moment, and the first moment is the earliest moment traced back according to the preset backtracking step length; The line windage dancing trajectory is constructed according to the line windage state vector at the first moment to the line windage state vector at the Mth moment.

[0056] In a preferred embodiment, the line windage state vector includes a series of windage state vectors from the first moment to the Mth moment. These moments are continuous, wherein the Mth moment is the current moment, and the first moment is the earliest moment determined by tracing back from the current moment according to the preset tracing back step. The preset tracing back step is the length of time, which determines the time span of the historical data based on which the line windage dancing trajectory is generated. For example, if the preset tracing back step is 10 minutes and the current moment is 12:00, then the first moment is 11:50, and the time span from the first moment to the Mth moment is 10 minutes. Within these 10 minutes, the state analysis unit 13 generates a line windage state vector at a certain time interval (such as 1 second), thereby obtaining a sequence containing 600 windage state vectors, i.e., the line windage state vector.

[0057] The trajectory construction unit 14 constructs a line wind deviation dancing trajectory based on the line wind deviation state vector. The line wind deviation dancing trajectory is a curve that changes with time, in which each point corresponds to the wind deviation state of the transmission line at a certain moment. By mapping the parameters such as the wind deviation angle and the dancing amplitude in the line wind deviation state vector to the corresponding positions of the trajectory curve, a complete wind deviation dancing trajectory is obtained, which intuitively shows the wind deviation changes of the transmission line over a period of time.

[0058] The trajectory construction unit 14 realizes dynamic tracking and prediction of the wind deviation state of the transmission line. The wind deviation dancing trajectory of the line, as an intuitive and comprehensive representation of the wind deviation state, provides an important basis for the safety monitoring and maintenance decision-making of the transmission line.

[0059] In summary, the binocular vision-driven transmission line windage and sway monitoring device provided in the embodiment of the present application has the following technical effects: The image acquisition unit is used to collect the two-dimensional image data of the laser profile of the line laser array through a binocular camera, obtain the original image data required for the wind deviation and dancing monitoring of the transmission line, and lay the foundation for subsequent processing. The feature extraction unit is used to extract the features of the two-dimensional image data of the laser profile through the laser profile vector analysis channel, obtain the laser profile vector matrix, realize the conversion of the laser profile image to the laser profile vector, reduce the data dimension, and provide a simplified data representation for the subsequent wind deviation state analysis. The state analysis unit is used to perform fusion analysis on the laser profile vector matrix through the line wind deviation feature analysis channel to obtain the line wind deviation state vector, realize the direct judgment of the wind deviation state, and avoid the three-dimensional modeling process. The trajectory construction unit is used to construct the line wind deviation dancing trajectory according to the line wind deviation state vector, realize the mapping of the wind deviation state to the wind deviation dancing trajectory, intuitively reflect the movement of the transmission line under the action of wind force, and provide a judgment basis for wind deviation alarm. The alarm generating unit is used to generate a line wind deviation alarm when the wind deviation dancing trajectory of the line exceeds the wind deviation safety range, thereby realizing the timely early warning function of wind deviation monitoring and improving the safety of transmission line operation.

[0060] Embodiment 2 Based on the same inventive concept as the binocular vision driven transmission line windage and dancing monitoring device in the aforementioned embodiment, Figure 2 As shown, the embodiment of the present application provides a binocular vision-driven transmission line windage and dancing monitoring method, the method comprising: The binocular camera is used to collect the two-dimensional image data of the laser profile of the linear laser array; Through a laser profile vector analysis channel, feature extraction is performed on the laser profile two-dimensional image data to obtain a laser profile vector matrix; Through the line windage feature analysis channel, the laser profile vector matrix is ​​fused and analyzed to obtain the line windage state vector; Constructing a line windage dancing trajectory according to the line windage state vector; When the line wind deviation dancing trajectory exceeds the wind deviation safety range, a line wind deviation alarm is generated.

[0061] Furthermore, the binocular camera is used to collect the two-dimensional image data of the laser profile of the line laser array, including: By means of the binocular camera, two-dimensional image data of a first laser beam of a first line laser of the line laser array on which a first electronic tag is installed is collected, wherein the first line laser is deployed at a first preset monitoring point of the transmission line, and when no wind deviation occurs, the first laser beam coincides with a preset area of ​​the first transmission line; Until the binocular camera is used to collect two-dimensional image data of the Nth laser beam of the Nth line laser of the line laser array equipped with the Nth electronic tag, wherein the Nth line laser is deployed at the Nth preset monitoring point of the transmission line, and when no wind deviation occurs, the Nth laser beam coincides with the preset area of ​​the Nth transmission line; adding the first laser beam two-dimensional image data to the Nth laser beam two-dimensional image data into the laser profile two-dimensional image data; Among them, the first transmission line preset area to the Nth transmission line preset area constitute a transmission line.

[0062] Furthermore, the laser profile two-dimensional image data is subjected to feature extraction through a laser profile vector analysis channel to obtain a laser profile vector matrix, including: The laser profile two-dimensional image data includes first laser beam two-dimensional image data, wherein the first laser beam two-dimensional image data includes a first eye laser beam image and a second eye laser beam image, the first eye laser beam image and the second eye laser beam image belong to the same plane, and the first eye camera optical axis of the first eye laser beam image and the second eye camera optical axis of the second eye laser beam image are parallel to each other, the first eye laser beam image and the second eye laser beam image have a one-to-one correspondence in pixel points, the first eye camera is a left camera, and the second eye camera is a right camera; The first image input channel of the laser profile vector analysis channel receives the first eye laser beam image and the first eye camera space position to extract horizontal coordinate features, and obtains the horizontal pixel value of the first eye laser beam image; The second image input channel of the laser profile vector analysis channel receives the second laser beam image and the second camera space position to extract horizontal coordinate features, and obtains the horizontal pixel value of the second laser beam image; Performing depth analysis through the horizontal pixel values ​​of the first laser beam image and the horizontal pixel values ​​of the second laser beam image to obtain a depth map of the first laser beam image; A first laser profile vector is constructed according to the first laser beam image depth map, the first eyepiece laser beam image and the second eyepiece laser beam image, added into the matrix position associated with the first electronic tag, and the laser profile vector matrix is ​​updated.

[0063] Further, a depth analysis is performed on the horizontal pixel values ​​of the first laser beam image and the horizontal pixel values ​​of the second laser beam image to obtain a depth map of the first laser beam image, including: Get the binocular camera distance, the focal length of the first eye camera and the focal length of the second eye camera; Obtain a disparity map of the first eye laser beam image by performing a one-to-one comparison of horizontal pixel values ​​of the first eye laser beam image and horizontal pixel values ​​of the second eye laser beam image; The binocular camera distance, the first binocular camera focal length, the second binocular camera focal length, and the first binocular laser beam image disparity map are fused through the depth map analysis channel of the laser profile vector analysis channel to obtain the first laser beam image depth map.

[0064] Furthermore, the binocular camera distance, the focal length of the first binocular camera, the focal length of the second binocular camera, and the disparity map of the first binocular laser beam image are fused through the depth map analysis channel of the laser profile vector analysis channel to obtain the first laser beam image depth map, including: Obtaining an acquisition timestamp of the first laser beam two-dimensional image data; When the acquisition timestamp does not belong to the preset correction time zone, the depth calculation formula is obtained: , in, Represents the depth value of the i-th pixel, Characterizes the focal length of the first eye camera, Characterizes the focal length of the second eye camera, Characterizes the disparity of the i-th pixel, Characterize the binocular camera distance; According to the depth calculation formula, a linear calculation is performed based on the binocular camera distance, the first binocular camera focal length, the second binocular camera focal length, and the first binocular laser beam image disparity map to obtain the first laser beam image depth map.

[0065] Furthermore, the embodiment of the present application also includes: When the acquisition timestamp belongs to the preset correction time zone, obtaining an ambient brightness monitoring value; By using the ambient brightness monitoring value and the acquisition timestamp, based on the binocular camera distance, the first binocular camera focal length, and the second binocular camera focal length, depth value sample acquisition is performed on the first binocular laser beam image disparity map to obtain a plurality of sample depth distribution maps; The mean values ​​of the depth distribution maps of the plurality of samples are counted to obtain the first laser beam image depth map.

[0066] Furthermore, the line wind deviation feature analysis channel construction step includes: Obtaining a laser profile vector matrix record data set and a line windage state vector identification array data set; Using the line windage state vector identification data set as supervision, the laser profile vector matrix record data set is trained to generate the line windage feature analysis channel; Among them, the wind deviation state vector identification array of any line includes wind deviation angle, dancing amplitude, vertical dancing amplitude, horizontal dancing amplitude, and elliptical tilt angle data.

[0067] Furthermore, according to the line windage state vector, a line windage dancing trajectory is constructed, including: The line windage state vector includes the line windage state vector at the first moment to the line windage state vector at the Mth moment, the first moment to the Mth moment are continuous moments, the Mth moment is the current moment, and the first moment is the earliest moment traced back according to the preset backtracking step length; The line windage dancing trajectory is constructed according to the line windage state vector at the first moment to the line windage state vector at the Mth moment.

[0068] Any step of the method described above can be stored as a computer instruction or program in an unlimited computer memory, and can be called and recognized by an unlimited computer processor to implement any method in the embodiments of the present application, without any unnecessary restrictions.

[0069] Furthermore, the first or second mentioned above may not only represent an order relationship, but may also represent a specific concept, and / or refer to the selection of multiple elements individually or in whole. Obviously, those skilled in the art may make various changes and modifications to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalents, the present application intends to include these changes and variations.

Claims

1. A binocular vision-driven transmission line windage and sway monitoring device, characterized in that: include: An image acquisition unit, the image acquisition unit is used to acquire two-dimensional image data of the laser profile of a line laser array through a binocular camera; A feature extraction unit, the feature extraction unit is used to extract features from the laser profile two-dimensional image data through a laser profile vector analysis channel to obtain a laser profile vector matrix; A state analysis unit, the state analysis unit is used to perform fusion analysis on the laser profile vector matrix through a line windage feature analysis channel to obtain a line windage state vector; A trajectory construction unit, the trajectory construction unit is used to construct a line wind deviation dancing trajectory according to the line wind deviation state vector; An alarm generating unit is used to generate a line wind deviation alarm when the line wind deviation dancing trajectory exceeds the wind deviation safety range.

2. The binocular vision driven transmission line windage and sway monitoring device according to claim 1, characterized in that: The image acquisition unit includes the following execution steps: By means of the binocular camera, two-dimensional image data of a first laser beam of a first line laser of the line laser array on which a first electronic tag is installed is collected, wherein the first line laser is deployed at a first preset monitoring point of the transmission line, and when no wind deviation occurs, the first laser beam coincides with a preset area of ​​the first transmission line; Until the binocular camera is used to collect two-dimensional image data of the Nth laser beam of the Nth line laser of the line laser array equipped with the Nth electronic tag, wherein the Nth line laser is deployed at the Nth preset monitoring point of the transmission line, and when no wind deviation occurs, the Nth laser beam coincides with the preset area of ​​the Nth transmission line; adding the first laser beam two-dimensional image data to the Nth laser beam two-dimensional image data into the laser profile two-dimensional image data; Among them, the first transmission line preset area to the Nth transmission line preset area constitute a transmission line.

3. The binocular vision driven transmission line windage and sway monitoring device according to claim 1, characterized in that: The feature extraction unit includes the following execution steps: The laser profile two-dimensional image data includes first laser beam two-dimensional image data, wherein the first laser beam two-dimensional image data includes a first eye laser beam image and a second eye laser beam image, the first eye laser beam image and the second eye laser beam image belong to the same plane, and the first eye camera optical axis of the first eye laser beam image and the second eye camera optical axis of the second eye laser beam image are parallel to each other, the first eye laser beam image and the second eye laser beam image have a one-to-one correspondence in pixel points, the first eye camera is a left camera, and the second eye camera is a right camera; The first image input channel of the laser profile vector analysis channel receives the first eye laser beam image and the first eye camera space position to extract horizontal coordinate features, and obtains the horizontal pixel value of the first eye laser beam image; The second image input channel of the laser profile vector analysis channel receives the second laser beam image and the second camera space position to extract horizontal coordinate features, and obtains the horizontal pixel value of the second laser beam image; Performing depth analysis through the horizontal pixel values ​​of the first laser beam image and the horizontal pixel values ​​of the second laser beam image to obtain a depth map of the first laser beam image; A first laser profile vector is constructed according to the first laser beam image depth map, the first eyepiece laser beam image and the second eyepiece laser beam image, added into the matrix position associated with the first electronic tag, and the laser profile vector matrix is ​​updated.

4. The binocular vision driven transmission line windage and sway monitoring device as claimed in claim 3, characterized in that: The feature extraction unit also includes the following execution steps: Get the binocular camera distance, the focal length of the first eye camera and the focal length of the second eye camera; Obtain a disparity map of the first eye laser beam image by performing a one-to-one comparison of horizontal pixel values ​​of the first eye laser beam image and horizontal pixel values ​​of the second eye laser beam image; The binocular camera distance, the first binocular camera focal length, the second binocular camera focal length, and the first binocular laser beam image disparity map are fused through the depth map analysis channel of the laser profile vector analysis channel to obtain the first laser beam image depth map.

5. The binocular vision driven transmission line windage and sway monitoring device as claimed in claim 4, characterized in that: The feature extraction unit also includes the following execution steps: Obtaining an acquisition timestamp of the first laser beam two-dimensional image data; When the acquisition timestamp does not belong to the preset correction time zone, the depth calculation formula is obtained: ， in, Represents the depth value of the i-th pixel, Characterizes the focal length of the first eye camera, Characterizes the focal length of the second eyepiece camera, Characterizes the disparity of the i-th pixel, Characterize the binocular camera distance; According to the depth calculation formula, a linear calculation is performed based on the binocular camera distance, the first binocular camera focal length, the second binocular camera focal length, and the first binocular laser beam image disparity map to obtain the first laser beam image depth map.

6. The binocular vision driven transmission line windage and sway monitoring device according to claim 5, characterized in that: The feature extraction unit also includes the following execution steps: When the acquisition timestamp belongs to the preset correction time zone, obtaining an ambient brightness monitoring value; By using the ambient brightness monitoring value and the acquisition timestamp, based on the binocular camera distance, the first binocular camera focal length, and the second binocular camera focal length, depth value sample acquisition is performed on the first binocular laser beam image disparity map to obtain a plurality of sample depth distribution maps; The mean values ​​of the depth distribution maps of the plurality of samples are counted to obtain the first laser beam image depth map.

7. The binocular vision driven transmission line windage and sway monitoring device according to claim 1, characterized in that: The line windage characteristic analysis channel construction step comprises: Obtaining a laser profile vector matrix record data set and a line windage state vector identification array data set; Using the line windage state vector identification data set as supervision, the laser profile vector matrix record data set is trained to generate the line windage feature analysis channel; Among them, the wind deviation state vector identification array of any line includes wind deviation angle, dancing amplitude, vertical dancing amplitude, horizontal dancing amplitude, and elliptical tilt angle data.

8. The binocular vision driven transmission line windage and sway monitoring device according to claim 1, characterized in that: The trajectory construction unit includes the following execution steps: The line windage state vector includes the line windage state vector at the first moment to the line windage state vector at the Mth moment, the first moment to the Mth moment are continuous moments, the Mth moment is the current moment, and the first moment is the earliest moment traced back according to the preset backtracking step length; The line windage dancing trajectory is constructed according to the line windage state vector at the first moment to the line windage state vector at the Mth moment.

9. A binocular vision-driven transmission line windage vibration monitoring method, characterized in that: include: The binocular camera is used to collect the two-dimensional image data of the laser profile of the linear laser array; Through a laser profile vector analysis channel, feature extraction is performed on the laser profile two-dimensional image data to obtain a laser profile vector matrix; Through the line windage feature analysis channel, the laser profile vector matrix is ​​fused and analyzed to obtain the line windage state vector; Constructing a line windage dancing trajectory according to the line windage state vector; When the wind deviation dancing trajectory of the line exceeds the wind deviation safety range, a line wind deviation alarm is generated.

Images