Transmission conductor sag measurement method based on binocular vision measurement device

Abstract

The invention discloses a transmission conductor sag measurement method based on a binocular vision measurement device. The method comprises the following steps: S1, fixing the binocular vision measurement device on one side of one transmission tower of a to-be-measured conductor sag; s2, collecting a left view and a right view, and performing internal reference distortion correction and stereo correction; s3, preprocessing the left view and the right view, performing stereo matching to obtain a disparity map, and calculating a horizontal span; s4, performing zero offset calibration and filtering on the original output of the tilt angle sensor to obtain tangential angles of the wires at the two ends; s5, inputting the horizontal span and the tangential angle into the catenary model to obtain a catenary constant, a lowest point, an end point height and a sag; according to the method, span and angle parameters are accurately obtained in a non-contact mode, the defects of terrain dependence, power failure requirements and precision in the prior art are overcome, conductor sag detection and inspection can be rapidly completed on site in a non-contact mode, and the method is high in measurement precision and robustness.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of power system transmission line equipment state monitoring, in particular to a transmission conductor sag measurement method based on a binocular vision measurement device. BACKGROUND

[0002] Transmission conductor sag refers to the vertical distance between the lowest point of the curve formed by the natural sag of the transmission conductor between two towers and the suspension point. It is a core mechanical parameter directly related to the safe operation of transmission lines, power supply reliability, and equipment life. Its precise control needs to be throughout the entire life cycle of line design, construction, and operation, which is specifically reflected in three aspects: ensuring the safe operation of the line, maintaining the reliability of the power grid, and prolonging the service life of the equipment. When the sag is too large, the safety distance of the conductor to the ground is insufficient, which can easily cause windage flashover and conductor interphase short circuit. When the sag is too small, the tension of the conductor increases significantly, and when extreme weather such as icing and strong winds occurs, the tension exceeds the tensile limit of the conductor, which can directly cause a conductor breakage accident, leading to a large area of power outage. If the conductor is in an unreasonable sag state for a long time, it will cause uneven stress on the suspension point fittings, leading to fitting wear and tear and conductor fatigue damage. At the same time, the continuous high tension caused by too small a sag can accelerate the fatigue of the conductor metal, reducing the service life of the conductor from the designed 25 years to less than 15 years, significantly increasing the operation and maintenance cost of the power grid. Therefore, real-time and accurate measurement of the transmission conductor sag is a key link to avoid line faults and ensure the safe and economic operation of the power grid.

[0003] The existing sag measurement techniques all have key defects: contact measurement (measuring rope method, tower climbing) requires power outage operation, which not only affects power supply reliability, but also faces safety risks such as high-altitude falling and induced electric shock, and is limited in complex terrain such as mountains and rivers, with low measurement efficiency. Laser range finder assisted single camera imaging measurement is a non-contact method, but the matching of the laser ranging point and the single camera image feature is easily disturbed by the perspective effect at large span, with a span measurement error of more than 5%, and the conductor inclination angle still needs to be analyzed through the image edge, with an angle error of 3° to 5° when there is shading or backlight, directly reducing the sag precision. Traditional non-vision indirect measurement has time synchronization error due to multi-parameter step-by-step acquisition, and the parabolic model deviates from the actual stress by more than 10% in a large height difference of more than 10m and large span scenes, which cannot meet the non-power outage, large span, and high-precision operation and maintenance requirements of transmission lines. SUMMARY

[0004] To overcome the above problems, the purpose of this invention is to provide a method for measuring the sag of power transmission conductors based on a binocular vision measurement device. This method relies on a binocular vision measurement device to measure the sag of power transmission conductors. This device accurately acquires the span and angle parameters in a non-contact manner, and combines the catenary equation to achieve high-precision calculation of the sag. It overcomes the terrain dependence, power outage requirements and accuracy defects of existing technologies, and can quickly and non-contactly complete the detection and inspection of conductor sag on site. It has high measurement accuracy, strong robustness and strong feasibility.

[0005] The technical solution adopted in this invention is:

[0006] A method for measuring the sag of power transmission lines based on a binocular vision measurement device includes the following steps:

[0007] S1: Fix the binocular vision measurement device to one side of one of the transmission towers of the conductor sag to be measured.

[0008] S2: Use a binocular vision measurement device to acquire left and right views and perform intrinsic parameter distortion correction and stereo correction.

[0009] S3: Preprocess the left and right views acquired in S2, perform stereo matching to obtain a disparity map, and calculate the horizontal span. .

[0010] S4: Perform zero-bias calibration and filtering on the raw output of the tilt sensor on the binocular measuring device to obtain the tangential angles of the two end wires. , .

[0011] S5: Calculate the horizontal clearance in S3. The tangential angle calculated in S4 , The input is fed into the catenary model to calculate the catenary constant. Lowest point Endpoint height sag .

[0012] As a further description of the present invention, the binocular vision measurement device includes a triangular support frame, an integrated bracket, a first monocular camera, a second monocular camera, and a tilt sensor. The integrated bracket is fixedly installed above the triangular support frame, and an adjustable slide rail is provided in the middle of the integrated bracket. The first monocular camera and the second monocular camera are installed in the slide rail, and the tilt sensor is fixedly installed on the integrated bracket in the middle position.

[0013] As a further description of the present invention, the specific process of intrinsic parameter distortion correction and stereo correction in S2 is as follows:

[0014] S21: select a chessboard grid of appropriate size, collect 100 images of different poses of the chessboard, covering the center and edges of the field of view.

[0015] S22: perform single target calibration, solve the intrinsic parameters of the left and right cameras respectively and remove distortion.

[0016] S23: solve the rotation matrix R and the translation matrix T by stereo calibration, and perform stereo correction.

[0017] S24: select according to the average re-projection error to meet the high-precision requirement.

[0018] As a further description of the application, the specific process of stereo matching and disparity map calculation in S3 is as follows:

[0019] Let the left image pixel be , and the corresponding right image pixel be , where is the disparity, and the disparity map is estimated by minimizing the pixel cost energy , which is approximately realized, and the calculation formula is:

[0020] .

[0021] where represents the pixel position ;

[0022] is the disparity of pixel ;

[0023] is the pixel matching cost, which is usually based on absolute disparity;

[0024] is the neighborhood of pixel ;

[0025] is the smoothing term.

[0026] As a further description of the application, the specific process of the lead tangent angle , in S4 is as follows:

[0027] S41: define the three-axis acceleration values of the tilt angle sensor as , which satisfy:

[0028] .

[0029] where is the gravitational acceleration.

[0030] S42: Tangential angle of wire , The calculation formula is:

[0031] .

[0032] S43: Filtering the output tangential angle in S42, the calculation formula is as follows:

[0033] .

[0034] Wherein, is the angular velocity integral term, is the inclination value obtained from acceleration, is the fusion coefficient.

[0035] As a further description of the application, the calculation formula of the catenary constant in S5 is:

[0036] .

[0037] The calculation formula of the lowest point

[0038] .

[0039] The calculation formula of the end point height

[0040] .

[0041] The calculation formula of the sag

[0042] .

[0043] Wherein, , are hyperbolic sine and hyperbolic cosine functions respectively.

[0044] As a further description of the application, the zero offset calibration and filtering of the original output of the inclination sensor on the binocular measuring device in S4 is specifically complementary filtering and sliding average processing, the proportional coefficient of the complementary filter is , and the sliding average window of the sliding average processing is .

[0045] ​​​​As further description of the present application, the tilt sensor adopts a three-axis MEMS inertial measurement unit, containing an accelerometer and a gyroscope, which are located inside the tilt sensor, and output the pitch angle at the wire suspension point in real time, and the tilt sensor is fixed through a magnetic base or a screw, and the installation height is consistent with the optical axis height of the first monocular camera and the second monocular camera, and the measurement angle is located in the same horizontal plane as the first monocular camera and the second monocular camera.

[0046] As further description of the present application, the slide rail is provided with a scale line.

[0047] As further description of the present application, the sag of the arc The calculation sets a threshold ,

[0048] When , the calculation formula of the sag of the arc is:

[0049] .

[0050] When , the calculation formula of the sag of the arc is:

[0051] .

[0052] Advantages of the present application:

[0053] The present application is a kind of transmission line sag measurement method based on binocular vision measuring device, and the measurement of transmission line sag relies on binocular vision measuring device, which accurately obtains span and angle parameters in a non-contact manner, and realizes high-precision calculation of sag by combining catenary equation, overcomes the terrain dependence, power outage demand and precision defects of prior art, and can quickly and non-contact complete wire sag detection and inspection on site, with high measurement precision, strong robustness and strong implementability. BRIEF DESCRIPTION OF DRAWINGS

[0054] Fig. 1 A kind of transmission line sag measurement method based on binocular vision measuring device flow chart is provided for the present application.

[0055] Fig. 2 A kind of transmission line sag measurement method based on binocular vision measuring device field installation and measurement schematic diagram is provided for the present application.

[0056] Fig. 3 A kind of transmission line sag measurement method based on binocular vision measuring device structure schematic diagram of binocular data measurement device is provided for the present application.

[0057] BRIEF DESCRIPTION OF DRAWINGS:​​

[0058] 1 - triangular support frame;

[0059] 2 - one-piece support, 21 - slide rail;

[0060] 3 - first monocular camera;

[0061] 4 - second monocular camera;

[0062] 5 - tilt sensor. DETAILED DESCRIPTION

[0063] In order to make the above objectives, features and advantages of the present application more obvious and easy to understand, the specific embodiments of the present application will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor should fall within the scope of protection of the present application.

[0064] In the following description, many specific details are set forth in order to provide a thorough understanding of the present application. However, the present application can be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the present application, so the present application is not limited to the specific embodiments disclosed below.

[0065] Secondly, the "one embodiment" or "embodiment" referred to herein means that the specific features, structures or characteristics can be included in at least one implementation of the present application. "In one embodiment" appearing in different places in the specification does not mean the same embodiment, nor is it an independent or alternative embodiment that excludes other embodiments.

[0066] The present application is described in detail in conjunction with the schematic drawings, and in the detailed description of the embodiments of the present application, the cross-sectional view of the device structure is partially enlarged without general proportion for the convenience of description, and the schematic drawings are only examples, which should not limit the scope of protection of the present application herein. In addition, three-dimensional spatial dimensions including length, width and depth should be included in actual manufacture.

[0067] Meanwhile, in the description of the present application, it should be noted that the terms "up, down, in and out" and the like indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the devices or elements referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as limiting the present application. In addition, the terms "first, second or third" are only for descriptive purposes, and cannot be understood as indicating or implying relative importance.

[0068] Unless otherwise defined, the terms "mounting, connecting, associating" in the present application should be interpreted broadly, for example: it can be fixed connection, detachable connection or integral connection; it can also be mechanical connection, electrical connection or direct connection, it can also be indirectly connected through intermediate medium, or it can be the internal communication of two elements. For those skilled in the art, the specific meaning of the above terms in the present application can be understood according to the specific circumstances.

[0069] As shown in Figs. 1-3 , which shows a specific embodiment of the present application:

[0070] Embodiment one:

[0071] A power transmission conductor sag measurement method based on a binocular vision measurement device, comprising the following steps:

[0072] S1: Fix the binocular vision measurement device on one side of one of the power transmission towers whose conductor sag is to be measured.

[0073] S2: Use the binocular vision measurement device to collect left and right views and perform internal parameter distortion correction and stereo correction.

[0074] S3: Preprocess the left and right views collected in S2, and perform stereo matching to obtain a parallax map, and calculate a horizontal span .

[0075] S4: Zero offset calibration and filtering are performed on the original output of the inclination sensor on the binocular measurement device, and the tangent angles of the conductor at both ends are obtained , .

[0076] S5: The horizontal span calculated in S3 and the tangent angle , calculated in S4 are input into the catenary model, and the catenary constant , the lowest point , the end point height , and the sag are calculated.

[0077] In this embodiment, as shown in Fig. 1 , Fig. 2 , the measurement method for measuring the sag of the power transmission conductor relies on the binocular vision measurement device, which accurately obtains the span and angle parameters in a non-contact manner, and combines the catenary equation to realize high-precision calculation of the sag. Overcome the terrain dependence, power outage requirement and precision defects of the prior art, and can quickly and non-contactly complete the conductor sag detection and inspection on site, with high measurement precision, strong robustness and strong implementability.

[0078] Embodiment two:

[0079] Specifically, the binocular vision measurement device comprises a triangular support frame 1, an integrated support 2, a first monocular camera 3, a second monocular camera 4, and an inclination sensor 5. The integrated support 2 is fixedly installed above the triangular support frame 1. An adjustable slide rail 21 is arranged at a middle position of the integrated support 2. The first monocular camera 3 and the second monocular camera 4 are installed in the slide rail 21. The inclination sensor 5 is fixedly installed on the integrated support 2 at the middle position.

[0080] In the embodiment, as shown in Fig. 2 , Fig. 3 In actual use, the binocular vision measurement device is fixed by the triangular support frame 1, and the stability is ensured, thereby providing a basis for obtaining accurate measurement results. The adjustable slide rail 21 arranged at the middle position of the integrated support 2 can facilitate adjustment of the distance between the first monocular camera 3 and the second monocular camera 4, that is, adjustment of the length of the baseline, thereby being applicable to measurement of voltage conductors of different levels. The inclination sensor 5 is used for measuring a tangential angle.

[0081] In actual use, taking the use of 8 million pixel cameras for the first monocular camera 3 and the second monocular camera 4 as an example (the camera photosensitive chip model is IMX317), the CMOS size is 1 / 2.5 inch, the focal length is 35 mm, and the resolution is 3840*2160. In this use scenario, the baseline length of the first monocular camera 3 and the second monocular camera 4 can be selected as follows:

[0082] The baseline of 150 mm can be selected for a 35kv voltage level.

[0083] The baseline of 300 mm can be selected for a 110kv voltage level.

[0084] The baseline of 500 mm can be selected for a 220kv voltage level.

[0085] Specifically, the slide rail 21 is provided with a scale line.

[0086] In the embodiment, the scale line is arranged on the slide rail 21. The installation positions of the first monocular camera 3 and the second monocular camera 4 can be adjusted according to the scale line, so as to change the baseline between the cameras to obtain a required parallax range. The baseline length can be clearly and intuitively recorded in real time through the scale line on the slide rail 21, and the required information can be quickly and conveniently obtained.

[0087] Specifically, the inclination sensor 5 adopts a three-axis MEMS inertial measurement unit, which contains an accelerometer and a gyroscope located inside the inclination sensor, and outputs the pitch angle at the wire suspension point in real time. The inclination sensor 5 is fixed by magnetic attraction or screw, and the installation height is consistent with the optical axis height of the first monocular camera 3 and the second monocular camera 4. The measurement angle is in the same horizontal plane as the first monocular camera 3 and the second monocular camera 4.

[0088] In this embodiment, the three-axis MEMS inertial measurement unit of the inclination sensor 5 is a way to achieve high precision, high dynamic, and high stability angle measurement in complex working conditions in dynamic inclination measurement. The accelerometer can accurately calculate the pitch angle and roll angle by measuring the components of the gravitational acceleration on the three axes in a static or low-speed state, and has good long-term stability and no cumulative error. The gyroscope can sensitively track the change of angle by measuring angular velocity. Since it is not sensitive to linear motion and vibration, it performs well in dynamic environment.

[0089] In actual use, the inclination sensor 5 combines the accelerometer and the gyroscope to effectively suppress vibration interference. Through information fusion, high-precision and high-stability angle measurement can be achieved in real and complex dynamic environments. It has the absolute reference and non-drift characteristics of the accelerometer, and the dynamic response and anti-vibration ability of the gyroscope, which can obtain accurate measurement results in the industrial application scene of wire inclination monitoring.

[0090] Embodiment three:

[0091] In this embodiment, in order to ensure the accuracy and robustness of binocular distance measurement, systematic calibration of the binocular vision system is required. The calibration content includes: camera intrinsic parameter calibration and distortion correction, binocular extrinsic parameter (rotation matrix R and translation matrix T) calculation, stereo correction, accurate determination of baseline B, calibration accuracy evaluation and error propagation analysis.

[0092] Specifically, the specific process of the intrinsic parameter distortion correction and stereo correction in S2 is as follows:

[0093] S21: Select a chessboard of appropriate size, collect 100 chessboard images of different poses, and cover the center and edge of the field of view.

[0094] S22: Single target calibration, solve the intrinsic parameters of the left and right cameras respectively and correct the distortion.

[0095] In this embodiment, the single target calibration in S22 is completed by Zhang Zhengyou chessboard, and the specific process is as follows:

[0096] Let the i-th chessboard corner point in the world coordinate system be ​ , the extrinsic parameters of the frame and , the intrinsic matrix of the camera is expressed as:

[0097] .

[0098] where are the normalized focal lengths of the camera in and axes, usually in pixels; , are the principal point (optical center) of the camera in and coordinates in the image coordinate system (usually in pixels); is the skew coefficient of the image between and axes, usually 0 in modern cameras.

[0099] The radial and tangential distortion of the camera is expressed by the following model (taking the first 2 orders of radial and 2 orders of tangential):

[0100] .

[0101] .

[0102] where is the normalized image plane coordinate, , are the distorted image coordinates normalized image plane coordinate, , is the radial distortion coefficient, is the tangential distortion coefficient.

[0103] The accuracy of the calibration is determined by minimizing the re-projection error:

[0104] .

[0105] where is the re-projection error, is the pixel coordinate of the th corner point observed in the th frame image, is the camera projection function that projects the world coordinate to the image plane through the camera intrinsic and extrinsic parameters , is the translation vector corresponding to the th frame.​

[0106] S23: Solve the rotation matrix R and translation matrix T using stereo calibration, and perform stereo correction.

[0107] In this embodiment, when the stereo calibration method is used in step S23 to solve for the rotation matrix R and translation matrix T between the left and right cameras, the extrinsic parameters satisfy:

[0108] .

[0109] in, Let be the three-dimensional coordinates of a point in space in the left camera coordinate system. These are the three-dimensional coordinates of the same point in the right camera coordinate system.

[0110] Baseline Accuracy is crucial to precision. In this embodiment, a dual calibration strategy is employed, using a scale or threaded mechanism on the slide rail to precisely measure the initial baseline value. After obtaining the actual baseline distance, observe the translation matrix in the calibration parameters. The parameters are compared, and the differences between the two are determined.

[0111] The original left and right images are distorted using calibration parameters to obtain a distortion-free image. Baseline alignment is then performed, constraining corresponding points to the same scan line to obtain a corrected image for disparity calculation. Before disparity calculation, it is recommended to perform grayscale equalization, histogram normalization, and small-range Gaussian filtering to suppress noise and improve feature stability.

[0112] S24: High precision requirements are achieved by screening based on the average reprojection error.

[0113] In this embodiment, the mean reprojection error (MRE) in S24 is used as a calibration quality index, and its calculation formula is as follows:

[0114] .

[0115] in, For actual corner pixels, For reprojected pixels.

[0116] In practical applications, industrial-grade calibration typically requires an MRE of <0.5 pixels. After initial calibration, all intrinsic and extrinsic parameters and 3D points are jointly optimized to further minimize the global reprojection error and improve system stability and robustness.

[0117] The basic relationship between parallax and depth is:

[0118] ,

[0119] in, Let the depth from the target point to the camera's optical center be denoted as . The focal length in pixels of the camera. As the baseline, For parallax.

[0120] parallax measurement error By performing error propagation estimation, an approximate depth error is obtained:

[0121] ,

[0122] in, This represents the depth measurement error.

[0123] Therefore, under conditions of large distance (small parallax), Follow To ensure depth accuracy, engineering measures can be taken to increase the baseline B (adjusted via a slide rail) and select a larger focal length. The lens, or by reducing subpixel parallax estimation and multi-view fusion Furthermore, if baseline B itself has measurement errors... Then the contribution to Z is:

[0124] ,

[0125] in Given the uncertainty of focal length, the above formula can be used to select appropriate B, f, and target parallax range.

[0126] Example 4:

[0127] In this embodiment, the goal of stereo matching is to find the pixel correspondence of the same spatial point in two calibrated images. To achieve high-precision distance L measurement, a semi-global stereo matching algorithm is used to estimate disparity in the calibrated images during stereo reconstruction.

[0128] The stereo matching algorithm (SGBM) strikes a balance between minimizing global energy and aggregating local costs, offering good accuracy and real-time performance, making it suitable for outdoor conductor scenarios.

[0129] Specifically, the process of stereo matching and disparity map calculation in S3 is as follows:

[0130] For left-right corrected images, let the left image have pixels. The corresponding right image pixel is ,in For parallax, parallax map The estimation is achieved by minimizing the pixel cost energy. An approximate implementation, its calculation formula is as follows:

[0131] ,

[0132] wherein, represents the pixel position ,

[0133] is the disparity of pixel ;

[0134] is the pixel matching cost, usually based on the absolute disparity;

[0135] is the neighborhood of pixel ;

[0136] is the smoothness term, with a two-level penalty, expressed as:

[0137] ,

[0138] wherein, encourages small-amplitude disparity jumps, suppresses large-amplitude jumps to maintain smoothness.

[0139] In the present embodiment, the key idea of SGBM is to accumulate the matching cost of each pixel in several directions (usually 8 or 16 paths) to obtain the path-aggregated cost and then sum all directions:

[0140] ,

[0141] Finally, the disparity corresponding to the minimum accumulated cost is selected at each pixel:

[0142] ,

[0143] The path-aggregated cost of each direction is calculated in a dynamic programming recursive form:

[0144] ,

[0145] wherein represents the previous pixel position on the current path, and the last term is a normalization for numerical stability.

[0146] In the present embodiment, a disparity confidence threshold can also be set, and when the disparity confidence is lower than the preset threshold, a degradation strategy or reacquisition is triggered.

[0147] Embodiment Five:

[0148] Specifically, the specific process of the wire tangential angle , in S4 is:

[0149] S41: define the three-axis acceleration values of the tilt sensor as , in the static state, satisfy:

[0150] ,

[0151] wherein, is the acceleration of gravity.

[0152] S42: the tangent angle of the guide wire , The calculation formula is:

[0153] .

[0154] In this embodiment, only the pitch angle is used to calculate the height difference, and when the guide wire has a lateral tilt, the roll angle can be calculated, and the calculation formula of the roll angle is:

[0155] .

[0156] S43: filter the output tangent angle of the guide wire in S42, and the calculation formula is as follows:

[0157] ,

[0158] wherein, is the angular velocity integral term, is the tilt angle value obtained from the acceleration, is a fusion coefficient.

[0159] In this embodiment, in order to reduce the influence of environmental vibration, wind disturbance and electromagnetic interference, the angle data output by the tilt sensor is subjected to first-order complementary filtering and time sliding average processing, and in general cases, The value range of is: 0.96~0.98, after filtering processing, the angle measurement noise can be reduced to ±0.05°, and the measurement accuracy is improved.

[0160] Specifically, the zero offset calibration and filtering of the original output of the tilt sensor (5) on the binocular measuring device in S4 are specifically complementary filtering and sliding average processing, the proportional coefficient of the complementary filter is , and the sliding average window of the sliding average processing is .

[0161] In this embodiment, complementary filtering and sliding average processing are used to suppress wind vibration and high-frequency noise, and to ensure the accuracy of the tilt sensor output data, so as to obtain accurate calculation results.

[0162] Embodiment six:

[0163] Specifically, the formula for calculating the catenary constant in S5 is:

[0164]

[0165] The formula for calculating the lowest point is:

[0166]

[0167] The formula for calculating the end point height is:

[0168]

[0169] The formula for calculating the sag is:

[0170]

[0171] wherein, are the hyperbolic sine and hyperbolic cosine functions, respectively.

[0172] In this embodiment, the sag is measured using the angle method. The inclination sensor 5 is fixedly connected to the first monocular camera 3 and the second monocular camera 4 through the integrated support 2, so that the pitch measurement axis of the inclination sensor 5 is coplanar and consistent with the camera optical axis. At this time, the pitch downward is positive. The height from the camera optical axis center to the local reference is recorded . The calibration can be repeatedly used once. If the integrated support or the inclination sensor 5 is replaced, it needs to be measured again. The horizontal span is obtained by stereo matching and triangulation of the binocular camera.

[0173] In the actual measurement process, the angle is measured as follows: the device is arranged near the center stake of the No. 1 pole, so that the camera optical axis is tangent to the guide line. First, the guide line tangent at the left end of the No. 1 pole suspension point is aligned, and is read. Second, the direction in which the optical axis is still tangent to the guide line and intersects with the vertical line of the No. 2 pole tower is adjusted, and is read.

[0174] The formula for calculating the vertical distance a of the No. 1 pole suspension point to the horizontal plane of the camera optical axis is:

[0175]

[0176] wherein, is the height of the pole tower; ​​​​​​​​​​​​For the length of the insulator string, For the height of the camera optical axis center to the local reference.

[0177] The slope difference of the two tangential visual lines is converted to The calculation formula is:

[0178] ,

[0179] To ensure geometric meaning, it is required that If negative values appear, it is usually a sign of non-uniform agreement, and the sign should be re-read or corrected.

[0180] For the accuracy of measurement, preprocessing and standardization can be performed first:

[0181] First, the consistency check and optimization processing of the input parameters, on the one hand, through the Kalman filter algorithm to the angle data , De-noising, eliminating abnormal values caused by instantaneous pulse interference and environmental vibration, ensuring that the angle measurement error is controlled within ±0.05°; On the other hand, the pixel coordinate conversion result of the span is unified to the actual distance unit of meters, and the parameter integrity is checked synchronously (if there is data missing, the automatic sampling mechanism of the module is triggered). At the same time, the target line design parameter library is called, the corresponding cross-sectional area , elastic modulus and other basic parameters of the conductor type are obtained, combined with the real-time temperature collected by the on-site weather sensor, the conductor specific load (load per unit length per unit cross section) is corrected, and the correction formula is:

[0182] .

[0183] Catenary equation selection and parameter substitution: Considering the actual stress characteristics of the transmission conductor under the action of its own weight, the traditional parabolic model (only applicable to small span scenarios with sag and span ratio <1 / 10) is abandoned, and the catenary equation that meets the specification requirements is used for calculation. The equation is based on the core assumption that the conductor load is uniformly distributed along the line length, and the parabolic model with high calculation precision is substituted into the optimized catenary sag calculation formula:

[0184] .

[0185] Among them, , are the hyperbolic sine and hyperbolic cosine functions, respectively.

[0186] The following parabolic approximation formula can also be used to calculate the sag (the maximum vertical distance from the conductor to the lowest point in the span): ​

[0187] .

[0188] Specifically, the sag of the arc The calculation sets a threshold value ,

[0189] When , the calculation formula of the sag of the arc is:

[0190] ,

[0191] When , the calculation formula of the sag of the arc is:

[0192] .

[0193] In this embodiment, the calculation formula of the sag when the setting is less than the threshold value is because the angle difference between the and the is small at this time, and the value of the at this time can be selected to solve the average value of the and the to simplify the calculation process.

[0194] The preferred embodiments of the application are described in detail above in combination with the drawings, but the application is not limited to the above-mentioned embodiments, and various changes can be made within the knowledge possessed by those skilled in the art without departing from the purpose of the application.

[0195] Many other changes and modifications can be made without departing from the spirit and scope of the application. It should be understood that the application is not limited to a particular embodiment, and the scope of the application is defined by the appended claims.​​

Claims

1. A method for measuring the sag of power transmission lines based on a binocular vision measurement device, characterized in that, Includes the following steps: S1: Fix the binocular vision measurement device to one side of one of the transmission towers of the conductor sag to be measured; S2: Use a binocular vision measurement device to acquire left and right views and perform intrinsic parameter distortion correction and stereo correction; S3: Preprocess the left and right views acquired in S2, perform stereo matching to obtain a disparity map, and calculate the horizontal span. ; S4: Perform zero-bias calibration and filtering on the raw output of the tilt sensor on the binocular measuring device to obtain the tangential angles of the two end wires. , ; S5: Calculate the horizontal clearance in S3. The tangential angle calculated in S4 , The input is fed into the catenary model to calculate the catenary constant. Lowest point Endpoint height sag .

2. The method for measuring the sag of power transmission lines based on a binocular vision measurement device according to claim 1, characterized in that, The binocular vision measurement device includes a triangular support frame (1), an integrated bracket (2), a first monocular camera (3), a second monocular camera (4), and a tilt sensor (5). The integrated bracket (2) is fixedly installed above the triangular support frame (1). An adjustable slide rail (21) is provided in the middle of the integrated bracket (2). The first monocular camera (3) and the second monocular camera (4) are installed in the slide rail (21). The tilt sensor (5) is fixedly installed on the integrated bracket (2) and located in the middle position.

3. The method for measuring the sag of power transmission lines based on a binocular vision measurement device according to claim 1, characterized in that, The specific processes for intrinsic parameter distortion correction and stereo correction in S2 are as follows: S21: Select a chessboard of appropriate size and collect 100 images of the chessboard in different poses, covering the center and edge of the field of view; S22: Perform single-target positioning, solve the intrinsic parameters of the left and right cameras respectively, and remove distortion; S23: Solve the rotation matrix R and translation matrix T using stereo calibration, and perform stereo correction; S24: High precision requirements are achieved by screening based on the average reprojection error.

4. The method for measuring the sag of power transmission lines based on a binocular vision measurement device according to claim 1, characterized in that, The specific process of stereo matching and disparity map calculation in S3 is as follows: Let the left image have pixels The corresponding right image pixel is ,in For parallax, parallax map The estimation is achieved by minimizing the pixel cost energy. An approximate implementation, its calculation formula is as follows: ， in, Indicates pixel position ; For pixels Parallax; The cost of pixel matching is typically based on absolute aberration; For pixels The neighborhood; This is a smoothing term.

5. The method for measuring the sag of power transmission lines based on a binocular vision measurement device according to claim 1, characterized in that, The conductor tangential angle in S4 , The specific process is as follows: S41: Define the triaxial acceleration values ​​of the tilt sensor as follows: In a static state, the following conditions are met: ， in, It is the acceleration due to gravity; S42: Wire tangential angle , The calculation formula is: ； S43: Filter the tangential angle of the output conductor in S42. The calculation formula is as follows: ， in, This is the integral term of angular velocity. The tilt angle value is obtained from acceleration. This is the fusion coefficient.

6. The method for measuring the sag of a power transmission line based on a binocular vision measurement device according to claim 1, characterized in that, The catenary constant in S5 The calculation formula is: ， Lowest point The calculation formula is: ， Endpoint height The calculation formula is: ， Sag The calculation formula is: ， in, , These are hyperbolic sine and hyperbolic cosine functions, respectively.

7. The method for measuring the sag of a power transmission line based on a binocular vision measurement device according to claim 1, characterized in that, In step S4, the zero-bias calibration and filtering of the raw output of the tilt sensor (5) on the binocular measuring device specifically involves performing complementary filtering and moving average processing on it. The scaling factor of the complementary filter is... The moving average window for the moving average processing is... .

8. The method for measuring the sag of a power transmission line based on a binocular vision measurement device according to claim 2, characterized in that, The tilt sensor (5) adopts a three-axis MEMS inertial measurement unit, which includes an accelerometer and a gyroscope. The accelerometer and gyroscope are located inside the tilt sensor and output the pitch angle at the suspension point of the wire in real time. The tilt sensor (5) is fixed by a magnetic base or thread. The installation height is consistent with the optical axis height of the first monocular camera (3) and the second monocular camera (4). Its measurement angle is located on the same horizontal plane as the first monocular camera (3) and the second monocular camera (4).

9. The method for measuring the sag of a power transmission line based on a binocular vision measurement device according to claim 2, characterized in that, The slide rail (21) is provided with scale lines.

10. A method for measuring the sag of a power transmission line based on a binocular vision measurement device according to claim 6, characterized in that, The sag Calculation setting threshold , when At that time, sag The calculation formula is: ， when At that time, sag The calculation formula is: 。

Images