Transmission tower deformation assessment method based on visual identification

Abstract

The invention provides a transmission tower deformation assessment method based on visual identification, and relates to the technical field of visual identification, and the method comprises the steps: obtaining tower building information through interaction, and obtaining a transmission tower model through restoration; executing structure deviation visual identification to obtain a tower deformation information set; connecting deviation visual identification is carried out on the positioning connecting nodes to obtain a risk connecting node set; presetting a flaw detection point array, and performing ultrasonic flaw detection to obtain an ultrasonic detection data set; performing internal defect identification to obtain an internal defect information set; fitting the internal defect information set, the risk connection node set and the tower deformation information set to a transmission tower model to obtain a tower defect analysis model; and carrying out deformation risk quantitative evaluation to obtain a deformation risk coefficient. According to the method and the device, the technical problems of omission or misjudgment of complex deformation caused by the fact that single appearance detection is generally adopted for deformation evaluation of the power transmission tower in the prior art, and relatively low evaluation accuracy and efficiency caused by relatively much manual intervention are solved.

Description

Technical Field

[0001] The present invention relates to the technical field of visual recognition, and in particular to a method for evaluating the deformation of a transmission tower based on visual recognition. Background Art

[0002] Transmission towers are crucial structural components in power systems. Their structural safety is directly related to the reliability and stability of power transmission. As time goes by and environmental conditions change, transmission towers may deform, their connection points may be damaged, or they may have internal defects. These problems lead to reduced structural safety. Existing technologies usually use a single appearance inspection method for transmission tower deformation assessment. On the one hand, existing technologies are difficult to effectively process and accurately analyze the large amount of data generated by the detection equipment, and lack unified deformation and defect assessment standards, resulting in poor accuracy of deformation assessment. On the other hand, a single inspection may lead to omissions or misjudgments of complex deformations, which still requires manual intervention when dealing with complex deformations, resulting in low automation levels and detection efficiency. Summary of the invention

[0003] The present application provides a transmission tower deformation assessment method based on visual recognition, aiming to solve the technical problems that the transmission tower deformation assessment in the prior art usually adopts a single appearance detection, resulting in omission or misjudgment of complex deformations, and there is a lot of manual intervention, resulting in low assessment accuracy and efficiency.

[0004] The present application discloses a method for evaluating deformation of a transmission tower based on visual recognition, the method comprising: interactively obtaining tower building information of a target transmission tower, and restoring a transmission tower model based on the tower building information; performing visual recognition of structural deviations on the target transmission tower based on the transmission tower model to obtain a tower deformation information set; locating connection nodes on the target transmission tower according to the tower building information, performing visual recognition of connection deviations on the located connection node set to obtain a risk connection node set; presetting a flaw detection point array, and performing ultrasonic flaw detection on the target transmission tower based on the flaw detection point array to obtain an ultrasonic detection data set; performing internal defect identification on the target transmission tower based on the ultrasonic detection data set to obtain an internal defect information set; fitting the internal defect information set, the risk connection node set and the tower deformation information set to the transmission tower model to obtain a tower defect analysis model; performing quantitative deformation risk evaluation on the target transmission tower based on the tower defect analysis model to obtain a deformation risk coefficient.

[0005] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0006] By interactively acquiring tower building information and restoring the transmission tower model, the digital model of the target transmission tower can be accurately constructed, providing a basis for subsequent visual recognition and analysis; by performing structural deviation visual recognition on the target tower based on the transmission tower model, the deformation of the tower can be fully detected, and a detailed deformation information set can be obtained, so as to accurately identify the deviation of the tower structure and timely discover potential deformation problems; by locating the connection nodes and performing connection deviation visual recognition, the risky connection node set can be accurately identified, which enables priority attention to those connection points with potential risks during maintenance and inspection, thereby improving the efficiency and accuracy of maintenance; the preset detection point array and ultrasonic flaw detection can deeply detect defects inside the transmission tower. This non-destructive testing method ensures accurate identification of internal defects and will not cause damage to the tower. Further damage is detected and detailed internal defect data is provided at the same time; internal defect identification is performed based on the ultrasonic detection data set to obtain the internal defect information set, so as to achieve a comprehensive grasp of the internal structural defects and provide a reliable basis for subsequent risk assessment and maintenance decisions; the internal defect information set, risk connection node set and tower deformation information set are fitted to the transmission tower model to obtain the tower defect analysis model. This comprehensive analysis model can integrate information from different sources, provide a comprehensive tower defect and risk assessment, and ensure the comprehensiveness and accuracy of the analysis results; based on the tower defect analysis model, the deformation risk of the target transmission tower is quantitatively assessed to obtain the deformation risk coefficient, to quantify various test results and analysis data, and to comprehensively judge the overall risk level of the tower, which provides a scientific basis for maintenance and management decisions and reduces operational risks.

[0007] 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

[0008] Figure 1 A schematic flow chart of a method for evaluating deformation of a transmission tower based on visual recognition is provided for an embodiment of the present application;

[0009] Figure 2 A schematic diagram of a process for obtaining a tower deformation information set in a transmission tower deformation assessment method based on visual recognition is provided for an embodiment of the present application. DETAILED DESCRIPTION

[0010] The embodiment of the present application provides a method for evaluating the deformation of a transmission tower based on visual recognition, thereby solving the technical problems that the deformation evaluation of a transmission tower in the prior art usually adopts a single appearance detection, resulting in omission or misjudgment of complex deformations, and there is a lot of manual intervention, resulting in low evaluation accuracy and efficiency.

[0011] 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.

[0012] like Figure 1 As shown, an embodiment of the present application provides a method for evaluating the deformation of a transmission tower based on visual recognition, the method comprising:

[0013] The tower building information of the target transmission tower is interactively obtained, and the transmission tower model is restored based on the tower building information.

[0014] The tower construction information of the target transmission tower is collected through data sources such as design drawings and construction records, including the tower's geometric structure, material properties, geographical location, environmental conditions, etc. The collected tower construction information is cleaned and preprocessed to ensure the standardization of the data. The key parameters of the tower are extracted based on the preprocessed construction information, and three-dimensional modeling tools are used for modeling to obtain the transmission tower model.

[0015] Based on the transmission tower model, visual recognition of structural deviation is performed on the target transmission tower to obtain a tower deformation information set.

[0016] According to the preset shooting point array, the target transmission tower is imaged at various angles to ensure that the key structures of the transmission tower are covered and the actual image is obtained. At the same point, the transmission tower model is imaged to obtain a virtual image. The virtual image is consistent with the actual image in terms of shooting point, viewing angle, etc. The actual image and the virtual image are spatially aligned and image registered, and deviation detection is performed on the aligned images to identify the difference areas between the actual image and the virtual image. These difference areas represent deformation or offset on the tower structure.

[0017] The identified multiple difference areas are analyzed and the local deformation data of each area are extracted. These data include but are not limited to the size, direction, and spatial position of the deformation area. Multiple local deformation data are aggregated to construct a complete tower deformation information set. This information set can reflect the structural deviation of the target transmission tower in the actual environment and provide data support for subsequent risk assessment.

[0018] The connection nodes are located on the target transmission tower according to the tower building information, and the connection deviation visual recognition is performed on the located connection node set to obtain the risk connection node set.

[0019] According to the tower building information, all key connection nodes on the target transmission tower are identified to obtain a set of connection nodes. These connection nodes are important supporting points of the tower structure, including the connection between the tower frame and the beam, the connection between the support rod and the main rod, etc.

[0020] According to the node positions in the connection node set, the flight path of the drone is planned. When planning the path, the visibility and acquisition angle of each connection node, as well as the flight safety of the drone and the performance of the image acquisition equipment are considered. On the basis of the planned path, the specific image acquisition path is determined to ensure that all connection nodes can be covered to achieve the best image coverage. The drone is controlled to collect images according to the image acquisition path to obtain images of the connection nodes.

[0021] Multiple sample connection deviation image sets of multiple sample connection nodes are obtained to train a deviation recognition model that can identify connection deviations. The collected connection node images are input into the model, and the model automatically identifies the connection nodes with deviations and locates the specific positions of these nodes on the tower to form a risky connection node set.

[0022] A flaw detection point array is preset, and ultrasonic flaw detection is performed on the target transmission tower based on the flaw detection point array to obtain an ultrasonic detection data set.

[0023] According to the structural characteristics and material properties of the tower, as well as the tower construction information obtained in the early stage, the distribution of the detection points is determined and an array of detection points is generated. The detection points are usually set at key stress-bearing parts and areas prone to defects, such as welds, connection nodes, support rods, etc.

[0024] Select appropriate ultrasonic flaw detection instruments and configure a variety of probe types to meet different detection needs. Place the probe at each preset flaw detection point, start the flaw detection instrument, and the ultrasonic probe sends ultrasonic pulses to the inside of the tower. When the ultrasonic wave encounters materials or defects of different densities during propagation, an echo will be generated. The probe receives the echo signal and converts it into an electrical signal and transmits it to the flaw detection instrument for processing. An ultrasonic waveform diagram, time domain diagram, spectrum diagram, etc. are generated and converted into specific detection parameters. The flaw detection points are scanned one by one, and the ultrasonic echo data of each point is recorded to obtain an ultrasonic detection data set.

[0025] Internal defects of the target transmission tower are identified based on the ultrasonic detection data set to obtain an internal defect information set.

[0026] Based on the ultrasonic detection data set, characteristic parameters related to internal defects are extracted, such as abnormal changes in echo amplitude, waveform distortion, wave velocity slowdown, changes in spectral characteristics, etc. The internal defects of the target transmission tower are identified according to the characteristic parameters. For example, the threshold range of the characteristic parameters is set, and signals exceeding or below the threshold are judged as possible defects. Combined with the analysis of the time domain and frequency domain characteristics of the ultrasonic signal, specific defect signal patterns are identified, such as cracks, holes, inclusions, desoldering, etc. The ultrasonic signal of each detection point is analyzed one by one, the existing internal defects are identified, and the specific location, size, depth and other characteristics of the defects are determined, and the identified defect information is integrated into an internal defect information set.

[0027] The internal defect information set, the risk connection node set and the tower deformation information set are fitted to the transmission tower model to obtain a tower defect analysis model.

[0028] Based on a unified coordinate system, internal defects, risky connection nodes and deformation information are aligned to the same spatial frame so that they are all represented in the same coordinate system for accurate positioning in the transmission tower model. The data after coordinate alignment is input into the transmission tower model and fitted into the structural framework of the model so that each internal defect, risky connection node and deformation information can be accurately reflected in the model. Based on the fitted data, a complete tower defect analysis model including internal defects, risky connection nodes and deformation information is generated.

[0029] A deformation risk quantitative assessment is performed on the target transmission tower based on the tower defect analysis model to obtain a deformation risk coefficient.

[0030] Data information related to deformation is extracted from the tower defect analysis model, including the size of the deformation variable, the deformation position, the deformation direction, and other influencing factors related to the deformation. Based on the deformation data information, the connection risk discrete coefficient, the internal defect risk coefficient, the local deformation risk coefficient, and the deformation risk discrete coefficient are analyzed and obtained. The above coefficients are integrated and calculated based on the tower deformation evaluation function to obtain the deformation risk coefficient. The obtained deformation risk coefficient provides a quantitative risk assessment basis for maintenance personnel, which can more scientifically evaluate the impact of tower deformation on structural safety and formulate corresponding management and maintenance strategies.

[0031] Furthermore, if Figure 2 As shown, based on the transmission tower model, visual recognition of structural deviation is performed on the target transmission tower to obtain a tower deformation information set. The method further includes:

[0032] A shooting point array is preset, and images of the target transmission tower are collected based on the shooting point array to obtain a tower entity image array; the shooting point array is fitted to the transmission tower model, and image collection and replication are performed on the transmission tower model to obtain a tower virtual image array; image deviation visual recognition is performed on the tower entity image array and the tower virtual image array according to the shooting point array to obtain the tower deformation information set.

[0033] According to the structural characteristics of the transmission tower, multiple shooting points are preset so that these points can cover all key parts of the transmission tower, including the tower base, tower body, tower head, connection nodes, etc., and the shooting point array is obtained by integration. According to the height and complexity of the tower, suitable shooting equipment is selected, such as a high-resolution camera, etc. The shooting equipment is controlled to collect images of the target transmission tower according to the preset shooting point array to ensure that the image can fully reflect the actual structure of the tower. The collected images are sorted according to the shooting points to generate a complete tower entity image array.

[0034] The shooting point array is fitted to the transmission tower model to ensure that the spatial position and angle of the shooting points are consistent with the actual acquisition situation. Based on the fitted shooting point array, the actual image acquisition process is simulated on the transmission tower model to generate a tower virtual image array corresponding to the tower physical image array.

[0035] According to the preset shooting point array, the corresponding points in the physical image array and the virtual image array are mapped. Each shooting point has a pair of corresponding image pairs. The physical image and the virtual image of the same shooting point are spatially aligned. The image deviation analysis method is used, such as calculating the pixel difference or shape difference of the image, to identify the deviation between the physical image and the virtual image. According to the deviation identification result, the deformation areas existing in the physical image are identified. These deformation areas are the parts that show significant differences in image comparison. The multiple deformation areas identified are refined, the specific deformation data are determined, and the tower deformation information set is obtained by integration.

[0036] Furthermore, according to the shooting point array, the pole tower physical image array and the pole tower virtual image array are mapped to perform image deviation visual recognition to obtain the pole tower deformation information set, and the method further includes:

[0037] Color feature extraction is performed based on the pole tower physical image array to obtain the pole tower color feature; image contrast optimization is performed on the pole tower physical image array with the pole tower color feature as a constraint to obtain an optimized physical image array; a first pole tower physical image and a first pole tower virtual image mapped to a first shooting point are extracted from the optimized physical image array and the pole tower virtual image array, wherein the first shooting point is any shooting point in the shooting point array; the first pole tower physical image and the first pole tower virtual image are spatially aligned to obtain an image registration result; deformation area identification is performed based on the image registration result to locate the first deformation area; and so on, the optimized physical image array and the pole tower virtual image array are traversed to identify the deformation area to obtain multiple deformation areas; the pole tower deformation is numerically quantified for the target transmission tower according to the multiple deformation areas to obtain the pole tower deformation information set, wherein the pole tower deformation information set includes multiple local deformation data of the multiple deformation areas.

[0038] Extract the color histogram of the pole tower entity image, and count the distribution ratio of each color in the image, including hue distribution, saturation distribution, etc., to form global color features. According to important areas in the image, such as key structural parts of the pole tower, extract local color features. The image can be divided into multiple regions using a regional segmentation algorithm, and color features are extracted for each region. The extracted global color features and local color features are integrated into pole tower color features for subsequent image processing.

[0039] Taking the color characteristics of the tower as the constraint condition, the contrast optimization goal is set, including enhancing the visual clarity of the key parts of the tower while maintaining the color consistency of the image, etc., to optimize the image contrast. Specifically, for the global image, the image equalization technology is applied to improve the image contrast, making the details of the dark and bright parts of the image more obvious. This method enhances the overall contrast of the image while maintaining the color characteristics; for the local image, the adaptive histogram equalization is applied to make the contrast optimization effect under different brightness conditions more uniform, avoiding the increase of image noise caused by excessive enhancement. After completing the contrast optimization, the optimized entity image array is obtained.

[0040] Any shooting point is selected from the shooting point array as the first shooting point as the object of analysis, the first tower physical image corresponding to the first shooting point is extracted from the optimized physical image array, and the first tower virtual image corresponding to the first shooting point is extracted from the tower virtual image array. The first tower physical image and the first tower virtual image are both from the same shooting point, namely, the first shooting point.

[0041] Spatial alignment is to align two images, namely the physical image of the first tower and the virtual image of the first tower, in the spatial dimension so that the corresponding structures in the two images are aligned at the same position, thereby laying the foundation for subsequent image registration. Specifically, key feature points are extracted from the physical image of the first tower and the virtual image of the first tower. These feature points can be corner points, edge points or other significant structures in the image. The feature points in the two images are matched using a feature matching algorithm to find the corresponding feature point pairs. According to the feature point matching results, a spatial transformation is performed between the two images so that the two are aligned in space to obtain the image registration result.

[0042] The deformation area refers to the area where there are significant differences between the physical image and the virtual image of the transmission tower after alignment. These differences may indicate that the tower has undergone physical deformation or structural damage. The first tower physical image and the first tower virtual image that have been spatially aligned are subjected to pixel-level or feature-level difference detection. The difference detection method includes calculating the difference of the images, using edge detection algorithms, etc. According to the detected difference area, the position and range of the first deformation area are determined and marked on the image. When the difference detection result shows that the two images are the same and there is no difference area, it means that there is no deformation area, and the pair of images is skipped.

[0043] The physical image array and the tower virtual image array are traversed and optimized, and the same deformation area recognition process is performed on each pair of images. During the recognition process, if the two images are completely consistent, that is, there is no difference, then the pair of images is skipped; if there is a difference, then the deformation area is continued to be marked. After the traversal is completed, the deformation areas identified each time are summarized to obtain multiple deformation areas, which represent the deformation of the entire transmission tower at different positions and different perspectives.

[0044] The numerical quantification of tower deformation refers to converting the degree and characteristics of deformation into specific numerical data by numerically analyzing the multiple deformation areas identified. Specifically, each deformation area is analyzed to extract characteristic data including deformation displacement distance, deformation angle, deformation area, etc., calculate the displacement distance of the deformation area, that is, the offset between the deformation point in the area and the normal structure, quantify the deformation direction of the deformation area, determine the directionality of the deformation, measure the area of ​​the deformation area to indicate the range of the deformation, and integrate the quantitative data of each deformation area to form a deformation information set containing multiple local deformation data.

[0045] Furthermore, numerically quantifying the deformation of the target transmission tower according to the multiple deformation areas to obtain the tower deformation information set, the method further includes:

[0046] Guided by the multiple deformation areas, laser point cloud data collection is performed on the target transmission tower to obtain multiple local point cloud data; geometric modeling is performed based on the multiple local point cloud data to obtain multiple local solid models; guided by the multiple deformation areas, the multiple local solid models are fitted to the transmission tower model to perform numerical quantification of tower deformation to obtain multiple local deformation data, and the multiple local deformation data constitute the tower deformation information set.

[0047] Based on the identified multiple deformation areas, the laser scanning equipment is precisely positioned in these areas to ensure that the focus of data collection is concentrated on the deformation areas. The laser scanning equipment is used to scan each deformation area to obtain the three-dimensional point cloud data of these areas, and multiple local point cloud data are integrated to obtain the surface shape, contour and any slight deformation of the area.

[0048] The local point cloud data is converted into a triangular mesh to form a continuous surface mesh model. The generated mesh model is smoothed and optimized to ensure the continuity and accuracy of the surface, especially for the details of the deformation area. According to the optimized mesh model, a three-dimensional solid model of the deformation area is generated as a local solid model. The model can accurately reflect the geometric features of the area, and the multiple local point cloud data are traversed to obtain multiple local solid models.

[0049] Locate the position of each deformation area in the overall transmission tower model, and determine the position correspondence between each local entity model and the overall transmission tower model based on the position positioning result, so as to match the local entity model with the overall transmission tower model to ensure that the local model of the deformation area is accurately aligned to the overall tower model. After the model fitting is completed, the geometric data of the deformation area is extracted from the local entity model, mainly including deformation amount, angle deviation, translation distance and other indicators, and the extracted geometric data is quantified to calculate the specific deformation value of each deformation area. For example, the specific deformation of the tower in these areas is determined by calculating the displacement or deviation value, and all local deformation data are integrated together to form a tower deformation information set, which covers the detailed deformation data of each deformation area of ​​the target transmission tower.

[0050] Furthermore, according to the tower building information, the connection nodes are located on the target transmission tower, and the connection deviations are visually identified on the located connection node set to obtain the risk connection node set. The method also includes:

[0051] The connection nodes are located on the transmission tower model to obtain the connection node set; a flight path is planned according to the connection node set to obtain an image acquisition path; a drone is navigated according to the image acquisition path to acquire images of the target transmission tower to obtain a connection node image sequence; connection deviations are visually identified on the connection node image sequence to locate the risky connection node set.

[0052] The structural information in the transmission tower model is used to identify the locations of connection nodes, which include key structural parts such as the intersections, support points, and bolt connections of the towers. The identified connection nodes are marked in the model to form a connection node set.

[0053] The set of connected nodes is input into the path planning algorithm, and the algorithm is used to generate the optimal flight path connecting all nodes. The algorithm comprehensively considers factors such as the shortest distance of the path, energy consumption, obstacle avoidance requirements, etc., and generates an image acquisition path that covers all connected nodes.

[0054] The planned image acquisition path is loaded into the navigation system of the UAV, and the UAV is controlled to automatically navigate and fly along the predetermined image acquisition path, arrive at the position of each connection node in turn, start the onboard camera to acquire images of the connection nodes, and integrate the image acquisition results according to the order of the connection nodes to obtain the connection node image sequence.

[0055] Multiple sample connection deviation image sets of multiple sample connection nodes are collected and used as training data for model training to obtain a connection deviation recognition model. The model is used to analyze the connection node image sequence, identify risky connection nodes, and integrate all risky connection nodes to obtain a risky connection node set.

[0056] Furthermore, a flight path is planned according to the set of connected nodes to obtain an image acquisition path, and the method further includes:

[0057] Interactively obtain the acquisition vector set of the connection node set; preset an image acquisition distance, and use the image acquisition distance as a constraint to locate the image acquisition node set according to the connection node set and the acquisition vector set; interactively obtain the image acquisition device parameters of the UAV, and perform field of view overlap analysis on the image acquisition node set according to the image acquisition device parameters and the image acquisition distance to obtain M groups of overlapping acquisition nodes; perform node aggregation on the M groups of overlapping acquisition nodes, and update the image acquisition node set based on the aggregation results to obtain an updated acquisition node set; perform shortest flight path fitting according to the updated acquisition node set to obtain the image acquisition path.

[0058] Access past transmission tower operation and maintenance logs, which record information such as image acquisition angles and directions used in different maintenance operations, especially effective acquisition vectors used to determine the status of connection nodes. Extract experience vectors from the operation and maintenance logs to obtain the acquisition vector set of the connection node set. These vectors describe the directions from which image acquisition can effectively determine the status of the connection nodes. These experience vectors are based on the actual situation and experience summary in historical operation and maintenance.

[0059] According to the performance of the drone and its camera equipment, the size and position of the target connection node, the image acquisition distance is preset to ensure that clear node images can be acquired in actual operation. According to the preset image acquisition distance, the acquisition vector set is applied to determine the image acquisition node of each connection node one by one to form an image acquisition node set, ensuring that all connection nodes can be fully covered under the preset acquisition distance and vector constraints.

[0060] By interacting with the UAV control system, the parameters of the image acquisition device currently carried by the UAV are obtained, including the camera's resolution, field of view angle, focal length, aperture size, sensor size, shooting mode, etc. The field of view range of each image acquisition node is calculated based on the acquired image acquisition device parameters and the preset image acquisition distance. The field of view range determines the area that the camera can cover at a specified position and distance. By simulating the field of view of each image acquisition node, the field of view ranges of adjacent nodes are compared and analyzed to identify the overlapping field of view areas between two or more image acquisition nodes. If the fields of view of two or more nodes overlap, the overlapping nodes are divided into a group to form overlapping acquisition nodes. All nodes are traversed to obtain M groups of overlapping acquisition nodes, where M is a positive integer.

[0061] A distance threshold is set according to the image acquisition requirements and actual operation conditions, which is defined as the maximum allowable distance between image acquisition nodes. If the distance between connected nodes is less than the threshold, they can be aggregated into one node. The clustering algorithm is used to aggregate M groups of overlapping acquisition nodes. According to the distance threshold and relative position, the close acquisition nodes are merged into a new aggregation node to represent the position of these original nodes. According to the aggregation result, the original M groups of overlapping acquisition nodes are replaced with the new aggregation node to obtain an updated acquisition node set.

[0062] Use the shortest path algorithm, such as the Dijkstra algorithm, to plan the path for updating the acquisition node set. Combined with factors such as obstacles, flight restrictions, and efficiency in the actual flight environment, ensure that the drone covers all acquisition nodes in the shortest time, and use the shortest flight path obtained as the image acquisition path.

[0063] Through node aggregation, the number of required acquisition nodes is reduced, thereby improving the efficiency of image acquisition. Through the shortest flight path planning, the flight time of the UAV can be greatly shortened, energy consumption can be reduced, and the efficiency of task completion can be improved.

[0064] Furthermore, the connection deviation visual recognition is performed on the connection node image sequence to locate the risk connection node set, and the method further includes:

[0065] Interactively obtain multiple sample connection deviation image sets of multiple sample connection nodes; use the multiple sample connection deviation image sets to construct a connection deviation recognition model; synchronize the connection node image sequence to the connection deviation recognition model to perform connection deviation visual recognition, and locate the risk connection node set, wherein N risk connection nodes in the risk connection node set have N node spatial position identifiers.

[0066] According to actual operational requirements, different categories of sample connection nodes are defined, including normal connection nodes and various deviation connection nodes, such as loose, misaligned, corroded, etc. The sample image data of these connection nodes in different states are obtained through the image acquisition system. The images should cover various possible connection deviation situations. Each sample image is annotated to mark the deviation type and degree of the connection node. The annotation information is used to train the recognition model. After the annotation is completed, multiple sample connection deviation image sets are obtained by integration.

[0067] Based on deep learning algorithms, such as convolutional neural networks, a connection deviation recognition model is constructed, and the architecture of the model is designed, including input layer, convolution layer, pooling layer, fully connected layer, etc. Multiple sample connection deviation image sets are divided into training set, validation set and test set. The model is trained with the training set, and the model parameters are adjusted so that it can accurately identify the deviation type of connection nodes. The training process includes hyperparameter tuning, regularization and other technologies to optimize model performance. The performance of the model is evaluated on the validation set, and indicators such as recognition accuracy are checked. If the model performs poorly, adjustments and optimizations are made. The performance of the final model is tested on the test set to ensure that it has good generalization ability and practical application effect. According to the evaluation results, the model is further optimized to finally obtain a connection deviation recognition model that meets the requirements, which is used to identify and classify the deviation of connection nodes in real time.

[0068] The connection node image sequence is input into the connection deviation recognition model. The model identifies various types of deviations in the connection nodes, such as looseness, misalignment, corrosion, etc., based on the features learned during the training process, and marks the connection nodes in each image as normal or with deviations. All connection nodes marked as deviated are extracted from the recognition results, and their specific spatial position identifiers are indicated to form a risk connection node set. The risk connection node set will include N risk connection nodes with spatial position identifiers.

[0069] Furthermore, based on the tower defect analysis model, a deformation risk quantitative assessment is performed on the target transmission tower to obtain a deformation risk coefficient, and the method further includes:

[0070] In the pole tower defect analysis model, the average point spacing calculation of N risk connection nodes in the risk connection node set is performed to obtain a connection risk dispersion coefficient; based on the internal defect information set, a defect extreme value call is performed to obtain an internal defect extreme value array; the internal defect extreme value array is normalized to obtain an internal defect risk coefficient; the multiple local deformation data in the pole tower deformation information set are normalized, and extreme value calls are performed on the normalized result to obtain a local deformation risk coefficient; the regional center coordinates of the multiple deformation areas are extracted from the pole tower defect analysis model to obtain a plurality of deformation risk nodes; in the pole tower defect analysis model, the average point spacing calculation of the multiple deformation risk nodes is performed to obtain a deformation risk dispersion coefficient; a pole tower deformation evaluation function is introduced, and based on the pole tower deformation evaluation function, a comprehensive evaluation is performed on the connection risk dispersion coefficient, the internal defect risk coefficient, the local deformation risk coefficient and the deformation risk dispersion coefficient to obtain the deformation risk coefficient.

[0071] The spatial coordinate information of N risk connection nodes in the risk connection node set is obtained to represent their positions in three-dimensional space. The N risk connection nodes are paired with each other. For each pair of risk connection nodes, the Euclidean distance between them is calculated according to the spatial coordinate information to obtain the point spacing of the node pair. All node pairs are traversed, and an N×N distance matrix is ​​constructed according to the point spacing of all node pairs. The average value of all non-diagonal elements in the distance matrix is ​​calculated to obtain the average point spacing between all risk connection nodes.

[0072] The connection risk dispersion coefficient is calculated based on the average point spacing. Specifically, the standard deviation of the point spacing of all node pairs is calculated, and then the standard deviation is divided by the average point spacing to obtain the connection risk dispersion coefficient. This coefficient reflects the degree of dispersion of risk connection nodes in spatial distribution. A high dispersion coefficient indicates that there is a large uneven distribution of risk nodes in space, indicating that risks are more concentrated in certain areas.

[0073] The extreme value parameters of each defect are extracted from the internal defect information set, such as the maximum length, maximum width, depth, etc. of the internal crack, to obtain an internal defect extreme value array. The internal defect extreme value array is normalized, for example, by using the maximum and minimum normalization method to map the data to the [0,1] interval. For each defect, the normalized extreme value parameters are integrated, for example, by using the weighted average method to calculate the overall internal defect risk coefficient. A high risk coefficient indicates a more serious defect.

[0074] Multiple local deformation data are quantified to obtain deformation angle, displacement distance, etc., and the maximum and minimum values ​​of all local deformation data are obtained. The deformation data are mapped to the [0,1] interval using the maximum and minimum normalization method to obtain normalized local deformation data. The normalized local deformation data is extreme valued, that is, the maximum value of all deformation data, such as the maximum deformation angle, maximum displacement distance, etc., is called. The maximum value of all deformation data is weighted averaged to obtain the local deformation risk coefficient. A high risk coefficient indicates a greater degree of deformation severity.

[0075] Extract information of multiple deformation areas from the tower defect analysis model, including their spatial position, shape, size, etc., analyze each deformation area, and calculate its regional center coordinates. For example, calculate the mean of the coordinates of all points in each deformation area to obtain the geometric center coordinates as the regional center coordinates. Integrate all regional center coordinates as multiple deformation risk nodes for subsequent risk analysis.

[0076] The Euclidean distance is used to calculate the distance between every two deformation risk nodes, and the mean of all distances is calculated. Based on the mean of the distance, the standard deviation of the distance between every two deformation risk nodes is calculated. The standard deviation is then divided by the mean of the distance to obtain the deformation risk dispersion coefficient, which is used to quantitatively analyze the risk degree of tower deformation.

[0077] A tower deformation evaluation function is introduced, and the connection risk discrete coefficient, internal defect risk coefficient, local deformation risk coefficient and deformation risk discrete coefficient are input into the tower deformation evaluation function to calculate the deformation risk coefficient. The calculated deformation risk coefficient reflects the overall risk level after comprehensive connection deviation, internal defect, local deformation and deformation discreteness. The higher the value, the more serious the use risk caused by deformation.

[0078] Furthermore, the tower deformation evaluation function is as follows:

[0079]

[0080] Among them, F is the deformation risk coefficient, D c is the connection risk dispersion coefficient, R i is the internal defect risk factor, R l is the local deformation risk factor, D d is the deformation risk dispersion coefficient, w1 is the connection risk weight, w2 is the associated risk weight, w3 is the deformation risk weight, w1+w2+w3+w4=1.

[0081] The tower deformation evaluation function is as follows:

[0082]

[0083] For the tower deformation evaluation function, the first part By taking the square root of the connection risk dispersion coefficient, its nonlinear effect on the deformation risk coefficient can be reduced, and the obtained value reflects the contribution of the connection risk dispersion coefficient to the overall risk; Part II Adjusting the influence of internal defect risk coefficient relative to local deformation risk coefficient ensures a reasonable assessment of internal defect risk taking into account local deformation; the third part is achieved by By calculating the relative impact of local deformation and considering the influence of deformation risk discrete coefficient, this combined method can comprehensively evaluate the risk of local deformation and adjust the weight according to the discreteness of deformation. In summary, the tower deformation evaluation function combines various risk factors through calculation, conducts scientific and reasonable risk assessment, and provides a basis for subsequent maintenance and overhaul decisions.

[0084] In summary, the method for evaluating the deformation of a transmission tower based on visual recognition provided by the embodiment of the present application has the following technical effects:

[0085] By interactively acquiring tower building information and restoring the transmission tower model, the digital model of the target transmission tower can be accurately constructed, providing a basis for subsequent visual recognition and analysis; by performing structural deviation visual recognition on the target tower based on the transmission tower model, the deformation of the tower can be fully detected, and a detailed deformation information set can be obtained, so as to accurately identify the deviation of the tower structure and timely discover potential deformation problems; by locating the connection nodes and performing connection deviation visual recognition, the risky connection node set can be accurately identified, which enables priority attention to those connection points with potential risks during maintenance and inspection, thereby improving the efficiency and accuracy of maintenance; the preset detection point array and ultrasonic flaw detection can deeply detect defects inside the transmission tower. This non-destructive testing method ensures accurate identification of internal defects and will not cause damage to the tower. Further damage is detected and detailed internal defect data is provided at the same time; internal defect identification is performed based on the ultrasonic detection data set to obtain the internal defect information set, so as to achieve a comprehensive grasp of the internal structural defects and provide a reliable basis for subsequent risk assessment and maintenance decisions; the internal defect information set, risk connection node set and tower deformation information set are fitted to the transmission tower model to obtain the tower defect analysis model. This comprehensive analysis model can integrate information from different sources, provide a comprehensive tower defect and risk assessment, and ensure the comprehensiveness and accuracy of the analysis results; based on the tower defect analysis model, the deformation risk of the target transmission tower is quantitatively assessed to obtain the deformation risk coefficient, to quantify various test results and analysis data, and to comprehensively judge the overall risk level of the tower, which provides a scientific basis for maintenance and management decisions and reduces operational risks.

[0086] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for evaluating the deformation of a transmission tower based on visual recognition, characterized in that: The method comprises: Interactively obtain tower building information of a target transmission tower, and restore a transmission tower model based on the tower building information; Performing visual recognition of structural deviation on the target transmission tower based on the transmission tower model to obtain a tower deformation information set; Locate the connection nodes on the target transmission tower according to the tower building information, perform connection deviation visual identification on the located connection node set, and obtain the risk connection node set; Presetting a flaw detection point array, and performing ultrasonic flaw detection on the target transmission tower based on the flaw detection point array to obtain an ultrasonic detection data set; Based on the ultrasonic detection data set, internal defects of the target transmission tower are identified to obtain an internal defect information set; Fitting the internal defect information set, the risk connection node set and the tower deformation information set to the transmission tower model to obtain a tower defect analysis model; A deformation risk quantitative assessment is performed on the target transmission tower based on the tower defect analysis model to obtain a deformation risk coefficient.

2. A method for evaluating transmission tower deformation based on visual recognition according to claim 1, characterized in that: Based on the transmission tower model, visual recognition of structural deviation is performed on the target transmission tower to obtain a tower deformation information set, and the method further includes: Preset a shooting point array, and based on the shooting point array, perform image acquisition on the target transmission tower to obtain a tower entity image array; Fitting the shooting point array to the transmission tower model, performing image acquisition replication on the transmission tower model, and obtaining a tower virtual image array; The pole tower physical image array and the pole tower virtual image array are mapped according to the shooting point array to perform image deviation visual recognition to obtain the pole tower deformation information set.

3. A method for evaluating transmission tower deformation based on visual recognition as claimed in claim 2, characterized in that: According to the shooting point array, the pole tower physical image array and the pole tower virtual image array are mapped to perform image deviation visual recognition to obtain the pole tower deformation information set, and the method further includes: Performing color feature extraction based on the pole tower entity image array to obtain pole tower color features; Optimizing the image contrast of the pole tower entity image array based on the pole tower color feature as a constraint to obtain an optimized entity image array; Extracting a first pole tower physical image and a first pole tower virtual image mapped to a first shooting point from the optimized physical image array and the pole tower virtual image array, wherein the first shooting point is any shooting point in the shooting point array; Performing spatial alignment on the first pole tower physical image and the first pole tower virtual image to obtain an image registration result; Performing deformation region identification based on the image registration result to locate the first deformation region; Similarly, the optimized entity image array and the tower virtual image array are traversed to identify deformation areas, and multiple deformation areas are obtained; The target transmission tower is numerically quantified for tower deformation according to the multiple deformation regions to obtain the tower deformation information set, wherein the tower deformation information set includes multiple local deformation data of the multiple deformation regions.

4. A method for evaluating transmission tower deformation based on visual recognition as claimed in claim 3, characterized in that: According to the multiple deformation areas, the target transmission tower is numerically quantified to obtain the tower deformation information set, and the method further includes: Using the multiple deformation areas as a guide, performing laser point cloud data collection on the target transmission tower to obtain multiple local point cloud data; Performing geometric modeling based on the multiple local point cloud data to obtain multiple local entity models; Guided by the multiple deformation areas, the multiple local entity models are fitted to the transmission tower model to numerically quantify the tower deformation, and a plurality of local deformation data are obtained. The plurality of local deformation data constitute the tower deformation information set.

5. The method for evaluating transmission tower deformation based on visual recognition according to claim 1, characterized in that: According to the tower building information, the connection nodes are located on the target transmission tower, and the connection deviation is visually identified on the connection node set obtained by the location to obtain the risk connection node set. The method further includes: Positioning connection nodes on the transmission tower model to obtain the connection node set; Perform flight path planning according to the connection node set to obtain an image acquisition path; Navigating the drone according to the image acquisition path to acquire images of the target transmission tower to obtain a connection node image sequence; Visual identification of connection deviations is performed on the connection node image sequence to locate the risky connection node set.

6. A method for evaluating transmission tower deformation based on visual recognition as claimed in claim 5, characterized in that: Performing flight path planning according to the connection node set to obtain an image acquisition path, the method further includes: interactively obtaining a collection vector set of the connection node set; Preset an image acquisition distance, and use the image acquisition distance as a constraint to locate the image acquisition node set according to the connection node set and the acquisition vector set; Interactively obtain image acquisition device parameters of the UAV, and perform field of view overlap analysis on the image acquisition node set according to the image acquisition device parameters and the image acquisition distance to obtain M groups of overlapping acquisition nodes; Performing node aggregation on the M groups of overlapping acquisition nodes, and updating the image acquisition node set based on the aggregation result to obtain an updated acquisition node set; The shortest flight path is fitted according to the updated acquisition node set to obtain the image acquisition path.

7. A method for evaluating transmission tower deformation based on visual recognition as claimed in claim 5, characterized in that: Performing visual identification of connection deviations on the connection node image sequence to locate the risky connection node set, the method further comprising: Interactively obtain multiple sample connection deviation image sets of multiple sample connection nodes; A connection deviation recognition model is constructed by using the plurality of sample connection deviation image sets; The connection node image sequence is synchronized to the connection deviation recognition model to perform connection deviation visual recognition to locate the risk connection node set, wherein N risk connection nodes in the risk connection node set have N node spatial position identifiers.

8. The method for evaluating transmission tower deformation based on visual recognition according to claim 4, characterized in that: Based on the tower defect analysis model, a deformation risk quantitative assessment is performed on the target transmission tower to obtain a deformation risk coefficient. The method further includes: In the tower defect analysis model, the average point spacing calculation of N risk connection nodes in the risk connection node set is performed to obtain a connection risk dispersion coefficient; Perform defect extreme value calling based on the internal defect information set to obtain an internal defect extreme value array; Normalizing the internal defect extreme value array to obtain an internal defect risk coefficient; Normalizing the plurality of local deformation data in the tower deformation information set, and performing extreme value calling on the normalized result to obtain a local deformation risk coefficient; Extracting and obtaining the regional center coordinates of the plurality of deformation regions from the tower defect analysis model to obtain a plurality of deformation risk nodes; Calculating the average point spacing of the plurality of deformation risk nodes in the tower defect analysis model to obtain a deformation risk dispersion coefficient; A pole tower deformation evaluation function is introduced, and based on the pole tower deformation evaluation function, a comprehensive evaluation is performed on the connection risk dispersion coefficient, the internal defect risk coefficient, the local deformation risk coefficient and the deformation risk dispersion coefficient to obtain the deformation risk coefficient.

9. A method for evaluating transmission tower deformation based on visual recognition as claimed in claim 8, characterized in that: The tower deformation evaluation function is as follows: Among them, F is the deformation risk coefficient, D c is the connection risk dispersion coefficient, R i is the internal defect risk factor, R l is the local deformation risk factor, D d is the deformation risk dispersion coefficient, w1 is the connection risk weight, w2 is the associated risk weight, w3 is the deformation risk weight, w1+w2+w3+w4=1.

Images