A method and device for unmanned inspection of power line facilities

Abstract

The present application discloses an unmanned inspection method and device for power line facilities, and the present application belongs to the field of power inspection technology. The method includes: obtaining a first point cloud and image data of a power line facility collected by a drone, determining a second point cloud of the power line facility based on the image data, performing point cloud registration on the first point cloud and the second point cloud, and establishing a three-dimensional model of the power line facility, determining an inspection plan for the drone based on the three-dimensional model and preset inspection requirements, and sending the inspection plan to the drone. The three-dimensional model generated by the point cloud registration in this scheme can accurately reflect the actual situation of the power line facility, thereby improving the accuracy and reliability of the inspection. Automatically formulate an inspection plan based on the three-dimensional model and preset inspection requirements, reduce manual intervention, reduce labor costs, and improve inspection efficiency and inspection safety.

Description

Technical Field

[0001] The present application relates to the technical field of electric power inspection, and in particular to an unmanned inspection method and device for electric power line facilities. Background Art

[0002] Power line facilities are the core components of the power system, and their operating status directly affects the transmission and distribution of electric energy. Through regular inspections, faults and hidden dangers in power line facilities can be discovered and handled in a timely manner to ensure the normal operation of the power system.

[0003] Nowadays, inspectors will arrive at designated inspection points to inspect power line facilities one by one according to task requirements. They will use measuring tools to measure key parameters, such as conductor tension and insulator resistance. During the inspection process, if the inspectors find any abnormalities, they will record the abnormalities and notify maintenance personnel to carry out repairs.

[0004] However, manual inspections are inefficient and costly, and inspectors need to face a variety of complex and dangerous environments, which reduces inspection safety. Summary of the invention

[0005] The present application provides an unmanned inspection method and device for power line facilities, which solves the problem that manual inspection of power line facilities is very inefficient and has high labor costs, and at the same time, inspectors need to face various complex and dangerous environments, which reduces the safety of inspection.

[0006] According to a first aspect of the present application, a method for unmanned inspection of power line facilities is provided, the method comprising:

[0007] Acquire a first point cloud and image data of a power line facility collected by a drone, and determine a second point cloud of the power line facility based on the image data;

[0008] The first centroid of the first point cloud is calculated according to a preset centroid calculation formula, and the second centroid of the second point cloud is calculated; wherein the preset centroid calculation formula is:

[0009] ;

[0010] in, is the first centroid; Q is the first point cloud; is the i-th point of the first point cloud; is the total number of points in the first point cloud; is the second centroid; P is the second point cloud; is the i-th point of the second point cloud, is the total number of points in the second point cloud;

[0011] Normalizing the first point cloud and the second point cloud according to the first centroid and the second centroid, determining a rotation matrix and a translation vector according to the normalized first point cloud and the normalized second point cloud, and transforming the normalized second point cloud according to the rotation matrix and the translation vector to obtain a third point cloud;

[0012] The evaluation function value between the normalized first point cloud and the third point cloud is calculated according to a preset evaluation function calculation formula to determine whether the evaluation function value reaches a preset convergence condition; wherein the preset evaluation function calculation formula is:

[0013] ;

[0014] Where E is the evaluation function value; q is the point of the first point cloud; is the point of the third point cloud; R is the rotation matrix; T is the translation vector;

[0015] If the evaluation function value reaches the preset convergence condition, a three-dimensional model of the power line facility is established based on the normalized first point cloud and the third point cloud, and an inspection plan for the UAV is determined based on the three-dimensional model and preset inspection requirements. The inspection plan is sent to the UAV so that the UAV can inspect the power line facilities according to the inspection plan.

[0016] Furthermore, normalizing the first point cloud and the second point cloud according to the first centroid and the second centroid includes:

[0017] The first centroid and the second centroid are input into a preset normalization formula to perform normalization processing on each point in the first point cloud and the second point cloud; wherein the preset normalization formula is:

[0018] ;

[0019] in, is the i-th point of the first point cloud; is the i-th point of the first point cloud after normalization; is the first centroid; is the i-th point of the second point cloud; is the i-th point of the second point cloud after normalization; is the second centroid.

[0020] Further, determining a rotation matrix and a translation vector according to the normalized first point cloud and the normalized second point cloud includes:

[0021] Determine corresponding point pairs between the normalized first point cloud and the normalized second point cloud;

[0022] The first matrix is ​​determined according to the corresponding point pairs and a preset first matrix construction formula; wherein the preset first matrix construction formula is:

[0023] ;

[0024] Among them, C is the first matrix; N is the point pair ( , ) for The transpose of , i.e. the row vector;

[0025] Perform SVD decomposition on the first matrix to obtain a singular value decomposition of the first matrix; wherein the process of performing SVD decomposition on the first matrix includes:

[0026] ;

[0027] Among them, U is the left singular vector matrix; V is the right singular vector matrix; is a diagonal matrix; is the transposed matrix of the right singular vector matrix;

[0028] The rotation matrix and the translation vector are calculated according to the singular value decomposition of the first matrix; wherein the calculation process includes:

[0029] .

[0030] Furthermore, after determining whether the evaluation function value reaches a preset convergence condition, the method further includes:

[0031] If the evaluation function value does not reach a preset convergence condition, determining corresponding point pairs between the normalized first point cloud and the third point cloud;

[0032] The first matrix is ​​re-determined according to the corresponding point pairs and a preset first matrix construction formula; wherein the preset first matrix construction formula is:

[0033] ;

[0034] Where C is the first matrix; N is the number of point pairs (pi, qi); is the transpose of qi, i.e., the row vector;

[0035] The first matrix is ​​re-performed with SVD decomposition to obtain a singular value decomposition of the first matrix; wherein the process of performing SVD decomposition on the first matrix includes:

[0036] ;

[0037] Among them, U is the left singular vector matrix; V is the right singular vector matrix; is the transposed matrix of the right singular vector matrix;

[0038] The rotation matrix and the translation vector are recalculated according to the singular value decomposition of the first matrix; wherein the calculation process includes:

[0039] ;

[0040] The third point cloud is updated according to the rotation matrix and the translation vector. If the evaluation function value between the normalized first point cloud calculated according to the preset evaluation function calculation formula and the updated third point cloud still does not reach the preset convergence condition, the corresponding point pairs between the normalized first point cloud and the updated third point cloud are recalculated, and the third point cloud is continuously updated until the evaluation function value between the normalized first point cloud calculated according to the preset evaluation function calculation formula and the updated third point cloud reaches the preset convergence condition.

[0041] Furthermore, the inspection plan of the drone is determined according to the three-dimensional model and the preset inspection requirements, including:

[0042] Determine the inspection area, each inspection component in the inspection area, and the inspection data collection task according to the preset inspection requirements; wherein the inspection area includes the inspection starting point and the inspection end point;

[0043] Determine the inspection component position of each inspection component and the obstacle position in the inspection area according to the three-dimensional model;

[0044] Determining an inspection path of the drone according to the inspection area, the location of the inspection component, and the location of the obstacle;

[0045] Determine the inspection method of the drone for each inspection component according to the inspection data collection task;

[0046] The inspection plan of the drone is determined according to the inspection method and the inspection path.

[0047] Further, determining the inspection path of the drone according to the inspection area, the inspection component position, and the obstacle position includes:

[0048] Identify terrain information of the inspection area, and divide the inspection area into at least two inspection layers according to the terrain information, the location of the inspection component, and the location of the obstacle;

[0049] Determine the track line of the drone at each inspection layer according to the inspection starting point, the inspection component position, the obstacle position and the inspection end point;

[0050] The inspection path of the drone is determined according to the track line of the drone at each inspection layer.

[0051] Furthermore, the inspection method of the drone for each inspection component is determined according to the inspection data collection task, including:

[0052] Determine the target collection task of each inspection component according to the inspection data collection task;

[0053] Determine the target acquisition components and acquisition parameters of the drone according to the target acquisition task;

[0054] The inspection method of the drone for each inspection component is determined according to the target collection component and the collection parameters.

[0055] Furthermore, after sending the inspection plan to the drone, the method further includes:

[0056] Receive inspection data of each inspection component transmitted by the drone, and determine the data type of the inspection data;

[0057] If the inspection data is image data, determining a corresponding target acquisition task according to the image data;

[0058] Determine a target recognition method according to the target acquisition task, recognize corresponding image data according to the target recognition method, and determine whether there is abnormal information;

[0059] If it is determined that there is abnormal information, the abnormal information is transmitted to the staff's handheld terminal so that the staff can inspect and repair the power line facilities based on the abnormal information.

[0060] Further, after determining the data type of the inspection data, the method further includes:

[0061] If the inspection data is non-image data, determining a corresponding target acquisition task according to the non-image data;

[0062] Determine a corresponding target threshold according to the target acquisition task, and determine whether there is abnormal information according to the non-image data and the corresponding target threshold;

[0063] If it is determined that there is abnormal information, the abnormal information is transmitted to the staff's handheld terminal so that the staff can inspect and repair the power line facilities based on the abnormal information.

[0064] According to a second aspect of the present application, there is provided an unmanned inspection device for power line facilities, which is used to implement the unmanned inspection method for power line facilities described in the first aspect. The device comprises:

[0065] A data acquisition module, used to acquire a first point cloud and image data of a power line facility acquired by a drone, and determine a second point cloud of the power line facility based on the image data;

[0066] The centroid determination module is used to calculate the first centroid of the first point cloud and the second centroid of the second point cloud according to a preset centroid calculation formula; wherein the preset centroid calculation formula is:

[0067] ;

[0068] in, is the first centroid; Q is the first point cloud; is the i-th point of the first point cloud; is the total number of points in the first point cloud; is the second centroid; P is the second point cloud; is the i-th point of the second point cloud, is the total number of points in the second point cloud;

[0069] a point cloud transformation module, configured to normalize the first point cloud and the second point cloud according to the first centroid and the second centroid, determine a rotation matrix and a translation vector according to the normalized first point cloud and the normalized second point cloud, and transform the normalized second point cloud according to the rotation matrix and the translation vector to obtain a third point cloud;

[0070] The condition determination module is used to calculate the evaluation function value between the first point cloud and the third point cloud after normalization according to a preset evaluation function calculation formula, and determine whether the evaluation function value reaches a preset convergence condition; wherein the preset evaluation function calculation formula is:

[0071] ;

[0072] Where E is the evaluation function value; q is the point of the first point cloud; is the point of the third point cloud; R is the rotation matrix; T is the translation vector;

[0073] The inspection plan determination module is used to establish a three-dimensional model of the power line facility based on the normalized first point cloud and the third point cloud if the evaluation function value reaches a preset convergence condition, determine the inspection plan of the drone based on the three-dimensional model and preset inspection requirements, and send the inspection plan to the drone so that the drone can inspect the power line facilities according to the inspection plan.

[0074] In an embodiment of the present application, a first point cloud and image data of a power line facility collected by a drone are obtained, and a second point cloud of the power line facility is determined based on the image data; a first center of mass of the first point cloud is calculated based on a preset center of mass calculation formula, and a second center of mass of the second point cloud is calculated; the first point cloud and the second point cloud are normalized based on the first center of mass and the second center of mass, a rotation matrix and a translation vector are determined based on the normalized first point cloud and the normalized second point cloud, and the normalized second point cloud is transformed based on the rotation matrix and the translation vector to obtain a third point cloud; an evaluation function value between the normalized first point cloud and the third point cloud is calculated based on a preset evaluation function calculation formula, and it is determined whether the evaluation function value reaches a preset convergence condition; if the evaluation function value reaches the preset convergence condition, a three-dimensional model of the power line facility is established based on the normalized first point cloud and the third point cloud, an inspection plan of the drone is determined based on the three-dimensional model and preset inspection requirements, and the inspection plan is sent to the drone for the drone to inspect the power line facility according to the inspection plan. Through the above-mentioned unmanned inspection method for power line facilities, the three-dimensional model generated by point cloud registration can accurately reflect the actual situation of the power line facilities, thereby improving the accuracy and reliability of the inspection. According to the three-dimensional model and the preset inspection requirements, the inspection plan is automatically formulated to reduce manual intervention, reduce labor costs, and improve inspection efficiency and inspection safety.

[0075] It should be understood that the contents described in the Summary of the Invention are not intended to limit the key or important features of the embodiments of the present application, nor are they intended to limit the scope of the present application. Other features of the present application will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0076] The above and other features, advantages and aspects of the embodiments of the present application will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. The accompanying drawings are used to better understand the present solution and do not constitute a limitation of the present application. In the accompanying drawings, the same or similar reference numerals represent the same or similar elements, among which:

[0077] Figure 1 It is a flowchart of an unmanned inspection method for power line facilities provided in Example 1 of the present application;

[0078] Figure 2 This is a flow chart of an unmanned inspection method for power line facilities provided in Embodiment 2 of the present application;

[0079] Figure 3 It is a flowchart of an unmanned inspection method for power line facilities provided in Embodiment 3 of the present application;

[0080] Figure 4It is a structural schematic diagram of an unmanned inspection device for power line facilities provided in Example 4 of the present application. DETAILED DESCRIPTION

[0081] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0082] In addition, the term "and / or" in this article is only a description of the association relationship between the associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship.

[0083] The following, in conjunction with the accompanying drawings, describes in detail the unmanned inspection method for power line facilities provided by the embodiment of the present application through specific embodiments and their application scenarios.

[0084] Embodiment 1: Figure 1 1 is a flow chart of an unmanned inspection method for power line facilities provided in the first embodiment of the present application. Figure 1 As shown, the specific steps include:

[0085] S101, obtaining a first point cloud and image data of a power line facility collected by a drone, and determining a second point cloud of the power line facility based on the image data.

[0086] First of all, the usage scenario of this solution can be that the control end obtains the first point cloud data and image data of the power line facilities collected by the drone, and determines the second point cloud data from the image data, and establishes a three-dimensional model of the power line facilities after point cloud registration of the first point cloud data and the second point cloud data, and formulates the drone inspection plan according to the three-dimensional model and preset inspection requirements.

[0087] Based on the above usage scenarios, it can be understood that the executor of the present application can be an unmanned inspection device for power line facilities that integrates data analysis functions, model building functions, and inspection plan determination functions, and no excessive limitations are made here.

[0088] In this solution, the control end can be the core of the entire system, responsible for the entire process from data reception, processing, analysis to model building and task assignment. It ensures that the drone can accurately perform the inspection task of power line facilities.

[0089] A drone is an unmanned aerial vehicle that can be remotely controlled or fly autonomously. Drones are equipped with various sensors and cameras to collect data, such as images, videos, and point cloud data, for inspection and monitoring of power facilities.

[0090] Power line facilities can include transmission lines, poles, towers, substations and their related equipment. During the flight, drones will collect images and point cloud data of these facilities for inspection and maintenance.

[0091] The first point cloud may refer to the three-dimensional point cloud data of the power line facilities collected by the drone during flight. The point cloud data may be composed of a large number of spatial points, which may be generated by laser radar or stereo vision technology, and these points represent the three-dimensional spatial information of the power line facilities and their surroundings.

[0092] Image data may refer to photos or videos collected by a drone during flight. Specifically, these images may be visible light images, infrared images, or images in other wavelengths.

[0093] The second point cloud may be a three-dimensional point cloud generated by processing image data collected by the drone.

[0094] During the flight, the drone can collect the first point cloud and image data of the power line facilities through its sensors and cameras. The first point cloud can be generated by lidar or stereo vision technology, representing the three-dimensional structure of the power line facilities and their surroundings. The image data can be a high-resolution image or video taken by the drone, providing detailed information about the power facilities. The collected first point cloud and image data are then transmitted to the control end through a wireless communication link, and the control end receives and stores the data. The control end can then process the image data using an image processing algorithm. Specifically, computer vision technology can be used to extract key feature points from the image. The feature points in the image are then matched, spatial relationships are established, and multiple image data are converted into a second point cloud based on multi-view geometry and dense matching technology.

[0095] S102, calculating a first centroid of the first point cloud and a second centroid of the second point cloud according to a preset centroid calculation formula; wherein the preset centroid calculation formula is:

[0096] ;

[0097] in, is the first centroid; Q is the first point cloud; is the i-th point of the first point cloud; is the total number of points in the first point cloud; is the second centroid; P is the second point cloud; is the i-th point of the second point cloud; N2 is the total number of points in the second point cloud.

[0098] The first centroid may be a centroid calculated based on the first point cloud, indicating a geometric center of the first point cloud.

[0099] The second centroid may be a centroid calculated based on the second point cloud, indicating a geometric center of the second point cloud.

[0100] For each point in the first point cloud Q , where i ranges from 1 to , sum up their coordinates. Finally, divide this sum by the total number of points , get the coordinates of the first centroid. For each point in the second point cloud P , where i ranges from 1 to , sum up their coordinates. Finally, divide this sum by the total number of points , and get the coordinates of the second centroid.

[0101] S103, normalizing the first point cloud and the second point cloud according to the first centroid and the second centroid, determining the rotation matrix and the translation vector according to the normalized first point cloud and the normalized second point cloud, and transforming the normalized second point cloud according to the rotation matrix and the translation vector to obtain a third point cloud.

[0102] The rotation matrix can be a 3x3 matrix used to describe the rotation in three-dimensional space. Specifically, it can be an orthogonal matrix that satisfies R^T * R = I (I is the identity matrix) and its determinant is 1 (det(R) = 1). Each column of the rotation matrix R is a unit vector, and these column vectors are orthogonal to each other.

[0103] The translation vector T is a 3x1 vector used to describe the translation in three-dimensional space. Specifically, the translation vector T can be a simple three-dimensional vector representing the translation amount in the x, y and z directions.

[0104] The third point cloud may be a point cloud obtained by applying a rotation matrix and a translation vector to the normalized second point cloud.

[0105] The two point clouds can be normalized using the calculated centroid. The purpose of normalization is to move the geometric center of the point cloud to the origin of the coordinate system. Specifically, the coordinates of the centroid of the corresponding point cloud can be subtracted from the coordinates of each point. In this way, the normalized point cloud will have its centroid as the origin. Then the iterative closest point algorithm can be used to find the best alignment between the two point clouds in an iterative manner, that is, to minimize the distance between corresponding points. In each iteration, the algorithm finds the closest point correspondence between the two point clouds and calculates a rotation matrix and a translation vector based on these corresponding points to align one point cloud to the other point cloud. Using the rotation matrix and translation vector, the normalized second point cloud is transformed. Specifically, the coordinates of each point can be multiplied by the rotation matrix and then the translation vector is added. In this way, the second point cloud is transformed to a position that is more aligned with the first point cloud, resulting in a third point cloud.

[0106] On the basis of the above technical solution, optionally, normalizing the first point cloud and the second point cloud according to the first centroid and the second centroid includes:

[0107] The first centroid and the second centroid are input into a preset normalization formula to perform normalization processing on each point in the first point cloud and the second point cloud; wherein the preset normalization formula is:

[0108] ;

[0109] in, is the i-th point of the first point cloud; is the i-th point of the first point cloud after normalization; is the first centroid; is the i-th point of the second point cloud; is the i-th point of the second point cloud after normalization; is the second centroid.

[0110] In this solution, the purpose of normalization is to move all points in the first point cloud and the second point cloud to a coordinate system with the first centroid and the second centroid as the origin, that is, to move the point cloud to the center point. After determining the first centroid and the second centroid, all points in the first point cloud and the second point cloud can be calculated through a preset normalization formula, and then the normalized first point cloud and the normalized second point cloud are obtained.

[0111] In this scheme, the point cloud data is normalized to eliminate the influence of translation transformation on subsequent algorithms.

[0112] On the basis of the above technical solution, optionally, determining a rotation matrix and a translation vector according to the normalized first point cloud and the normalized second point cloud includes:

[0113] Determine corresponding point pairs between the normalized first point cloud and the normalized second point cloud;

[0114] The first matrix is ​​determined according to the corresponding point pairs and a preset first matrix construction formula; wherein the preset first matrix construction formula is:

[0115] ;

[0116] Among them, C is the first matrix; N is the point pair ( , ) for The transpose of , i.e. the row vector;

[0117] Perform SVD decomposition on the first matrix to obtain a singular value decomposition of the first matrix; wherein the process of performing SVD decomposition on the first matrix includes:

[0118] ;

[0119] Among them, U is the left singular vector matrix; V is the right singular vector matrix; is a diagonal matrix; is the transposed matrix of the right singular vector matrix;

[0120] The rotation matrix and the translation vector are calculated according to the singular value decomposition of the first matrix; wherein the calculation process includes:

[0121] .

[0122] In this solution, the corresponding point pair may be a point pair at the same or similar three-dimensional position between the normalized first point cloud and the normalized second point cloud.

[0123] For each point in the normalized first point cloud, a point with the shortest Euclidean distance to the point can be found in the normalized second point cloud to determine a corresponding point pair between the normalized first point cloud and the normalized second point cloud.

[0124] The first matrix may be a transformation matrix for aligning the normalized second point cloud to the normalized first point cloud.

[0125] The default first matrix construction formula computes a matrix where each column is With the corresponding The accumulation of the results of multiplication.

[0126] Singular value decomposition may be a method of decomposing a matrix into its singular vectors and singular values.

[0127] The left singular vector matrix can be an orthogonal matrix, whose column vectors are called left singular vectors.

[0128] The right singular vector matrix can be an orthogonal matrix, whose column vectors are called right singular vectors.

[0129] The elements on the diagonal of the diagonal matrix are the singular values ​​of C. These singular values ​​are non-negative and are arranged in descending order. The size of the singular value reflects the importance of the corresponding singular vector in the matrix C, and larger singular values ​​correspond to more important eigendirections.

[0130] The transpose matrix of the right singular vector matrix may be the transpose of the right singular vector matrix. In the SVD decomposition, it is used to convert the row vectors in the right singular vector matrix into column vectors so as to be multiplied with the diagonal matrix and the left singular vector matrix to reconstruct the first matrix.

[0131] The goal of SVD is to decompose the first matrix into the product of three matrices, namely, the product of U, Σ, and V^T. In order to find U and V, we need to calculate the transpose of C multiplied by C ( ·C) and C multiplied by the transpose of C (C· ). The two results are n x n and m x m matrices, where n and m are the number of rows and columns of C. Next, we can calculate C and C These eigenvectors will form the columns of U and V. C, its eigenvectors will form the columns of U, and the square roots of the corresponding eigenvalues ​​will form the diagonal elements of Σ. , its eigenvectors will form the columns of V, and the square roots of the corresponding eigenvalues ​​will also form the diagonal elements of Σ (matching the ordering in U). Then construct U, Σ, and V. Specifically, U is an mxm orthogonal matrix whose columns are ·Eigenvectors of C. Σ is an m x n matrix, where all elements except the diagonal elements, i.e., the singular values, are 0. The singular values ​​on the diagonal are C and C The square root of the eigenvalue of , and if m > n, the bottom mn rows of Σ are all 0. V is an nxn orthogonal matrix whose columns are C· Finally, we can verify that U,Σ, Whether it is equal to the original matrix C, specifically, it can be done by calculating the difference between the two matrices and checking whether its norm is small enough.

[0132] By connecting V and , we can get the rotation matrix R, which can rotate the point set P to a position that is as aligned as possible with the point set Q. The translation vector T describes how to translate the rotated point set P to the position of the point set Q. Specifically, it can be achieved by calculating the difference between the first centroid and the second centroid, that is, the translation vector T is such that R +T As close as possible The vector of .

[0133] In this solution, point clouds or objects in two different coordinate systems can be accurately aligned to the same coordinate system through rotation matrices and translation vectors.

[0134] S104, calculating the evaluation function value between the normalized first point cloud and the third point cloud according to a preset evaluation function calculation formula, and determining whether the evaluation function value reaches a preset convergence condition; wherein the preset evaluation function calculation formula is:

[0135] ;

[0136] Where E is the evaluation function value; q is the point of the first point cloud; is the point of the third point cloud; R is the rotation matrix; T is the translation vector.

[0137] The evaluation function value can be used to quantify the degree of alignment between two point clouds. Specifically, it can be the average value of the sum of squares of the Euclidean distances between corresponding points. The smaller the evaluation function value, the better the degree of alignment between the two point clouds, that is, the closer the distance between the corresponding points.

[0138] In the preset convergence condition, if the evaluation function value is less than the preset threshold, it is considered that the convergence condition is met. Specifically, the threshold can be set according to the application requirements and the characteristics of the point cloud.

[0139] For each target point q and its corresponding transformed source point p', the square of the Euclidean distance between them can be calculated, and the squares of the distances between all corresponding point pairs are added up to obtain the total sum of squared distances. Finally, the total sum of squared distances is divided by the number of points and then multiplied by 1 / 2 to obtain the evaluation function value. After obtaining the evaluation function value, the evaluation function value can be compared with the preset convergence condition.

[0140] S105: If the evaluation function value reaches the preset convergence condition, a three-dimensional model of the power line facility is established based on the normalized first point cloud and the third point cloud, and an inspection plan for the UAV is determined based on the three-dimensional model and preset inspection requirements. The inspection plan is sent to the UAV so that the UAV can inspect the power line facilities according to the inspection plan.

[0141] The three-dimensional model of the power line facilities can be a digital three-dimensional model created based on point cloud data, which represents the detailed structure of the power line and its related facilities. Specifically, it can include various components of the power line such as transmission towers, conductors, insulators, cross arms and fasteners.

[0142] The preset inspection requirements can be pre-defined specific requirements and goals for inspection. Specifically, they can include detection targets, that is, the parts of the power facilities that need to be inspected, such as the connection points of the conductors, whether the insulators are damaged, and whether the tower structure is intact. Accuracy requirements, that is, the accuracy requirements for data collection, can include image resolution and point cloud density.

[0143] The inspection plan can be a specific drone inspection task plan formulated based on the 3D model and preset inspection requirements. Specifically, it can include the flight path, that is, the flight route planning of the drone, to ensure that all parts of the power facilities that need to be inspected are covered. The mission plan is the specific tasks performed by the drone during the inspection process, such as taking photos, recording videos, and collecting point cloud data. Technical parameters are the technical parameter settings of the drone, such as flight altitude, speed, shooting angle, and data collection frequency.

[0144] If the evaluation function value reaches the preset convergence condition, the normalized first point cloud and the third point cloud can be merged to remove redundant and noise points, and the surface reconstruction algorithm can be used to generate a three-dimensional surface model of the power facility, and the details of the model can be enhanced by the first and third point clouds. Then plan the flight path of the drone according to the three-dimensional model to ensure that all parts that need to be inspected are covered. Formulate specific tasks for the drone to perform during the inspection process, such as taking photos, recording videos, and collecting point cloud data. And set the drone's technical parameters such as flight altitude, speed, shooting angle, and data collection frequency. Finally, merge them into an inspection plan and send them to the drone through wireless communication technology to ensure that the drone can perform tasks according to the inspection plan.

[0145] In an embodiment of the present application, a first point cloud and image data of a power line facility collected by a drone are obtained, and a second point cloud of the power line facility is determined based on the image data; a first center of mass of the first point cloud is calculated based on a preset center of mass calculation formula, and a second center of mass of the second point cloud is calculated; the first point cloud and the second point cloud are normalized based on the first center of mass and the second center of mass, a rotation matrix and a translation vector are determined based on the normalized first point cloud and the normalized second point cloud, and the normalized second point cloud is transformed based on the rotation matrix and the translation vector to obtain a third point cloud; an evaluation function value between the normalized first point cloud and the third point cloud is calculated based on a preset evaluation function calculation formula, and it is determined whether the evaluation function value reaches a preset convergence condition; if the evaluation function value reaches the preset convergence condition, a three-dimensional model of the power line facility is established based on the normalized first point cloud and the third point cloud, an inspection plan of the drone is determined based on the three-dimensional model and preset inspection requirements, and the inspection plan is sent to the drone for the drone to inspect the power line facility according to the inspection plan. Through the above-mentioned unmanned inspection method for power line facilities, the three-dimensional model generated by point cloud registration can accurately reflect the actual situation of the power line facilities, thereby improving the accuracy and reliability of the inspection. According to the three-dimensional model and the preset inspection requirements, the inspection plan is automatically formulated to reduce manual intervention, reduce labor costs, and improve inspection efficiency and inspection safety.

[0146] Embodiment 2: Figure 2 is a flow chart of an unmanned inspection method for power line facilities provided in the second embodiment of the present application, such as Figure 2 As shown, the specific method includes the following steps:

[0147] S201, obtaining a first point cloud and image data of a power line facility collected by a drone, and determining a second point cloud of the power line facility based on the image data.

[0148] S202, calculating a first centroid of the first point cloud and a second centroid of the second point cloud according to a preset centroid calculation formula; wherein the preset centroid calculation formula is:

[0149] ;

[0150] in, is the first centroid; Q is the first point cloud; is the i-th point of the first point cloud; is the total number of points in the first point cloud; is the second centroid; P is the second point cloud; is the i-th point of the second point cloud; N2 is the total number of points in the second point cloud.

[0151] S203, normalizing the first point cloud and the second point cloud according to the first centroid and the second centroid, determining the rotation matrix and the translation vector according to the normalized first point cloud and the normalized second point cloud, and transforming the normalized second point cloud according to the rotation matrix and the translation vector to obtain a third point cloud.

[0152] S204, calculating the evaluation function value between the normalized first point cloud and the third point cloud according to a preset evaluation function calculation formula, and determining whether the evaluation function value reaches a preset convergence condition; wherein the preset evaluation function calculation formula is:

[0153] ;

[0154] Where E is the evaluation function value; q is the point of the first point cloud; is the point of the third point cloud; R is the rotation matrix; T is the translation vector.

[0155] S205, if the evaluation function value reaches the preset convergence condition, a three-dimensional model of the power line facility is established based on the normalized first point cloud and the third point cloud, and the inspection area, the inspection components in the inspection area and the inspection data collection tasks are determined according to the preset inspection requirements; wherein the inspection area includes the inspection starting point and the inspection end point.

[0156] The inspection area can refer to the specific geographical scope where drones are required to conduct inspections, which can include the coverage area of ​​power lines and their ancillary facilities. Specifically, it can be the entire power line corridor, including transmission towers, conductors, insulators, and substations.

[0157] Each inspection component in the inspection area may be a specific facility or component in the inspection area that needs to be inspected in particular, and may include a transmission tower, a conductor, an insulator, a cross arm, and fasteners.

[0158] The inspection data collection task may refer to the data collection activities that the UAV needs to perform during the inspection process. Specifically, it may include taking photos, recording videos, collecting point cloud data, and collecting infrared thermal imaging.

[0159] The inspection starting point can be the location where the drone starts the inspection mission, and the inspection end point can be the location where the drone returns after completing the mission.

[0160] According to the inspection requirements, you can define specific inspection areas and select appropriate starting and ending points to ensure that the drone can successfully complete the task. Identify key components in the inspection area, such as transmission towers, wires, and insulators, as inspection components. According to the characteristics of each component and the inspection requirements, determine the type of data to be collected, such as photos, videos, point clouds, and thermal imaging, and set the parameters for data collection, such as shooting angle, resolution, and collection frequency.

[0161] S206: Determine the inspection component position of each inspection component and the position of obstacles in the inspection area according to the three-dimensional model.

[0162] The inspection component position may refer to the specific coordinates of the power facility components that need to be inspected in detail in three-dimensional space, and may specifically include the positions of components such as transmission towers, conductors, insulators, connectors, and fasteners.

[0163] The obstacle location may refer to the specific coordinates of objects in the inspection area that may affect the flight or data collection of the drone in three-dimensional space. Specifically, it may include the location of trees, buildings, and other infrastructure.

[0164] In the three-dimensional model, the key components of the power facilities can be identified, and the three-dimensional coordinate information of each component can be extracted and stored as the location data of the inspection component. In addition, obstacles that may affect the flight of the drone can be identified in the three-dimensional model. The three-dimensional coordinate information of each obstacle is extracted and stored as the location data of the obstacle.

[0165] S207: Determine an inspection path of the drone according to the inspection area, the location of the inspection component, and the location of the obstacle.

[0166] The inspection path can refer to the flight trajectory of the drone when performing the inspection task. The path planning needs to ensure that the drone can effectively cover all parts that need to be inspected, avoid obstacles, and ensure flight safety. Specifically, it can include the starting point, the end point, and the flight route between the inspection parts, and clearly define the three-dimensional coordinates and flight order of each flight point.

[0167] The three-dimensional model of the inspection area can be obtained to determine the position coordinates of all inspection components, and the position coordinates of all obstacles can be identified and determined. Then, a suitable path planning algorithm is selected for path planning, such as the Dijkstra algorithm, the A* algorithm, the fast random tree algorithm, and the path planning based on graph theory. After path planning, the path can be optimized. Specifically, it can include obstacle avoidance, that is, ensuring that the drone path avoids all obstacles, and it may be necessary to increase the buffer zone; smooth the path, that is, through interpolation or other methods, make the path smoother and reduce the sharp turns of the drone. Cover all components, that is, ensure that the path passes through the positions of all components that need to be inspected. Then verify the path, specifically, it can include simulated flight, that is, simulating the flight of the drone in a simulation environment to verify the feasibility and safety of the path; adjust the path, that is, adjust the path planning according to the simulation results to ensure the rationality of the path.

[0168] On the basis of the above technical solution, optionally, determining the inspection path of the drone according to the inspection area, the position of the inspection component and the position of the obstacle includes:

[0169] Identify terrain information of the inspection area, and divide the inspection area into at least two inspection layers according to the terrain information, the location of the inspection component, and the location of the obstacle;

[0170] Determine the track line of the drone at each inspection layer according to the inspection starting point, the inspection component position, the obstacle position and the inspection end point;

[0171] The inspection path of the drone is determined according to the track line of the drone at each inspection layer.

[0172] In this solution, the terrain information may refer to the terrain features of the inspection area, specifically, may include height, slope, and surface type.

[0173] Inspection layers can be different height levels that divide the inspection area according to terrain information, inspection component locations, and obstacle locations. Each level corresponds to a different flight altitude to avoid obstacles and ensure the best inspection effect. Specifically, it can be divided according to the undulations of the terrain, the height of the obstacles, and the distribution of the inspection components.

[0174] The flight path can be the path line that the drone flies in the inspection layer, which determines the specific flight route of the drone from the inspection starting point to the end point. The flight path should take into account the terrain undulations, obstacle locations, and inspection component locations.

[0175] The terrain data of the inspection area can be obtained by using drone aerial photography or satellite images, and digital elevation models can be generated through data processing software such as ArcGIS and QGIS, and terrain features such as height and slope can be analyzed. Then the terrain undulations and obstacle heights are analyzed, that is, the terrain undulations and obstacle heights are determined, and different flight altitude layers are divided according to the location of the inspection components. According to the analysis results, the inspection layers are divided to ensure that each layer can avoid obstacles and cover the inspection components. In each inspection layer, the path planning algorithm is used to determine the flight path of the drone to ensure that obstacles are avoided and all inspection components are covered. Specifically, path planning algorithms such as the Dijkstra algorithm and the A* algorithm can be used to combine terrain and obstacle information for path planning. The flight paths of each inspection layer are integrated into a complete inspection path to ensure that the drone can perform inspections according to the predetermined flight path.

[0176] In this solution, by subdividing the inspection layer, the inspection components can be located more accurately, ensuring that the drone can inspect from the optimal angle and distance. The path is planned according to the terrain and obstacle location to ensure the safety of the drone flight and to be able to approach difficult-to-reach inspection points.

[0177] S208: Determine the inspection method of the drone for each inspection component according to the inspection data collection task.

[0178] Inspection methods can refer to the specific operations and methods that drones perform on different components during inspection missions. Different inspection components require different inspection methods to ensure that the required detailed data is obtained. Specifically, it can include specific operation methods and technical means such as taking photos, recording videos, point cloud scanning, and infrared thermal imaging.

[0179] You can analyze the inspection needs, that is, understand the characteristics and importance of each inspection component, and determine the data type and accuracy requirements that need to be collected based on the preset inspection needs. For example, insulators need to detect insulation conditions, which requires infrared thermal imaging data. Then choose the appropriate inspection method, among which taking photos is used for visual inspection and is suitable for most components, such as transmission towers and conductor connection points. Recording videos is used for dynamic monitoring and is suitable for components that require continuous monitoring. Point cloud scanning is used to generate three-dimensional models and is suitable for components with complex structures. Infrared thermal imaging is used to detect temperature anomalies and is suitable for components with thermal imaging requirements such as insulators. For example, conductor connection points require photos to be taken with a high-resolution camera.

[0180] On the basis of the above technical solution, optionally, determining the inspection method of each inspection component by the drone according to the inspection data collection task includes:

[0181] Determine the target collection task of each inspection component according to the inspection data collection task;

[0182] Determine the target acquisition components and acquisition parameters of the drone according to the target acquisition task;

[0183] The inspection method of the drone for each inspection component is determined according to the target collection component and the collection parameters.

[0184] In this solution, the target acquisition task may refer to specific data acquisition requirements for the inspection components, such as taking high-definition images and obtaining thermal imaging data, etc. For example, checking whether the insulator has cracks and detecting the temperature of the conductor.

[0185] The target acquisition components can be various sensors and devices used for data acquisition on drones. Specifically, they can include cameras, including high-resolution cameras, which are used to capture high-definition images and detect physical damage to equipment, such as cracks and corrosion. Thermal imaging cameras are used to detect temperature anomalies and identify hot and cold spots. LiDAR sensors can be used to generate high-precision three-dimensional point cloud data, measure distances and angles, and create three-dimensional models of power facilities. Spectral imagers are used to detect specific wavelengths of light and identify material properties and chemical composition, such as detecting contamination of insulators. Ultrasonic sensors can be used to detect the distance and position of objects to help with navigation and obstacle avoidance.

[0186] Collection parameters can be the specific settings and conditions required for the drone to complete the target collection task. Specifically, they can include shooting angle, that is, the shooting angle of the camera or sensor. The right angle helps to obtain clear and comprehensive images or data. Image resolution, that is, the resolution of the image captured by the camera. High-resolution images can capture finer details. Sensor settings, that is, the working parameters of specific sensors, such as the temperature range of thermal imaging cameras and the scanning frequency of LiDAR sensors. Collection frequency, that is, the frequency of data collection, that is, the interval between the drone collecting data during the inspection process.

[0187] The inspection data collection tasks can be reviewed to determine the data types and contents that need to be collected. According to the task requirements, the specific inspection components that need to be inspected can be determined, and specific target collection tasks can be formulated for each inspection component, which can include shooting images and obtaining sensor data. According to the target collection tasks, select the appropriate collection components on the drone, such as cameras and LiDAR sensors. And according to the requirements of the target collection tasks, set the parameters of the drone collection components, such as shooting angles and image resolution. Finally, the target collection components and corresponding collection parameters are combined to form the inspection mode of the drone for each inspection component. For example, when the drone shoots high-definition images above the cable, the selected target collection component is a camera, and the shooting parameters are a field of view of 150° or 155°; an equivalent focal length of 14.66 mm; an aperture of f / 2.8, a focus mode of FF; and a focus range of 0.6 meters.

[0188] In this solution, selecting acquisition components and setting parameters according to specific task requirements can improve resource utilization and avoid unnecessary data collection and waste. Each acquisition component is set according to specific task requirements, which can ensure that the collected data is of higher quality and more suitable for subsequent analysis and application.

[0189] S209, determining an inspection plan for the drone according to the inspection method and the inspection path, and sending the inspection plan to the drone, so that the drone can inspect the power line facilities according to the inspection plan.

[0190] It can determine the flight path of the drone from the starting point to the end point, cover all parts that need to be inspected, plan the specific coordinates and flight altitude of each inspection point, determine the drone's hovering time and data collection tasks at each inspection point, and finally generate the drone's inspection plan, which will be sent to the drone through wireless communication technology.

[0191] In this embodiment, the flight route and inspection steps are planned in advance, which can reduce the drone's stay and search time in the air and improve inspection efficiency.

[0192] Embodiment three: Figure 3 is a flow chart of an unmanned inspection method for power line facilities provided in the third embodiment of the present application, such as Figure 3 As shown, the specific method includes the following steps:

[0193] S301, obtaining a first point cloud and image data of a power line facility collected by a drone, and determining a second point cloud of the power line facility based on the image data.

[0194] S302, calculating a first centroid of the first point cloud and a second centroid of the second point cloud according to a preset centroid calculation formula; wherein the preset centroid calculation formula is:

[0195] ;

[0196] in, is the first centroid; Q is the first point cloud; is the i-th point of the first point cloud; is the total number of points in the first point cloud; is the second centroid; P is the second point cloud; is the i-th point of the second point cloud; N2 is the total number of points in the second point cloud.

[0197] S303, normalizing the first point cloud and the second point cloud according to the first centroid and the second centroid, determining the rotation matrix and the translation vector according to the normalized first point cloud and the normalized second point cloud, and transforming the normalized second point cloud according to the rotation matrix and the translation vector to obtain a third point cloud.

[0198] S304, calculating the evaluation function value between the normalized first point cloud and the third point cloud according to a preset evaluation function calculation formula, and determining whether the evaluation function value reaches a preset convergence condition; wherein the preset evaluation function calculation formula is:

[0199] ;

[0200] Where E is the evaluation function value; q is the point of the first point cloud; is the point of the third point cloud; R is the rotation matrix; T is the translation vector.

[0201] S305: If the evaluation function value reaches the preset convergence condition, a three-dimensional model of the power line facility is established based on the normalized first point cloud and the third point cloud, and an inspection plan for the UAV is determined based on the three-dimensional model and preset inspection requirements. The inspection plan is sent to the UAV so that the UAV can inspect the power line facilities according to the inspection plan.

[0202] S306, receiving the inspection data of each inspection component transmitted by the drone, and determining the data type of the inspection data.

[0203] Inspection data can be various data transmitted by drones, such as images, videos, and sensor measurement data.

[0204] The data type may be determined according to the content of the transmitted data, and specifically may include image data and non-image data.

[0205] The inspection data of each inspection component transmitted by the drone can be received through wireless communication technology, the received data can be decoded and processed to identify the data type, and then the type of data can be determined based on the format, content and characteristics of the data.

[0206] Based on the above technical solution, optionally, after determining the data type of the inspection data, the method further includes:

[0207] If the inspection data is non-image data, determining a corresponding target acquisition task according to the non-image data;

[0208] Determine a corresponding target threshold according to the target acquisition task, and determine whether there is abnormal information according to the non-image data and the corresponding target threshold;

[0209] If it is determined that there is abnormal information, the abnormal information is transmitted to the staff's handheld terminal so that the staff can inspect and repair the power line facilities based on the abnormal information.

[0210] In this solution, non-image data can be other types of data that do not contain images in the inspection data transmitted by the drone. Specifically, it can include temperature data, that is, temperature information measured by infrared sensors or thermal imaging equipment; vibration data, that is, vibration information obtained by accelerometers or vibration sensors. Current and voltage data, that is, current and voltage values ​​measured by current sensors or voltage sensors. Distance and altitude data, that is, distance or altitude information measured by laser rangefinders or ultrasonic sensors. Meteorological data can include environmental parameters such as wind speed, humidity, and air pressure.

[0211] The target threshold can be a numerical standard pre-set for a specific inspection task to determine whether there is an abnormality. For example, the temperature threshold can be a temperature range for normal operation of the device. Exceeding or falling below this range is an abnormality. The vibration threshold can be a frequency and amplitude range for normal vibration. Exceeding this range is an abnormality. The current and voltage thresholds can be normal current and voltage ranges. Exceeding this range is an abnormality.

[0212] The non-image data transmitted by the drone can be received through wireless communication technology, and the corresponding target collection task can be determined according to the type of non-image data received and the inspection task requirements. According to the target collection task, the pre-set target threshold is queried in the database. For example, for temperature data, the temperature range for normal operation of the equipment is set, such as 20°C to 80°C. The non-image data is compared with the corresponding target threshold. If the data exceeds the target threshold range, it is determined that abnormal information exists. When it is determined that there is abnormal information, the abnormal information is sent to the staff's handheld terminal through wireless communication technology.

[0213] In this solution, automatic analysis and judgment of data can quickly identify anomalies, reduce the time spent on manual analysis, and improve inspection efficiency. In addition, abnormal information can be transmitted to staff in real time, enabling them to take timely actions, reduce fault response time, and avoid potential losses.

[0214] S307: If the inspection data is image data, determine a corresponding target acquisition task according to the image data.

[0215] Image data may be a type of inspection data, specifically, may include visible light images and infrared images. When image data is identified, the target acquisition task associated with the image data may be queried at that time.

[0216] S308, determining a target recognition method according to the target acquisition task, identifying corresponding image data according to the target recognition method, and determining whether there is abnormal information.

[0217] Target recognition methods can be divided into recognition methods for visible light images and recognition methods for infrared images. Target recognition methods for visible light images may include edge detection, shape recognition, color segmentation, feature extraction, classifiers, and convolutional neural networks, etc. Target recognition methods for infrared images may include temperature threshold method, hot spot detection, infrared feature extraction and machine learning, and infrared image enhancement.

[0218] Anomaly information can be problems or anomalies detected during the object recognition process. For example, in a crack detection task, anomaly information can be the location, size, and severity of the detected cracks.

[0219] For each target acquisition task, a suitable target recognition method can be determined, such as using image processing algorithms or machine learning models for crack detection or temperature anomaly detection. The target recognition method is applied to the image data to identify abnormal information, such as the detected crack location or temperature anomaly area.

[0220] S309: If it is determined that there is abnormal information, the abnormal information is transmitted to the handheld terminal of the staff so that the staff can inspect and repair the power line facilities according to the abnormal information.

[0221] The handheld terminal may refer to a portable device carried by a staff member for receiving and viewing abnormal information, and may be a smart phone, a tablet computer or a dedicated mobile terminal device.

[0222] When abnormal information is identified, it can be transmitted to the staff's handheld terminal through wireless communication technology.

[0223] In this embodiment, the target recognition method is determined according to the target acquisition task, which can improve the accuracy and reliability of recognition. If a method is not effective, it can be supplemented or verified by other methods to improve the accuracy of recognition.

[0224] Embodiment 4: Figure 4 is a schematic diagram of the structure of an unmanned inspection device for power line facilities provided in the fourth embodiment of the present application, such as Figure 4 As shown, the device is used to implement an unmanned inspection method for power line facilities provided in embodiments 1, 2, and 3, and the device specifically includes the following:

[0225] The data acquisition module 401 is used to acquire a first point cloud and image data of a power line facility acquired by a drone, and determine a second point cloud of the power line facility according to the image data;

[0226] The centroid determination module 402 is used to calculate the first centroid of the first point cloud and the second centroid of the second point cloud according to a preset centroid calculation formula; wherein the preset centroid calculation formula is:

[0227] ;

[0228] in, is the first centroid; Q is the first point cloud; is the i-th point of the first point cloud; is the total number of points in the first point cloud; is the second centroid; P is the second point cloud; is the i-th point of the second point cloud;

[0229] A point cloud transformation module 403 is used to normalize the first point cloud and the second point cloud according to the first centroid and the second centroid, determine a rotation matrix and a translation vector according to the normalized first point cloud and the normalized second point cloud, and transform the normalized second point cloud according to the rotation matrix and the translation vector to obtain a third point cloud;

[0230] The condition determination module 404 is used to calculate the evaluation function value between the first point cloud and the third point cloud after normalization according to a preset evaluation function calculation formula, and determine whether the evaluation function value reaches a preset convergence condition; wherein the preset evaluation function calculation formula is:

[0231] ;

[0232] Where E is the evaluation function value; q is the point of the first point cloud; is the point of the third point cloud; R is the rotation matrix; T is the translation vector;

[0233] The inspection plan determination module 405 is used to establish a three-dimensional model of the power line facility based on the normalized first point cloud and the third point cloud if the evaluation function value reaches a preset convergence condition, determine the inspection plan of the drone based on the three-dimensional model and preset inspection requirements, and send the inspection plan to the drone so that the drone can inspect the power line facilities according to the inspection plan.

[0234] In an embodiment of the present application, a data acquisition module is used to obtain a first point cloud and image data of a power line facility collected by a drone, and determine a second point cloud of the power line facility according to the image data; a centroid determination module is used to calculate a first centroid of the first point cloud according to a preset centroid calculation formula, and calculate a second centroid of the second point cloud; a point cloud transformation module is used to normalize the first point cloud and the second point cloud according to the first centroid and the second centroid, determine a rotation matrix and a translation vector according to the normalized first point cloud and the normalized second point cloud, and transform the normalized second point cloud according to the rotation matrix and the translation vector. The point cloud is transformed to obtain a third point cloud; a condition judgment module is used to calculate the evaluation function value between the normalized first point cloud and the third point cloud according to a preset evaluation function calculation formula, and determine whether the evaluation function value reaches the preset convergence condition; an inspection plan determination module is used to establish a three-dimensional model of the power line facility according to the normalized first point cloud and the third point cloud if the evaluation function value reaches the preset convergence condition, determine the inspection plan of the drone according to the three-dimensional model and the preset inspection requirements, and send the inspection plan to the drone for the drone to inspect the power line facility according to the inspection plan. Through the above-mentioned unmanned inspection device for power line facilities, the three-dimensional model generated by point cloud registration can accurately reflect the actual situation of the power line facilities, thereby improving the accuracy and reliability of the inspection. According to the three-dimensional model and the preset inspection requirements, the inspection plan is automatically formulated to reduce manual intervention, reduce labor costs, and improve inspection efficiency and inspection safety.

[0235] The above are only preferred embodiments of the present application and the technical principles used. The present application is not limited to the specific embodiments described herein, and various obvious changes, readjustments and substitutions that can be made by those skilled in the art will not deviate from the scope of protection of the present application. Therefore, although the present application is described in more detail through the above embodiments, the present application is not limited to the above embodiments, and may include more other equivalent embodiments without departing from the concept of the present application, and the scope of the present application is determined by the scope of the claims.

[0236] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps recorded in this application can be executed in parallel, sequentially or in different orders, as long as the expected results of the technical solution disclosed in this application can be achieved, and this document is not limited here.

Claims

1. A method for unmanned inspection of power line facilities, characterized in that: The method is executed by a control end, and the method includes: Acquire a first point cloud and image data of a power line facility collected by a drone, and determine a second point cloud of the power line facility based on the image data; The first centroid of the first point cloud is calculated according to a preset centroid calculation formula, and the second centroid of the second point cloud is calculated; wherein the preset centroid calculation formula is: ； in, is the first centroid; Q is the first point cloud; is the i-th point of the first point cloud; is the total number of points in the first point cloud; is the second centroid; P is the second point cloud; is the i-th point of the second point cloud; is the total number of points in the second point cloud; Normalizing the first point cloud and the second point cloud according to the first centroid and the second centroid, determining a rotation matrix and a translation vector according to the normalized first point cloud and the normalized second point cloud, and transforming the normalized second point cloud according to the rotation matrix and the translation vector to obtain a third point cloud; The evaluation function value between the normalized first point cloud and the third point cloud is calculated according to a preset evaluation function calculation formula to determine whether the evaluation function value reaches a preset convergence condition; wherein the preset evaluation function calculation formula is: ； Where E is the evaluation function value; is the i-th point of the first point cloud after normalization; is the i-th point of the third point cloud; R is the rotation matrix; T is the translation vector; If the evaluation function value reaches the preset convergence condition, a three-dimensional model of the power line facility is established based on the normalized first point cloud and the third point cloud, and an inspection plan for the UAV is determined based on the three-dimensional model and preset inspection requirements. The inspection plan is sent to the UAV so that the UAV can inspect the power line facilities according to the inspection plan.

2. The unmanned inspection method for power line facilities according to claim 1, characterized in that: Normalizing the first point cloud and the second point cloud according to the first centroid and the second centroid includes: The first centroid and the second centroid are input into a preset normalization formula to perform normalization processing on each point in the first point cloud and the second point cloud; wherein the preset normalization formula is: ； in, is the i-th point of the first point cloud; is the i-th point of the first point cloud after normalization; is the first centroid; is the i-th point of the second point cloud; is the i-th point of the second point cloud after normalization; is the second centroid.

3. The unmanned inspection method for power line facilities according to claim 1, characterized in that: Determining a rotation matrix and a translation vector according to the normalized first point cloud and the normalized second point cloud includes: Determine corresponding point pairs between the normalized first point cloud and the normalized second point cloud; The first matrix is ​​determined according to the corresponding point pairs and a preset first matrix construction formula; wherein the preset first matrix construction formula is: ； Among them, C is the first matrix; N is the point pair ( , ) for The transpose of Perform SVD decomposition on the first matrix to obtain a singular value decomposition of the first matrix; wherein the process of performing SVD decomposition on the first matrix includes: ； Among them, U is the left singular vector matrix; V is the right singular vector matrix; is a diagonal matrix; is the transposed matrix of the right singular vector matrix; The rotation matrix and the translation vector are calculated according to the singular value decomposition of the first matrix; wherein the calculation process includes: 。 4. The unmanned inspection method for power line facilities according to claim 1, characterized in that: After determining whether the evaluation function value reaches a preset convergence condition, the method further includes: If the evaluation function value does not reach a preset convergence condition, determining corresponding point pairs between the normalized first point cloud and the third point cloud; Re-determine the first matrix according to the corresponding point pairs and a preset first matrix construction formula; Re-perform SVD decomposition on the first matrix to obtain the singular value decomposition of the first matrix; Recalculate the rotation matrix and the translation vector according to the singular value decomposition of the first matrix; The third point cloud is updated according to the rotation matrix and the translation vector. If the evaluation function value between the normalized first point cloud calculated according to the preset evaluation function calculation formula and the updated third point cloud still does not reach the preset convergence condition, the corresponding point pairs between the normalized first point cloud and the updated third point cloud are recalculated, and the third point cloud is continuously updated until the evaluation function value between the normalized first point cloud calculated according to the preset evaluation function calculation formula and the updated third point cloud reaches the preset convergence condition.

5. The unmanned inspection method for power line facilities according to claim 1, characterized in that: The inspection plan of the drone is determined according to the three-dimensional model and the preset inspection requirements, including: Determine the inspection area, each inspection component in the inspection area, and the inspection data collection task according to the preset inspection requirements; wherein the inspection area includes the inspection starting point and the inspection end point; Determine the inspection component position of each inspection component and the obstacle position in the inspection area according to the three-dimensional model; Determining an inspection path of the drone according to the inspection area, the location of the inspection component, and the location of the obstacle; Determine the inspection method of the drone for each inspection component according to the inspection data collection task; The inspection plan of the drone is determined according to the inspection method and the inspection path.

6. The method for unmanned inspection of power line facilities according to claim 5, characterized in that: Determining the inspection path of the drone according to the inspection area, the inspection component position, and the obstacle position includes: Identify terrain information of the inspection area, and divide the inspection area into at least two inspection layers according to the terrain information, the location of the inspection component, and the location of the obstacle; Determine the track line of the drone at each inspection layer according to the inspection starting point, the inspection component position, the obstacle position and the inspection end point; The inspection path of the drone is determined according to the track line of the drone at each inspection layer.

7. The unmanned inspection method for power line facilities according to claim 5, characterized in that: The inspection method of the drone for each inspection component is determined according to the inspection data collection task, including: Determine the target collection task of each inspection component according to the inspection data collection task; Determine the target acquisition components and acquisition parameters of the drone according to the target acquisition task; The inspection method of the drone for each inspection component is determined according to the target collection component and the collection parameters.

8. The unmanned inspection method for power line facilities according to claim 1, characterized in that: After sending the inspection plan to the drone, the method further includes: Receive inspection data of each inspection component transmitted by the drone, and determine the data type of the inspection data; If the inspection data is image data, determining a corresponding target acquisition task according to the image data; Determine a target recognition method according to the target acquisition task, recognize corresponding image data according to the target recognition method, and determine whether there is abnormal information; If it is determined that there is abnormal information, the abnormal information is transmitted to the staff's handheld terminal so that the staff can inspect and repair the power line facilities based on the abnormal information.

9. The unmanned inspection method for power line facilities according to claim 8, characterized in that: After determining the data type of the inspection data, the method further includes: If the inspection data is non-image data, determining a corresponding target acquisition task according to the non-image data; Determine a corresponding target threshold according to the target acquisition task, and determine whether there is abnormal information according to the non-image data and the corresponding target threshold; If it is determined that there is abnormal information, the abnormal information is transmitted to the staff's handheld terminal so that the staff can inspect and repair the power line facilities based on the abnormal information.

10. An unmanned inspection device for power line facilities, used to implement an unmanned inspection method for power line facilities as claimed in any one of claims 1 to 9, characterized in that: The device comprises: A data acquisition module, used to acquire a first point cloud and image data of a power line facility acquired by a drone, and determine a second point cloud of the power line facility based on the image data; The centroid determination module is used to calculate the first centroid of the first point cloud and the second centroid of the second point cloud according to a preset centroid calculation formula; wherein the preset centroid calculation formula is: ； in, is the first centroid; Q is the first point cloud; is the i-th point of the first point cloud; is the total number of points in the first point cloud; is the second centroid; P is the second point cloud; is the i-th point of the second point cloud; a point cloud transformation module, configured to normalize the first point cloud and the second point cloud according to the first centroid and the second centroid, determine a rotation matrix and a translation vector according to the normalized first point cloud and the normalized second point cloud, and transform the normalized second point cloud according to the rotation matrix and the translation vector to obtain a third point cloud; is the total number of points in the second point cloud; The condition determination module is used to calculate the evaluation function value between the first point cloud and the third point cloud after normalization according to a preset evaluation function calculation formula, and determine whether the evaluation function value reaches a preset convergence condition; wherein the preset evaluation function calculation formula is: ； Where E is the evaluation function value; is the i-th point of the first point cloud after normalization; is the i-th point of the third point cloud; R is the rotation matrix; T is the translation vector; The inspection plan determination module is used to establish a three-dimensional model of the power line facility based on the normalized first point cloud and the third point cloud if the evaluation function value reaches a preset convergence condition, determine the inspection plan of the drone based on the three-dimensional model and preset inspection requirements, and send the inspection plan to the drone so that the drone can inspect the power line facilities according to the inspection plan.

Images