Distribution network tower complete model generation method and system based on combination construction

By preprocessing and structurally decomposing the laser point cloud data of 3D power distribution network towers, an abstract template is constructed to generate a complete power distribution network tower model. This solves the problems of traditional manual modeling being time-consuming, labor-intensive, and prone to errors, and achieves efficient and accurate model construction.

CN119047130APending Publication Date: 2024-11-29GUIZHOU POWER GRID CO LTD
View PDF 0 Cites 1 Cited by

Patent Information

Application Number
CN202410971106.X
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-19
Publication Date
2024-11-29

AI Technical Summary

Technical Problem

Traditional manual construction of distribution network tower models is time-consuming, labor-intensive, and prone to errors, failing to meet the efficiency and accuracy requirements of modern power grid management.

Method used

By preprocessing the laser point cloud data of the three-dimensional power distribution network towers, structural decomposition is performed using a point cloud cutting algorithm to extract feature points and geometric attributes, an abstract template is constructed, and a complete model is generated by combining it with the actual power grid topology.

Benefits of technology

It improves modeling efficiency and accuracy, reduces the possibility of human error, meets the actual needs of power grid design and maintenance, and improves the efficiency of the power industry.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119047130A_ABST
    Figure CN119047130A_ABST
Patent Text Reader

Abstract

The invention discloses a distribution network tower complete model generation method and system based on combination construction. The method comprises the steps of preprocessing acquired three-dimensional distribution network tower laser point cloud data; performing structure decomposition on different parts of the tower in the preprocessed distribution network tower laser point cloud data by using a point cloud cutting algorithm, and extracting corresponding feature points and geometric attributes according to different decomposed structure types to obtain a structure decomposition result; constructing corresponding abstract templates according to different structure types, and assisting in constructing a single structure model based on a structure decomposition result and the corresponding abstract templates; and according to an actual power grid topological structure and design requirements, the single structure models are combined to generate a complete distribution network tower model. The complete distribution network tower model is constructed through combination of different models, the modeling efficiency and precision are improved, the possibility of human errors is reduced, and powerful support is provided for development of the power industry.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of construction of complete models of distribution network pole towers, and in particular to a method and system for generating complete models of distribution network pole towers based on combined construction. Background Art

[0002] In the field of power systems, the complete model of distribution towers plays an irreplaceable role in the safe and efficient operation, accurate planning and timely maintenance of the power grid. With the continuous growth of power demand and the continuous advancement of power grid technology, the scale of the power grid is expanding and the structure is becoming more complex. The traditional method of manually constructing distribution tower models can no longer meet the efficiency and accuracy requirements of modern power grid management. Manual modeling is not only time-consuming and labor-intensive, but also prone to errors due to human factors, which in turn affects the overall performance and safety of the power grid.

[0003] Therefore, the method of generating a complete model of distribution network towers based on combination construction has become a hot topic and trend in current research. This method uses advanced computer technology and automation tools to achieve fast and accurate modeling of distribution network towers through standardized material libraries and component libraries. With the continuous development of power grid technology and the emergence of new tower designs, the system can easily update the material library and component library to adapt to new modeling needs. Summary of the invention

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

[0005] Therefore, the present invention provides a method and system for generating a complete model of a distribution network tower based on combined construction to solve the problems of low efficiency and accuracy in constructing distribution network tower models in the prior art, and time-consuming and error-prone manual construction.

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

[0007] In a first aspect, the present invention provides a method for generating a complete model of a distribution network tower based on combined construction, comprising: preprocessing the acquired three-dimensional distribution network tower laser point cloud data; using a point cloud cutting algorithm to structurally decompose different parts of the tower in the preprocessed distribution network tower laser point cloud data, and extracting corresponding feature points and geometric attributes according to different decomposed structural types to obtain a result of structural decomposition; constructing corresponding abstract templates according to different structural types, and assisting in constructing a single structural model based on the result of the structural decomposition and the corresponding abstract template; and combining the single structural models to generate a complete distribution network tower model according to the actual power grid topology and design requirements.

[0008] As a preferred solution of the method for generating a complete model of a distribution network tower based on combined construction described in the present invention, the preprocessing of the acquired three-dimensional distribution network tower laser point cloud data includes:

[0009] De-noising the three-dimensional distribution network tower laser point cloud data using a three-dimensional Gaussian kernel function of a Gaussian filter;

[0010] According to the characteristics of the distribution network tower and the distribution of the three-dimensional distribution network tower laser point cloud data, a cube is selected as a structural element, the denoised data is processed using the corrosion operation of the morphological filter, the features in the distribution network tower laser point cloud data are enhanced using the dilation operation of the morphological filter, the distribution network tower laser point cloud data are further refined using the opening operation, and the corrosion, dilation, and opening operation steps are repeated according to different structural types, and the processed distribution network tower laser point cloud data meets the expected structural characteristics through multiple iterative processing;

[0011] The least squares method is used to align the distribution network tower laser point cloud data processed by the morphological filter. In the alignment process, a reference point cloud and a target point cloud are first selected from the distribution network tower laser point cloud data processed by the morphological filter. For each point in the reference point cloud, the distance between each point in the reference point cloud and each point in the target point cloud is calculated using the Euclidean distance. Then, a transformation matrix is ​​applied to the target point cloud to align the target point cloud with the reference point cloud, thereby obtaining the aligned three-dimensional distribution network tower laser point cloud data.

[0012] As a preferred solution of the method for generating a complete model of a distribution network tower based on combined construction described in the present invention, the structural decomposition using a point cloud cutting algorithm includes:

[0013] Using nearest neighbor search to determine the vector of each point in the registered three-dimensional distribution network tower laser point cloud data;

[0014] Calculate the curvature of each point in the registered three-dimensional distribution network pole tower laser point cloud data using Gaussian curvature according to the normal vector, and calculate the linear shape feature of each point in the registered three-dimensional distribution network pole tower laser point cloud data using eigenvalues;

[0015] Calculate the shape descriptor of each point in the registered three-dimensional distribution network tower laser point cloud data based on the curvature and the linearity shape features;

[0016] According to the differences in the shape descriptors, a clustering algorithm is used to divide the registered three-dimensional distribution network tower laser point cloud data into different clusters, and the volume and surface area of ​​each cluster are calculated;

[0017] The ratio of the volume to the surface area of ​​each cluster is compared with a maximum threshold and a minimum threshold set in advance. Clusters with a volume greater than the maximum threshold are judged to be columnar structures, clusters with a volume less than the minimum threshold are judged to be inverted triangular pyramid structures, and clusters between the minimum threshold and the maximum threshold are judged to be quadrangular pyramid structures.

[0018] As a preferred solution of the method for generating a complete model of a distribution network tower based on combined construction described in the present invention, for a cluster determined to be a columnar structure, extracting the height and diameter of the columnar structure includes:

[0019] Determining the height of the columnar structure by calculating the difference between the maximum and minimum Z coordinates in the point cloud data of the cluster of the columnar structure;

[0020] The diameter of the bottom of the columnar structure is obtained by projecting the coordinates of the bottom point cloud data of the cluster of the columnar structure onto the XY plane, calculating the distance between each point in the projected point cloud data and the origin (0,0), and adding the maximum distance and the minimum distance.

[0021] As a preferred solution of the method for generating a complete model of a distribution network tower based on combined construction described in the present invention, for a cluster determined to be an inverted triangular pyramid structure, extracting the vertex coordinates and angles of the inverted triangular pyramid structure includes:

[0022] Traversing all point cloud data in the cluster of the inverted triangular pyramid structure, for each point i, using the Euclidean distance formula to calculate the distance dij between point i and other points j, accumulating the calculated distances and selecting the coordinates of the minimum point from the distance sum as the vertex coordinates of the inverted triangular pyramid structure;

[0023] The principal component analysis is used to perform plane fitting on the bottom surface point cloud data in the cluster of the inverted triangular pyramid structure to obtain the normal vector of the fitting plane, and the dot product operation is performed on the normal vector of the fitting plane and the unit vector [0,0,1] of the Z axis to obtain the cosine value of the angle between the two vectors. The inverse cosine function is used to convert the cosine value of the angle into the angle value of the inverted triangular pyramid structure.

[0024] As a preferred solution of the method for generating a complete model of a distribution network tower based on combined construction described in the present invention, for a cluster determined to be a tetrahedral structure, extracting the normal vectors and vertex coordinates of each face of the tetrahedral structure includes:

[0025] The steps of extracting the vertex coordinates of the tetrahedral structure are:

[0026] Collect the point cloud data of the cluster of the tetrahedral structure, assuming that there are n points, the coordinates of each point are (x_i, y_i, z_i), set 0.3 times the average distance as the distance threshold, and calculate the Euclidean distance between each point (x_i, y_i, z_i) and all other points;

[0027] For each point (x_i, y_i, z_i), count the number of points whose distance is less than the distance threshold. If the number of points whose distance is less than the distance threshold is greater than or equal to the set distance threshold, it means that there are enough neighboring points around the point, that is, the density is high, and the point is considered to be one of the vertex candidates;

[0028] Select the point with the largest number of candidates as the vertex, and repeat the above judgment steps until all vertices are found;

[0029] The steps of extracting the normal vectors of each face of the tetrahedral structure are:

[0030] Collecting point cloud data of the bottom surface of the cluster of the tetrahedral structure, a point set P = {(x_i, y_i, z_i)}, where each point represents a point on the bottom surface;

[0031] The principal component analysis is used to perform plane fitting on the point set P to obtain a normal vector of the fitting plane, and the normal vector κ=(x, y, z) of the fitting plane is normalized to obtain a normalized normal vector κ′=(x′, y′, z′).

[0032] As a preferred solution of the method for generating a complete distribution network tower model based on combined construction described in the present invention, the generation of the complete distribution network tower model includes:

[0033] For each structure type, define the topological properties of the corresponding structure, the topological properties include the connection mode, support mode and material of the structure, use computer CAD software to build an abstract template according to the columnar structure, inverted triangular pyramid structure and quadrangular pyramid structure, and assist in building a single columnar model, inverted triangular pyramid model and quadrangular pyramid model based on the result of the structure decomposition and the corresponding abstract template;

[0034] Acquire the actual grid topology structure, and combine the single columnar model, inverted triangular pyramid model, and quadrangular pyramid model according to the combination rule of tower head + tower body + tower base according to the actual grid topology structure and design requirements to generate a distribution network tower model framework;

[0035] Add conductors and equipment, including crossarms, insulators, and transformers, to the distribution network tower model framework; determine the position, length, and angle of the conductors according to actual power transmission requirements; and connect them to the corresponding distribution network tower model framework to generate a complete distribution network tower model.

[0036] In a second aspect, the present invention provides a distribution network tower complete model generation system based on combination construction, comprising:

[0037] A data processing unit, used for preprocessing the acquired three-dimensional distribution network tower laser point cloud data;

[0038] A structural decomposition unit is used to structurally decompose different parts of the tower in the pre-processed distribution network tower laser point cloud data using a point cloud cutting algorithm, and extract corresponding feature points and geometric attributes according to different decomposed structural types to obtain a structural decomposition result;

[0039] A single model building unit is used to build corresponding abstract templates according to different structural types, and to assist in building a single structural model based on the result of the structural decomposition and the corresponding abstract template;

[0040] The complete model generation unit is used to combine the individual structural models to generate a complete distribution network tower model according to the actual power grid topology and design requirements.

[0041] In a third aspect, the present invention provides a computing device, comprising:

[0042] Memory and processor;

[0043] The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, the steps of the method are implemented.

[0044] In a fourth aspect, the present invention provides a computer-readable storage medium storing computer-executable instructions, which implement the steps of the method when executed by a processor.

[0045] Compared with the prior art, the beneficial effects of the present invention are as follows: the present invention decomposes different parts of the distribution network tower in the laser point cloud data into a columnar structure, an inverted triangular pyramid structure, and a tetrahedral pyramid structure, and then constructs a columnar model, an inverted triangular pyramid model, and a tetrahedral pyramid model with the assistance of an abstract template, and finally constructs a complete distribution network tower model through a combination of different models, thereby improving modeling efficiency and accuracy, reducing the possibility of human errors, meeting actual needs, and improving the design, construction, and maintenance efficiency of distribution network towers, thereby providing strong support for the development of the power industry. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0047] Figure 1 The present invention is an overall flow chart of a method for generating a complete distribution network tower model based on combined construction according to an embodiment of the present invention. DETAILED DESCRIPTION

[0048] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without creative work should fall within the scope of protection of the present invention.

[0049] Example 1

[0050] Reference Figure 1 According to an embodiment of the present invention, a method for generating a complete model of a distribution network tower based on combined construction is provided, which specifically comprises the following steps:

[0051] S100: pre-processing the acquired three-dimensional distribution network tower laser point cloud data;

[0052] In a feasible implementation, the laser point cloud data of the distribution network tower is obtained by a laser scanner. The laser scanner uses laser radar technology to obtain the three-dimensional coordinate information of the object surface by emitting laser pulses to the target object and measuring the time of reflection. Specifically, the laser scanner rotates in the vertical direction and emits laser pulses in all directions at the same time. Each laser pulse will be reflected on the surface of the object and return to the laser scanner. The laser scanner measures the round-trip time of each laser pulse, combined with the rotation angle and distance information of the scanner, to obtain the three-dimensional coordinates of the object surface and form point cloud data. The calculation process and principle are common knowledge known to those skilled in the art, so they will not be described in detail here. In addition, the laser scanner can be installed on a drone or other mobile platform to perform high-altitude or ground scanning of the distribution network tower to obtain complete point cloud data;

[0053] In the embodiment of the present application, the specific steps of preprocessing the acquired three-dimensional distribution network tower laser point cloud data include:

[0054] ① For each point P (x, y, z) in the distribution network tower laser point cloud data, the three-dimensional Gaussian kernel function of Gaussian filtering is used to denoise the three-dimensional distribution network tower laser point cloud data. The formula is expressed as:

[0055]

[0056] Among them, σ represents the standard deviation of the three-dimensional Gaussian function, which determines the smoothness of the filter;

[0057] ②Use morphological filters to enhance the image of the denoised three-dimensional distribution network tower laser point cloud data, specifically:

[0058] According to the characteristics of the distribution network tower and the distribution of the 3D distribution network tower laser point cloud data, the cube is selected as the structural element to adapt to the structure of the tower;

[0059] The denoised data is processed using the corrosion operation of the morphological filter. The corrosion operation can eliminate small particle noise in the data and smooth surface details. By comparing each point with the value of its neighboring points and taking the smaller value, the outliers caused by noise can be removed. The calculation formula of the corrosion operation is:

[0060] Erosion(A,B)=(AB)∩B

[0061] Among them, A represents the input denoised distribution network tower laser point cloud data, and B represents the structural element;

[0062] The expansion operation of the morphological filter is used to enhance the features in the laser point cloud data of the distribution network tower and highlight the structural features of the tower. By comparing the value of each point with its neighboring points and taking the larger value, the edge and protrusion of the tower can be highlighted. The calculation formula of the expansion operation is:

[0063] Dilation(A,B)=(A∪B)-(Erosion(A,B))

[0064] The opening operation is used to further refine the distribution network tower laser point cloud data. The calculation formula of the opening operation is:

[0065] Opening(A,B)=Erosion(A,B)∪(BA)

[0066] According to different structural types, including columnar structure, inverted triangular pyramid structure, and quadrangular pyramid structure, the above corrosion, expansion, and opening operation steps are repeated. Through multiple iterative processes, the processed distribution network tower laser point cloud data meets the expected structural characteristics and can clearly distinguish different structural types.

[0067] ③ The least squares method is used to register the distribution network tower laser point cloud data processed by the morphological filter. In the registration process, the reference point cloud and the target point cloud are first selected from the distribution network tower laser point cloud data processed by the morphological filter. The reference point cloud is a point cloud with known position and direction, and the target point cloud is a point cloud that needs to be aligned with the reference point cloud. For each point in the reference point cloud, the Euclidean distance is used to calculate the distance between each point in the reference point cloud and each point in the target point cloud. The Euclidean distance calculation formula is:

[0068] d = sqrt((x2-x1) 2 +(y2-y1) 2 +(z2-z1) 2 )

[0069] Among them, (x1, y1, z1) represents the coordinates of the reference point cloud, (x2, y2, z2) represents the coordinates of the target point cloud, and d represents the Euclidean distance between two points. According to the calculated distance, an original distance matrix C is constructed, and the least squares method is used to optimize the original distance matrix C;

[0070] The calculation formula of the least squares method is:

[0071] w=(C^TC)^{-1}C^TN

[0072] Where C represents the original distance matrix, C^T represents the transpose of the original distance matrix C, (C^TC) represents the product of the original matrix C transposed and itself, (C^TC)^{-1} represents the inverse matrix of the matrix (C^TC), N represents the vector from the target point to the reference point, and w represents the transformation matrix to be solved;

[0073] ④ Apply the transformation matrix to the target point cloud to align the target point cloud with the reference point cloud to obtain the registered three-dimensional distribution network tower laser point cloud data.

[0074] It should be noted that the acquired three-dimensional distribution network tower laser point cloud data is preprocessed, including denoising, image enhancement and registration, to obtain high-quality distribution network tower laser point cloud data. This step is to improve the accuracy and availability of the data and provide a high-quality data foundation for subsequent model construction.

[0075] S200: using a point cloud cutting algorithm to structurally decompose different parts of the tower in the pre-processed distribution network tower laser point cloud data, and extracting corresponding feature points and geometric attributes according to different decomposed structural types to obtain a result of structural decomposition;

[0076] In the embodiment of the present application, the nearest neighbor search is used to determine the vector of each point in the registered three-dimensional distribution network tower laser point cloud data. The specific formula of the nearest neighbor search is:

[0077]

[0078] eigvals,eigvecs=np.linalg.eig(C)

[0079] Wherein, C represents the covariance matrix, k represents the number of points in the neighborhood, which is used to perform weighted averaging on the calculation results, q represents the points in the neighborhood, p represents the center point, (qp) represents the vector between the points in the neighborhood and the center point, (qp)^T represents the transpose of the vector (qp), (qp)*(qp)^T represents a two-dimensional matrix constructed with the vector (qp) as the column vector, eigvals represents a one-dimensional array containing the eigenvalues ​​of C, eigvecs represents a two-dimensional array containing the eigenvectors of C, each column corresponds to an eigenvector, the eigenvector corresponding to the minimum eigenvalue of C is selected as the normal vector of the point, and the eigenvector corresponding to the largest eigenvalue among the eigenvalues ​​is selected as the tangent vector;

[0080] In the embodiment of the present application, the curvature of each point in the registered three-dimensional distribution network tower laser point cloud data is calculated using Gaussian curvature according to the normal vector. The curvature can reflect the curvature of the point cloud surface. The calculation formula of Gaussian curvature is:

[0081]

[0082] Where det(C) represents the determinant of the covariance matrix C, eigvals[0] and eigvecs[1] represent the two eigenvalues ​​of C respectively, and G represents the Gaussian curvature index of point p, which is used to reflect the curvature of the point cloud surface at that point. The larger the curvature value, the more curved the surface near the point, and the smaller the curvature value, the flatter the surface near the point.

[0083] In the embodiment of the present application, the linear shape feature of each point in the laser point cloud data of the three-dimensional distribution network tower after registration is calculated using the eigenvalue. The linearity can reflect the straightness of the point cloud surface. The linearity calculation formula is:

[0084] Linearity = min(eigvals)

[0085] Among them, min(eigvals) represents the minimum value in the eigenvalue array eigvals. By calculating the linearity of each point, the straightness information of the point cloud surface can be obtained. The closer the linearity is to 0, the closer the point cloud surface is to a straight line. Conversely, the larger the linearity is, the stronger the curve of the point cloud surface is.

[0086] In an embodiment of the present application, a shape descriptor of each point in the registered three-dimensional distribution network tower laser point cloud data is calculated based on the curvature and linearity shape features;

[0087] Specifically, a normal vector and a set of tangent vectors are used to represent the local reference frame so as to convert the shape descriptor into a descriptor in the local coordinate system. Based on the local reference frame, the curvature and linearity shape features are encoded by the histogram to calculate the PFH shape descriptor. The calculation formula of the PFH shape descriptor is:

[0088]

[0089] Among them, k represents the number of neighborhood points around point p, n i ,n j represents the normal vector of the i-th and j-th neighboring points, t i ,t j represents the tangent vector of the i-th and j-th neighboring points at p, h(n i ,n j ,t i ,t j ) represents a function for calculating the relative angle between the i-th and j-th neighborhood points and the angle difference between the normal vectors. Finally, PFH returns a vector of a fixed size to represent the shape descriptor of point p, and uses the PFH descriptor of each point as its feature vector;

[0090] In the embodiment of the present application, a clustering algorithm is used to divide the registered three-dimensional distribution network tower laser point cloud data into different clusters according to the differences in shape descriptors;

[0091] Specifically, the number of clusters selected is 3, which are the tower head, tower body and tower base; L points are randomly selected from the registered distribution network tower laser point cloud data as the initial cluster centers; for each sample point, the Euclidean distance formula is used to calculate its distance from all cluster centers and assign it to the cluster center with the nearest distance. The Euclidean distance formula is:

[0092]

[0093] Among them, n represents the dimension of the feature vector, x i and i are the i-th elements of the eigenvectors x and y respectively;

[0094] For each cluster center, calculate the mean of all sample points in the cluster and use the mean as the new cluster center. The mean calculation formula is:

[0095]

[0096] in, It means summing all sample points in the cluster.

[0097] Repeat the above steps until the stopping condition is reached, and the stopping condition is the maximum number of iterations 12; get the final clustering result, each point is assigned to a cluster, and according to the clustering result, each point is marked as a tower head, tower body or tower base cluster;

[0098] In the embodiment of the present application, the volume and surface area of ​​each cluster are calculated according to the clustering results, and further shape analysis is performed to distinguish different structural types, which are divided into columnar structure, inverted triangular pyramid structure, and quadrangular pyramid structure. For each cluster, its bounding box is calculated, that is, the smallest cube or cuboid surrounding all points of the cluster, and then the volume of the cluster is obtained by calculating the length, width, and height of the bounding box. The volume calculation formula is:

[0099] V=L*W*H

[0100] Among them, L, W, and H represent the length, width, and height of the bounding box respectively;

[0101] Use the triangular mesh reconstruction algorithm in the point cloud data to convert the point cloud data of each cluster into a triangular mesh model. For the triangular mesh model of each cluster, calculate its surface area. The surface area calculation formula is:

[0102]

[0103] Among them, A represents the number of vertices of the triangular mesh model, E represents the number of edges, and F represents the number of faces;

[0104] In the embodiment of the present application, the ratio of the volume to the surface area of ​​each cluster is compared with a maximum threshold and a minimum threshold set in advance, and a cluster greater than the maximum threshold is determined to be a columnar structure, a cluster less than the minimum threshold is determined to be an inverted triangular pyramid structure, and a cluster between the minimum threshold and the maximum threshold is determined to be a quadrangular pyramid structure;

[0105] It should be noted that for columnar structures, due to their relatively regular shapes, the ratio of volume to surface area is usually higher, for inverted triangular pyramid structures, the ratio of volume to surface area is usually lower, and for quadrangular pyramid structures, the ratio of volume to surface area is between columnar structures and inverted triangular pyramid structures. In addition, the value of the threshold is obtained by statistically analyzing the point cloud data of known types of structures to obtain the range of the ratio of volume to surface area of ​​different types of structures, and then this range is used as the threshold reference value for judgment during the classification process;

[0106] Furthermore, for a cluster determined to be a columnar structure, extracting the height and diameter of the columnar structure includes:

[0107] The height of the columnar structure is determined by calculating the difference between the maximum and minimum Z coordinates in the point cloud data of the cluster of the columnar structure. The height calculation formula is:

[0108] Height = Z max-Z min

[0109] The coordinates of the bottom point cloud data of the columnar cluster are projected onto the XY plane, and the distance between each point in the projected point cloud data and the origin (0,0) is calculated. The maximum distance and the minimum distance are added together to obtain the diameter of the bottom of the columnar structure. The diameter calculation formula is:

[0110] Diameter = Q max +Q min

[0111] Further, for a cluster determined to be an inverted triangular pyramid structure, extracting the vertex coordinates of the inverted triangular pyramid structure includes:

[0112] Traverse all point cloud data in the cluster of the inverted triangular pyramid structure, and for each point i, use the Euclidean distance formula to calculate the distance dij between point i and other points j. The Euclidean distance formula is:

[0113]

[0114] Among them, (xi,yi,zi) represents the coordinates of point i, (xj,yj,zj) represents the coordinates of point j;

[0115] For each point i, add up the sum of its distances to all other points j to get the "distance sum" si of the point. The calculation formula for the "distance sum" si is:

[0116] si = ∑dij, where j! = i

[0117] Among them, j!=i means that j is not equal to i, which is used to ensure that when calculating the distance between point i and other points j, the case where the distance to itself is zero is excluded;

[0118] Select the coordinates of the minimum point from the sum of the distances as the vertex coordinates of the inverted triangular pyramid structure;

[0119] Furthermore, for a cluster determined to be an inverted triangular pyramid structure, the angle of extracting the inverted triangular pyramid structure includes:

[0120] The principal component analysis is used to perform plane fitting on the bottom surface point cloud data of the inverted triangular pyramid structure cluster to obtain the normal vector of the fitting plane.

[0121] Specifically, collect the bottom surface point cloud data in the cluster. Assuming there are n points, the coordinates of each point are (x_i, y_i, z_i), and calculate the centroid of the point cloud data. The calculation formula is:

[0122]

[0123] Subtract the centroid from the coordinates of each point in the point cloud data to obtain the decentralized point cloud coordinates. The calculation formula is:

[0124] x_i′=x_i-x_c

[0125] y_i′=y_i-y_c

[0126] z_i′=z_i-z_c

[0127] Construct the covariance matrix:

[0128]

[0129] Perform eigenvalue decomposition on the covariance matrix C to obtain eigenvalues ​​and eigenvectors. The eigenvector corresponding to the minimum eigenvalue in the eigenvector is used as the normal vector of the fitting plane, denoted as [nx, ny, nz];

[0130] Perform a dot product operation on the normal vector of the fitting plane and the unit vector [0,0,1] of the Z axis to obtain the cosine value of the angle between the two vectors. The formula for the cosine value of the angle is:

[0131] cosine=dot(n,[0,0,1])=nx*0+ny*0+nz*1=nz

[0132] Use the arc cosine function acos to convert the cosine value of the angle into an angle value. The calculation formula is:

[0133] angle = acos(cosine)

[0134] Wherein, angle represents the angle between the normal vector of the bottom surface of the inverted triangular pyramid and the Z axis, and the unit of the angle is radians or degrees;

[0135] Furthermore, for the cluster determined to be a tetrahedral structure, the normal vectors and vertex coordinates of each face of the tetrahedral structure are extracted, including:

[0136] The steps to extract the vertex coordinates of the tetrahedral structure are:

[0137] Collect the point cloud data of the cluster of tetrahedral structures. Assume that there are n points, and the coordinates of each point are (x_i, y_i, z_i). Set 0.3 times the average distance as the distance threshold, and calculate the Euclidean distance between each point (x_i, y_i, z_i) and all other points. The calculation formula is:

[0138] distance = sqrt((x_i - x_j) ^2 +(y_i-y_j) ^2 +(z_i-z_j) ^2 )

[0139] Among them, j≠i;

[0140] For each point (x_i, y_i, z_i), count the number of points whose distance is less than the distance threshold. If the number of points whose distance is less than the distance threshold is greater than or equal to the set distance threshold, it means that there are enough neighboring points around the point, that is, the density is high, and the point is considered to be one of the vertex candidates. By counting the number of points whose distance is less than the threshold, it is possible to better distinguish between local dense areas and sparse areas, and screen out vertex candidates with a certain neighborhood density, thereby achieving more accurate judgment of vertices in point cloud data and improving the accuracy of vertex estimation;

[0141] Select the point with the largest number of candidates as the vertex, and repeat the above judgment steps until all vertices are found;

[0142] The steps to extract the normal vectors of each face of the tetrahedral structure are:

[0143] Collect the point cloud data of the bottom surface of the cluster of the tetrahedral structure, and the point set P = {(x_i, y_i, z_i)}, where each point represents a point on the bottom surface;

[0144] The principal component analysis is used to perform plane fitting on the point set P to obtain the normal vector of the fitting plane, which includes:

[0145] Calculate the centroid of set P, also known as the center point, using the following formula:

[0146]

[0147] Where n represents the size of the point set P;

[0148] Calculate the decentralized point set P′, the calculation formula is:

[0149]

[0150] Construct the covariance matrix C, the calculation formula is:

[0151]

[0152] By performing eigenvalue decomposition on the covariance matrix C, we can obtain the eigenvalue λ and the corresponding eigenvector δ. The calculation formula is:

[0153] C*δ=λ*δ

[0154] Select the eigenvector δ corresponding to the smallest eigenvalue as the required normal vector κ = (x, y, z);

[0155] Normalize the normal vector κ = (x, y, z) of the fitting plane to obtain the normalized normal vector κ′ = (x′, y′, z′), which is:

[0156] Calculate the length of the normal vector using the formula:

[0157]

[0158] Among them, x represents the component of the vector on the x-axis, y represents the component of the vector on the y-axis, and z represents the component of the vector on the z-axis;

[0159] Normalize by dividing by the length, and the calculation formula is:

[0160]

[0161]

[0162] The normalized normal vector κ′=(x′, y′, z′) is obtained.

[0163] It should be explained that the point cloud cutting algorithm is used to decompose the different parts of the tower in the pre-processed distribution network tower laser point cloud data, including the tower head, tower body and tower base, according to the columnar structure, inverted triangular pyramid structure and tetrahedron structure. This step is to decompose the complex distribution network tower into simple and easy-to-process substructures to facilitate the subsequent model construction. Then, according to different structural types, the corresponding feature points and geometric properties of different parts of the tower in the decomposed distribution network tower laser point cloud data are extracted to obtain the result of structural decomposition. This step is to extract key feature points and geometric properties to provide necessary information for subsequent model construction.

[0164] S300: construct corresponding abstract templates according to different structural types, and assist in constructing a single structural model based on the result of structural decomposition and the corresponding abstract templates;

[0165] In the embodiment of the present application, for each structure type, the topological attributes of the corresponding structure are defined, wherein the topological attributes include the connection method, support method and material of the structure, and computer CAD software is used to construct an abstract template according to the columnar structure, the inverted triangular pyramid structure and the quadrangular pyramid structure, and a single columnar model, inverted triangular pyramid model and quadrangular pyramid model are constructed based on the result of the structural decomposition and the corresponding abstract template;

[0166] It should be explained that, according to different structural types, corresponding abstract templates are constructed. This step is to use abstract templates to assist in the construction of various models, making model construction more efficient and accurate. According to the results of structural decomposition and the corresponding abstract templates, the feature points and geometric properties of different parts of the tower in the distribution network tower laser point cloud data are used to construct a single columnar model, inverted triangular pyramid model, and tetrahedral pyramid model. This step is to use abstract templates to construct various models based on the extracted feature points and geometric properties to achieve preliminary construction of the model.

[0167] S400: Combine individual structural models to generate a complete distribution network tower model based on the actual grid topology and design requirements;

[0168] In the embodiment of the present application, the actual grid topology is obtained, and according to the actual grid topology and design requirements, a single columnar model, an inverted triangular pyramid model, and a quadrangular pyramid model are combined according to the combination rule of tower head + tower body + tower base to generate a distribution network tower model framework;

[0169] Add conductors and equipment, including crossarms, insulators, and transformers, to the distribution network tower model framework. Determine the position, length, and angle of the conductors based on actual power transmission requirements, and connect them to the corresponding distribution network tower model framework to generate a complete distribution network tower model.

[0170] It should be noted that according to the actual grid topology and design requirements, the single columnar model, inverted triangular pyramid model and tetrahedron model are combined to generate a complete distribution network tower model. This step is to combine the single models to form a complete distribution network tower model to meet the actual application requirements.

[0171] Therefore, it can be seen from the above technical scheme that the present invention decomposes the different parts of the distribution network tower in the laser point cloud data according to the columnar structure, inverted triangular pyramid structure and tetrahedral pyramid structure, and then constructs the columnar model, inverted triangular pyramid model and tetrahedral pyramid model with the assistance of abstract templates. Finally, a complete distribution network tower model is constructed by combining different models, thereby improving the modeling efficiency and accuracy, reducing the possibility of human errors, meeting actual needs, and improving the design, construction and maintenance efficiency of distribution network towers, providing strong support for the development of the power industry.

[0172] The above is a schematic scheme of a method for generating a complete model of a distribution network tower based on combined construction in this embodiment. It should be noted that the technical scheme of the system for generating a complete model of a distribution network tower based on combined construction and the technical scheme of the method for generating a complete model of a distribution network tower based on combined construction belong to the same concept. For details not described in detail in the technical scheme of the system for generating a complete model of a distribution network tower based on combined construction in this embodiment, please refer to the description of the technical scheme of the method for generating a complete model of a distribution network tower based on combined construction.

[0173] This embodiment also provides a distribution network tower complete model generation system based on combination construction, including:

[0174] A data processing unit, used for preprocessing the acquired three-dimensional distribution network tower laser point cloud data;

[0175] The structural decomposition unit is used to structurally decompose different parts of the tower in the pre-processed distribution network tower laser point cloud data using a point cloud cutting algorithm, and extract corresponding feature points and geometric attributes according to different decomposed structural types to obtain the result of structural decomposition;

[0176] A single model building unit is used to build corresponding abstract templates according to different structural types, and to assist in building a single structural model based on the results of structural decomposition and the corresponding abstract templates;

[0177] The complete model generation unit is used to combine individual structural models to generate a complete distribution network tower model according to the actual power grid topology and design requirements.

[0178] This embodiment also provides a computing device, including:

[0179] Memory and processor;

[0180] The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the method proposed in the above embodiment.

[0181] This embodiment also provides a storage medium on which a computer program is stored. When the program is executed by a processor, the method proposed in the above embodiment is implemented.

[0182] The storage medium proposed in this embodiment and the method proposed in the above embodiment belong to the same inventive concept. The technical details not fully described in this embodiment can be referred to the above embodiment, and this embodiment has the same beneficial effects as the above embodiment.

[0183] Through the above description of the implementation methods, the technicians in the relevant field can clearly understand that the present invention can be implemented by means of software and necessary general hardware, and of course can also be implemented by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory (ReadOnly, Memory, ROM), random access memory (RandomAccess Memory, RAM), flash memory (FLASH), hard disk or optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, server, or network device, etc.) to perform the methods of various embodiments of the present invention.

[0184] Example 2

[0185] Based on the previous embodiment, this embodiment provides an application example of a method and system for generating a complete model of a distribution network tower based on combined construction, in order to verify and illustrate the technical effects adopted in this method.

[0186] Firstly, the acquired three-dimensional distribution network tower laser point cloud data is preprocessed, that is, the three-dimensional Gaussian kernel function of Gaussian filtering is used to denoise the three-dimensional distribution network tower laser point cloud data; the morphological filter is used to enhance the image of the denoised distribution network tower laser point cloud data; the least squares method is used to register the distribution network tower laser point cloud data processed by the morphological filter to obtain the registered three-dimensional distribution network tower laser point cloud data;

[0187] Then, the point cloud cutting algorithm is used to decompose the different parts of the tower in the preprocessed distribution network tower laser point cloud data, and the corresponding feature points and geometric attributes are extracted according to the different decomposed structure types to obtain the result of structural decomposition. Specifically, the nearest neighbor search is used to determine the vector of each point in the registered three-dimensional distribution network tower laser point cloud data; the curvature of each point in the registered three-dimensional distribution network tower laser point cloud data is calculated using Gaussian curvature according to the normal vector, and the linear shape features of each point in the registered three-dimensional distribution network tower laser point cloud data are calculated using the eigenvalue; based on the curvature and line The shape descriptor of each point in the registered three-dimensional distribution network tower laser point cloud data is calculated based on the shape feature; according to the difference in shape descriptors, the clustering algorithm is used to divide the registered three-dimensional distribution network tower laser point cloud data into different clusters, and the volume and surface area of ​​each cluster are calculated; the ratio of the volume and surface area of ​​each cluster is compared with the maximum and minimum thresholds set in advance, and the clusters greater than the maximum threshold are judged as columnar structures, the clusters less than the minimum threshold are judged as inverted triangular pyramid structures, and the clusters between the minimum and maximum thresholds are judged as quadrangular pyramid structures. The specific data are shown in Table 1.

[0188] Table 1: Structural type determination table.

[0189]

[0190] Finally, corresponding abstract templates are constructed according to different structural types, and a single structural model is constructed based on the results of structural decomposition and the corresponding abstract templates. According to the actual power grid topology and design requirements, the single structural models are combined to generate a complete distribution network tower model, which greatly improves the modeling efficiency and accuracy. The specific experimental data are shown in Table 2.

[0191] Table 2: Experimental data table.

[0192]

[0193] It can be seen from the data in Table 2 that the method provided by our invention can achieve an accuracy rate of more than 98% in judging the structure type and the processing time is short. Compared with the prior art, this embodiment decomposes the different parts of the tower in the laser point cloud data of the distribution network tower according to the columnar structure, inverted triangular pyramid structure, and tetrahedral pyramid structure, and then constructs the columnar model, inverted triangular pyramid model, and tetrahedral pyramid model with the assistance of abstract templates. Finally, a complete distribution network tower model is constructed through the combination of different models, so as to improve the modeling efficiency and accuracy, reduce the possibility of human errors, meet actual needs, and improve the design, construction and maintenance efficiency of distribution network towers, providing strong support for the development of the power industry.

[0194] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A method for generating a complete model of a distribution network tower based on combined construction, characterized in that: include: Preprocess the acquired three-dimensional distribution network tower laser point cloud data; Using a point cloud cutting algorithm to structurally decompose different parts of the tower in the preprocessed distribution network tower laser point cloud data, and extracting corresponding feature points and geometric attributes according to different decomposed structural types to obtain a result of structural decomposition; Construct corresponding abstract templates according to different structural types, and assist in constructing a single structural model based on the result of the structural decomposition and the corresponding abstract templates; According to the actual grid topology and design requirements, the individual structural models are combined to generate a complete distribution network tower model.

2. The method for generating a complete model of a distribution network tower based on combined construction according to claim 1, characterized in that: The preprocessing of the acquired three-dimensional distribution network tower laser point cloud data includes: De-noising the three-dimensional distribution network tower laser point cloud data using a three-dimensional Gaussian kernel function of a Gaussian filter; According to the characteristics of the distribution network tower and the distribution of the three-dimensional distribution network tower laser point cloud data, a cube is selected as a structural element, the denoised data is processed using the corrosion operation of the morphological filter, the features in the distribution network tower laser point cloud data are enhanced using the dilation operation of the morphological filter, the distribution network tower laser point cloud data are further refined using the opening operation, and the corrosion, dilation, and opening operation steps are repeated according to different structural types, and the processed distribution network tower laser point cloud data meets the expected structural characteristics through multiple iterative processing; The least squares method is used to align the distribution network tower laser point cloud data processed by the morphological filter. In the alignment process, a reference point cloud and a target point cloud are first selected from the distribution network tower laser point cloud data processed by the morphological filter. For each point in the reference point cloud, the distance between each point in the reference point cloud and each point in the target point cloud is calculated using the Euclidean distance. Then, a transformation matrix is ​​applied to the target point cloud to align the target point cloud with the reference point cloud, thereby obtaining the aligned three-dimensional distribution network tower laser point cloud data.

3. The method for generating a complete model of a distribution network tower based on combined construction according to claim 2, characterized in that: The structure decomposition using point cloud cutting algorithm includes: Using nearest neighbor search to determine the vector of each point in the registered three-dimensional distribution network tower laser point cloud data; Calculate the curvature of each point in the registered three-dimensional distribution network pole tower laser point cloud data using Gaussian curvature according to the normal vector, and calculate the linear shape feature of each point in the registered three-dimensional distribution network pole tower laser point cloud data using eigenvalues; Calculate the shape descriptor of each point in the registered three-dimensional distribution network tower laser point cloud data based on the curvature and the linearity shape features; According to the differences in the shape descriptors, a clustering algorithm is used to divide the registered three-dimensional distribution network tower laser point cloud data into different clusters, and the volume and surface area of ​​each cluster are calculated; The ratio of the volume to the surface area of ​​each cluster is compared with a maximum threshold and a minimum threshold set in advance. Clusters with a volume greater than the maximum threshold are judged to be columnar structures, clusters with a volume less than the minimum threshold are judged to be inverted triangular pyramid structures, and clusters between the minimum threshold and the maximum threshold are judged to be quadrangular pyramid structures.

4. The method for generating a complete model of a distribution network tower based on combined construction according to claim 3, characterized in that: For a cluster determined to be a columnar structure, extracting the height and diameter of the columnar structure includes: Determining the height of the columnar structure by calculating the difference between the maximum and minimum Z coordinates in the point cloud data of the cluster of the columnar structure; The diameter of the bottom of the columnar structure is obtained by projecting the coordinates of the bottom point cloud data of the cluster of the columnar structure onto the XY plane, calculating the distance between each point in the projected point cloud data and the origin (0,0), and adding the maximum distance and the minimum distance.

5. The method for generating a complete model of a distribution network tower based on combined construction according to claim 3, characterized in that: For a cluster determined to be an inverted triangular pyramid structure, extracting the vertex coordinates and angles of the inverted triangular pyramid structure includes: Traversing all point cloud data in the cluster of the inverted triangular pyramid structure, for each point i, using the Euclidean distance formula to calculate the distance dij between point i and other points j, accumulating the calculated distances and selecting the coordinates of the minimum point from the distance sum as the vertex coordinates of the inverted triangular pyramid structure; The principal component analysis is used to perform plane fitting on the bottom surface point cloud data in the cluster of the inverted triangular pyramid structure to obtain the normal vector of the fitting plane, and the dot product operation is performed on the normal vector of the fitting plane and the unit vector [0,0,1] of the Z axis to obtain the cosine value of the angle between the two vectors. The inverse cosine function is used to convert the cosine value of the angle into the angle value of the inverted triangular pyramid structure.

6. The method for generating a complete model of a distribution network tower based on combined construction according to claim 3, characterized in that: For a cluster determined to be a tetrahedral structure, the normal vectors and vertex coordinates of each face of the tetrahedral structure are extracted, including: The steps of extracting the vertex coordinates of the tetrahedral structure are: Collect the point cloud data of the cluster of the tetrahedral structure, assuming that there are n points, the coordinates of each point are (x_i, y_i, z_i), set 0.3 times the average distance as the distance threshold, and calculate the Euclidean distance between each point (x_i, y_i, z_i) and all other points; For each point (x_i, y_i, z_i), count the number of points whose distance is less than the distance threshold. If the number of points whose distance is less than the distance threshold is greater than or equal to the set distance threshold, it means that there are enough neighboring points around the point, that is, the density is high, and the point is considered to be one of the vertex candidates; Select the point with the largest number of candidates as the vertex, and repeat the above judgment steps until all vertices are found; The steps of extracting the normal vectors of each face of the tetrahedral structure are: Collecting point cloud data of the bottom surface of the cluster of the tetrahedral structure, a point set P = {(x_i, y_i, z_i)}, where each point represents a point on the bottom surface; The principal component analysis is used to perform plane fitting on the point set P to obtain a normal vector of the fitting plane, and the normal vector κ=(x, y, z) of the fitting plane is normalized to obtain a normalized normal vector κ′=(x′, y′, z′).

7. The method for generating a complete model of a distribution network tower based on combined construction according to claim 1, characterized in that: The generation of the complete distribution network tower model includes: For each structure type, define the topological properties of the corresponding structure, the topological properties include the connection mode, support mode and material of the structure, use computer CAD software to build an abstract template according to the columnar structure, inverted triangular pyramid structure and quadrangular pyramid structure, and assist in building a single columnar model, inverted triangular pyramid model and quadrangular pyramid model based on the result of the structure decomposition and the corresponding abstract template; Acquire the actual grid topology structure, and combine the single columnar model, inverted triangular pyramid model, and quadrangular pyramid model according to the combination rule of tower head + tower body + tower base according to the actual grid topology structure and design requirements to generate a distribution network tower model framework; Add conductors and equipment, including crossarms, insulators, and transformers, to the distribution network tower model framework; determine the position, length, and angle of the conductors according to actual power transmission requirements; and connect them to the corresponding distribution network tower model framework to generate a complete distribution network tower model.

8. A system using the method for generating a complete model of a distribution network tower based on combined construction as claimed in any one of claims 1 to 7, characterized in that: include: A data processing unit, used for preprocessing the acquired three-dimensional distribution network tower laser point cloud data; A structural decomposition unit is used to structurally decompose different parts of the tower in the pre-processed distribution network tower laser point cloud data using a point cloud cutting algorithm, and extract corresponding feature points and geometric attributes according to different decomposed structural types to obtain a structural decomposition result; A single model building unit is used to build corresponding abstract templates according to different structural types, and to assist in building a single structural model based on the result of the structural decomposition and the corresponding abstract template; The complete model generation unit is used to combine the individual structural models to generate a complete distribution network tower model according to the actual power grid topology and design requirements.

9. An electronic device, comprising: Memory and processor; The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, the steps of the method described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having computer program instructions stored thereon, wherein the computer-executable instructions, when executed by a processor, implement the steps of the method according to any one of claims 1 to 7.

Citation Information

Cited By

  • Method for intelligently monitoring structural deformation of civil engineering of transformer substation

    CN119984085A