Power system inspection registration method based on BIM model and three-dimensional point cloud model

Abstract

The invention relates to the technical field of electric power system maintenance, in particular to an electric power system inspection registration method based on a BIM model and a three-dimensional point cloud model, and the method comprises the steps: obtaining and analyzing a BIM model file and a three-dimensional point cloud file of an electric power system, so as to construct a first registration descriptor and a second registration descriptor; calculating a rough transformation matrix through the first registration descriptor and the second registration descriptor; converting a space coordinate system of the BIM into a space coordinate system of the point cloud according to the rough transformation matrix so as to obtain a BIM model file after rough registration; performing triangulation on the BIM model file after coarse registration to obtain three-dimensional point cloud models of different scales, and organizing the three-dimensional point cloud models in a KD-tree level to iteratively calculate a fine transformation matrix; and performing fine registration on the BIM model file after coarse registration according to the fine transformation matrix. Therefore, the problems that in existing electric power line inspection, unmanned aerial vehicle operation highly depends on pilot experience excessively, safety risks are brought to an electric power system, and the inspection efficiency is low are solved.

Description

Technical Field

[0001] The present invention relates to the technical field of power system maintenance, and particularly relates to a power system inspection registration method based on a BIM model and a three-dimensional point cloud model. Background Art

[0002] With the continuous development of power system maintenance work, the power system maintenance work has gradually shifted from manual operation to unattended operation. Currently, the mainstream unmanned aerial vehicle (UAV) technology not only has the advantages of small size, controllable speed, simple operation, etc., but also can achieve remote shooting, providing a new means for power inspection. In the past few years, emerging UAVs equipped with sensors (such as light detection and ranging / laser radar, optical cameras, infrared cameras, and ultraviolet cameras) have provided rich data sources for comprehensive and accurate power line inspection. One challenge that still hinders the use of UAVs in power line inspection is that their operation highly depends on the experience of the pilot, which may pose risks to the safety of the power system and reduce the inspection efficiency. Therefore, it is necessary to research and develop a method for guiding the inspection of power system components using rich data sources (three-dimensional point cloud model, BIM model). Summary of the Invention

[0003] The present invention provides a power system inspection registration method based on a BIM model and a three-dimensional point cloud model to solve the problems in existing power line inspection that the operation of UAVs highly depends on the experience of the pilot, which may pose safety risks to the power system and has low inspection efficiency.

[0004] An embodiment of the first aspect of the present invention provides a power system inspection registration method based on a BIM model and a three-dimensional point cloud model, including the following steps: obtaining a BIM model, a BIM model file, a three-dimensional point cloud model, and a three-dimensional point cloud file of a target power system; extracting a geometric representation and a geometric coordinate system from the BIM model file, and generating a first registration descriptor of the BIM model according to the geometric representation and the geometric coordinate system; extracting an intersection line direction feature and a normal vector feature of the three-dimensional point cloud file, and generating a second registration descriptor of the three-dimensional point cloud model according to the intersection line direction feature and the normal vector feature; calculating a rough transformation matrix between the BIM model and the three-dimensional point cloud model through the first registration descriptor and the second registration descriptor; converting a first spatial coordinate system of the BIM model to a second spatial coordinate system of the three-dimensional point cloud model according to the rough transformation matrix to obtain a roughly registered BIM model file; triangulating the roughly registered BIM model file to generate a number of triangular Mesh patches; downsampling the triangular Mesh patches to obtain three-dimensional point cloud models of different scales; organizing the three-dimensional point cloud model and the three-dimensional point cloud models of different scales in a KD-tree hierarchy to obtain KD-tree node units, and iteratively calculating a fine transformation matrix according to the KD-tree node units; converting the roughly registered BIM model file to a finely registered BIM model file according to the fine transformation matrix; generating an inspection route according to the finely registered BIM model file, and sending the inspection route to a target unmanned aerial vehicle.

[0005] Optionally, the extracting a geometric representation and a geometric coordinate system from the BIM model file, and generating a first registration descriptor of the BIM model according to the geometric representation and the geometric coordinate system includes: extracting the geometric coordinate system and shape definition of each element in the BIM model file; extracting a geometric representation under the shape definition and parsing the type of the geometric representation; calculating the normal vector information, boundary information, and intersection line information of each face according to the type of the geometric representation; combining the normal vector information, boundary information, and intersection line information of each face based on the geometric coordinate system to obtain a series of first quadruple representations or first triple representations of non-parallel planes as the first registration descriptor.

[0006] Optionally, the extracting an intersection line direction feature and a normal vector feature of the three-dimensional point cloud file, and generating a second registration descriptor of the three-dimensional point cloud model according to the intersection line direction feature and the normal vector feature includes: Randomly select three points in the three-dimensional point cloud file to construct a temporary plane, and calculate the distance from each point to the temporary plane; According to the distances, distinguish the three points into inliers and outliers, and extract the inliers and the temporary plane from the three-dimensional point cloud file as the plane boundary information of the current surface; Iteratively execute the process of extracting the plane boundary information from the three-dimensional point cloud file until all points in the three-dimensional point cloud file have been randomly selected, and obtain the plane boundary information of multiple surfaces; Calculate the intersection line direction feature and the normal vector feature according to the plane boundary information of the multiple surfaces; Combine the intersection line direction feature and the normal vector feature to obtain a series of second quadruple representations or second triple representations of non-parallel planes as the second registration descriptor.

[0007] Optionally, the calculating the rough transformation matrix between the BIM model and the three-dimensional point cloud model through the first registration descriptor and the second registration descriptor includes: Construct a KD tree of the first registration descriptor and the second registration descriptor; Calculate the Euclidean distance between each descriptor in the first registration descriptor and each descriptor in the second registration descriptor according to the KD tree; Compare the Euclidean distance with a preset threshold. If the Euclidean distance is less than the preset threshold, use the line segment of two matching descriptors extracted from the BIM model and the three-dimensional point cloud model to calculate the rough transformation matrix corresponding to each pair of matching descriptors.

[0008] Optionally, the converting the first space coordinate system of the BIM model to the second space coordinate system of the three-dimensional point cloud model according to the rough transformation matrix to obtain a roughly registered BIM model file includes: Calculate the probability that the points in the three-dimensional point cloud model fall into the first space coordinate system according to the rough transformation matrix corresponding to each pair of matching descriptors; Convert the first space coordinate system to the second space coordinate system according to the rough transformation matrix with the highest probability to obtain the roughly registered BIM model file.

[0009] Optionally, the organizing the three-dimensional point cloud model and the three-dimensional point cloud models of different scales in the hierarchy of a KD-tree to obtain KD-tree node units, and iteratively calculating the fine transformation matrix according to the KD-tree node units includes: Organize the three-dimensional point cloud model and the three-dimensional point cloud models of different scales in the hierarchy of a KD-tree to obtain the KD-tree node units; Solve the objective function of the KD-tree node unit based on the maximum likelihood method and the Gaussian function; Use the Newton method to iteratively solve the transformation of the KD-tree node unit until the objective function converges to obtain the fine transformation matrix.

[0010] An embodiment of the second aspect of the present invention provides a power system inspection registration device based on a BIM model and a three-dimensional point cloud model, including: an acquisition module for acquiring the BIM model, BIM model file, three-dimensional point cloud model, and three-dimensional point cloud file of the target power system; a first extraction module for extracting the geometric representation and geometric coordinate system in the BIM model file, and generating a first registration descriptor of the BIM model according to the geometric representation and the geometric coordinate system; a second extraction module for extracting the intersection line direction feature and normal vector feature of the three-dimensional point cloud file, and generating a second registration descriptor of the three-dimensional point cloud model according to the intersection line direction feature and the normal vector feature; a matrix calculation module for calculating a rough transformation matrix between the BIM model and the three-dimensional point cloud model through the first registration descriptor and the second registration descriptor; a rough registration module for converting the first space coordinate system of the BIM model to the second space coordinate system of the three-dimensional point cloud model according to the rough transformation matrix to obtain a roughly registered BIM model file; a triangulation module for triangulating the roughly registered BIM model file to generate a plurality of triangular Mesh patches; a downsampling module for downsampling the triangular Mesh patches to obtain three-dimensional point cloud models of different scales; an iterative calculation module for organizing the three-dimensional point cloud model and the three-dimensional point cloud models of different scales in a KD-tree hierarchy to obtain KD-tree node units, and iteratively calculating a fine transformation matrix according to the KD-tree node units; a fine registration module for converting the roughly registered BIM model file to a finely registered BIM model file according to the fine transformation matrix; a generation module for generating an inspection route according to the finely registered BIM model file and sending the inspection route to the target unmanned aerial vehicle.

[0011] An embodiment of the third aspect of the present invention provides an electronic device, including: a memory, a processor, and a computer program stored on the memory and executable on the processor, where the processor executes the program to implement the power system inspection registration method based on the BIM model and the three-dimensional point cloud model as described in the above embodiments.

[0012] An embodiment of the fourth aspect of the present invention provides a computer program product, where the computer program / instructions implement the power system inspection registration method based on the BIM model and the three-dimensional point cloud model as described above when executed by a processor.

[0013] In a fifth aspect embodiment of the present invention, a computer-readable storage medium is provided. The computer-readable storage medium stores a computer program, which when executed by a processor implements the above-mentioned power system inspection registration method based on a BIM model and a three-dimensional point cloud model.

[0014] The power system inspection registration method based on a BIM model and a three-dimensional point cloud model proposed in the embodiments of the present invention can plan a more accurate inspection route for the unmanned aerial vehicle (UAV) by registering the BIM model and the three-dimensional point cloud model. This method can effectively reduce the inspection accuracy problems caused by differences in pilots' experience and improve the efficiency of power system inspection; using a high-performance server for route planning reduces the requirements for hardware devices and computing capabilities on the UAV side, enabling users to more conveniently and at low cost perform component inspections of the power system without relying on high-performance UAV-side hardware; it can comprehensively use three-dimensional point cloud model files and BIM model files to ensure multi-source fusion of inspection data and provide real-time feedback, making it an effective intelligent power system maintenance tool that enhances the scientificity and accuracy of inspection decisions; this solution has strong adaptability and can be adjusted and optimized according to different power inspection scenarios and can also handle power inspection tasks of different scales and complexities.

[0015] Additional aspects and advantages of the present invention will be given in part in the following description, will become apparent in part from the following description, or will be understood through the practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The above and / or additional aspects and advantages of the present invention will become apparent and be readily understood from the following description of the embodiments in conjunction with the drawings, where: Figure 1 is a flowchart of a power system inspection registration method based on a BIM model and a three-dimensional point cloud model provided by an embodiment of the present invention; Figure 2 is a specific execution schematic diagram of a power system inspection registration method based on a BIM model and a three-dimensional point cloud model provided by an embodiment of the present invention; Figure 3 is a schematic diagram of the geometric feature representation of a BIM model provided by an embodiment of the present invention; Figure 4 is a block schematic diagram of a power system inspection registration device based on a BIM model and a three-dimensional point cloud model provided by an embodiment of the present invention; Figure 5 is a structural schematic diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0017] Embodiments of the present invention will be described in detail below. Examples of the embodiments are shown in the accompanying drawings, where the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to explain the present invention, and should not be construed as a limitation of the present invention.

[0018] The power system inspection registration method based on the BIM model and the 3D point cloud model according to the embodiments of the present invention will be described below with reference to the accompanying drawings.

[0019] Specifically, Figure 1 It is a schematic flowchart of a power system inspection registration method based on the BIM model and the 3D point cloud model provided by the embodiments of the present invention.

[0020] As Figure 1 shown, the power system inspection registration method based on the BIM model and the 3D point cloud model includes the following steps: In step S101, obtain the BIM model, BIM model file, 3D point cloud model, and 3D point cloud file of the target power system. In the actual execution process, a lidar is carried on the target unmanned aerial vehicle, and the 3D point cloud model and 3D point cloud file of the target power system are obtained through the lidar. At the same time, the BIM model and the BIM model file in the IFC file format preset for the target power system are read.

[0021] In step S102, extract the geometric representation and geometric coordinate system in the BIM model file, and generate the first registration descriptor of the BIM model according to the geometric representation and geometric coordinate system.

[0022] In some embodiments, extracting the geometric representation and geometric coordinate system in the BIM model file and generating the first registration descriptor of the BIM model according to the geometric representation and geometric coordinate system includes: Extract the geometric coordinate system and shape definition of each element in the BIM model file; Extract the geometric representation under the shape definition and parse the type of the geometric representation; Calculate the normal vector information, boundary information, and intersection line information of each face according to the type of the geometric representation; Based on the geometric coordinate system, combine the normal vector information, boundary information, and intersection line information of each face to obtain a series of first quadruple representations or first triple representations of non-parallel planes as the first registration descriptor.

[0023] As Figure 2 and 3As shown, during the actual execution process, the geometric coordinate system IfcLocalPlacement and the shape definition IfcProductDefinitionShape of each element IfcElement are extracted from the BIM model file. Under the shape definition, the geometric representation Representations is extracted, and the type Item of the geometric representation is parsed. If the type is an extruded solid IfcExtrudedAreaSolid, then the extrusion length Depth, the extrusion direction ExtrusionDirection, the coordinate position Position, and the attributes of the swept area SweptArea of the extruded solid are parsed. Then, the normal vector information and the intersection line information of each face are calculated based on the attributes. If the type is an extruded solid IfcFaseBasedSurfaceModel, a polyhedron IfcFaseTedBrep, etc., then the boundary information and the normal vector information of each face are extracted. Subsequently, the above-extracted normal vector information, boundary information, and intersection line information are combined to obtain a quadruple representation of a series of non-parallel planes or a triple representation , and a first registration descriptor is obtained based on the multiple tuple representation.

[0024] In step S103, the intersection line direction feature and the normal vector feature of the three-dimensional point cloud file are extracted, and a second registration descriptor of the three-dimensional point cloud model is generated according to the intersection line direction feature and the normal vector feature.

[0025] In some embodiments, extracting the intersection line direction feature and the normal vector feature of the three-dimensional point cloud file, and generating a second registration descriptor of the three-dimensional point cloud model according to the intersection line direction feature and the normal vector feature includes:[[]] Randomly select three points in the three-dimensional point cloud file to construct a temporary plane, and calculate the distance from each point to the temporary plane; According to the distance, the three points are classified into inliers and outliers, and the inliers and the temporary plane are extracted from the three-dimensional point cloud file as the plane boundary information of the current face; Iteratively execute the process of extracting the plane boundary information from the three-dimensional point cloud file until all the points in the three-dimensional point cloud file have been randomly selected, and the plane boundary information of multiple faces is obtained; Calculate the intersection line direction feature and the normal vector feature according to the plane boundary information of multiple faces; Combine the intersection line direction feature and the normal vector feature to obtain a second quadruple representation or a second triple representation of a series of non-parallel planes as the second registration descriptor.

[0026] During the actual execution process, based on the method of Random Sample Consensus (RANSAC), randomly select three points in the three-dimensional point cloud file 、 、 , construct a temporary plane equation , and calculate the distance from each point to the temporary plane; compare the distance with a preset threshold to distinguish the three points into inliers and outliers. If the proportion of inliers meets the preset requirements, extract the inliers and the temporary plane from the 3D point cloud file as the plane boundary information of the current surface; iteratively execute the process of extracting the plane boundary information from the 3D point cloud file until all points in the 3D point cloud file have been randomly selected, obtaining the plane boundary information of multiple surfaces; calculate the intersection line direction feature and the normal vector feature based on the plane boundary information of multiple surfaces; combine the intersection line direction feature and the normal vector feature to obtain a series of second quadruple representations or second triple representations of non-parallel planes as the second registration descriptor.

[0027] For example, in steps S102 and S103, the registration descriptor is obtained based on the multiple group representation of non-parallel planes, and the steps can be as follows: Assume plane , plane , represents the distance between the skew lines and , represents the angle between the lines and , is and 's included angle, represents the angle between the line and the plane , then the registration descriptor of the quadruple is described as follows. If ; otherwise, change the plane order:

[0028] In very unlikely cases, fewer than two pairs of non-parallel planes can be found. Among them, only two parallel horizontal planes and two parallel vertical planes are extracted, and there is no non-parallel plane quadruple that can uniquely define a rigid transformation. Similarly, the descriptor based on the plane and the intersection line is a 6D vector, defined as:

[0029] For the digital model (BIM model or 3D point cloud model) , the registration descriptor of the model is defined as .

[0030] In step S104, calculate the rough transformation matrix between the BIM model and the 3D point cloud model through the first registration descriptor and the second registration descriptor.

[0031] In some embodiments, calculating a rough transformation matrix between a BIM model and a three-dimensional point cloud model through a first registration descriptor and a second registration descriptor includes: Constructing KD-trees for the first registration descriptor and the second registration descriptor; Calculating the Euclidean distance between each descriptor in the first registration descriptor and each descriptor in the second registration descriptor respectively according to the KD-trees; Comparing the Euclidean distance with a preset threshold. If the Euclidean distance is less than the preset threshold, the straight line segments of two matching descriptors extracted from the BIM model and the three-dimensional point cloud model are used to calculate the rough transformation matrix corresponding to each pair of matching descriptors.

[0032] During the actual execution process, constructing the first registration descriptor corresponding to the BIM model and the second registration descriptor corresponding to the three-dimensional point cloud model of the KD-tree to calculate the Euclidean distance between two descriptors. For an eight-dimensional descriptor , excluding the descriptor vectors of two extra dimensions therefrom and downsampling to for calculation; If the Euclidean distance between descriptors is less than the threshold, these two descriptors are called matching descriptors; after obtaining the matching descriptor vectors, the straight line segments extracted from the BIM model and the three-dimensional point cloud model are used to calculate the rough transformation matrix corresponding to each registration descriptor , where represents the rotation part, represents the translation part.

[0033] In step S105, the first spatial coordinate system of the BIM model is converted to the second spatial coordinate system of the three-dimensional point cloud model according to the rough transformation matrix to obtain a roughly registered BIM model file.

[0034] In some embodiments, converting the first spatial coordinate system of the BIM model to the second spatial coordinate system of the three-dimensional point cloud model according to the rough transformation matrix to obtain a roughly registered BIM model file includes: Calculating the probability that the points in the three-dimensional point cloud model fall into the first spatial coordinate system according to the rough transformation matrix corresponding to each pair of matching descriptors; Converting the first spatial coordinate system to the second spatial coordinate system according to the rough transformation matrix with the highest probability to obtain a roughly registered BIM model file.

[0035] In the actual execution process, calculate the 3D point cloud model according to the rough transformation matrix corresponding to each pair of matching descriptors the probability that the points in fall into the first spatial coordinate system of the BIM model, select the transformation matrix with the highest probability as the result of rough registration, and output it as the BIM model of the BIM model file after rough registration. This process realizes the preliminary registration of the BIM model and the 3D point cloud with less computational effort, and can also be used for the preliminary generation of the UAV flight path.

[0036] In step S106, triangulate the BIM model file after rough registration to generate a number of triangular Mesh patches.

[0037] In step S107, downsample the triangular Mesh patches to obtain 3D point cloud models of different scales.

[0038] In the actual execution process, triangulate the BIM model file after rough registration, downsample it into a number of Mesh triangular patches, and then perform downsampling with different spatial resolutions to obtain 3D point cloud models of different scales. For example, the BIM model file after rough registration is downsampled with different voxel units to obtain 3D point cloud models with different resolutions , and the spatial resolution satisfies .

[0039] In step S108, organize the 3D point cloud model and the 3D point cloud models of different scales in the hierarchy of the KD-tree to obtain KD-tree node units, and iteratively calculate the fine transformation matrix according to the KD-tree node units.

[0040] In some embodiments, organizing the 3D point cloud model and the 3D point cloud models of different scales in the hierarchy of the KD-tree to obtain KD-tree node units, and iteratively calculating the fine transformation matrix according to the KD-tree node units includes: Organize the 3D point cloud model and the 3D point cloud models of different scales in the hierarchy of the KD-tree to obtain KD-tree node units; Solve the objective function of the KD-tree node unit based on the maximum likelihood method and the Gaussian function; Use the Newton method to iteratively solve the transformation of the KD-tree node unit until the objective function converges to obtain the fine transformation matrix.

[0041] In the actual execution process, apply the rough transformation matrix to the 3D point cloud models of different scales to obtain the 3D point cloud models of different scales after rough transformation, and the 3D point cloud model The 3D point cloud models of different scales after the rough transformation are organized hierarchically by a KD-tree. Each point in the point cloud model is located in a unique KD-tree node unit, and each node of the KD-tree represents a voxel. The characteristics of the node are mainly represented by the mean value of the points therein. and the covariance matrix are described as follows:

[0042]

[0043] In the formula, is the number of points, is the coordinate of the point in the voxel; The coordinate distribution of the points in the voxel follows a Gaussian distribution with noise level corrected, and the density function is:

[0044] where , are the coordinates of any two points in the voxel.

[0045] The objective function obtained based on the maximum likelihood method is:

[0046] where represents the coordinate after applying the transformation to . Using Gaussian function approximation to accelerate the processing, the objective function is obtained:

[0047]

[0048]

[0049] where is the node feature of the KD-tree where the downsampled point cloud is located after transformation , , , are the coordinate values of the coordinates of two points in the voxel in the first dimension, the second dimension, and the third dimension, respectively.

[0050] The Newton method is used to optimize the objective function for iterative solution of the transformation : , denoted as The obtained transformation matrix is , then the current solution transformation 。

[0051] Repeat for point clouds of different resolutions and perform the above process until the objective function converges to obtain the fine transformation matrix.

[0052] In step S109, the coarsely registered BIM model file is converted into a finely registered BIM model file according to the fine transformation matrix.

[0053] In step S110, an inspection route is generated according to the finely registered BIM model file and sent to the target UAV.

[0054] In the actual execution process, the coarsely registered BIM model file is converted into a finely registered BIM model file according to the fine transformation matrix. After the initial registration of the BIM model and the 3D point cloud, the details of the 3D point cloud are combined with the details of the downsampled point cloud of the BIM, making the alignment between the BIM model and the 3D point cloud more accurate; Furthermore, the A* algorithm (i.e., the A-star search algorithm) can be used to process the finely registered BIM model file to generate an inspection route and send the inspection route to the target UAV, so that the target UAV can inspect the target power system according to the inspection route.

[0055] In summary, the power system inspection registration method based on the BIM model and the 3D point cloud model proposed in the embodiment of the present invention includes the following steps: (1) By registering the BIM model and the 3D point cloud model, a more accurate inspection route can be planned for the UAV. This method can effectively reduce the inspection accuracy problems caused by differences in pilot experience and improve the efficiency of power system inspection; (2) Using a high-performance server for route planning reduces the requirements for hardware devices and computing capabilities on the UAV side. Users can perform component inspections of the power system more conveniently and at low cost without relying on high-performance UAV-side hardware; (3) It can comprehensively use the 3D point cloud model file and the BIM model file to ensure the multi-source fusion of inspection data and provide real-time feedback. It is an effective intelligent power system maintenance tool that improves the scientificity and accuracy of inspection decisions; (4) This solution has strong adaptability and can be adjusted and optimized according to different power inspection scenarios. It can also handle power inspection tasks of different scales and complexities.

[0056] Next, a power system inspection registration device based on the BIM model and the 3D point cloud model proposed in the embodiment of the present invention is described with reference to the accompanying drawings.

[0057] Figure 4 The block diagram of a power system inspection registration device based on a BIM model and a 3D point cloud model provided by an embodiment of the present invention.

[0058] As Figure 4 shown, the power system inspection registration device 40 based on the BIM model and the 3D point cloud model includes: an acquisition module 401, a first extraction module 402, a second extraction module 403, a matrix calculation module 404, a rough registration module 405, a dissection module 406, a downsampling module 407, an iterative calculation module 408, a fine registration module 409, and a generation module 410.

[0059] Among them, the acquisition module 401 is used to acquire the BIM model, BIM model file, 3D point cloud model, and 3D point cloud file of the target power system. The first extraction module 402 is used to extract the geometric body representation and geometric body coordinate system in the BIM model file, and generate the first registration descriptor of the BIM model according to the geometric body representation and geometric body coordinate system. The second extraction module 403 is used to extract the intersection line direction feature and normal vector feature of the 3D point cloud file, and generate the second registration descriptor of the 3D point cloud model according to the intersection line direction feature and normal vector feature. The matrix calculation module 404 is used to calculate the rough transformation matrix between the BIM model and the 3D point cloud model through the first registration descriptor and the second registration descriptor. The rough registration module 405 is used to convert the first space coordinate system of the BIM model to the second space coordinate system of the 3D point cloud model according to the rough transformation matrix to obtain the roughly registered BIM model file. The dissection module 406 is used to perform triangular dissection on the roughly registered BIM model file to generate a number of triangular Mesh patches. The downsampling module 407 is used to downsample the triangular Mesh patches to obtain 3D point cloud models of different scales. The iterative calculation module 408 is used to organize the 3D point cloud model and the 3D point cloud models of different scales in the hierarchy of the KD-tree to obtain the KD-tree node units, and iteratively calculate the fine transformation matrix according to the KD-tree node units. The fine registration module 409 is used to convert the roughly registered BIM model file to a finely registered BIM model file according to the fine transformation matrix. The generation module 410 is used to generate an inspection route according to the finely registered BIM model file and send the inspection route to the target unmanned aerial vehicle.

[0060] In some embodiments, the first extraction module 402 includes: A first extraction unit for extracting the geometric body coordinate system and shape definition of each element in the BIM model file; A second extraction unit for extracting the geometric representation under the shape definition and parsing the type of the geometric representation; A first calculation unit for calculating the normal vector information, boundary information, and intersection line information of each face according to the type of the geometric representation; The first combination unit is configured to combine the normal vector information, boundary information, and intersection line information of each face based on the geometric body coordinate system to obtain a first quadruple representation or a first triple representation of a series of non-parallel planes as the first registration descriptor.

[0061] In some embodiments, the second extraction module 403 includes: A random selection unit for randomly selecting three points in the three-dimensional point cloud file to construct a temporary plane and calculating the distance from each point to the temporary plane; A third extraction unit for classifying the three points into inliers and outliers according to the distance and extracting the inliers and the temporary plane from the three-dimensional point cloud file as the plane boundary information of the current face; An iterative extraction unit for iteratively executing the process of extracting the plane boundary information from the three-dimensional point cloud file until all the points in the three-dimensional point cloud file have been randomly selected to obtain the plane boundary information of multiple faces; A second calculation unit for calculating the intersection line direction feature and the normal vector feature according to the plane boundary information of multiple faces; A second organization unit for combining the intersection line direction feature and the normal vector feature to obtain a second quadruple representation or a second triple representation of a series of non-parallel planes as the second registration descriptor.

[0062] In some embodiments, the rough registration module 405 includes: A construction unit for constructing a KD tree of the first registration descriptor and the second registration descriptor; A third calculation unit for calculating the Euclidean distance between each descriptor in the first registration descriptor and each descriptor in the second registration descriptor according to the KD tree; A comparison unit for comparing the Euclidean distance with a preset threshold. If the Euclidean distance is less than the preset threshold, the straight line segments of the two matching descriptors extracted from the BIM model and the three-dimensional point cloud model are used to calculate the rough transformation matrix corresponding to each pair of matching descriptors.

[0063] In some embodiments, the iterative calculation module 408 includes: A structure organization unit for organizing the three-dimensional point cloud model and the three-dimensional point cloud models of different scales in the hierarchy of the KD-tree to obtain the KD-tree node unit; A solution unit for solving the objective function of the KD-tree node unit based on the maximum likelihood method and the Gaussian function; An iterative solution transformation unit for iteratively solving the transformation of the KD-tree node unit by using the Newton method until the objective function converges to obtain the fine transformation matrix.

[0064] It should be noted that the foregoing explanatory description of the embodiments of the power system inspection registration method based on the BIM model and the three-dimensional point cloud model is also applicable to the power system inspection registration device based on the BIM model and the three-dimensional point cloud model of this embodiment, and will not be elaborated here.

[0065] The power system inspection registration device based on the BIM model and the three-dimensional point cloud model according to the embodiment of the present invention includes the following steps: (1) By registering the BIM model and the three-dimensional point cloud model, a more accurate inspection route can be planned for the unmanned aerial vehicle (UAV). This method can effectively reduce the inspection accuracy problems caused by differences in pilots' experience and improve the efficiency of power system inspection. (2) Using a high-performance server for route planning reduces the requirements of the UAV side for hardware devices and computing capabilities. Users can more conveniently and at low cost perform component inspections of the power system without relying on high-performance UAV-side hardware. (3) It can comprehensively use the three-dimensional point cloud model file and the BIM model file to ensure the multi-source fusion of inspection data and provide real-time feedback. It is an effective intelligent power system maintenance tool that improves the scientificity and accuracy of inspection decisions. (4) This solution has strong adaptability and can be adjusted and optimized according to different power inspection scenarios, and can also handle power inspection tasks of different scales and complexities.

[0066] Figure 5 The structural schematic diagram of the electronic device provided by the embodiment of the present invention. The electronic device may include: A memory 501, a processor 502, and a computer program stored on the memory 501 and executable on the processor 502.

[0067] When the processor 502 executes the program, it implements the power system inspection registration method based on the BIM model and the three-dimensional point cloud model provided in the above embodiment.

[0068] Furthermore, the electronic device further includes: A communication interface 503 for communication between the memory 501 and the processor 502.

[0069] The memory 501 is used to store the computer program executable on the processor 502.

[0070] The memory 501 may include a high-speed RAM memory and may also include non-volatile memory, such as at least one disk memory.

[0071] If the memory 501, the processor 502, and the communication interface 503 are implemented independently, the communication interface 503, the memory 501, and the processor 502 can be interconnected through a bus and communicate with each other. The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, an Extended Industry Standard Architecture (EISA) bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc. For the sake of simplicity in representation, Figure 5 only a thick line is used to represent it in Figure 5 , but it does not mean that there is only one bus or one type of bus.

[0072] Optionally, in a specific implementation, if the memory 501, the processor 502, and the communication interface 503 are integrated on a single chip, the memory 501, the processor 502, and the communication interface 503 can communicate with each other through an internal interface.

[0073] The processor 502 may be a Central Processing Unit (CPU), or an Application Specific Integrated Circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of the present invention.

[0074] The embodiments of the present invention also provide a computer program product. When the computer program / instructions are executed by a processor, the power system inspection registration method based on a BIM model and a three-dimensional point cloud model as described above is implemented.

[0075] The embodiments of the present invention also provide a computer-readable storage medium, on which a computer program is stored. When the program is executed by a processor, the power system inspection registration method based on a BIM model and a three-dimensional point cloud model as described above is implemented.

[0076] In the description of this specification, the descriptions referring to terms such as "one embodiment", "some embodiments", "examples", "specific examples", or "some examples" etc. mean that the specific features, structures, materials, or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described can be combined in any one or N embodiments or examples in a suitable manner. In addition, without contradiction, those skilled in the art can combine and combine the different embodiments or examples described in this specification and the features of different embodiments or examples.

[0077] In addition, the terms "first" and "second" are used only for descriptive purposes and cannot be construed as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include at least one of such features. In the description of the present invention, the meaning of "N" is at least two, such as two, three, etc., unless otherwise specifically defined.

[0078] Any process or method description shown in the flowchart or described in other ways herein can be understood to represent a module, segment, or portion of code including one or N executable instructions for implementing a customized logical function or process, and the scope of the preferred embodiments of the present invention includes additional implementations, where the functions can be executed in a substantially simultaneous manner or in a reverse order according to the functions involved, rather than in the order shown or discussed, which should be understood by those skilled in the art to which the embodiments of the present invention pertain.

[0079] The logic and / or steps represented in the flowchart or otherwise described herein can, for example, be considered a defined sequence list of executable instructions for implementing logical functions, which can be embodied in any computer-readable medium for use by an instruction execution system, apparatus, or device (such as a computer-based system, a system including a processor, or other systems that can fetch and execute instructions from the instruction execution system, apparatus, or device), or used in conjunction with these instruction execution systems, apparatuses, or devices. For the purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. More specific examples (non-exhaustive list) of computer-readable media include the following: an electrical connection part (electronic device) having one or N wirings, a portable computer diskette (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). Additionally, the computer-readable medium can even be paper or other suitable media on which the program can be printed, because the program can be obtained electronically by optically scanning the paper or other media, followed by editing, interpretation, or otherwise processing as appropriate, and then stored in a computer memory.

[0080] It should be understood that various parts of the present invention can be implemented by hardware, software, firmware, or a combination thereof. In the above-described embodiments, the N steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. If implemented in hardware, as in another embodiment, any one or a combination of the following techniques well known in the art can be used: discrete logic circuits having logic gate circuits for implementing logical functions on data signals, application-specific integrated circuits having suitable combinational logic gate circuits, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0081] Those of ordinary skill in the art of this technology can understand that all or part of the steps carried by the method of the above embodiments can be completed by a program instructing relevant hardware, and the program can be stored in a computer-readable storage medium. When the program is executed, it includes one or a combination of the steps of the method embodiments.

[0082] In addition, in each embodiment of the present invention, each functional unit may be integrated into a processing module, may exist separately as individual physical units, or two or more units may be integrated into one module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of a software functional module. When the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it may also be stored in a computer-readable storage medium.

[0083] The above-mentioned storage medium may be a read-only memory, a magnetic disk, an optical disc, etc. Although the embodiments of the present invention have been shown and described above, it can be understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those of ordinary skill in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of the present invention.

Claims

1. A power system inspection registration method based on a BIM model and a three-dimensional point cloud model, characterized in that, Including the following steps: Obtain the BIM model, BIM model file, 3D point cloud model, and 3D point cloud file of the target power system; Extract the geometric representation and geometric coordinate system from the BIM model file, and generate the first registration descriptor of the BIM model according to the geometric representation and the geometric coordinate system; Extract the intersection line direction feature and normal vector feature of the 3D point cloud file, and generate the second registration descriptor of the 3D point cloud model according to the intersection line direction feature and the normal vector feature; Calculate the rough transformation matrix between the BIM model and the 3D point cloud model through the first registration descriptor and the second registration descriptor; Convert the first spatial coordinate system of the BIM model to the second spatial coordinate system of the 3D point cloud model according to the rough transformation matrix to obtain the BIM model file after rough registration; Triangulate the BIM model file after rough registration to generate a number of triangular Mesh patches; Downsample the triangular Mesh patches to obtain 3D point cloud models of different scales; Organize the 3D point cloud model and the 3D point cloud models of different scales in the hierarchy of a KD-tree to obtain KD-tree node units, and iteratively calculate the fine transformation matrix according to the KD-tree node units; Convert the BIM model file after rough registration to the BIM model file after fine registration according to the fine transformation matrix; Generate an inspection flight path according to the BIM model file after fine registration, and send the inspection flight path to the target UAV.

2. The power system inspection registration method based on a BIM model and a three-dimensional point cloud model according to claim 1, wherein, The extracting the geometric representation and geometric coordinate system from the BIM model file, and generating the first registration descriptor of the BIM model according to the geometric representation and the geometric coordinate system includes: Extract the geometric coordinate system and shape definition of each element in the BIM model file; Extract the geometric representation under the shape definition and analyze the type of the geometric representation; Calculate the normal vector information, boundary information, and intersection line information of each face according to the type of the geometric representation; Based on the geometric coordinate system, combine the normal vector information, boundary information, and intersection line information of each face to obtain a series of first quadruple representations or first triple representations of non-parallel planes as the first registration descriptor.

3. The power system inspection registration method based on the BIM model and the three-dimensional point cloud model according to claim 1, wherein The extracting the intersection line direction feature and normal vector feature of the 3D point cloud file, and generating the second registration descriptor of the 3D point cloud model according to the intersection line direction feature and the normal vector feature includes: Randomly select three points in the 3D point cloud file to construct a temporary plane, and calculate the distance from each point to the temporary plane; Distinguish the three points into inliers and outliers according to the distance, and extract the inliers and the temporary plane from the 3D point cloud file as the plane boundary information of the current face; Iteratively execute the process of extracting the plane boundary information from the 3D point cloud file until all points in the 3D point cloud file have been randomly selected, and obtain the plane boundary information of multiple faces; Calculate the intersection line direction feature and the normal vector feature according to the plane boundary information of the multiple faces; Combining the intersection line direction feature and the normal vector feature to obtain a series of second quadruple representations or second triple representations of non-parallel planes as the second registration descriptor.

4. The power system inspection registration method based on the BIM model and the three-dimensional point cloud model according to claim 1, wherein, Calculating the rough transformation matrix between the BIM model and the three-dimensional point cloud model through the first registration descriptor and the second registration descriptor includes: Constructing a KD tree for the first registration descriptor and the second registration descriptor; Calculating the Euclidean distance between each descriptor in the first registration descriptor and each descriptor in the second registration descriptor according to the KD tree; Comparing the Euclidean distance with a preset threshold. If the Euclidean distance is less than the preset threshold, use the line segments of the two matching descriptors extracted from the BIM model and the three-dimensional point cloud model to calculate the rough transformation matrix corresponding to each pair of matching descriptors.

5. The power system inspection registration method based on the BIM model and the three-dimensional point cloud model according to claim 4, wherein, Converting the first spatial coordinate system of the BIM model to the second spatial coordinate system of the three-dimensional point cloud model according to the rough transformation matrix to obtain a roughly registered BIM model file, including: Calculating the probability that the points in the three-dimensional point cloud model fall into the first spatial coordinate system according to the rough transformation matrix corresponding to each pair of matching descriptors; Converting the first spatial coordinate system to the second spatial coordinate system according to the rough transformation matrix with the highest probability to obtain the roughly registered BIM model file.

6. The power system inspection registration method based on a BIM model and a three-dimensional point cloud model according to claim 1, characterized in that Organizing the three-dimensional point cloud model and the three-dimensional point cloud models of different scales in the hierarchy of the KD-tree to obtain KD-tree node units, and iteratively calculating the fine transformation matrix according to the KD-tree node units, including: Organizing the three-dimensional point cloud model and the three-dimensional point cloud models of different scales in the hierarchy of the KD-tree to obtain the KD-tree node units; Solving the objective function of the KD-tree node unit based on the maximum likelihood method and the Gaussian function; Using the Newton method to iteratively solve the transformation of the KD-tree node unit until the objective function converges to obtain the fine transformation matrix.

7. A power system inspection registration device based on a BIM model and a three-dimensional point cloud model, characterized in that, Including: An acquisition module for acquiring the BIM model, the BIM model file, the three-dimensional point cloud model, and the three-dimensional point cloud file of the target power system; A first extraction module for extracting the geometric body representation and the geometric body coordinate system in the BIM model file, and generating the first registration descriptor of the BIM model according to the geometric body representation and the geometric body coordinate system; A second extraction module for extracting the intersection line direction feature and the normal vector feature of the three-dimensional point cloud file, and generating the second registration descriptor of the three-dimensional point cloud model according to the intersection line direction feature and the normal vector feature; A matrix calculation module for calculating the rough transformation matrix between the BIM model and the three-dimensional point cloud model through the first registration descriptor and the second registration descriptor; A rough registration module for converting the first spatial coordinate system of the BIM model to the second spatial coordinate system of the three-dimensional point cloud model according to the rough transformation matrix to obtain a roughly registered BIM model file; A dissection module, configured to perform triangulation on the coarsely registered BIM model file to generate a number of triangular Mesh patches; A downsampling module, configured to downsample the triangular Mesh patches to obtain 3D point cloud models of different scales; An iterative calculation module, configured to organize the 3D point cloud model and the 3D point cloud models of different scales in a KD-tree hierarchy to obtain KD-tree node units, and iteratively calculate a fine transformation matrix based on the KD-tree node units; A fine registration module, configured to convert the coarsely registered BIM model file into a finely registered BIM model file according to the fine transformation matrix; A generation module, configured to generate an inspection route based on the finely registered BIM model file and send the inspection route to a target UAV.

8. An electronic device, characterized in that, Comprising: A memory, a processor, and a computer program stored on the memory and executable on the processor, where the processor executes the program to implement the power system inspection registration method based on a BIM model and a 3D point cloud model according to any one of claims 1-6.

9. A computer program product, characterized in that, When the computer program / instructions are executed by the processor, the power system inspection registration method based on a BIM model and a 3D point cloud model according to any one of claims 1-6 is implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, The program is executed by the processor to be used for implementing the power system inspection registration method based on a BIM model and a 3D point cloud model according to any one of claims 1-6.

Images