Transformer substation twinborn model construction method based on VR virtual reality technology

Abstract

The invention relates to the technical field of twinborn model construction, in particular to a transformer substation twinborn model construction method based on a VR virtual reality technology, and the method comprises the steps: obtaining dense point clouds of transformer equipment in a transformer substation scene, carrying out the meshing of all dense point clouds, and recording the vertexes of all obtained triangular grids as triangular point clouds; calculating a steep coefficient, a coplanar representation value, local complexity, a degradation evaluation value and a degradation contribution degree of each triangular grid; determining a distortion evaluation value; and for the distribution position of each trigonometric point cloud on the transformer equipment, determining the reconstruction weight of each trigonometric point cloud, performing three-dimensional reconstruction on the transformer equipment in the transformer substation scene, and constructing a transformer substation twinborn model. The transformer equipment twinborn model in the transformer substation scene is closer to the original condition, small and medium-sized equipment in the transformer substation scene is modeled more finely, and the visualization effect of the transformer substation twinborn model is improved.

Description

Technical Field

[0001] The present application relates to the technical field of twin model construction, and specifically to a method for constructing a substation twin model based on VR virtual reality technology. Background Art

[0002] Switching operations are a key part of daily substation maintenance, requiring complex and demanding procedures. Traditional substation switching operation training primarily relies on diagrams and videos. On-site training is constrained by factors such as site, equipment, and safety, preventing trainees from fully engaging in practical exercises. However, diagrams and video tutorials lack realism and interactivity, making it difficult for trainees to gain a deep understanding of the equipment structure and operating procedures. Virtual reality (VR) technology offers a new solution for substation switching operation training by digitally recreating a 500kV substation in a virtual space. Using VR equipment, standardized 500kV ultra-high voltage switching operations can be performed in a digital twin environment, enabling the implementation of standardized 500kV switching operations.

[0003] Traditional substation modeling uses lidar for three-dimensional reconstruction. Data collection for the large, real-world 500kV substation is time-consuming and labor-intensive, and restoration in a virtual environment is slow. Furthermore, the degradation of equipment in the substation scenario due to long-term operation is not considered, resulting in differences between the constructed substation twin model and the initial conditions of the substation. This leads to low modeling accuracy for small equipment in the substation scenario, which in turn affects the effectiveness of constructing the substation twin model. Summary of the Invention

[0004] In order to solve the above technical problems, a substation twin model construction method based on VR virtual reality technology is provided to solve the existing problems.

[0005] The solution to the technical problem in this application is to provide a method for constructing a substation twin model based on VR virtual reality technology, including the following steps:

[0006] Obtain dense point clouds of transformer equipment in the substation scene, mesh all dense point clouds, obtain triangular meshes, and record the vertices of each triangular mesh as triangular point clouds;

[0007] Analyze the distribution and discreteness of the triangular point cloud in each triangular mesh, and calculate the steepness coefficient of each triangular mesh based on the area of the triangular mesh; calculate the coplanarity value of each triangular mesh based on the angle between each triangular mesh and the other triangular meshes connected to it;

[0008] The local complexity of each triangular mesh is calculated based on the distribution of the triangular point cloud in the neighborhood and the clutter of the triangular point cloud in the neighborhood in each dimension of the three-dimensional space. The degradation evaluation value of each triangular mesh is determined by combining the steepness coefficient and the coplanarity characterization value. The degradation contribution of each triangular point cloud is determined using the degradation evaluation value of the triangular mesh connected to each triangular point cloud.

[0009] A deep learning algorithm is used to extract all triangular point clouds corresponding to each heat sink in the transformer equipment and divide them into blocks. The convex hull algorithm is used to analyze the geometric distribution of the triangular point clouds within each block on each heat sink to determine the distortion assessment value of each block.

[0010] According to the distribution position of each triangulated point cloud on the transformer equipment, the reconstruction weight of each triangulated point cloud is determined based on the degradation contribution and distortion evaluation value. The transformer equipment in the substation scene is reconstructed in three dimensions to construct a substation twin model.

[0011] Preferably, the calculating of the steepness coefficient of each triangular mesh includes:

[0012] Calculate the discrete degree of the position coordinates of all triangular point clouds in each triangular mesh on the X axis, Y axis and Z axis respectively;

[0013] The sum of the discrete degrees of each triangular mesh on the X-axis, Y-axis and Z-axis is recorded as the dispersion coefficient;

[0014] The steepness coefficient is the ratio of the dispersion coefficient to the area of each triangular mesh.

[0015] Preferably, the calculating of the coplanarity characterization value of each triangular mesh includes:

[0016] The remaining triangular meshes that share a common edge with each triangular mesh are recorded as adjacent meshes; the cosine value of the angle between the normal vector of each triangular mesh and each adjacent mesh is calculated, and the cosine value is negatively mapped;

[0017] The coplanar characterization value is the sum of the negative mapping results of each triangle mesh and all its adjacent meshes.

[0018] Preferably, the calculating the local complexity of each triangular mesh includes:

[0019] Cluster all triangle point clouds on all triangular meshes to obtain the local density of each triangle point cloud, where the cutoff distance of the clustering algorithm is the mean of all edge lengths on all triangular meshes;

[0020] Calculating the average of the local densities of all triangular point clouds on each triangular mesh as the local density of each triangular mesh;

[0021] Analyze the clutter of the position coordinates of different triangle point clouds in the neighborhood of all triangle point clouds in each triangle mesh on the X-axis, Y-axis, and Z-axis, and calculate the distribution clutter of each triangle mesh;

[0022] The local complexity is the product of the distribution disorder and the local density.

[0023] Preferably, the calculating the distribution clutter of each triangular mesh includes:

[0024] Taking each triangular point cloud as the center, all triangular point clouds within a spherical neighborhood with the cutoff distance as the radius are combined into a first point cloud set; the first point cloud set of all triangular point clouds on each triangular mesh is unioned to form a second point cloud set;

[0025] For each triangular mesh, calculating the variance of the position coordinates of all triangular point clouds in the second point cloud set on the X-axis, Y-axis, and Z-axis respectively;

[0026] The distribution disorder is the sum of the variances on the X-axis, the Y-axis, and the Z-axis.

[0027] Preferably, determining the degradation assessment value of each triangular mesh includes:

[0028] The ratio of the local complexity to the coplanar characterization value is recorded as a degradation coefficient;

[0029] The degradation evaluation value is a product of the steepness coefficient and the degradation coefficient.

[0030] Preferably, the degradation contribution V of the kth triangle point cloud is k The calculation method is: Among them, P q is the degradation evaluation value of the qth triangle mesh, n q is the number of all triangle point clouds in the qth triangle mesh, Q k is the number of all triangle meshes with the kth triangle point cloud as vertices.

[0031] Preferably, extracting all triangulated point clouds corresponding to each heat sink in the transformer device and dividing them into blocks includes:

[0032] All triangulated point clouds on the transformer device are used as input to the 3D semantic segmentation deep learning algorithm to extract all triangulated point clouds corresponding to each heat sink;

[0033] All triangular point clouds corresponding to each heat sink are divided into multiple blocks according to the corresponding Z-axis coordinates in their position coordinates.

[0034] Preferably, determining the distortion evaluation value of each block includes:

[0035] Based on all the triangular point clouds in each block of each heat sink, the convex hull algorithm is used to obtain the convex hull of each block and calculate the convex hull volume;

[0036] Calculate the average value of the degradation contribution of all triangular point clouds corresponding to each heat sink; record the heat sink corresponding to the minimum average value as the standard heat sink; and take the average value of the convex hull volume corresponding to all blocks on the standard heat sink as the standard volume;

[0037] The difference between the convex hull volume of each block on each heat sink and the standard volume is recorded as the relative deviation;

[0038] The distortion evaluation value is a ratio of the relative deviation to the standard volume.

[0039] Preferably, determining the reconstruction weight of each triangulated point cloud includes:

[0040] For all triangular point clouds distributed on the heat sink, calculate the normalized result of the product of the degradation contribution of each triangular point cloud and the distortion evaluation value of the block to which it belongs; perform negative mapping on the normalized result to obtain the reconstruction weight of each triangular point cloud;

[0041] For the remaining triangular point clouds not distributed on the heat sink, the degradation contribution is normalized, and the normalized result is negatively mapped to obtain the reconstruction weight of each triangular point cloud.

[0042] This application has at least the following beneficial effects:

[0043] This application triangulates the dense point cloud on the transformer equipment, analyzes the distribution of the triangular point cloud on each triangular mesh and the mesh volume, and calculates the steepness coefficient of each triangular mesh. Its beneficial effect is that it takes into account the flatness of the transformer equipment in the corresponding areas of different triangular meshes to preliminarily reflect the degree of degradation of the area; analyzes the angle between each triangular mesh and the other triangular meshes connected to it, and calculates the coplanar characterization value of each triangular mesh. Its beneficial effect is that it takes into account the coplanar characteristics between adjacent triangular meshes, and further reflects the flatness of the corresponding adjacent areas; calculates the local complexity of each triangular mesh. Its beneficial effect is that it evaluates the structural complexity of the local space of the triangular mesh through the distribution of the triangular point cloud in the neighborhood, and explains the degradation of the transformer equipment surface in the corresponding area; determines the degradation evaluation value of each triangular mesh. Its beneficial effect is that it analyzes the local characteristics of the triangular mesh from multiple scales to reflect the local structural complexity caused by the degradation of the transformer surface, avoids misjudging the angular parts of the transformer as degradation conditions, and improves the recognition accuracy of the degradation condition of the transformer equipment surface; and then determines the degradation contribution of each triangular point cloud. Its beneficial effect is that it divides the degradation degree of the triangular mesh equally into each triangular point cloud to reflect the degradation influence of each triangular point cloud on the triangular mesh, so that different triangular point clouds can be assigned according to the degradation condition of each triangular point cloud. 3D reconstruction is performed with different weights; the distortion evaluation value of each block is determined. The beneficial effect is that the distortion phenomenon of the triangulated point cloud on the heat sink in the transformer equipment is taken into account, which causes the shape of the area to be stretched or compressed. By quantifying the distortion degree of different triangulated point clouds on the heat sink, different weights can be assigned to different triangulated point clouds according to the distortion of each triangulated point cloud for subsequent 3D reconstruction, thereby weakening the influence of the distortion phenomenon on the 3D reconstruction; by judging whether each triangulated point cloud is distributed on the heat sink in the transformer equipment, the reconstruction weights of the triangulated point clouds at different positions are determined. The beneficial effect is that By using a stereo camera to take pictures of the 500kV substation in a real ultra-large space, the data collection speed is fast. Secondly, for the triangulated point cloud on the heat sink, the distortion and degradation degree on the three-dimensional reconstruction are reduced; for the triangulated point cloud in other positions, the degradation degree on the three-dimensional reconstruction is reduced, so as to eliminate the degradation condition of the transformer equipment as much as possible, avoid bending, stretching and other distortion phenomena of the heat sink in the constructed twin model, make the twin model of the transformer equipment closer to the original condition, strengthen the refinement of the modeling of small and medium-sized equipment in the substation scene, and improve the visualization effect of the twin model. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] The substation twin model construction method based on VR virtual reality technology of the present application is further described in detail below with reference to the accompanying drawings.

[0045] Figure 1A flowchart of the steps of the substation twin model construction method based on VR virtual reality technology provided in an embodiment of the present application;

[0046] Figure 2 A flowchart of the steps of a method for obtaining a degradation evaluation value of each triangular mesh provided in an embodiment of the present application;

[0047] Figure 3 A flowchart of the steps of the method for obtaining reconstruction weights provided in an embodiment of the present application. DETAILED DESCRIPTION

[0048] To make the objectives, technical solutions, and advantages of this application more clearly understood, the following, in conjunction with the accompanying drawings and implementation examples, further details the substation twin model construction method based on VR virtual reality technology proposed in this application. It should be understood that the specific embodiments described herein are only used to explain this application and are not intended to limit this application.

[0049] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs.

[0050] See also Figure 1 , which shows a flowchart of a method for constructing a substation twin model based on VR virtual reality technology provided by an embodiment of the present application, the method comprising the following steps:

[0051] Step 1: Obtain dense point clouds of transformer equipment in the substation scene, mesh all dense point clouds, obtain triangular meshes, and record the vertices of each triangular mesh as triangular point clouds.

[0052] In the process of building a twin model of a substation, the core is to digitally map the physical entity, operating environment and other information of the real substation to build a virtual model that is highly similar to the real substation. Substation equipment may be deformed or modified during long-term operation, resulting in differences between the actual equipment and the original state. Reverse engineering is used to build a three-dimensional model of the original substation scene to achieve the construction and support of standardized operation scenarios.

[0053] A multi-view depth camera is used to take multi-angle photos of all equipment in a 500kV substation. The offline Structure from Motion (SfM) technology is used to construct a sparse point cloud of the equipment in the 500kV substation, and the trajectory pose of the camera is obtained. The Multi-View Stereo (MVS) technology is used to estimate the dense depth maps of the equipment in the substation from multiple viewpoints. By fusing the dense depth maps from multiple viewpoints, the dense point clouds of all equipment in the substation are constructed.

[0054] In this embodiment, a binocular depth camera is used to take multi-angle photos of all equipment in the 500kV substation. As other implementation methods, the implementer can set them according to actual conditions. Secondly, SfM technology and MVS technology are well-known technologies and will not be described in detail here.

[0055] At this point, the dense point clouds of all equipment in the substation are obtained.

[0056] Step 2: Analyze the distribution and discreteness of the triangular point cloud in each triangular mesh, and calculate the steepness coefficient of each triangular mesh based on the area of the triangular mesh; calculate the coplanarity characterization value of each triangular mesh based on the angle between each triangular mesh and the other triangular meshes connected to it.

[0057] Power transformers, as electrical equipment that changes the voltage of alternating current (AC), are critical components of substations. They feature vertical heat sinks, whose primary function is to dissipate heat during operation to maintain the equipment's normal operating temperature. However, these fins inevitably degrade over time. For example, these fins can easily accumulate impurities such as dust, dirt, and oil. Some even experience coating loss and corrosion over time.

[0058] Secondly, in order to ensure uniform heat distribution and efficient heat dissipation, the surface of the transformer equipment should be flat and smooth. The dense point cloud on the heat sink surface should show a high degree of flatness. However, the surface of the heat sink in a deteriorated state is rough, and the dense point cloud structure is highly complex. Therefore, the distribution of the dense point cloud of the transformer equipment is analyzed and the steepness coefficient is calculated, which is:

[0059] All dense point clouds on transformer equipment in the substation are triangulated to obtain triangular meshes, and the vertices of each triangular mesh are recorded as triangular point clouds;

[0060] In this embodiment, the Delaunay triangulation algorithm is used for processing, wherein the Delaunay triangulation algorithm is a well-known technology and will not be described in detail here.

[0061] Calculate the discrete degree of the position coordinates of all triangular point clouds in each triangular mesh on the X axis, Y axis and Z axis respectively;

[0062] In this embodiment, the degree of discreteness is measured by calculating the standard deviation of the position coordinates of all triangular point clouds in each triangular mesh on the X-axis, Y-axis and Z-axis respectively. As other implementation methods, the implementer can adopt other methods of the existing technology, such as variance, etc. This embodiment does not impose any special restrictions on this.

[0063] The sum of the discrete degrees of each triangular mesh on the X-axis, Y-axis and Z-axis is used as the dispersion coefficient of each triangular mesh;

[0064] The ratio between the dispersion coefficient and the area of each triangular mesh is used as the steepness coefficient of each triangular mesh;

[0065] It should be noted that the more significant the degradation condition of the transformer equipment, the higher the structural complexity of the dense point cloud in the degraded area, and the easier it is to form a small-area triangular mesh. That is, the smaller the area of the triangular mesh, and the greater the difference between the triangular point clouds in the same dimension, that is, the larger the dispersion coefficient, the worse the flatness of the triangular mesh, the larger the resulting steepness coefficient, and the more significant the degradation condition of the transformer equipment.

[0066] Secondly, in the triangular mesh of the transformer equipment, due to the high flatness of the transformer surface and the uniform and sparse distribution of the point cloud, it is easy to form triangles with larger areas. Within the local neighborhood of the triangle, the area between the triangular meshes is larger, the spatial distribution is highly flat, the local spatial structure complexity is low, and the coplanarity is strong. However, the deteriorated area on the transformer surface is prone to form small-area dense triangles. Within the local neighborhood of the triangle, the area between the triangular meshes is smaller, the spatial distribution is highly discrete, the local spatial structure complexity is high, and the coplanarity is weak. Based on the above analysis, the coplanarity characterization value is calculated by the angle between the triangular meshes, specifically:

[0067] The remaining triangular meshes that have common edges with each triangular mesh are recorded as adjacent meshes;

[0068] Calculating the cosine value of the angle between the normal vector of each triangular mesh and each adjacent mesh, and performing negative mapping on the cosine value;

[0069] Calculating the sum of the negative mapping results of each triangular mesh and all its adjacent meshes as a coplanar characterization value of each triangular mesh;

[0070] In this embodiment, the negative mapping process is: calculating the difference between the value 1 and the cosine value as the result of the negative mapping; therefore, the calculation formula for the coplanarity characterization value of each triangular mesh is:

[0071]

[0072] Among them, λ r is the coplanar representation value of the r-th triangle mesh, M r is the number of all adjacent grids corresponding to the r-th triangle grid, is the normal vector of the r-th triangle mesh, is the normal vector of the i-th adjacent mesh corresponding to the r-th triangle mesh, and cos() is the calculated cosine value.

[0073] It should be noted that the closer the angle between the normal vectors of two triangular meshes is to 180°, the more coplanar they are. The smaller the cosine value, the larger the coplanar characterization value, the stronger the coplanar feature between the two triangular meshes, and the higher the flatness of the triangular meshes in the local neighborhood.

[0074] At this point, the steepness coefficient and coplanarity characterization value of each triangular mesh are obtained.

[0075] Step 3: Calculate the local complexity of each triangular mesh based on the distribution of triangular point clouds in the neighborhood and the clutter of triangular point clouds in the neighborhood in each dimension of three-dimensional space. Combined with the steepness coefficient and the coplanarity characterization value, determine the degradation evaluation value of each triangular mesh. Determine the degradation contribution of each triangular point cloud using the degradation evaluation value of the triangular meshes connected to each triangular point cloud.

[0076] Furthermore, the flowchart of the method for obtaining the degradation evaluation value of each triangular mesh provided in the embodiment of the present application is as follows: Figure 2 shown.

[0077] First, the coplanar eigenvalues of the triangular mesh at the corners of the transformer equipment are also weak, but the structural complexity is lower than that of the deteriorated parts. Therefore, a cluster analysis is performed on the triangular point cloud to quantify the complexity of the deteriorated parts, specifically:

[0078] The mean of all side lengths on all triangular meshes is used as the cutoff distance of the clustering algorithm, and all triangular point clouds on all triangular meshes are clustered to obtain the local density of each triangular point cloud.

[0079] In this embodiment, a density peak clustering algorithm (DPC) is used for clustering. The metric distance of the DPC clustering algorithm is the Euclidean distance between any two triangular point clouds. The DPC clustering algorithm is a well-known technology and will not be described in detail here.

[0080] Calculating the average of the local densities of all triangular point clouds on each triangular mesh as the local density of each triangular mesh;

[0081] It should be noted that, the greater the local density, the greater the number of triangle point clouds within the local neighborhood of the triangular mesh and the higher the density.

[0082] Taking each triangular point cloud as the center and all triangular point clouds within a spherical neighborhood with the cutoff distance as the radius, form a first point cloud set;

[0083] Calculate the union of the first point cloud sets of all triangular point clouds on each triangular mesh, and record it as the second point cloud set;

[0084] Calculating the variance of the position coordinates of all triangular point clouds in the second point cloud set on the X-axis, Y-axis, and Z-axis respectively;

[0085] Calculate the sum of the variances of each triangular mesh on the X-axis, Y-axis, and Z-axis as the distribution clutter of each triangular mesh;

[0086] The product of the distribution disorder and the local density is used as the local complexity of each triangular mesh;

[0087] It should be noted that when the number of triangular meshes within the local neighborhood is greater, the local density is greater, and at the same time, the triangular point cloud within the local neighborhood of the triangular mesh is more scattered and irregular, and the greater the distribution disorder, the greater the local complexity, indicating that the local spatial structure complexity of the triangular mesh within the local neighborhood is higher.

[0088] Secondly, based on the coplanar characterization value, the local complexity and the steepness coefficient, a degradation assessment value is determined, specifically:

[0089] Using the ratio of the local complexity to the coplanar characterization value as the degradation coefficient of each triangular mesh;

[0090] Taking the product of the steepness coefficient and the degradation coefficient as the degradation evaluation value of each triangular mesh;

[0091] It should be noted that when the coplanar characterization value is 0, the degradation coefficient is equal to the local complexity. Secondly, the greater the local complexity, the more complex the distribution of the triangular point cloud in the local neighborhood range. The smaller the coplanar characterization value, the lower the flatness of the transformer surface corresponding to the triangular mesh, and the more likely it is that dust and dirt have accumulated on the surface of the transformer, resulting in more obvious stratification of the triangular point cloud on the triangular mesh. The greater the degradation coefficient, the greater the obtained degradation assessment value, indicating that the degradation condition of the local area corresponding to the triangular mesh is more significant. By analyzing the local characteristics of the triangular mesh at multiple scales, it is possible to avoid misjudging the angular parts as degradation conditions, thereby improving the recognition accuracy of the surface degradation condition of the transformer equipment.

[0092] Furthermore, different triangular meshes of the transformer equipment are connected to each other. Therefore, each triangular point cloud is connected to multiple triangular meshes. By analyzing the degradation assessment values of the triangular meshes connected to each triangular point cloud, the degradation contribution is calculated. Specifically,

[0093] All triangle meshes with each triangle point cloud as vertices are combined into a mesh set;

[0094] The calculation formula for the degradation contribution of each triangular point cloud is:

[0095]

[0096] Among them, V k is the degradation contribution of the kth triangle point cloud, P q is the degradation evaluation value of the qth triangle mesh, n q is the number of all triangle point clouds in the qth triangle mesh, Q k is the number of all triangular meshes in the mesh set corresponding to the kth triangular point cloud.

[0097] It should be noted that n q The value of is 3. The purpose of dividing the degradation assessment value by 3 is to divide the degradation degree of the entire triangular mesh equally into each triangular point cloud to measure the influence of each triangular point cloud on the degradation of the triangular mesh. Therefore, the greater the degradation contribution, the more serious the degradation condition of the transformer equipment at the kth triangular point cloud. Therefore, when the transformer equipment is subsequently reconstructed in three dimensions, the triangular point cloud should be given a smaller weight for reconstruction to eliminate the degradation condition of the transformer equipment as much as possible and make the transformer equipment twin model closer to the original condition.

[0098] At this point, the degradation contribution of each triangulated point cloud is obtained.

[0099] Step 4: Use a deep learning algorithm to extract all the triangulated point clouds corresponding to each heat sink in the transformer equipment and divide them into blocks. Use the convex hull algorithm to analyze the geometric distribution of the triangulated point clouds in each block on each heat sink and determine the distortion evaluation value of each block. Based on the distribution position of each triangulated point cloud on the transformer equipment, the reconstruction weight of each triangulated point cloud is determined based on the degradation contribution and distortion evaluation value, and the transformer equipment in the substation scene is three-dimensionally reconstructed to construct a substation twin model.

[0100] When using a camera to take photos of transformer equipment from multiple angles, the point cloud data corresponding to the heat sink on the transformer equipment is prone to distortion due to complex issues such as reflections, weak textures, and lighting changes in the power industry scene. The distortion of the triangular point cloud corresponding to the heat sink will cause the shape of the triangular point cloud of the heat sink to deviate from the actual geometric shape of the heat sink, causing the shape of some areas to be stretched or compressed, making the originally vertical heat sink become curved. The more severely distorted the triangular point cloud, the smaller the weight should be given to the triangular point cloud for reconstruction to avoid bending, stretching, and other phenomena in the heat sink in the twin model. Therefore, the triangular point cloud on the heat sink is identified, and the distortion evaluation value is calculated by analyzing the deformation and distortion of the triangular point cloud on the heat sink, specifically:

[0101] All triangular point clouds on the transformer device are used as input to the 3D semantic segmentation deep learning algorithm (Point-VoxelCNN, PVCNN) to extract all triangular point clouds corresponding to each heat sink;

[0102] It should be noted that the PVCNN algorithm is a well-known technology and will not be described in detail here.

[0103] Calculating an average value of the degradation contribution of all triangular point clouds corresponding to each heat sink;

[0104] The heat sink corresponding to the minimum average value is recorded as the standard heat sink;

[0105] Divide all triangular point clouds corresponding to each heat sink into multiple blocks according to the corresponding Z-axis coordinates in its position coordinates;

[0106] In this embodiment, the height of each block is 1 cm. As for other implementation methods, the implementer can set it according to actual conditions.

[0107] Based on all the triangular point clouds in each block, the convex hull algorithm is used to obtain the convex hull of each block and calculate the convex hull volume;

[0108] In this embodiment, the Quick Hull algorithm is used to obtain the convex hull of each block. The Quick Hull algorithm is a well-known technology and will not be described in detail here. As other implementation methods, implementers can adopt other methods of the existing technology, such as the Graham Scan algorithm, etc. This embodiment does not impose any special restrictions on this.

[0109] The mean of the convex hull volumes corresponding to all blocks on the standard heat sink is taken as the standard volume;

[0110] The difference between the convex hull volume of each block on each heat sink and the standard volume is recorded as the relative deviation;

[0111] In this embodiment, the absolute value of the difference between the convex hull volume of each block on each heat sink and the standard volume is calculated and recorded as the relative deviation.

[0112] The ratio of the relative deviation to the standard volume is used as the distortion evaluation value of each block on each heat sink;

[0113] It should be noted that the larger the relative deviation is, the more likely the triangular point cloud on the heat sink is to be distorted, causing the shape of the corresponding block position on the heat sink to be stretched or compressed. The larger the resulting distortion evaluation value is, the more likely the weight of the triangular point cloud corresponding to the severe distortion should be reduced during three-dimensional reconstruction in order to improve the visual effect of the twin model.

[0114] Furthermore, based on the degradation contribution and the distortion evaluation value, a reconstruction weight is determined, specifically:

[0115] For all triangular point clouds distributed on the heat sink, calculate the normalized result of the product of the degradation contribution of each triangular point cloud and the distortion evaluation value of the block to which it belongs; perform negative mapping on the normalized result to obtain the reconstruction weight of each triangular point cloud;

[0116] For the remaining triangular point clouds not distributed on the heat sink, the degradation contribution is normalized, and the normalized result is negatively mapped to obtain the reconstruction weight of each triangular point cloud;

[0117] In this embodiment, the sigmoid function is used for normalization processing, wherein the sigmoid function is a well-known technology and will not be described in detail here. As other implementation methods, the implementer can adopt other methods of the existing technology, such as the tanh function, etc. This embodiment does not impose any special restrictions on this; secondly, the negative mapping process is: calculate the difference between the value 1 and the normalized result as the reconstruction weight of the triangulated point cloud.

[0118] It should be noted that the higher the degree of degradation of each triangulated point cloud or the more serious the distortion phenomenon, the smaller the corresponding reconstruction weight, so as to eliminate the degradation condition of the transformer equipment as much as possible, make the twin model of the transformer equipment closer to the original condition, and avoid distortion phenomena such as bending and stretching of the heat sink in the twin model, and improve the visualization effect of the twin model. The flowchart of the step of the method for obtaining the reconstruction weight provided in the embodiment of the present application is as follows: Figure 3 shown.

[0119] Based on the reconstruction weights, three-dimensionally reconstruct the transformer device through all triangulated point clouds to obtain a three-dimensional model;

[0120] In this embodiment, a screened Poisson surface reconstruction algorithm is used to perform three-dimensional reconstruction. The screened Poisson surface reconstruction algorithm is a well-known technology and will not be described in detail here.

[0121] Texture mapping is performed on the 3D model, and the 3D model of the transformer equipment in the substation scenario is loaded into the 3D rendering engine for model rendering to obtain a panoramic view of the substation. Content is produced through visual editing tools, including placing virtual nameplates for the transformer equipment. A unified perception network is built through IoT technology to collect the operating parameters of the transformer equipment in the substation. The logical model of the transformer equipment in each substation is obtained through the equipment instruction manual. Using data mapping technology, the logical model and operating parameters are synchronized to the corresponding 3D model to build a substation twin model.

[0122] It should be noted that texture mapping is a well-known technology and will not be described in detail here; texture mapping is intended to enhance the visual experience.

[0123] The substation twin model is uploaded to the cloud server and integrated with physics engines such as Havok and PhysX to simulate gravity, collision and other physical effects to enhance the realism of the virtual world. The gesture recognition software of the VR device or the operation of the handle can be used to realize the substation interactive training control and accident simulation.

[0124] It should be understood that although Figure 1 The steps in the flowchart are shown in sequence as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified in this document, there is no strict order restriction for the execution of these steps, and these steps can be executed in other orders. In addition, Figure 1 At least part of the steps may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily executed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least part of the sub-steps or stages of other steps.

[0125] The technical features of the above embodiments can be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0126] The above-described embodiments merely represent several implementation methods of the present application. While the descriptions are relatively specific and detailed, they should not be construed as limiting the present application. It should be noted that a person skilled in the art can make various modifications and improvements without departing from the spirit of the present application. Therefore, any simple modifications, equivalent variations, and modifications to the above embodiments made in accordance with the technical essence of the present application without departing from the content of the present application's technical solution fall within the scope of protection of the present application's technical solution.

Claims

1. A substation twin model construction method based on VR virtual reality technology is characterized by: The method comprises the following steps: Obtain dense point clouds of transformer equipment in the substation scene, mesh all dense point clouds, obtain triangular meshes, and record the vertices of each triangular mesh as triangular point clouds; Analyze the distribution and discreteness of the triangular point cloud in each triangular mesh, and calculate the steepness coefficient of each triangular mesh based on the area of the triangular mesh; calculate the coplanarity value of each triangular mesh based on the angle between each triangular mesh and the other triangular meshes connected to it; The local complexity of each triangular mesh is calculated based on the distribution of the triangular point cloud in the neighborhood and the clutter of the triangular point cloud in the neighborhood in each dimension of the three-dimensional space. The degradation evaluation value of each triangular mesh is determined by combining the steepness coefficient and the coplanarity characterization value. The degradation contribution of each triangular point cloud is determined using the degradation evaluation value of the triangular mesh connected to each triangular point cloud. A deep learning algorithm is used to extract all triangular point clouds corresponding to each heat sink in the transformer equipment and divide them into blocks. The convex hull algorithm is used to analyze the geometric distribution of the triangular point clouds within each block on each heat sink to determine the distortion assessment value of each block. According to the distribution position of each triangulated point cloud on the transformer equipment, the reconstruction weight of each triangulated point cloud is determined based on the degradation contribution and distortion evaluation value. The transformer equipment in the substation scene is reconstructed in three dimensions to construct a substation twin model.

2. The method for constructing a substation twin model based on VR virtual reality technology according to claim 1, characterized in that: The step of calculating the steepness coefficient of each triangular mesh includes: Calculate the discrete degree of the position coordinates of all triangular point clouds in each triangular mesh on the X axis, Y axis and Z axis respectively; The sum of the discrete degrees of each triangular mesh on the X-axis, Y-axis and Z-axis is recorded as the dispersion coefficient; The steepness coefficient is the ratio of the dispersion coefficient to the area of each triangular mesh.

3. The substation twin model construction method based on VR virtual reality technology according to claim 1 is characterized in that: Calculating the coplanarity characterization value of each triangular mesh includes: The remaining triangular meshes that share a common edge with each triangular mesh are recorded as adjacent meshes; the cosine value of the angle between the normal vector of each triangular mesh and each adjacent mesh is calculated, and the cosine value is negatively mapped; The coplanar characterization value is the sum of the negative mapping results of each triangle mesh and all its adjacent meshes.

4. The method for constructing a substation twin model based on VR virtual reality technology according to claim 1, characterized in that: Calculating the local complexity of each triangular mesh includes: Cluster all triangle point clouds on all triangular meshes to obtain the local density of each triangle point cloud, where the cutoff distance of the clustering algorithm is the mean of all edge lengths on all triangular meshes; Calculating the average of the local densities of all triangular point clouds on each triangular mesh as the local density of each triangular mesh; Analyze the clutter of the position coordinates of different triangle point clouds in the neighborhood of all triangle point clouds in each triangle mesh on the X-axis, Y-axis, and Z-axis, and calculate the distribution clutter of each triangle mesh; The local complexity is the product of the distribution disorder and the local density.

5. The method for constructing a substation twin model based on VR virtual reality technology according to claim 4 is characterized in that: The calculation of the distribution clutter of each triangular mesh includes: Taking each triangular point cloud as the center, all triangular point clouds within a spherical neighborhood with the cutoff distance as the radius are combined into a first point cloud set; the first point cloud set of all triangular point clouds on each triangular mesh is unioned to form a second point cloud set; For each triangular mesh, calculating the variance of the position coordinates of all triangular point clouds in the second point cloud set on the X-axis, Y-axis, and Z-axis respectively; The distribution disorder is the sum of the variances on the X-axis, the Y-axis, and the Z-axis.

6. The substation twin model construction method based on VR virtual reality technology according to claim 1 is characterized in that: Determining the degradation assessment value of each triangular mesh includes: The ratio of the local complexity to the coplanar characterization value is recorded as a degradation coefficient; The degradation evaluation value is a product of the steepness coefficient and the degradation coefficient.

7. The method for constructing a substation twin model based on VR virtual reality technology according to claim 1, characterized in that: The degradation contribution V of the kth triangular point cloud k The calculation method is: Among them, P q is the degradation evaluation value of the qth triangle mesh, n q is the number of all triangle point clouds in the qth triangle mesh, Q k is the number of all triangle meshes with the kth triangle point cloud as vertices.

8. The method for constructing a substation twin model based on VR virtual reality technology according to claim 1, characterized in that: The step of extracting all triangulated point clouds corresponding to each heat sink in the transformer device and dividing the triangulated point clouds into blocks includes: All triangulated point clouds on the transformer device are used as input to the 3D semantic segmentation deep learning algorithm to extract all triangulated point clouds corresponding to each heat sink; All triangular point clouds corresponding to each heat sink are divided into multiple blocks according to the corresponding Z-axis coordinates in their position coordinates.

9. The method for constructing a substation twin model based on VR virtual reality technology according to claim 1, characterized in that: Determining the distortion evaluation value of each block includes: Based on all the triangular point clouds in each block of each heat sink, the convex hull algorithm is used to obtain the convex hull of each block and calculate the convex hull volume; Calculate the average value of the degradation contribution of all triangular point clouds corresponding to each heat sink; record the heat sink corresponding to the minimum average value as the standard heat sink; and take the average value of the convex hull volume corresponding to all blocks on the standard heat sink as the standard volume; The difference between the convex hull volume of each block on each heat sink and the standard volume is recorded as the relative deviation; The distortion evaluation value is a ratio of the relative deviation to the standard volume.

10. The method for constructing a substation twin model based on VR virtual reality technology according to claim 1, characterized in that: Determining the reconstruction weight of each triangulated point cloud includes: For all triangular point clouds distributed on the heat sink, calculate the normalized result of the product of the degradation contribution of each triangular point cloud and the distortion evaluation value of the block to which it belongs; perform negative mapping on the normalized result to obtain the reconstruction weight of each triangular point cloud; For the remaining triangular point clouds not distributed on the heat sink, the degradation contribution is normalized, and the normalized result is negatively mapped to obtain the reconstruction weight of each triangular point cloud.

Images