Method, device and system for identifying key components of power grid equipment and medium

Abstract

The invention provides a method, a device and a system for identifying key components of power grid equipment, and a medium. According to the implementation scheme, point cloud data of key components of power grid equipment are recognized, and a plurality of candidate component point cloud clusters with the same component type are obtained; searching a component related rule corresponding to the component type from a preset knowledge rule base to obtain a target geometric rule and a target topological rule; on the basis of a target geometric rule, performing integrity verification of a component geometric structure on each candidate component point cloud cluster to obtain a plurality of incomplete component point cloud clusters; based on a target topology rule, reasoning a connection relationship between the incomplete component point cloud clusters to perform point cloud cluster connection on the incomplete component point cloud clusters to obtain a target component point cloud cluster; and determining the component type of the key component of the power grid equipment based on the component type corresponding to the target component point cloud cluster. According to the embodiment of the invention, the accuracy of power grid equipment part identification can be improved.

Description

Technical Field

[0001] This invention relates to the field of power grid equipment technology, and in particular to a method, device, system and medium for identifying key components of power grid equipment. Background Technology

[0002] In recent years, with the widespread application of sensing devices such as 3D LiDAR, target recognition technology based on point cloud data has gradually become a core research direction in fields such as autonomous driving, robot perception, intelligent manufacturing, and digital twins. Point cloud data, as a discrete set of points describing a 3D spatial target, can relatively completely reflect the geometric structure and spatial distribution characteristics of an object. Deep learning models, such as PointNet and PointNet++, are widely used in point cloud data recognition due to their powerful feature extraction capabilities.

[0003] However, in real-world applications, due to limitations in scanning angle, occluded objects, and ranging accuracy, the target surface may have local missing parts or overlapping occlusions. Existing PointNet-based models are prone to insufficient feature representation, misidentification, or omission when dealing with these complex situations, resulting in low recognition accuracy. Summary of the Invention

[0004] This invention provides a method, apparatus, system, and medium for identifying key components of power grid equipment, which can solve at least one of the above-mentioned technical problems.

[0005] In a first aspect, embodiments of the present invention provide a method for identifying key components of power grid equipment, comprising:

[0006] Point cloud data of key components of power grid equipment are identified to obtain multiple candidate component point cloud clusters with the same component type;

[0007] The target geometric rules and target topology rules are obtained by searching for the component-related rules corresponding to the component type from the preset knowledge rule base.

[0008] Based on the target geometric rules, the integrity of the component geometric structure of each candidate component point cloud cluster is verified to obtain multiple incomplete component point cloud clusters.

[0009] Based on the target topology rules, the connection relationship between each incomplete component point cloud cluster is inferred, so as to connect each incomplete component point cloud cluster to obtain the target component point cloud cluster;

[0010] Based on the component type corresponding to the target component point cloud cluster, the component type of the key component of the power grid equipment is determined.

[0011] Secondly, embodiments of the present invention provide an identification device for key components of power grid equipment, comprising:

[0012] The identification module is used to identify point cloud data of key components of power grid equipment and obtain multiple candidate component point cloud clusters with the same component type;

[0013] The search module is used to search for component-related rules corresponding to the component type from a preset knowledge rule base to obtain target geometric rules and target topological rules;

[0014] The integrity verification module is used to perform integrity verification of the component geometry structure of each candidate component point cloud cluster based on the target geometry rules, so as to obtain multiple incomplete component point cloud clusters.

[0015] The reasoning module is used to reason about the connection relationship between the point cloud clusters of each incomplete component based on the target topology rules, so as to connect the point cloud clusters of each incomplete component to obtain the point cloud cluster of the target component.

[0016] The component type determination module is used to determine the component type of the key component of the power grid equipment based on the component type corresponding to the target component point cloud cluster.

[0017] Thirdly, embodiments of the present invention also provide an identification system for key components of power grid equipment, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the method described in any one of the embodiments of the present invention.

[0018] Fourthly, embodiments of the present invention also provide a non-transitory computer-readable storage medium storing computer instructions, wherein the computer instructions are used to cause a computer to perform the method described in any one of the embodiments of the present invention.

[0019] This invention employs a method where, after identifying point cloud data of key components of power grid equipment and obtaining multiple candidate component point cloud clusters of the same component type, relevant component rules corresponding to the component type are retrieved from a pre-defined knowledge rule base to obtain target geometric rules and target topology rules. Based on the target geometric rules, the integrity of the component geometric structure of each candidate component point cloud cluster is verified to obtain multiple incomplete component point cloud clusters. Furthermore, based on the target topology rules, the connection relationships between the incomplete component point cloud clusters are inferred to connect them, resulting in the target component point cloud cluster. This method eliminates the influence of local missing or overlapping occlusions in the obtained target component point cloud clusters. Therefore, when determining the component type of key power grid equipment based on the component type corresponding to the target component point cloud cluster, misidentification or omission due to complex scenarios such as local missing or overlapping occlusions can be avoided, thereby improving the accuracy of key component identification in power grid equipment.

[0020] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0021] The accompanying drawings are provided for a better understanding of this solution and do not constitute a limitation of the invention. Wherein:

[0022] Figure 1 This is a flowchart of a method for identifying key components of power grid equipment according to an embodiment of the present invention;

[0023] Figure 2 This is a structural block diagram of a key component identification device for power grid equipment according to an embodiment of the present invention;

[0024] Figure 3 This is a block diagram of an electronic device used to implement the methods of embodiments of the present invention. Detailed Implementation

[0025] The following description, in conjunction with the accompanying drawings, illustrates exemplary embodiments of the present invention, including various details to aid understanding. These details should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope of the invention. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0026] Figure 1 This is a flowchart of a method for identifying key components of power grid equipment according to an embodiment of the present invention.

[0027] like Figure 1As shown, the method for identifying key components of the power grid equipment may include:

[0028] S110, Identify the point cloud data of key components of power grid equipment to obtain multiple candidate component point cloud clusters with the same component type;

[0029] S120: Search for the component-related rules corresponding to the component type from the preset knowledge rule base to obtain the target geometric rules and target topology rules;

[0030] S130, based on the target geometric rules, the integrity of the component geometric structure of each candidate component point cloud cluster is verified to obtain multiple incomplete component point cloud clusters;

[0031] S140, Based on the target topology rules, reason about the connection relationship between the point cloud clusters of each incomplete component, so as to connect the point cloud clusters of each incomplete component to obtain the point cloud cluster of the target component;

[0032] S150, based on the component type corresponding to the target component point cloud cluster, determine the component type of the key components of the power grid equipment.

[0033] In this embodiment of the invention, after identifying point cloud data of key components of power grid equipment and obtaining multiple candidate component point cloud clusters of the same component type, the relevant component rules corresponding to the component type are searched from a preset knowledge rule base to obtain target geometric rules and target topology rules. First, based on the target geometric rules, the integrity of the component geometric structure of each candidate component point cloud cluster is verified to obtain multiple incomplete component point cloud clusters. Then, based on the target topology rules, the connection relationship between the incomplete component point cloud clusters is inferred to connect the incomplete component point cloud clusters to obtain the target component point cloud cluster. Therefore, the target component point cloud cluster obtained by this method can effectively offset the interference caused by problems such as local point cloud missingness, component occlusion, or overlap. Thus, in the process of determining the specific type of key components of power grid equipment according to the component type corresponding to the target component point cloud cluster, the misidentification or omission caused by complex scenarios such as local missingness, occlusion, and overlap can be avoided, thereby significantly improving the accuracy of the identification results of key components of power grid equipment.

[0034] For example, key components of power grid equipment may include circuit breakers, disconnect switches, instrument transformers, insulators, bushings, conductor clamps, switch blades, support columns, etc.

[0035] For example, point cloud data is a collection of three-dimensional coordinate points acquired by a 3D laser scanner, structured light camera, or binocular vision device. Each point cloud data point includes three-dimensional spatial coordinates and optional attributes (such as reflection intensity and color information).

[0036] For example, point cloud data of key components of power grid equipment is preprocessed to reduce data size while retaining key geometric features. This preprocessing includes noise filtering (such as based on statistical filtering, radius filtering, or cloth simulation filtering algorithms) and voxel downsampling techniques.

[0037] In this example, noise filtering can effectively remove outliers and noise points caused by environmental interference or equipment vibration. Voxel downsampling technology can reduce the density of point cloud data while preserving the key geometric features of components, thereby significantly improving the efficiency of subsequent processing.

[0038] For example, preprocessed point cloud data collected from different batches and different devices are unified to the local coordinate system of the target substation through coordinate transformation to ensure that all point cloud data are analyzed under the same spatial reference.

[0039] For example, the point cloud data after coordinate transformation is spatially segmented to obtain multiple point cloud clusters. The geometric features of each point cloud cluster (such as bounding box size, voxel shape, and principal direction vector) are calculated and compared with the geometric features of known component types in a preset knowledge rule base. Point cloud clusters with the same component type are selected from these clusters to obtain multiple candidate component point cloud clusters.

[0040] In this example, a segmentation method based on region growing or a clustering method based on Euclidean distance are used to spatially segment the point cloud data.

[0041] For example, point cloud data of key components of power grid equipment are identified by a deep learning network model, thereby obtaining multiple candidate component point cloud clusters with the same component type.

[0042] In this example, deep learning network models, such as PointNet, PointNet++, Dynamic Graph CNN (DGCNN), Point Transformer (point self-attention mechanism), or Point-BERT (point cloud self-attention model based on mask modeling).

[0043] For example, the pre-set knowledge rule base refers to the rule-based description of the geometric structure and topological connection of various components in power grid equipment, which is a database pre-built based on historical data, design drawings or expert experience.

[0044] For example, geometric rules (Rule-G Insulator):

[0045] {Overall shape: chain-like or rod-like};

[0046] {Overall length: [1.5, 2.5], unit: meter (m)} (for 220 kV voltage level);

[0047] {Main components: Consists of multiple disc-shaped bodies (insulator discs) with diameters of [0.2m, 0.3m] connected in series}.

[0048] Topology rules (Rule-T-Insulator):

[0049] {Connection relationship: One end must be connected to a point cloud cluster of the "Wire" category, and the other end must be connected to a point cloud cluster of the "Tower Crossarm" category};

[0050] {Spatial orientation: The overall axis direction is roughly perpendicular to the ground (the angle with the Z-dimensional coordinate axis is less than 20°)}.

[0051] For example, target geometric rules are used to describe the standardized characteristics of components in terms of shape, size, curvature, symmetry, etc., and are used to verify whether the point cloud clusters of candidate components meet the complete geometric structure.

[0052] For example, the target geometry rules for surge arresters are: {overall shape: columnar, axisymmetric}; {overall height: [1m, 2m]}; {overall diameter: [0.05m, 0.15m]}.

[0053] For example, target topology rules refer to rules that describe topological features such as proximity, connection direction, relative angle, and functional connection between components, and are used to infer reasonable connections between incomplete component point cloud clusters.

[0054] For example, the target topology rule for disconnecting switches (knife switch arms):

[0055] {Connection relationship: One end of the knife switch arm must be connected to a point cloud cluster of the "support column" category};

[0056] {Connection relationship: The other end of the switch arm must be connected to a point cloud cluster of the "contact" category};

[0057] {Spatial angle: The included angle between the switch arm and the support column is [0°, 90°]}.

[0058] For example, a target point cloud cluster refers to a point cloud cluster obtained by reasoning and connecting several candidate point cloud clusters, which can correspond to a complete key component of a power grid device.

[0059] For example, for an insulator string, an incomplete point cloud cluster may only contain some insulator pieces. By using the target topology rules (such as "multiple insulator pieces are connected in sequence along the vertical direction"), the position of the missing piece can be deduced and filled in, thus obtaining a complete insulator string point cloud cluster, i.e., the target point cloud cluster.

[0060] For example, the component types of key components of power grid equipment are determined based on the component types of the target point cloud cluster.

[0061] In this example, firstly, three-dimensional geometric features are extracted from the target point cloud cluster, such as size features (length, height, diameter), shape features (symmetry, curvature distribution), and spatial distribution features (topological relationships between adjacent point cloud clusters). Secondly, the extracted geometric and topological features are compared with the geometric and topological rules of various components in a preset knowledge rule base to determine whether the target point cloud cluster satisfies the feature constraints of a certain component type (feature constraints are component-related rules). If the target point cloud cluster satisfies the geometric and topological rules of a certain component type, it is identified as the corresponding key component of the power grid equipment. If the target point cloud cluster satisfies multiple component rules simultaneously, the optimal component type can be selected based on confidence calculation (e.g., minimum geometric deviation, maximum topological matching degree). Finally, the key component type of the power grid equipment corresponding to the target point cloud cluster is output, such as "insulator string," "disconnect switch," or "surge arrester."

[0062] According to the above implementation method, firstly, the point cloud data of key components of power grid equipment is identified to obtain multiple candidate component point cloud clusters of the same type. Then, the component-related rules corresponding to the component type are searched from a preset knowledge rule base to obtain target geometric rules and target topological rules. The completeness of the component geometric structure of each candidate component point cloud cluster is verified according to the target geometric rules, resulting in multiple incomplete component point cloud clusters. Next, the connection relationships between these incomplete component point cloud clusters are inferred based on the target topological rules, and the point cloud clusters are connected to form the target component point cloud cluster. Through the dual rule constraints of geometric and topological rules, not only can the geometric completeness of a single candidate component point cloud cluster be verified, but the reasonable connections between candidate component point cloud clusters can also be inferred using topological relationships, thereby reducing false identification and missed identification. Thus, even when there are local missing points, occlusions, or overlaps between components in the point cloud data, the complete target component point cloud cluster can still be obtained through topological rules for completion and inference, enhancing the robustness of the algorithm. Finally, the component type of the key components of the power grid equipment is determined according to the component type corresponding to the target component point cloud cluster. By outputting accurate information on the key component types of power grid equipment, reliable basic data support can be provided for subsequent applications such as automatic inspection, digital twin modeling, and condition assessment of power grid equipment.

[0063] In one implementation, based on the target geometric rules, the integrity of the component geometry structure of each candidate component point cloud cluster is verified, including: if the component length of the candidate component point cloud cluster is less than the minimum value of the component length interval threshold in the target geometric rules, then the candidate component point cloud cluster is determined to be an incomplete component point cloud cluster; if the component length of the candidate component point cloud cluster is greater than the maximum value of the component length interval threshold in the target geometric rules, then the candidate component point cloud cluster is determined to be an incorrectly identified component point cloud cluster.

[0064] For example, the candidate component point cloud clusters are filtered and denoised to remove isolated and outlier points, ensuring computational accuracy. Principal Component Analysis (PCA) is used to perform eigenvalue decomposition on the candidate component point cloud clusters, obtaining the principal direction vector, which serves as the geometric principal axis direction. All points of the candidate component point cloud cluster are projected onto the geometric principal axis direction, and the maximum and minimum values ​​after projection are obtained, corresponding to the two endpoints of the candidate component point cloud cluster on the geometric principal axis direction, respectively. The distance between the maximum and minimum projected values ​​is calculated, yielding the geometric length of the candidate component point cloud cluster. The calculated length is compared with the interval threshold of component length in the target geometric rules to determine whether the point cloud cluster belongs to an incomplete component point cloud cluster, an incorrectly identified point cloud cluster, or a complete component point cloud cluster.

[0065] For example, the target geometry rules for insulator strings at a voltage level of 220 kV specify a length range threshold of [1.5m, 2.5m]. If a candidate point cloud cluster has a calculated length of 1.2m, since 1.2m < 1.5m, it is determined to be an incomplete component point cloud cluster (possibly due to incomplete acquisition caused by occlusion). If a candidate point cloud cluster has a length of 3.0m, since 3.0m > 2.5m, it is determined to be an incorrectly identified component point cloud cluster (possibly misidentifying the insulator string and part of the conductor as the same cluster).

[0066] According to the above implementation method, the candidate component point cloud cluster is compared with the component length interval threshold in the target geometric rules to determine whether the candidate component point cloud cluster is an incomplete component point cloud cluster. First, this effectively distinguishes between complete components, incomplete components, and incorrectly identified components, avoiding misclassification of abnormal point cloud clusters as critical components of power grid equipment, thereby improving the overall accuracy of identification. Second, by identifying incomplete component point cloud clusters, input conditions can be provided for subsequent reasoning and connection based on topology rules, thus ensuring that the generated target component point cloud cluster is more complete and reliable.

[0067] In one implementation, based on target topology rules, the connection relationships between various incomplete component point cloud clusters are inferred to connect the various incomplete component point cloud clusters to obtain the target component point cloud cluster. This includes: determining the neighboring component point cloud clusters connected to each incomplete component point cloud cluster; if the component types of the first neighboring component point cloud cluster connected to the first end of the first incomplete component point cloud cluster and the second neighboring component point cloud cluster connected to the second end of the second incomplete component point cloud cluster are consistent with the component types of the neighboring components connected to the first and second ends of the target component as specified by the target topology rules, then based on the axial angle and gap between the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster, the connection relationship between the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster is determined; if the connection relationship between the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster is that a connection relationship exists, then the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster are connected to obtain the target component point cloud cluster.

[0068] For example, a neighboring component point cloud cluster refers to a component point cloud cluster that is spatially connected to the endpoints of an incomplete component point cloud cluster and is within a preset range.

[0069] For example, the component point cloud clusters within the range of 0 meters to 0.6 meters at the upper end of the insulator string are candidate component point cloud clusters connected to the upper end of the insulator string; similarly, the component point cloud clusters within the range of 0 meters to 0.6 meters at the lower end of the insulator string are candidate component point cloud clusters connected to the lower end of the insulator string.

[0070] For example, a principal direction analysis (such as PCA) is performed on the incomplete component point cloud cluster to obtain the principal axis direction, and two endpoint candidate regions are determined along the principal axis direction. Within each endpoint candidate region, other point cloud clusters within a certain range (which can be set according to actual needs, for example, [0.5m, 1m]) are retrieved using a K-Dimensional Tree (KD-Tree) spatial index or a spherical neighborhood search method, and the retrieved point cloud clusters are used as neighboring component point cloud clusters connected to the incomplete component point cloud cluster.

[0071] For example, the first incomplete component point cloud cluster and the second incomplete component point cloud cluster refer to point cloud clusters in which, although the geometric structure of the component is incomplete, its connection with neighboring component point cloud clusters can be determined. The first incomplete component point cloud cluster and the second incomplete component point cloud cluster differ only in their naming; they refer to the same thing.

[0072] For example, a target component refers to a component specified in the target topology rule, and the component type of the component connected to the neighboring component is consistent with the component type of the point cloud cluster of the neighboring component.

[0073] For example, the target topology rule specifies that the component type of the neighboring component connected to the first end of component A is component type B, and the component type of the neighboring component connected to the second end of component A is component type C. If the component type of the first neighboring component point cloud cluster connected to the first incomplete component point cloud cluster at its first end is component type B, and the component type of the second neighboring component point cloud cluster connected to the second incomplete component point cloud cluster at its second end is component type C, it can be seen that both of their first ends are component type B, and both of their second ends are component type C. Therefore, the component types of the first neighboring component point cloud cluster connected to the first end of the first incomplete component point cloud cluster and the second neighboring component point cloud cluster connected to the second end of the second incomplete component point cloud cluster are consistent with the component types of the neighboring components connected to the first and second ends of the target component as specified by the target topology rule. Then, component A is the target component.

[0074] For example, a 220kV insulator string is used as an example. The target topology rule corresponding to the insulator string is specified as follows: {Connection relationship: one end must be connected to the point cloud cluster of the "conductor" category, and the other end must be connected to the point cloud cluster of the "tower crossarm" category}.

[0075] If the analysis results obtained through the main axis analysis are as follows: the component type of the first neighboring component point cloud cluster A1 connected to the first incomplete component point cloud cluster A at its upper end (first end) is "conductor", and the component type of the second neighboring component point cloud cluster B1 connected to the second incomplete component point cloud cluster B at its lower end (second end) is "pole crossarm".

[0076] Therefore, the analysis results are consistent with the component types specified in the target topology rules. Subsequently, the included angle and gap between the first end of the first incomplete component point cloud cluster A and the second end of the second incomplete component point cloud cluster B can be calculated. Based on this included angle and gap, the connection relationship between the first end of the first incomplete component point cloud cluster A and the second end of the second incomplete component point cloud cluster B can be determined. If a connection relationship exists between the first end of the first incomplete component point cloud cluster A and the second end of the second incomplete component point cloud cluster B, then point cloud cluster connections are made between the first end of the first incomplete component point cloud cluster A and the second end of the second incomplete component point cloud cluster B to obtain the target component point cloud cluster (insulator string).

[0077] According to the above implementation method, the connection relationship between incomplete component point cloud clusters is determined by comprehensively considering the type constraints, axial angles, and gaps of adjacent component point cloud clusters. This ensures that splicing or connecting point cloud clusters only occurs under a reasonable topology, thereby reducing the occurrence of incorrect connections. Thus, in complex power grid scenarios, even if components have local deficiencies, occlusions, or overlaps, the true component structure can be restored through rule-based reasoning and connection operations.

[0078] In one embodiment, determining the connection relationship between the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster based on the axial angle and gap between the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster includes: if the axial angle between the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster is less than a preset axial angle threshold, and the gap between the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster is equal to the component thickness of the target component, then the connection relationship between the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster is determined to exist; if the axial angle between the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster is greater than or equal to a preset axial angle threshold, and the gap between the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster is less than / or greater than the component thickness of the target component, then the connection relationship between the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster is determined to not exist.

[0079] For example, the axial angle threshold can be set according to actual needs, for example, 5°.

[0080] For example, the thickness of the target component can be predefined in a knowledge rule base according to national standards and equipment manuals. For instance, if the target component is a crossarm, its thickness can be set to 0.4 meters.

[0081] According to the above implementation method, the connection between the first incomplete component point cloud cluster and the second incomplete component point cloud cluster is determined based on the axial angle and gap between the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster. This method provides a clear quantitative standard for the computer to automatically determine whether incomplete components are connected. It can automatically determine the connection relationship of components without manual intervention, improving the automation level of point cloud recognition processing. Specifically, if the axial angle between the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster is less than a preset axial angle threshold and the gap between them is equal to the thickness of the target component, then the connection relationship between them is determined to exist. Conversely, if the axial angle between the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster is greater than or equal to the preset axial angle threshold and the gap between them is less than / or greater than the thickness of the target component, then the connection relationship between them is determined to not exist. By combining the axial angle and gap for determination, erroneous splicing caused by excessively large or small distances can be effectively prevented, reducing the probability of identifying incorrect component point cloud clusters.

[0082] In one embodiment, the method further includes: performing decentralization calculations on the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster to obtain the decentralized points of the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster; calculating the decentralized points of the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster based on the covariance calculation formula to obtain the covariance matrix of the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster; and performing decentralization calculations on the covariance matrix of the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster. Eigenvalue decomposition is performed to obtain the axial directions of the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster. The axial directions of the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster are normalized to obtain the normalized direction vectors of the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster. Based on the inverse cosine function, the absolute value of the product of the normalized direction vectors of the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster is processed to obtain the angle between the axial directions of the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster.

[0083] For example, the point set at the first end of the first incomplete component point cloud cluster A is represented as: Where, p A，i This represents the three-dimensional coordinate vector of the i-th point in the first end of the point cloud cluster A of the first incomplete component, i = 1, 2, ..., n. A ;nA Let represent the total number of points at the first end of the point cloud cluster A, which is the first incomplete component. The point set at the second end of the point cloud cluster B, which is the second incomplete component, is represented as . Where, p B，j This represents the three-dimensional coordinate vector of the i-th point in the second end of the point cloud cluster B of the second incomplete component, j = 1, 2, ..., n. B ;n B This represents the total number of midpoints at the second end of the point cloud cluster B, which is the second incomplete component.

[0084] For example, the decentralized computation of the first end of the first incomplete component point cloud cluster A can be represented by a function expression as follows:

[0085]

[0086] q A，i =p A，i -c A .

[0087] In the formula, c A q represents the decentralized point (centroid) at the first end of the first incomplete component point cloud cluster A; A，i This represents the i-th point after the first end of the point cloud cluster A of the first incomplete component is decentralized, i.e., the origin p. A，i Relative to the center of mass c A The offset vector.

[0088] For example, the decentralized computation of the second end of the second incomplete component point cloud cluster B can be represented by a function expression as follows:

[0089]

[0090] q B，j =p B，j -c B .

[0091] In the formula, c B q represents the decentralized point (centroid) at the second end of the point cloud cluster B of the second incomplete component; B，i This represents the i-th point after the second end of the point cloud cluster B of the second incomplete component is decentered, i.e., the origin p. B，i Relative to the center of mass c B The offset vector.

[0092] For example, the covariance matrix of the first end of the first incomplete component point cloud cluster A is determined by summing and averaging the outer products of the decentralized points at the first end of the first incomplete component point cloud cluster A. Similarly, the covariance matrix of the second end of the second incomplete component point cloud cluster B is determined by summing and averaging the outer products of the decentralized points at the second end of the second incomplete component point cloud cluster B.

[0093] For example, eigenvalue decomposition is performed on the covariance matrix of the first end of the first incomplete component point cloud cluster A, and the eigenvector corresponding to the largest eigenvalue is taken as the axial direction of the first end of the first incomplete component point cloud cluster A. Similarly, eigenvalue decomposition is performed on the covariance matrix of the second end of the second incomplete component point cloud cluster B, and the eigenvector corresponding to the largest eigenvalue is taken as the axial direction of the second end of the second incomplete component point cloud cluster B.

[0094] For example, the angle between the axes of the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster can be expressed by a function as follows:

[0095]

[0096] In the formula, v′ A v represents the unit vector indicating the axial direction of the first end of the first incomplete component point cloud cluster (i.e., the eigenvector corresponding to the largest eigenvalue); A Indicates the axial direction of the first end of the first incomplete component point cloud cluster; ||v A || represents the eigenvector v A The modulus (length); v′ B The unit vector representing the axial direction of the second end of the point cloud cluster of the second incomplete component (i.e., the eigenvector corresponding to the largest eigenvalue); v B Indicates the axial direction of the second end of the point cloud cluster of the second incomplete component; ||v B || represents the eigenvector v B The modulus (length); θ represents the angle between the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster; |v′ A ·v′ B That is, the absolute value of the dot product of two unit vectors represents the absolute value of the cosine similarity of the two axes.

[0097] According to the above implementation method, firstly, decentralization calculations are performed on the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster to obtain their respective decentralized points. Then, based on the covariance calculation formula, calculations are performed on these two decentralized points to obtain their covariance matrices. Next, eigenvalue decomposition is performed on the two covariance matrices to obtain the axial directions of the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster. Decentralization and covariance matrix calculation can eliminate the offset effect of the end point cloud distribution. Subsequently, these two axial directions are normalized to obtain the corresponding normalized direction vectors. Finally, the absolute value of the product of these two normalized direction vectors is calculated using the inverse cosine function to obtain the axial angle between the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster. Normalizing the axis direction eliminates the influence of vector "length differences" on angle calculation (only direction information is retained). By using the inverse cosine function to calculate the absolute value of vector product, the abstract vector direction relationship can be transformed into an intuitive and quantifiable angle value (range of 0-90°). This not only conforms to geometric logic but also facilitates subsequent judgment based on "angle thresholds" to determine whether two components meet the topological rules of the connection target (such as whether they are collinear or at a specific angle), providing a clear basis for accurate connection of incomplete point cloud clusters.

[0098] In one embodiment, the method further includes: obtaining the maximum and minimum values ​​of the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster on each dimensional coordinate axis in a three-dimensional coordinate system; for each dimensional coordinate axis, determining a first distance between the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster on the dimensional coordinate axis based on the difference between the minimum value of the first end of the first incomplete component point cloud cluster on the dimensional coordinate axis and the maximum value of the second end of the second incomplete component point cloud cluster on the corresponding dimensional coordinate axis; and determining the gap between the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster based on the maximum value of the first distance between the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster on each dimensional coordinate axis.

[0099] For example, suppose for the first end of the first incomplete component point cloud cluster A:

[0100] The minimum value on the x-axis is have Millimeter (mm), with a maximum value on the x-axis of 1. have

[0101] The minimum value on the y-axis is have The maximum value on the y-axis is have

[0102] The minimum value on the z-axis is have The maximum value on the z-axis is have

[0103] For the second end of the second incomplete component point cloud cluster B:

[0104] The minimum value on the x-axis is have The maximum value on the x-axis is millimeters. have

[0105] The minimum value on the y-axis is have The maximum value on the y-axis is have

[0106] The minimum value on the z-axis is have The maximum value on the z-axis is have

[0107] For example, the first distance between the two (the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster) on each dimensional coordinate axis is determined.

[0108] In this example, for the x-axis:

[0109] In the formula, gap x This represents the first distance on the x-dimensional coordinate axis between the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster.

[0110] Similarly, for the y-axis:

[0111] In the formula, gap y This represents the first distance on the y-dimensional coordinate axis between the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster.

[0112] For the z-axis:

[0113] In the formula, gap zThis represents the first distance on the z-dimensional coordinate axis between the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster.

[0114] For example, the gap between the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster is determined based on the maximum value of the first distance between the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster on each dimensional coordinate axis.

[0115] In this example, if the gap between the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster is gap, then gap = max[gap]. x gap y gap z ] = max[5, 0, 0] = 5.

[0116] According to the above implementation method, firstly, the maximum and minimum values ​​of the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster on each dimensional coordinate axis are obtained. Then, for each dimensional coordinate axis, the difference between the minimum value of the first end of the first incomplete component point cloud cluster on that axis and the maximum value of the second end of the second incomplete component point cloud cluster on the corresponding axis is calculated to obtain the first distance between them on that dimensional coordinate axis. Finally, the maximum value of the first distance on each dimensional coordinate axis is taken as the gap between the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster. By processing each dimension of the 3D coordinate system separately, the problem of ignoring gaps in other directions in single-dimensional calculation is avoided, and the positional relationship between the endpoints of the two components in 3D space can be completely covered.

[0117] In one embodiment, the method further includes: identifying the component type, target geometric rules, and target topology rules corresponding to the target component point cloud cluster, and encapsulating the bounding box of the target component point cloud cluster in the local coordinate system to obtain standard structure data; and determining the identification results of key components of the power grid equipment based on the standard structure data.

[0118] For example, standard structure data refers to a unified data format formed after standardizing the description of the point cloud clusters of the target component.

[0119] According to the above implementation method, the component type corresponding to the target component point cloud cluster, the identifiers of the target geometric rules and the target topological rules, and the bounding box encapsulation of the target component point cloud cluster in the local coordinate system are packaged into a standard JSON (JavaScript Object Notation) object (i.e., standard structured data). This standard JSON object is then used as the identification result of key components of power grid equipment. In this way, the standardized expression of point cloud data of key components of power grid equipment is realized, and it also facilitates unified calling between different algorithm modules and systems in the future.

[0120] Figure 2 This is a structural block diagram of a device for identifying key components of power grid equipment according to an embodiment of the present invention.

[0121] like Figure 2 As shown, the identification device for key components of the power grid equipment may include:

[0122] The identification module 510 is used to identify the point cloud data of key components of power grid equipment and obtain multiple candidate component point cloud clusters with the same component type.

[0123] The search module 520 is used to search for the component-related rules corresponding to the component type from a preset knowledge rule base to obtain the target geometric rules and the target topological rules.

[0124] The integrity verification module 530 is used to perform integrity verification of the component geometry structure of each candidate component point cloud cluster based on the target geometry rules, so as to obtain multiple incomplete component point cloud clusters.

[0125] The reasoning module 540 is used to reason about the connection relationship between the incomplete component point cloud clusters based on the target topology rules, so as to connect the incomplete component point cloud clusters to obtain the target component point cloud clusters.

[0126] The component type determination module 550 is used to determine the component type of the key component of the power grid equipment based on the component type corresponding to the target component point cloud cluster.

[0127] In one embodiment, the integrity verification module includes:

[0128] The incomplete component point cloud cluster determination unit is used to determine the candidate component point cloud cluster as an incomplete component point cloud cluster if the component length of the candidate component point cloud cluster is less than the minimum value of the component length interval threshold in the target geometric rule.

[0129] The error component point cloud cluster identification unit is used to determine the candidate component point cloud cluster as an error component point cloud cluster if the component length of the candidate component point cloud cluster is greater than the maximum value of the component length interval threshold in the target geometric rule.

[0130] In one implementation, the inference module includes:

[0131] The neighboring component point cloud cluster determination unit is used to determine the neighboring component point cloud clusters connected to each of the incomplete component point cloud clusters.

[0132] The connection relationship determination unit is used to determine the connection relationship between the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster if the component types of the first neighboring component point cloud cluster connected to the first end of the first incomplete component point cloud cluster and the second neighboring component point cloud cluster connected to the second end of the second incomplete component point cloud cluster in each of the incomplete component point cloud clusters are consistent with the component types of the neighboring components connected to the first end and the second end of the target component as specified by the target topology rule.

[0133] The point cloud cluster connection unit is used to connect the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster to obtain the target component point cloud cluster if there is a connection relationship between the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster.

[0134] In one embodiment, determining the connection relationship between the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster based on the included angle and gap between the axes of the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster is specifically used for:

[0135] If the axial angle between the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster is less than a preset axial angle threshold, and the gap between the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster is equal to the component thickness of the target component, then the connection relationship between the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster is determined to exist.

[0136] If the axial angle between the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster is greater than or equal to a preset axial angle threshold, and the gap between the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster is less than / or greater than the component thickness of the target component, then the connection relationship between the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster is determined to be non-existent.

[0137] In one embodiment, the identification device for key components of the power grid equipment further includes:

[0138] A decentralized computing unit is used to perform decentralized computing on the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster respectively, to obtain the decentralized points of the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster.

[0139] The covariance matrix unit is used to calculate the covariance matrix of the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster based on the covariance calculation formula.

[0140] The eigenvalue decomposition unit is used to perform eigenvalue decomposition on the covariance matrices of the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster, respectively, to obtain the axial directions of the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster.

[0141] The normalization unit is used to normalize the axial directions of the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster, respectively, to obtain the normalized direction vectors of the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster.

[0142] The axis angle unit is used to process the absolute value of the normalized direction vector product of the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster based on the inverse cosine function, so as to obtain the axis angle between the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster.

[0143] In one embodiment, the identification device for key components of the power grid equipment further includes:

[0144] The acquisition unit is used to acquire the maximum and minimum values ​​of the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster on each dimensional coordinate axis in a three-dimensional coordinate system.

[0145] The first distance determination unit is used to determine, for each of the dimensional coordinate axes, the first distance between the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster on the dimensional coordinate axis based on the difference between the minimum value of the first end of the first incomplete component point cloud cluster on the dimensional coordinate axis and the maximum value of the second end of the second incomplete component point cloud cluster on the corresponding dimensional coordinate axis.

[0146] The gap determination unit is used to determine the gap between the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster based on the maximum value of the first distance between the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster on each dimensional coordinate axis.

[0147] In one embodiment, the identification device for key components of the power grid equipment further includes:

[0148] The encapsulation module is used to encapsulate the component type corresponding to the target component point cloud cluster, the identifiers of the target geometric rules and the target topology rules, and the bounding box of the target component point cloud cluster in the local coordinate system to obtain standard structure data.

[0149] The identification result determination module is used to determine the identification result of the key components of the power grid equipment based on the standard structure data.

[0150] The specific functions and examples of each module and submodule of the system in this embodiment of the invention can be found in the relevant descriptions of the corresponding steps in the above method embodiments, and will not be repeated here.

[0151] The acquisition, storage, and application of user personal information involved in the technical solution of this invention all comply with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0152] This invention also provides a system for identifying key components of power grid equipment, comprising:

[0153] At least one processor; and a memory communicatively connected to said at least one processor;

[0154] The memory stores instructions that can be executed by the at least one processor, which, when executed by the at least one processor, enables the at least one processor to perform the method described in any one of the embodiments of the present invention.

[0155] The beneficial effects of the identification system for key components of power grid equipment in this embodiment of the invention are equivalent to the beneficial effects of the identification method for key components of power grid equipment described above, and will not be repeated here.

[0156] This invention also provides a non-transitory computer-readable storage medium storing computer instructions, wherein the computer instructions are used to cause a computer to perform the method described in any one of the embodiments of this invention.

[0157] The beneficial effects of the storage medium of the present invention are equivalent to the beneficial effects of the above-described method for identifying key components of power grid equipment, and will not be repeated here.

[0158] Figure 3 A schematic block diagram of an example electronic device 800 that can be used to implement embodiments of the present invention is shown. Electronic device 800 is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. Electronic device 800 may also represent various forms of mobile devices, such as personal digital assistants, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0159] like Figure 3 As shown, the electronic device 800 includes a computing unit 801, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 802 or a computer program loaded from a storage unit 808 into a random access memory (RAM) 803. The RAM 803 may also store various programs and data required for the operation of the electronic device 800. The computing unit 801, ROM 802, and RAM 803 are interconnected via a bus 804. An input / output (I / O) interface 805 is also connected to the bus 804.

[0160] Multiple components in electronic device 800 are connected to I / O interface 805, including: input unit 806, such as keyboard, mouse, etc.; output unit 807, such as various types of displays, speakers, etc.; storage unit 808, such as disk, optical disk, etc.; and communication unit 809, such as network card, modem, wireless transceiver, etc. Communication unit 809 allows electronic device 800 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0161] The computing unit 801 can be various general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 801 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit 801 performs the various methods and processes described above, such as the method for identifying critical components of power grid equipment. For example, in some embodiments, the method for identifying critical components of power grid equipment can be implemented as a computer software program, which is tangibly contained in a machine-readable medium, such as storage unit 808. In some embodiments, part or all of the computer program can be loaded and / or installed on the electronic device 800 via ROM 802 and / or communication unit 809. When the computer program is loaded into RAM 803 and executed by the computing unit 801, one or more steps of the method for identifying critical components of power grid equipment described above can be performed. Alternatively, in other embodiments, the computing unit 801 may be configured by any other suitable means (e.g., by means of firmware) to perform a method for identifying critical components of the power grid equipment.

[0162] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0163] The program code used to implement the methods of the present invention can be written in any combination of one or more programming languages. This program code can be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable data processing device, such that when executed by the processor or controller, the program code causes the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code can be executed entirely on the machine, partially on the machine, as a standalone software package partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0164] In the context of this invention, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. Machine-readable media can include, but are not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0165] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device for displaying information to the user (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor); and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the computer. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0166] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as a data server), or computing systems that include middleware components (e.g., an application server), or computing systems that include frontend components (e.g., a user computer with a graphical user interface or web browser through which a user can interact with embodiments of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., a communication network). Examples of communication networks include local area networks (LANs), wide area networks (WANs), and the Internet.

[0167] Computer systems can include clients and servers. Clients and servers are generally located far apart and typically interact via communication networks. Client-server relationships are created by computer programs running on the respective computers and having a client-server relationship with each other. Servers can be cloud servers, servers in distributed systems, or servers incorporating blockchain technology.

[0168] It should be understood that the various forms of processes shown above can be used to reorder, add, or delete steps. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution disclosed in this invention can be achieved, and this is not limited herein.

[0169] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the principles of this invention should be included within the scope of protection of this invention.

Claims

1. A method for identifying key components of power grid equipment, characterized in that, include: Point cloud data of key components of power grid equipment are identified to obtain multiple candidate component point cloud clusters with the same component type; The target geometric rules and target topology rules are obtained by searching for the component-related rules corresponding to the component type from the preset knowledge rule base. Based on the target geometric rules, the integrity of the component geometric structure of each candidate component point cloud cluster is verified to obtain multiple incomplete component point cloud clusters. Based on the target topology rules, the connection relationship between each incomplete component point cloud cluster is inferred, so as to connect each incomplete component point cloud cluster to obtain the target component point cloud cluster; Based on the component type corresponding to the target component point cloud cluster, the component type of the key component of the power grid equipment is determined.

2. The method according to claim 1, characterized in that, The step of verifying the integrity of the component geometry structure for each candidate component point cloud cluster based on the target geometry rules includes: If the component length of the candidate component point cloud cluster is less than the minimum value of the component length interval threshold in the target geometric rule, then the candidate component point cloud cluster is determined to be an incomplete component point cloud cluster. If the component length of the candidate component point cloud cluster is greater than the maximum value of the component length interval threshold in the target geometric rule, then the candidate component point cloud cluster is determined to be an incorrectly identified component point cloud cluster.

3. The method according to claim 1, characterized in that, The step of reasoning about the connection relationships between the incomplete component point cloud clusters based on the target topology rules, and connecting the point cloud clusters of the incomplete components to obtain the target component point cloud cluster, includes: Determine the neighboring component point cloud clusters connected to each of the incomplete component point cloud clusters; If the component types of the first neighboring component point cloud cluster connected to the first end of the first incomplete component point cloud cluster and the second neighboring component point cloud cluster connected to the second end of the second incomplete component point cloud cluster in each of the incomplete component point cloud clusters are consistent with the component types of the neighboring components connected to the first and second ends of the target component as specified by the target topology rule, then the connection relationship between the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster is determined based on the axial angle and gap between the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster. If there is a connection between the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster, then the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster are connected to form a point cloud cluster to obtain the target component point cloud cluster.

4. The method according to claim 3, characterized in that, The determination of the connection relationship between the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster based on the included angle and gap between the axes of the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster includes: If the axial angle between the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster is less than a preset axial angle threshold, and the gap between the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster is equal to the component thickness of the target component, then the connection relationship between the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster is determined to exist. If the axial angle between the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster is greater than or equal to a preset axial angle threshold, and the gap between the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster is less than / or greater than the component thickness of the target component, then the connection relationship between the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster is determined to be non-existent.

5. The method according to claim 4, characterized in that, Also includes: Decentralization calculations are performed on the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster to obtain the decentralized points of the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster. Based on the covariance calculation formula, the points after decentering the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster are calculated to obtain the covariance matrix of the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster. Eigenvalue decomposition is performed on the covariance matrices of the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster to obtain the axial directions of the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster. The axial directions of the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster are normalized respectively to obtain the normalized direction vectors of the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster. Based on the inverse cosine function, the absolute value of the product of the normalized direction vectors of the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster is processed to obtain the included angle of the axis between the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster.

6. The method according to claim 4, characterized in that, Also includes: Obtain the maximum and minimum values ​​of the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster on each coordinate axis in a three-dimensional coordinate system. For each of the aforementioned dimensional coordinate axes, based on the difference between the minimum value of the first end of the first incomplete component point cloud cluster on the dimensional coordinate axis and the maximum value of the second end of the second incomplete component point cloud cluster on the corresponding dimensional coordinate axis, a first distance on the dimensional coordinate axis between the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster is determined. Based on the maximum value of the first distance between the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster on each dimensional coordinate axis, the gap between the first end of the first incomplete component point cloud cluster and the second end of the second incomplete component point cloud cluster is determined.

7. The method according to claim 1, characterized in that, Also includes: The component type corresponding to the target component point cloud cluster, the identifiers of the target geometric rules and the target topology rules, and the bounding box of the target component point cloud cluster in the local coordinate system are encapsulated to obtain standard structural data; Based on the standard structure data, the identification results of the key components of the power grid equipment are determined.

8. A device for identifying key components of power grid equipment, characterized in that, include: The identification module is used to identify point cloud data of key components of power grid equipment and obtain multiple candidate component point cloud clusters with the same component type; The search module is used to search for component-related rules corresponding to the component type from a preset knowledge rule base to obtain target geometric rules and target topological rules; The integrity verification module is used to perform integrity verification of the component geometry structure of each candidate component point cloud cluster based on the target geometry rules, so as to obtain multiple incomplete component point cloud clusters. The reasoning module is used to reason about the connection relationship between the point cloud clusters of each incomplete component based on the target topology rules, so as to connect the point cloud clusters of each incomplete component to obtain the point cloud cluster of the target component. The component type determination module is used to determine the component type of the key component of the power grid equipment based on the component type corresponding to the target component point cloud cluster.

9. A system for identifying key components of power grid equipment, characterized in that, include: At least one processor; and a memory communicatively connected to the at least one processor; The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-7.

10. A non-transitory computer-readable storage medium storing computer instructions, characterized in that, The computer instructions are used to cause the computer to perform the method according to any one of claims 1-7.

Images