Search

CN-122023651-A - Post-type insulator point cloud modeling system and method

CN122023651ACN 122023651 ACN122023651 ACN 122023651ACN-122023651-A

Abstract

The invention discloses a post-type insulator point cloud modeling system and method, which belong to the technical field of three-dimensional detection of power equipment, wherein a plurality of groups of acquisition units which are distributed annularly and symmetrically are controlled by a microcontroller to acquire original scanning data of a post-type insulator, standardized data are obtained through coordinate transformation, noise elimination preprocessing and umbrella skirt feature extraction, and an ICP algorithm which introduces umbrella skirt constraint is adopted for registration and splicing, and finally, the post-type insulator point cloud model is obtained through visual processing.

Inventors

  • LIU YAN
  • LI HAOYANG
  • Cai Chenbin
  • HU ZEKAI
  • GUO WEIHANG

Assignees

  • 陕西科技大学

Dates

Publication Date
20260512
Application Date
20260123

Claims (10)

  1. 1. The modeling method for the pillar insulator point cloud is characterized by comprising the following steps of: the method comprises the steps that a plurality of acquisition units are controlled by a microcontroller to acquire original scanning data, each acquisition unit comprises a 2D laser radar which is slidably arranged on a vertical sliding table module, the acquisition units are distributed in an annular symmetrical mode by taking the central axis of a pillar-type insulator as a reference, and the original scanning data are acquired by scanning along the height direction of the pillar-type insulator; Preprocessing and feature extraction are sequentially carried out on the original scanning data to obtain standardized data, registration splicing is carried out on the standardized data to obtain point cloud data after registration splicing, the preprocessing comprises coordinate transformation and denoising processing, the feature extraction comprises extraction of umbrella skirt features of a pillar insulator, the registration splicing adopts an ICP registration algorithm, and in the registration process, the umbrella skirt features of the pillar insulator are introduced to construct constraint items; And performing point cloud visualization processing on the point cloud data after registration and splicing to obtain a pillar insulator point cloud model.
  2. 2. The method for modeling the point cloud of the pillar insulator according to claim 1, wherein the plurality of acquisition units is three, and the three acquisition units are annularly and symmetrically distributed at 120 degrees with respect to the central axis of the pillar insulator.
  3. 3. The method for modeling a point cloud of a post-type insulator according to claim 1, wherein the coordinate transformation is to convert polar coordinate data acquired by a 2D laser radar into three-dimensional rectangular coordinate system data; and the noise elimination processing adopts a statistical filtering method, neighborhood points of the three-dimensional rectangular coordinate system data are built layer by layer, and outliers exceeding a set threshold value are removed by calculating neighborhood distances and distance averages to complete noise elimination.
  4. 4. The method for modeling a point cloud of a pillar insulator according to claim 1, wherein the umbrella skirt feature of the pillar insulator comprises an umbrella skirt curvature, an umbrella skirt edge point, an umbrella skirt radius and an umbrella skirt average radius; The constraint item is constructed based on the average radius of the umbrella skirt and the central axis direction vector of the pillar-type insulator; the umbrella skirt curvature is obtained by constructing a point cloud data vector, a neighborhood local centroid vector and a covariance matrix, calculating a characteristic value of the covariance matrix and solving based on a characteristic value duty ratio, wherein the characteristic value is represented by the following formula: Wherein, the Is the curvature of the i-th umbrella skirt, Respectively 3 eigenvalues of covariance matrix moment, and ; The umbrella skirt edge points are obtained by setting an umbrella skirt curvature threshold value and a height range and screening point cloud data meeting the conditions; the umbrella skirt radius is obtained by solving and calculating the average value of the distance from the edge point of the umbrella skirt to the center of the same layer by layer, wherein the average value is shown in the following formula: Wherein, the Is the radius of the ith layer of umbrella skirt, For the number of the edge points of the ith layer of umbrella skirt, To the edge of the umbrella skirt distance of center of circle of the same layer; The average umbrella skirt radius is obtained by solving the average value of the umbrella skirt radius of each layer.
  5. 5. The method for modeling the point cloud of the pillar insulator according to claim 1, wherein the ICP registration splicing based on the constraint term comprises the steps of constructing an optimization target, wherein the optimization target comprises a point cloud distance term, an umbrella skirt average radius constraint term and a central axis direction vector constraint term, and the method is shown in the following formula: Wherein, the For a table-by-table registration splice optimization objective, In order to rotate the matrix is rotated, , Is a regularized weight coefficient which, ,(i=1,2,3), ,(i=2,3,1), Is that Is used to determine the constraint terms of (1), Is that Is used to determine the constraint terms of (1), Is the central axis direction vector of the pillar type insulator, , Is the central axis vector of the data after the i-th noise elimination, ; Carrying out minimum solution on the optimization target by adopting a least square method to obtain an optimized rotation matrix and an optimized translation matrix; And transforming the standardized data based on the rotation matrix and the translation matrix to complete registration and splicing, wherein the registration and splicing is shown in the following formula: 。
  6. 6. the method for modeling a point cloud of a post insulator according to claim 1, wherein the 2D lidar is an MS200 lidar and the vertical slipway module is an FSL40 ball screw slipway module.
  7. 7. The method of modeling a post insulator point cloud of claim 1, wherein the microcontroller is an STM32F103ZET6 microcontroller.
  8. 8. A post insulator point cloud modeling system, comprising: The data acquisition unit is used for controlling a plurality of acquisition units to acquire original scanning data through a microcontroller, wherein a single acquisition unit comprises a 2D laser radar which is slidably arranged on a vertical sliding table module, the acquisition units are symmetrically distributed in a ring shape by taking the central axis of a pillar insulator as a reference, and the original scanning data are acquired by scanning along the height direction of the pillar insulator; The data processing unit is used for sequentially preprocessing the original scanning data and extracting features to obtain standardized data, and carrying out registration splicing on the standardized data to obtain point cloud data after registration splicing, wherein the preprocessing comprises coordinate transformation and denoising processing, the feature extraction comprises extracting the umbrella skirt features of the pillar insulator, the registration splicing adopts an ICP registration algorithm, and an umbrella skirt feature construction constraint item of the pillar insulator is introduced in the registration process; And the data visualization unit is used for performing point cloud visualization processing on the point cloud data after registration and splicing to obtain a pillar insulator point cloud model.
  9. 9. An electronic device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, the processor implementing a pillar insulator point cloud modeling method as claimed in any one of claims 1-7 when the computer program is executed.
  10. 10. A computer readable storage medium, characterized in that the computer readable storage medium stores a computer program which, when executed by a processor, implements a pillar insulator point cloud modeling method as claimed in any one of claims 1 to 7.

Description

Post-type insulator point cloud modeling system and method Technical Field The invention belongs to the technical field of three-dimensional detection of power equipment, and particularly relates to a point cloud modeling system and method for a pillar insulator. Background The pillar insulator is a key component for bearing the dual functions of support and insulation in a power transmission and transformation system, and the geometric quality of the pillar insulator directly influences the electrical performance and operation safety of a circuit. At present, although the standardized production of the pillar insulator can improve the manufacturing efficiency, the current quality system has triple faults in the stage of 'post production-pre use': (1) And detecting faults. When the strut insulator is produced in batches, microscopic geometric variation can be caused by casting stress, die abrasion and material discreteness, and the production line spot inspection only covers part and cannot scan the whole surface; (2) Responsible for faults. The post insulator may be damaged by external force during transportation or storage, and the transportation loss cannot be traced; (3) And (5) data fault. CAD models at the design stage are ideal geometric bodies, but actual products have systematic errors due to process fluctuation. Therefore, if the post insulator is subjected to point cloud modeling in the stage of 'post production-pre use', geometric quality data missing in spot inspection can be captured, a traceable digital twin body can be constructed, comparison reference data can be provided for subsequent operation and maintenance, actual geometric quality parameters can be extracted, and a CAD design library can be reversely calibrated. At present, the system schemes commonly used for modeling the point cloud on the surface of the object include a scheme based on a laser radar and a scheme based on a structured light camera, but the system schemes are in technical and economic paradox when applied to the surface modeling scene of the support insulator. If a low-cost scheme (a single 3D laser radar and a structured light camera) is adopted, 360-degree scanning cannot be performed, and if a high-precision scheme (a 3D laser radar and a structured light camera) is adopted, the high-cost of the scheme is against the economical principle, so that a modeling system and a modeling method with low-cost, 360-degree scanning and high-precision characteristics are needed to be invented. At present, although the existing patent ZL202420717653.0 solves the system appearance problem, the data processing flow and algorithm core of the unprotected system are needed to fill the gap in the technology. Disclosure of Invention The invention aims to solve the problem of how to model a pillar insulator and has the characteristics of low cost, 360-degree scanning and high precision, and provides a pillar insulator point cloud modeling method. In order to achieve the above purpose, the invention adopts the following technical scheme: in a first aspect, the present invention provides a method for modeling a point cloud of a pillar insulator, comprising the steps of: The method comprises the steps that a plurality of acquisition units are controlled by a microcontroller to acquire original scanning data, each acquisition unit comprises a 2D laser radar which is slidably arranged on a vertical sliding table module, the acquisition units are symmetrically distributed in an annular mode by taking the central axis of a pillar-type insulator as a reference, and the acquisition units are scanned along the height direction of the pillar-type insulator to acquire the original scanning data; Preprocessing and feature extraction are sequentially carried out on the original scanning data to obtain standardized data, registration and splicing are carried out on the standardized data to obtain point cloud data after registration and splicing, wherein the preprocessing comprises coordinate transformation and noise elimination processing; And performing point cloud visualization processing on the point cloud data after registration and splicing to obtain a pillar insulator point cloud model. Further, the number of the plurality of acquisition units is three, and the three acquisition units are annularly and symmetrically distributed at 120 degrees by taking the central axis of the pillar insulator as a reference. Further, the coordinate transformation is to convert polar coordinate data acquired by the 2D laser radar into three-dimensional rectangular coordinate system data; the noise elimination process adopts a statistical filtering method, neighborhood points of three-dimensional rectangular coordinate system data are built layer by layer, and outliers exceeding a set threshold value are removed by calculating neighborhood distances and distance averages, so that noise elimination is completed. Further, the umbrella skirt characteristics of the