An insulator size deviation analysis method and system based on point cloud data

Abstract

The application belongs to the technical field of image processing, and particularly relates to an insulator size deviation analysis method and system based on point cloud data, which comprises the following steps: acquiring insulator point clouds and performing slicing processing, and calculating the morphological distortion index of each slice; based on a gravity torque distribution model, the distortion index is normalized into a gravity impedance coefficient, and the gravity compliance confidence is determined according to the difference value of the impedance coefficients of adjacent slices; the confidence is used to drive the standard CAD model skeleton to perform weighted deformation, a deformed reference model is generated, and the point clouds are registered to obtain the size deviation. The application can distinguish the natural falling of the insulator caused by gravity from structural defects, eliminate the gravity bending interference by constructing an adaptive reference model, and realize high-precision detection of abnormal conditions such as insulator pollution, icing or damage.

Description

Technical Field

[0001] This invention relates to the field of image processing technology. More specifically, this invention relates to a method and system for analyzing insulator size deviations based on point cloud data. Background Technology

[0002] Insulators are critical insulation and support components in high-voltage transmission lines, and their physical condition directly affects the safe operation of the power grid. With the development of 3D vision technology, acquiring high-precision point cloud data of insulator strings using laser scanning equipment and comparing it with standard CAD models to analyze dimensional deviations such as thickened dirt accumulation, damaged skirts, or missing materials has become an important technical means for intelligent inspection. In performing this type of analysis, to spatially align the standard model with the measured data, existing mainstream technologies typically employ deformation algorithms based on skeleton extraction or iterative nearest-point registration algorithms to calculate the central skeleton line of the point cloud and then bend and deform the standard model accordingly to fit the point cloud data.

[0003] However, in actual service scenarios, insulator strings, influenced by their own gravity, exhibit a non-linear, catenary-like natural droop. Existing algorithms have the following drawbacks when handling such long, flexible objects: they typically fit only the similarity of geometric shapes, lacking consideration of the distribution of physical torques, thus failing to effectively distinguish between normal elastic bending caused by gravity and abnormal distortion caused by structural damage. Specifically, if the deformation algorithm is too compliant with the geometric features of the point cloud, the standard model may incorrectly deform along with the defective area, causing the true dimensional deviation to be masked by model deformation; conversely, if the compliance is insufficient, the overall natural bending caused by gravity will be misjudged as a large-scale dimensional deviation, leading to numerous false alarms and difficulty in accurately locating the true fault point. Summary of the Invention

[0004] To address the technical problem of poor analysis results for insulator size deviations, the present invention provides solutions in the following aspects.

[0005] In a first aspect, the present invention provides a method for analyzing insulator size deviations based on point cloud data, comprising:

[0006] Point cloud data of the insulator string is acquired and sliced ​​to obtain multiple point cloud slices. The centroid coordinates of each point cloud slice are obtained, and the morphological distortion index of each point cloud slice is calculated based on the centroid coordinates of adjacent point cloud slices and the distribution of data points within the point cloud slice. Based on the gravity torque distribution model of the insulator string, the morphological distortion index of each point cloud slice is normalized to obtain the gravity impedance coefficient of each point cloud slice. The gravity compliance confidence of each point cloud slice is determined based on the difference in the gravity impedance coefficients of adjacent point cloud slices. In response to the gravity compliance confidence of each point cloud slice, the skeleton of the standard CAD model is subjected to weighted deformation to generate a deformable reference model. The point cloud data is registered with the deformable reference model to obtain the dimensional deviation of the insulator string.

[0007] This invention introduces a gravity moment distribution model to calculate the gravity impedance coefficient, which can effectively distinguish between normal physical bending caused by gravity and abnormal distortion caused by structural damage. By calculating the gravity compliance confidence level, the skeleton of the standard CAD model is driven to undergo adaptive weighted deformation, so that it actively follows the bending of the point cloud in the normal region to eliminate pose errors, while maintaining rigidity in the defect region to retain detection margin. This generates a deformation reference model that conforms to the macroscopic gravity posture and is not skewed by local defects, thus achieving high-precision insulator size deviation analysis.

[0008] Preferably, the slicing process for the point cloud data includes:

[0009] Calculate the total projection length of the point cloud data on the main axis; obtain the total number of skirts of the standard insulator string; obtain the distance from the edge of the skirt to the central axis, and record it as the standard skirt radius of the insulator; record the ratio of the total projection length to the total number of skirts as the slicing step size;

[0010] Based on the slicing step size, the point cloud data is initially divided into multiple candidate slice intervals along the main axis;

[0011] For any candidate slice interval, calculate its point cloud density histogram along the principal axis, search for the peak coordinate of the histogram, and define the peak coordinate as the calibration center of the slice; using the calibration center as a reference, extend the slice step length to both sides. Determine the final slice boundaries; obtain multiple point cloud slices.

[0012] This invention calculates the slicing step size by determining the ratio of the total projected length to the number of skirts, effectively compensating for the axial compression effect caused by gravity. Simultaneously, it uses the peak value of the point cloud density histogram as a calibration center, eliminating nonlinear errors caused by uneven local curvature. This ensures that the slicing window can accurately align with each actual skirt unit, providing a reliable spatial reference for the subsequent accurate extraction of the geometric features of each slice.

[0013] Preferably, the method for calculating the morphological distortion index includes:

[0014] Calculate the first The centroid of a point cloud slice to the first The line vector connecting the centroids of the nth point cloud slice is denoted as the nth line vector. The forward connection vector of the nth point cloud slice is calculated; The centroid of a point cloud slice to the first The line vector connecting the centroids of the nth point cloud slice is denoted as the nth line vector. The backward connection vectors of each point cloud slice are calculated; the cosine of the angle between the forward connection vector and the backward connection vector is calculated.

[0015] Statistics All data points within a point cloud slice up to the first... The Euclidean distance of the centroids of the nth point cloud slices constitutes the nth point cloud slice. The distance set of a point cloud slice is calculated, and the standard deviation of the distance set is calculated; the morphological distortion index is negatively correlated with the cosine value of the included angle and positively correlated with the standard deviation.

[0016] This invention integrates the curvature of the macroscopic axis and the dispersion of the microscopic cross-section to construct a morphological distortion index. By combining the angle variation between adjacent slices and the dispersion of data points within the slices, it can comprehensively capture abnormal changes in the geometry of the insulator, providing fundamental data support for subsequently distinguishing between the effects of gravity and actual defects.

[0017] Preferably, the morphological distortion index satisfies the expression:

[0018] ;

[0019] In the formula, Indicates the first The morphological distortion index of a point cloud slice; Indicates the first Forward connection vectors of point cloud slices; Indicates the first The backward connection vector of a point cloud slice; Indicates the first The standard deviation of the distance set of a point cloud slice; Indicates the standard shed radius of the insulator; Indicates the dot product symbol; The symbol representing the magnitude of a vector.

[0020] Preferably, the method for obtaining the gravitational impedance coefficient includes:

[0021] The product of the total number of umbrella skirts and the slice step size is denoted as the theoretical total length of the insulator string;

[0022] Multiply the absolute difference between half the total number of umbrella skirts and i by the slice step size to obtain the i-th slice. The gravitational moment arm length of each point cloud slice; calculate the arithmetic mean of the morphological distortion indices of all point cloud slices, denoted as the global morphological distortion factor;

[0023] The first The product of the morphological distortion index of the i-th point cloud slice and the theoretical total length of the insulator string is denoted as the first product; the product of the gravitational moment arm length of the i-th point cloud slice and the global morphological distortion factor is denoted as the second product; the gravitational impedance coefficient of the i-th point cloud slice is positively correlated with the first product and negatively correlated with the second product.

[0024] This invention utilizes the moment distribution characteristics of insulators as flexible beams to construct a gravity reference model that varies with position. By calculating the gravity impedance coefficient, the observed morphological distortion is compared with the theoretical gravity moment. This coefficient eliminates the influence of gravity trends, ensuring its stability in normal gravity-induced sagging regions while exhibiting drastic fluctuations in abnormal defect regions. This allows for the verification and decomposition of the distortion index using physical moment laws.

[0025] Preferably, the acquisition of the gravity compliance confidence level includes:

[0026] Calculate the difference between the gravitational impedance coefficient of the i-th point cloud slice and the gravitational impedance coefficient of the (i-1)-th point cloud slice, and denote it as the impedance difference value of the i-th point cloud slice; count the impedance difference values ​​of all point cloud slices of the insulator string, calculate the standard deviation of all impedance difference values, and denote it as the adaptive impedance bandwidth.

[0027] The gravity compliance confidence of any point cloud slice satisfies the expression:

[0028] ;

[0029] In the formula, This represents the gravity compliance confidence of the i-th point cloud slice; This represents the impedance difference value of the i-th point cloud slice; Indicates the adaptive impedance bandwidth; Represents the natural exponential function; Represents the absolute value symbol.

[0030] This invention determines the adaptive impedance bandwidth by statistically analyzing the standard deviation of the difference in gravity impedance coefficients and constructs a gravity compliance confidence level using a Gaussian function, thus realizing a soft-switching control mechanism. This mechanism can automatically adjust the sensitivity to anomalies based on the overall stability of the current insulator string, distinguishing whether each slice position is more in line with gravity laws or defect characteristics, providing a precise weighting basis for the differentiated deformation of subsequent models.

[0031] Preferably, the weighted deformation of the skeleton of the standard CAD model includes:

[0032] Obtain the coordinates of the center axis at each slice location in the standard CAD model as ideal skeleton points; take the weighted average of the centroid coordinates of the corresponding point cloud slice and the ideal skeleton points as deformable skeleton points, where the weight of the centroid coordinates is the gravity compliance confidence of the corresponding point cloud slice, and the weight of the ideal skeleton points is the difference between 1 and the gravity compliance confidence.

[0033] Preferably, obtaining the dimensional deviation of the insulator string includes:

[0034] Calculate the shortest Euclidean distance from each data point in the point cloud data to the surface of the deformable reference model; obtain the projection points of the data points on the surface of the deformable reference model, and determine the surface normal vector at the projection points; take the projection length of the vector formed by the data points and the projection points onto the surface normal vector as the size deviation value of the corresponding data points.

[0035] Preferably, obtaining the dimensional deviation of the insulator string further includes:

[0036] The data points are classified based on the sign of the dimensional deviation value to obtain a set of volume redundancy points and a set of volume deficit points.

[0037] Secondly, the present invention provides an insulator size deviation analysis system based on point cloud data, including a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the above-mentioned insulator size deviation analysis method based on point cloud data is implemented.

[0038] By adopting the above technical solution, a computer program is generated from the above-mentioned method for analyzing insulator size deviation based on point cloud data and stored in a memory so that it can be loaded and executed by a processor. This allows for the creation of a terminal device based on the memory and processor, making it convenient to use.

[0039] The beneficial effects of this invention are as follows: By constructing a gravitational moment model to calculate the gravitational impedance coefficient, this invention distinguishes between normal gravitational bending and abnormal structural defects. Based on this, the invention uses the gravity compliance confidence level to drive the standard model for adaptive weighted deformation, generating a reference model that both fits the macroscopic natural drooping posture and retains local defect characteristics. This enables high-precision dimensional deviation measurement and defect identification under complex gravitational deformation backgrounds. Attached Figure Description

[0040] Figure 1 This is a flowchart illustrating an insulator size deviation analysis method based on point cloud data according to the present invention;

[0041] Figure 2 This is a schematic diagram illustrating the point cloud data of an insulator string;

[0042] Figure 3 This is a schematic diagram illustrating the morphological distortion index of each point cloud slice. Detailed Implementation

[0043] This invention discloses a method for analyzing insulator size deviations based on point cloud data, referring to... Figure 1 This includes steps S1-S5:

[0044] S1: Obtain point cloud data of the insulator string; construct the slice step size based on the point cloud distribution characteristics, and calibrate the slice boundary in combination with the local point cloud density to obtain multiple point cloud slices, and obtain the centroid coordinates of each point cloud slice.

[0045] It should be noted that during long-term service, insulator strings will naturally sag in a catenary-like manner due to gravity. Although the physical spacing between individual insulators is a fixed standard value, their projected spacing in the direction of gravity will be compressed due to the bending angle. If a fixed standard skirt spacing is directly used for linear slicing, cumulative errors will occur, causing misalignment between the slicing window and the actual skirt. Therefore, this invention employs a dual strategy of global compression compensation and local density calibration to ensure that the slicing window corresponds to each skirt unit.

[0046] Specifically, point cloud data of the insulator string is acquired; a slicing step size is constructed based on the point cloud distribution characteristics, and the slice boundaries are calibrated in conjunction with local point cloud density to obtain multiple point cloud slices. The centroid coordinates of each point cloud slice are then obtained, including:

[0047] Point cloud data of insulator strings is acquired using a 3D scanning device. Each data point is represented by its 3D coordinates in a spatial coordinate system, which includes the origin and the X, Y, and Z axes. It should be noted that... Figure 2 This is a schematic diagram of the point cloud data of an insulator string, showing the spatial location of all point cloud data of the insulator string.

[0048] Calculate the total length of the projection of the point cloud data onto the principal axis. For example, the Z-axis of the spatial coordinate system is used as the principal axis direction.

[0049] Read the standard CAD model of the insulator to obtain the total number of skirts in the standard insulator string; obtain the distance from the edge of the skirt to the central axis, and record it as the standard skirt radius of the insulator.

[0050] It should be noted that, considering the axial compression effect caused by gravity, this invention first calculates the average projection step size instead of the standard spacing. The average projection step size reflects the average length occupied by each skirt on the main axis under the current bending state.

[0051] The ratio of the total projected length to the total number of umbrella skirts is denoted as the slicing step size.

[0052] Based on the slicing step size, the point cloud data is initially divided into multiple candidate slice intervals along the main axis.

[0053] It should be noted that, in order to eliminate the nonlinear error caused by uneven local curvature, the center of the slice needs to be fine-tuned. Since the surface area of ​​the umbrella skirt edge is the largest, its point cloud density exhibits a significant peak characteristic along the principal axis. Therefore, this invention corrects the candidate slices based on the point cloud density distribution to obtain the final slice.

[0054] For any candidate slice interval, calculate its point cloud density histogram along the principal axis, search for the peak coordinate of the histogram, and define the peak coordinate as the calibration center of the slice; using the calibration center as a reference, extend the slice step length to both sides. The final slice boundary is determined; multiple point cloud slices are obtained, each containing several data points. The arithmetic mean of the coordinates of all data points in each point cloud slice is recorded as the centroid coordinates of the corresponding point cloud slice.

[0055] This completes the data acquisition and slicing process.

[0056] S2: Based on the centroid connection vector of adjacent point cloud slices and the distribution dispersion of data points within the point cloud slices relative to the centroid, obtain the morphological distortion index of each point cloud slice.

[0057] It should be noted that in order to identify potential defects in insulator strings, the geometry of these point cloud slices must first be analyzed. This includes the smoothness of the arrangement between slices, i.e., whether the macroscopic axis is curved, and the dispersion of the point cloud within each slice, i.e., whether the microscopic cross-section is damaged. The curvature of the macroscopic axis is reflected in the change of the angle between the lines connecting adjacent point cloud slices; the larger the angle, the more severe the curvature of the point cloud slice. Damage to the microscopic cross-section is reflected in the sparsity of the point cloud; when damage or contamination occurs, the point cloud no longer tightly surrounds the centroid, and the standard deviation increases. Therefore, this invention combines the geometric distribution and density variation of each point cloud slice to obtain the morphological distortion index of each point cloud slice.

[0058] Specifically, based on the centroid connection vectors of adjacent point cloud slices and the distribution dispersion of data points within a point cloud slice relative to the centroid, the morphological distortion index of each point cloud slice is obtained, including:

[0059] Calculate the first The centroid of a point cloud slice to the first The line vector connecting the centroids of the nth point cloud slice is denoted as the nth line vector. The forward connection vector of the nth point cloud slice is calculated; The centroid of a point cloud slice to the first The line vector connecting the centroids of the nth point cloud slice is denoted as the nth line vector. The backward connection vector of a point cloud slice.

[0060] Statistics All data points within a point cloud slice up to the first... The Euclidean distance of the centroids of the nth point cloud slices constitutes the nth point cloud slice. The standard deviation of the distance set of a point cloud slice is calculated.

[0061] The morphological distortion index of any point cloud slice satisfies the expression:

[0062] ;

[0063] In the formula, Indicates the first The morphological distortion index of a point cloud slice; Indicates the first Forward connection vectors of point cloud slices; Indicates the first The backward connection vector of a point cloud slice; Indicates the first The standard deviation of the distance set of a point cloud slice; Indicates the standard shed radius of the insulator; Indicates the dot product symbol; The symbol representing the magnitude of a vector.

[0064] In the formula, Representing vectors and The cosine of the included angle; The geometric curvature components of the point cloud slices are characterized, and the angle change is mapped to the monotonically decreasing property of the cosine function. The value of this term is 0 when the insulator is straight (angle is 0) and increases when it is bent. The discrete deformation components of the point cloud slices were characterized. By dividing by the standard insulator skirt radius, the dimensional influence of physical dimensions was eliminated, allowing it to be directly superimposed with the geometric bending components. Comprehensively reflects the first The morphological distortion index of a point cloud slice.

[0065] It should be noted that, as Figure 3 This is a schematic diagram of the morphological distortion index for each point cloud slice, showing the distribution of the morphological distortion index calculated for each point cloud slice.

[0066] At this point, the morphological distortion index of each point cloud slice was obtained.

[0067] S3: Based on the torque distribution characteristics of the insulator string and the morphological distortion index of all point cloud slices, obtain the gravitational impedance coefficient of each point cloud slice.

[0068] It should be noted that while the morphological distortion index reflects the degree of deformation of the insulator at various locations, its value does not determine whether the insulator is defective. This is because the insulator string naturally bends under gravity, and this normal distortion caused by gravity results in a higher morphological distortion index in the middle of the insulator string. To identify true structural defects, it is necessary to separate the normal bending caused by gravity from the abnormal distortion caused by defects. Therefore, this invention introduces a gravitational impedance coefficient, utilizing the laws of physical torque to verify the distortion index.

[0069] Specifically, based on the torque distribution characteristics of the insulator string and the morphological distortion index of all point cloud slices, the gravitational impedance coefficient of each point cloud slice is obtained, including:

[0070] The product of the total number of umbrella skirts and the slice step size is denoted as the theoretical total length of the insulator string.

[0071] It should be noted that an insulator string can be approximated as a flexible beam hinged at both ends, and the gravitational torque it experiences exhibits a characteristic of being large in the middle and small at both ends. Therefore, based on the geometric symmetry of the insulator string, this invention constructs a distribution model with the insulator string center as the peak value and linearly decreasing towards both ends to simulate the relative intensity of the gravitational torque at different slice locations, and calculates the gravitational torque arm length of any point cloud slice accordingly.

[0072] The gravitational moment arm length of any point cloud slice satisfies the expression:

[0073] ;

[0074] In the formula, Indicates the first The gravitational moment arm length of a point cloud slice; Indicates the total number of umbrella skirts; i represents the index of the point cloud slice, ranging from 1 to... ; Indicates the slice step size; Represents the absolute value symbol.

[0075] In the formula, This represents the distance from each point on the insulator string to the point of maximum sag. The smaller this value, the more significantly it is affected by the gravitational torque.

[0076] Calculate the arithmetic mean of the morphological distortion indices of all point cloud slices, and denote it as the global morphological distortion factor.

[0077] The gravitational impedance coefficient of any point cloud slice satisfies the following expression:

[0078] ;

[0079] In the formula, This represents the gravitational impedance coefficient of the i-th point cloud slice; Indicates the first The morphological distortion index of a point cloud slice; This represents the theoretical total length of the insulator string; Represents the gravitational torque arm length of the i-th point cloud slice; Represents the global morphological distortion factor; This represents a very small positive value, used to avoid a denominator of 0. For example, .

[0080] In the formula, A gravity baseline model representing how the position changes; This indicates the distortion produced by a unit torque; This represents the global equivalent distortion amplitude of the i-th point cloud slice, used to measure the weight of the distortion at that point relative to the theoretical total length. The influence of gravitational tendency was eliminated by the ratio relationship, if The stability between adjacent slices indicates a stable morphological distortion index. The change conforms to the gravitational torque The pattern of change is within the normal downward gravity. If... The occurrence of drastic fluctuations indicates that the distortion is an abnormal defect.

[0081] At this point, the gravitational impedance coefficients of each point cloud slice were obtained.

[0082] S4: Calculate the gravity compliance confidence of any point cloud slice based on the difference in the gravity impedance coefficients of all adjacent point cloud slices.

[0083] It should be noted that, in order to achieve high-precision dimensional deviation measurement, a standard reference model capable of adaptive deformation needs to be constructed. The behavioral logic of this model during registration is as follows: in areas determined to be under normal gravitational sag, the model should exhibit flexibility, actively following the bending of the point cloud to eliminate pose errors; while in areas determined to be abnormal defects, the model should exhibit rigidity, refusing to follow the bending, thus preserving the differences between the point cloud and the model for detection. Therefore, this invention constructs a dynamic gravity compliance confidence level based on the stability of the gravitational impedance coefficient of each point cloud slice to implement this control logic.

[0084] Specifically, based on the differences in the gravitational impedance coefficients of all adjacent point cloud slices, the gravity compliance confidence of any point cloud slice is calculated, including:

[0085] The difference between the gravitational impedance coefficient of the i-th point cloud slice and the gravitational impedance coefficient of the (i-1)-th point cloud slice is calculated and denoted as the impedance difference value of the i-th point cloud slice. The impedance difference values ​​of all point cloud slices of the insulator string are statistically analyzed, and the standard deviation of all impedance difference values ​​is calculated and denoted as the adaptive impedance bandwidth. It should be noted that the adaptive impedance bandwidth reflects the overall stability of the current insulator string and is used to automatically adjust the algorithm's sensitivity to anomalies.

[0086] The gravity compliance confidence of any point cloud slice satisfies the expression:

[0087] ;

[0088] In the formula, This represents the gravity compliance confidence of the i-th point cloud slice; This represents the impedance difference value of the i-th point cloud slice; Indicates the adaptive impedance bandwidth; Represents the natural exponential function; Represents the absolute value symbol.

[0089] In the formula, Soft-switching control was implemented using the decay characteristics of a Gaussian function, which measures the impedance change of the i-th point cloud slice. The degree of deviation from the normal global fluctuation range. The smaller the value, the more the point cloud slice is within the normal noise range. The larger the value, the more significant the abnormal performance of the point cloud slice.

[0090] At this point, the gravity compliance confidence scores for each point cloud slice were obtained.

[0091] S5: Perform weighted skeleton deformation operation on the standard CAD model according to the gravity compliance confidence level to obtain the deformed reference model; calculate the Euclidean distance between the point cloud data of the insulator string and the deformed reference model, and obtain the dimensional deviation of each part of the insulator by combining the projection normal vector.

[0092] It should be noted that after obtaining the gravity compliance confidence level of each point cloud slice, the standard CAD model can be driven to undergo intelligent deformation. By performing weighted interpolation between the measured centroid and the ideal straight line center, a reference model can be generated that fits the macroscopic gravity curve without being skewed by local defects. Comparing the original point cloud with the reference model can eliminate the interference of gravity bending on the measurement, directly read the true dimensional deviation of each part, and thus determine the condition of dirt accumulation, icing, or damage.

[0093] Specifically, a weighted skeleton deformation operation is performed on the standard CAD model using gravity compliance confidence to obtain a deformed reference model; the Euclidean distance between the point cloud data of the insulator string and the deformed reference model is calculated, and the dimensional deviations of various parts of the insulator are obtained by combining the projection normal vectors, including:

[0094] The centroid coordinates of the i-th point cloud slice are marked as the target position; the center axis coordinates of the standard CAD model at the i-th slice are marked as the ideal position.

[0095] The deformable skeleton position of an arbitrary point cloud slice satisfies the expression:

[0096] ;

[0097] In the formula, This indicates the position of the deformable skeleton of the model corresponding to the i-th point cloud slice; This represents the gravity compliance confidence of the i-th point cloud slice; Represents the centroid coordinates of the i-th point cloud slice; This represents the ideal position corresponding to the i-th point cloud slice.

[0098] In the formula, This indicates the degree of confidence in the gravity compliance score based on the measured data. This indicates the degree to which the gravity compliance confidence level preserves the ideal design; This reflects the skeletal position in regions with high confidence levels of gravity compliance. Tend to Eliminate gravity pose error; in defect areas with low confidence in gravity compliance, the skeleton position... Tend to Maintaining a standard shape allows for geometric margins to be left for subsequent defect identification.

[0099] Based on the model deformation skeleton positions of all point cloud slices, the standard CAD model is driven to perform mesh deformation to generate a deformation reference model: the standard CAD model is divided into mesh segments along the principal axis direction, corresponding one-to-one with the point cloud slices; the deformation displacement vector of the i-th mesh segment is calculated, which is the vector difference between the model deformation skeleton position and the ideal position; all mesh vertices in the i-th mesh segment are traversed, and the deformation displacement vector is superimposed on the original vertex coordinates to update the mesh vertex positions; after all mesh segments have been updated, they are combined to obtain the deformation reference model.

[0100] Calculate the shortest Euclidean distance from any data point in the point cloud data of the insulator string to the surface of the deformed reference model; find the normal vector of the tangent plane of the perpendicular projection point of the data point on the surface of the deformed reference model, and denote it as the model normal vector of the data point.

[0101] The dimensional deviation of any data point satisfies the expression:

[0102] ;

[0103] In the formula, This represents the size deviation of the u-th data point; This represents the shortest Euclidean distance from the u-th data point to the surface of the deformable reference model; This represents the symbol extraction function; This represents the model normal vector of the u-th data point; Represents the coordinates of the u-th data point; Indicates the coordinates of the projection point; This represents the dot product symbol.

[0104] In the formula, Used to determine the interior / exterior orientation of the u-th data point relative to the surface of the deformable reference model. If the dot product is positive, it means the point is outside the model. This represents volume redundancy caused by dirt or ice buildup; if the dot product is negative, it indicates that the point is inside the model. This is characterized as a volume loss caused by missing or damaged material.

[0105] This completes the insulator size deviation analysis based on point cloud data.

[0106] This invention also discloses an insulator size deviation analysis system based on point cloud data, including a processor and a memory. The memory stores computer program instructions, which, when executed by the processor, implement an insulator size deviation analysis method based on point cloud data according to the present invention.

[0107] The system also includes other components well known to those skilled in the art, such as communication buses and communication interfaces, the settings and functions of which are known in the art and will not be described in detail here.

[0108] While this specification has shown and described numerous embodiments of the invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Many modifications, alterations, and alternatives will occur to those skilled in the art without departing from the spirit and essence of the invention. It should be understood that various alternatives to the embodiments of the invention described herein may be employed in the practice of this invention.

Claims

1. A method for analyzing size deviation of an insulator based on point cloud data, characterized in that, The method comprises the following steps: Obtain the point cloud data of the insulator string, and slice the point cloud data to obtain a plurality of point cloud slices; Obtain the centroid coordinates of each point cloud slice, and calculate the morphological distortion index of each point cloud slice based on the centroid coordinates of adjacent point cloud slices and the distribution of data points within the point cloud slices; the morphological distortion index satisfies the expression: ; Indicates the first The morphological distortion index of a point cloud slice; Indicates the first Forward connection vectors of point cloud slices; Indicates the first The backward connection vector of a point cloud slice; Indicates the first The standard deviation of the distance set of a point cloud slice; Indicates the standard shed radius of the insulator; Indicates the dot product symbol; The sign representing the magnitude of a vector; Based on the gravitational torque distribution model of the insulator string, the morphological distortion index of each point cloud slice is normalized to obtain the gravitational impedance coefficient of each point cloud slice. This includes: taking the product of the total number of umbrella skirts and the slice step size as the theoretical total length of the insulator string; and multiplying the absolute difference between half of the total number of umbrella skirts and i by the slice step size to obtain the i-th... The gravitational moment arm length of each point cloud slice; calculate the arithmetic mean of the morphological distortion indices of all point cloud slices, denoted as the global morphological distortion factor; the gravitational impedance coefficient satisfies the expression: ; This represents the gravitational impedance coefficient of the i-th point cloud slice; This represents the theoretical total length of the insulator string; Represents the gravitational torque arm length of the i-th point cloud slice; Represents the global morphological distortion factor; Indicates a very small positive value; According to the difference of the gravity impedance coefficients of adjacent point cloud slices, a gravity compliance confidence of each point cloud slice is determined; the gravity compliance confidence satisfies an expression: ; represents the gravity compliance confidence of the i-th point cloud slice; represents the impedance difference value of the i-th point cloud slice; represents the adaptive impedance bandwidth; represents the natural exponential function; represents the absolute value symbol; In response to the gravity-conforming confidence of each point cloud slice, the skeleton of the standard CAD model is weighted and deformed to generate a deformed reference model, the point cloud data is registered with the deformed reference model to obtain the size deviation of the insulator string, including: marking the centroid coordinate of the i th point cloud slice as a target position; marking the center axis coordinate of the standard CAD model at the i th slice as an ideal position; the model deformation skeleton position of an arbitrary point cloud slice satisfies the expression: ; represents the model deformation skeleton position corresponding to the i th point cloud slice; represents the centroid coordinate of the i th point cloud slice; represents the ideal position corresponding to the i th point cloud slice; based on the model deformation skeleton positions of all point cloud slices, the standard CAD model is driven to perform mesh deformation to generate a deformed reference model: the standard CAD model is divided into mesh segments corresponding to the point cloud slices along the main axis direction; the deformation displacement vector of the i th mesh segment is calculated, which is the vector difference between the model deformation skeleton position and the ideal position; traverse all mesh vertices in the i th mesh segment, and superimpose the deformation displacement vector on the original vertex coordinates to update the mesh vertex position; when all mesh segments are updated, the deformed reference model is obtained by combination; the shortest Euclidean distance from any data point in the point cloud data of the insulator string to the surface of the deformed reference model is calculated; the tangent plane normal vector of the vertical projection point of the data point on the surface of the deformed reference model is found, which is denoted as the model normal vector of the data point; the size deviation of any data point satisfies the expression: ; represents the size deviation of the u th data point; represents the shortest Euclidean distance from the u th data point to the surface of the deformed reference model; represents a symbol extraction function; represents the model normal vector of the u th data point; represents the coordinate of the u th data point; represents the coordinate of the projection point.

2. The insulator size deviation analysis method based on point cloud data according to claim 1, characterized in that, The slicing of the point cloud data comprises the following steps: Calculate the total length of the projection of the point cloud data on the main shaft; obtain the total number of shed of the standard insulator string; obtain the distance from the shed edge to the center axis, denoted as the standard shed radius of the insulator; and calculate the ratio of the total length of the projection to the total number of the shed, denoted as the slice step length; Based on the slice step length, the point cloud data is preliminarily divided into a plurality of candidate slice intervals along the main shaft; For any candidate slice interval, the point cloud density histogram along the main axis direction is counted, the peak coordinate of the histogram is searched, and the peak coordinate is defined as the calibration center of the slice; taking the calibration center as the reference, each extending the slice step to both sides , determine the final slice boundary; obtain a plurality of point cloud slices.

3. The insulator size deviation analysis method based on point cloud data according to claim 1, characterized in that, The acquisition of the gravity conformance confidence comprises the following steps: Calculate the difference between the gravity impedance coefficient of the i-th point cloud slice and the gravity impedance coefficient of the i-1-th point cloud slice, denoted as the impedance difference value of the i-th point cloud slice; count the impedance difference values of all point cloud slices of the insulator string, calculate the standard deviation of all the impedance difference values, and denote the adaptive impedance bandwidth as the standard deviation.

4. The insulator size deviation analysis method based on point cloud data according to claim 1, characterized in that, The weighted deformation of the skeleton of the standard CAD model comprises the following steps: Obtain the center axis coordinates at each slice position in the standard CAD model as ideal skeleton points; and calculate the weighted average of the centroid coordinates of the corresponding point cloud slice and the ideal skeleton points as the deformed skeleton points, wherein the weight of the centroid coordinates is the gravity conformance confidence of the corresponding point cloud slice, and the weight of the ideal skeleton points is the difference between 1 and the gravity conformance confidence.

5. The insulator size deviation analysis method based on point cloud data according to claim 1, characterized in that, The method for obtaining the size deviation of the insulator string further comprises the following steps: Classify the data points based on the positive and negative of the size deviation value to obtain a volume redundant point set and a volume loss point set.

6. An insulator size deviation analysis system based on point cloud data, characterized by, The method comprises the following steps: A processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, a method for analyzing the size deviation of an insulator based on point cloud data according to any one of claims 1-5 is realized.

Images