A bolted steel truss beam member measurement data preprocessing method and system

Abstract

Provided are a bolted steel truss beam member measurement data preprocessing method and system, including the following steps: S1, defining a measurement reference, S2, checking data positions, S3, checking data numbers, S4, checking port characteristics and processing. The application innovates a reference construction mode, realizes unified data reference, innovates an automatic verification logic, reduces manual dependence, innovates a special member abnormality processing flow, covers more scenarios, and realizes a standardized and automated measurement data preprocessing system, which has significant advantages in engineering practicability, technical adaptability and industry promotion, and is expected to promote the standardization upgrade of bolted truss steel truss beam measurement data preprocessing technology.

Description

Technical Field

[0001] This invention belongs to the field of computer-aided design technology, and in particular relates to a method and system for preprocessing measurement data of bolted steel truss members. Background Technology

[0002] Currently, in the field of bolted steel truss girder engineering, the accuracy of member measurement data directly determines the precision of subsequent pre-assembly. Existing technologies revolve around the stages of "measurement data acquisition - error correction - coordinate integration - data application," forming a technical system centered on improving data reliability and adapting to digital processes. A detailed overview is as follows:

[0003] I. Current Status of Virtual Pre-assembly and Precision Testing Technology

[0004] Patent publication number CN112016147A proposes a method for achieving full-process digitalization of "measurement-modeling-inspection-reporting" based on secondary development of the Revit platform. This method collects the center points of bolt holes, corrects coordinate records based on temperature differences, establishes a 3D measured model and a theoretical model on the Revit platform, and achieves accuracy detection and outputs a report through visual programming and custom nodes combined with relevant algorithms.

[0005] Its core innovation lies in: deeply integrating "error correction" with "data standardization" to adapt to the digital needs of the BIM platform; introducing a temperature compensation mechanism to reduce the impact of ambient temperature on measurement accuracy; and reducing workload through "fewer but more precise" assembly point selection (prioritizing coplanar nodes and adjacent holes in the same row / column).

[0006] The limitations of this technology are as follows: data processing relies on subsequent modeling requirements of the Revit platform, and data cleaning rules are not designed for the dense characteristics of bolted truss beams with multiple members and bolt holes; furthermore, temperature correction only considers linear growth and does not take into account the temperature coefficient differences of different member materials, resulting in limited adaptability to complex members. The benchmark system lacks standardization, leading to poor data correlation. Anomaly handling is insufficiently adaptable, and coverage of special members is incomplete. All existing technologies lack specific processing procedures for members with special shapes, making parameter calculations for special members prone to errors and failing to reflect true dimensions.

[0007] II. Current Status of Point Cloud Measurement Technology for Large-Segment Steel Truss Beams

[0008] Patent publication number CN120063120A proposes a method for first measuring the three-dimensional coordinates (feature points) of bolt holes at end corners using a laser coordinate measuring machine, and then acquiring regional data collection of the planar coordinates and hole diameters of all bolt holes using a planar scanner. This method fits the planar coordinates of the bolt holes at the corners into three-dimensional coordinates, establishes coordinate transformation parameters through unconstrained nonlinear optimization, and converts the planar information of all bolt holes into a three-dimensional point cloud, forming the point cloud information of the steel truss manufacturing configuration.

[0009] Its core innovation lies in breaking through the workload bottleneck of traditional full-area 3D scanning, measuring only the 3D coordinates of corner feature points, and combining planar scanning data to achieve point cloud fusion, improving measurement efficiency by more than 50%; avoiding the errors caused by traditional target ball fitting, directly using the geometric features of bolt holes (hole diameter, planar coordinates) as the registration benchmark, improving the measurement accuracy from the traditional 3mm to 1mm level, meeting the high precision requirements of large-segment steel truss beams.

[0010] The limitations of this technology are as follows: it relies on the initial accuracy of the bolt holes at the corner points. If there are processing deviations in the corner holes (such as irregular hole diameters or center offsets), it will directly lead to amplified errors in subsequent fitting and transformation. Point cloud registration requires the establishment of a nonlinear optimization model, which places high demands on the professionalism of data processing algorithms. On-site construction personnel need to additionally master the algorithm principles and parameter debugging, increasing the difficulty of technology promotion. Data verification relies on manual labor and professional skills, resulting in low efficiency. Point cloud registration requires professionals to master nonlinear optimization algorithms.

[0011] III. Current Status of Steel Truss Girder Simulation Assembly Technology

[0012] Patent publication number CN120395734A proposes a method for using a sub-millimeter level handheld scanner to scan point clouds of multiple regions in a steel truss assembly segment, preprocessing (cropping, removing outliers) to extract screw hole coordinates, and then using a total station to measure the coordinates of characteristic screw holes to establish a reference splicing coordinate system. This method employs a four-point consistency algorithm to map the discrete screw hole information acquired by the scanner to the total station's reference coordinates. Then, through rigid registration (between the splicing plate and the first assembly segment) and Protodyakonov analysis (between the splicing plate and the second assembly segment), it simulates the assembly of the two steel truss segments, simultaneously identifying abnormal screw holes.

[0013] Its core innovation lies in solving the shortcomings of single devices, making up for the problems of scanners being "high-precision but discrete" (unable to unify the coordinate system) and total stations being "global coordinates but single-point measurement" (low efficiency), and achieving "high-precision + global" measurement.

[0014] The limitations of this technology are as follows: it relies on multi-device collaboration (scanner + total station), resulting in high equipment purchase and maintenance costs; the preprocessing process involves multiple stages such as scanner point cloud cleaning, total station coordinate registration, and multi-source data mapping, making the process complex and dependent on specialized software, requiring highly skilled operators. Relying on scanner and total station collaboration can easily lead to data fragmentation due to differences in reference standards between devices, making it impossible to establish a unified reference for data acquired from different stages and instruments, which can easily cause deviations in subsequent calculations. The preprocessing involves complex processes such as multi-source data mapping, all of which require manual intervention in parameter adjustment and result judgment, resulting in low verification efficiency and potential delays in construction progress due to human error.

[0015] It is evident from the in-depth research on the aforementioned patents that existing preprocessing technologies for measurement data of bolted steel truss members have significant shortcomings in three core aspects: benchmark system, processing efficiency, and anomaly adaptation, making it difficult to meet the high-efficiency and high-precision requirements of large-scale engineering projects. Therefore, the following improved technical solutions are proposed. Summary of the Invention

[0016] The purpose of this invention is to provide a method and system for preprocessing measurement data of bolted steel truss members, so as to achieve a standardized and automated measurement data preprocessing system.

[0017] The technical problems to be solved by this invention are: to innovate the benchmark construction mode and achieve unified data reference; to innovate the automated verification logic and reduce reliance on manual labor; and to innovate the anomaly handling process for special rods, covering more scenarios and solving the parameter calculation of rods with special shapes.

[0018] The technical solution adopted in this invention is: a method for preprocessing measurement data of bolted steel truss members, including the following steps: S1, defining the measurement benchmark; S2, checking the data position; S3, checking the data number; S4, checking the port characteristics and processing.

[0019] S1. Define the measurement benchmark: This includes the measurement point extraction method and point coding system. First, define the measurement point extraction method, using the ports and nodes of steel structure members as the extraction carriers for bolt hole positions. Establish a local coordinate system based on the geometric parameters of the member cross-section. Define four reference holes that can form a rectangle on the functional surface of the port and node box openings. Use a laser scanner to collect the center of the hole to obtain the coordinates of the four measurement points. Establish a geometric benchmark using the "three-point surface determination, four-point optimization" mechanism. Then, construct a matching point numbering system. Within each functional surface, the points are numbered clockwise. Different functional surfaces are numbered according to view priority. Different ports or nodes are numbered according to the priority of the axis from left to right. For axes with the same orientation, the points are numbered according to the priority of the view orientation.

[0020] S2. Check data location: Divide the measurement data into groups of 4 points, and use the geometric feature constraint method to iteratively verify the compliance of each group of points:

[0021] First, overlap or nearest neighbor anomalies are identified by calculating the Euclidean distance between any two points.

[0022] Secondly, calculate the cosine of the angle between the vectors of opposite sides, compare it with the maximum error of the theoretical parallel condition, and verify the parallelism of the opposite side directions.

[0023] Finally, the absolute difference between the lengths of the opposite side and the diagonal is calculated. First, the opposite sides are verified to be of equal length to determine the parallelogram. Then, the diagonals are verified to be of equal length to filter the rectangles. Finally, the compliance of the rectangle composition of each group of points is output with binary identifiers.

[0024] S3. Check the data numbering: Each group of 4 points constitutes a functional surface. Based on spatial geometric projection and surrounding direction analysis, determine the correctness of the point numbering on each functional surface:

[0025] First, the normal vector of the plane is obtained by the cross product of the three-point vectors, and then the main projection plane is determined, realizing the orthogonal projection of the three-dimensional point set into two dimensions.

[0026] Secondly, the surrounding value is calculated for the four projected points, and the viewpoint influence is eliminated by combining the preset observation direction vector to determine the surrounding direction of the point number of each functional surface.

[0027] Finally, output the compliance of the numbering of each group of points in binary format.

[0028] S4. Check and process port characteristics: Establish a standardized port characteristic processing flow.

[0029] First, distinguish the front and rear ports according to the port numbering order of the axis, and extract the measurement points in combination with the functional surface numbering rules.

[0030] Secondly, calculate the distance difference in the Z-axis direction between the measurement points at corresponding positions on the front surfaces of the front and rear ports, and set a threshold to identify "uneven height rods".

[0031] Then, the principle of "high-side port coordinate assimilation" is adopted to unify the Z coordinates of the low-side port points with the Z coordinates of the high-side port points as the reference, thereby eliminating height deviation.

[0032] Finally, the Euclidean distance between the feature points at both ends is calculated along the axis to obtain the actual measured value of the rod length.

[0033] In the above technical solution, step S1 further includes a measurement point extraction method and the construction of a matching point numbering system;

[0034] Measurement point extraction method: First, the ports and nodes of the steel structure members are used as the carriers for extracting bolt hole positions; the center point of the member is determined by the geometric parameters of the member section as the origin of the local coordinate system, and a local coordinate system is established based on this: the X-axis is parallel to the length direction of the member, the Y-axis is perpendicular to the member section, and the Z-axis is parallel to the functional surface of the port; on each functional surface of the port or node box opening, four reference holes that can form a rectangle within the threshold range are defined, and the center of the reference hole is designated as the measurement point. The coordinates of the four measurement points are collected using a laser scanner. The four points provide a unique and reproducible geometric reference for hole position measurement through the "three-point plane determination and four-point optimization" method.

[0035] Construct a corresponding point numbering system: For a single functional surface of a port or node opening, take the upper left point of the functional surface as the starting point for numbering, and assign consecutive numbers to the four measurement points within the surface in a clockwise direction; For multiple functional surfaces contained in a single port or node opening, use the rod view as a spatial reference, and number the points on different functional surfaces of the opening sequentially according to the view orientation priority of the front surface, upper surface, rear surface, and lower surface; For multiple ports, openings, or node openings contained in a rod, use the rod's length axis as a reference, and number all ports, openings, and node openings sequentially according to the axis orientation priority from left to right. If the axis orientations are the same, then number them sequentially according to the view orientation priority of front, upper, rear, and lower.

[0036] In the above technical solution, further: in step S2, the measurement data is divided into groups of four points, let the four points in the first group be A, B, C, and D, and the position of each group of points is checked using a method based on geometric feature constraints.

[0037] S201. Calculate the Euclidean distance between any two points in a set to determine whether there are overlapping or nearest-neighbor outliers.

[0038]

[0039] S202. If the discreteness of the four points meets the requirements, calculate the cosine of the angle between the opposite side vectors AB and CD, and AD and BC.

[0040]

[0041] Set the cosine value of the opposite side to the theoretical parallel condition, and the maximum reference error with an absolute value of 1 for the cosine value, to determine whether the opposite side satisfies the directional parallelism constraint.

[0042] S203. If the four points satisfy the condition of parallel opposite sides, calculate the absolute differences between the lengths of opposite sides AB and CD, BC and DA, and the lengths of diagonals AC and BD. Set the maximum length error, first verify whether opposite sides are of equal length. If they are, determine that the four points can form a parallelogram. Then verify whether the diagonals are of equal length. If they are, further filter out rectangles from the parallelograms.

[0043] S204. Calculate each group of points in a loop, and finally output the preset requirements for whether each group of points can form a rectangle, with 1 for yes and 0 for no, using binary identifiers. This realizes the compliance check of the measurement data position and provides a geometric basis for subsequent data calculation.

[0044] In the above technical solution, further: in step S3, a method based on spatial geometric projection and surrounding direction analysis is used to determine whether the numbering of each group of points is correct.

[0045] S301. Calculate the cross product of any three points in a set of points to obtain the surface normal vector. Determine the main projection plane (YZ, XZ, or XY plane) based on the maximum absolute value of each component of the normal vector to achieve orthogonal projection transformation of the three-dimensional point set to a two-dimensional plane.

[0046] S302. Project a set of 4 points according to the main projection plane, and calculate the wrap value by adding the cross product of the coordinates of the adjacent points to the 4 projected points (P1(x1y1), P2(x2,y2), P3(x3,y3), P4(x4,y4)).

[0047]

[0048] S303. Preset observation direction vectors for different projection planes. Use the direction coefficient ±1, where 1 indicates observation from the positive direction and -1 indicates observation from the negative direction to unify the observation angle reference. Couple the orbit value with the observation direction coefficient to eliminate the influence of the observation angle on the orbit direction determination. If the calculation result is less than 0, determine that a group of 4 points is clockwise numbered; otherwise, determine that the 4 points are counterclockwise numbered.

[0049] S304. Calculate each group of points in a loop, and finally output whether the number of each group of points meets the preset requirements with binary identifiers, 1 for yes and 0 for no, so as to realize the automatic verification of the numbering order of points on each functional surface.

[0050] In the above technical solution, further: In step S4, during the data preprocessing stage, a feature inspection system for the rod ends is constructed according to the numbering rules. This system can determine whether there are any special cases in the rods, laying a standardized foundation for subsequent geometric calculations. The specific process is as follows:

[0051] S401. Based on the preset port numbering sequence of the member axis direction, the measurement point numbers of the front and rear ports of the member can be effectively distinguished; combined with the numbering sequence of a single port and a single functional surface, the rules for extracting measurement points can be clearly defined.

[0052] S402. For the identified front and rear ports, extract the measurement points at the corresponding positions on the front surface of the ports, and calculate the normal distances D1 and D2 between the two points in the Z-axis direction.

[0053] S403. Set a distance difference judgment threshold, compare the normal distances D1 and D2 between the front and rear port measurement points. When the measured distance difference d exceeds the threshold, it is determined that the current member is a "member with unequal height at the front and rear", and the coordinates of the port measurement points need to be standardized.

[0054] S404. Adopting the principle of "high port coordinate reference assimilation", the Z coordinate of the port measurement point on the side with the larger height value between the front and rear ports is extracted as the reference, and the Z coordinate of the feature point on the side with the smaller height value is assigned to the reference value, that is, moved to the position of point J, so as to realize the coordinate reference of the two port measurement points in the height direction and eliminate the spatial coordinate deviation caused by the inconsistency of port height.

[0055] S405. After completing the port benchmark unification, calculate the Euclidean distance between the port feature points along the axis of the member, and use it as the final measured value of the member length L. This avoids the length calculation deviation caused by the height difference of the port and ensures that the member length calculation result reflects its true size, rather than the oblique distance affected by spatial height misalignment.

[0056] In the above technical solution, the measurement point extraction process in step S401 is further as follows:

[0057] S4011. Ensure that the number of measurement points at each port is fixed and the same.

[0058] S4012. When extracting measurement points at the front port, take the upper left side of the port as the starting point and follow the numbering principle of different functional surfaces of a single port to extract the continuous measurement points corresponding to the front port in ascending order.

[0059] S4013. When extracting measurement points at the rear port, take the measurement point number corresponding to the total number of holes n of the rod as the starting point, and extract the continuous measurement point numbers corresponding to the rear port in reverse order. The extraction range matches the total number of measurement points at a single port.

[0060] The present invention also claims protection for a system comprising a processor and a memory, the memory storing at least one instruction, at least one program, code set, or instruction set, the at least one instruction, the at least one program, the code set, or the instruction set being loaded and executed by the processor to implement the preprocessing method as described in any of the above.

[0061] The beneficial effects of this invention are:

[0062] 1. Innovate the benchmark construction model to achieve unified data reference. Abandon the existing benchmark methods that rely on platforms or multiple devices, and establish a geometric benchmark through the "three-point surface determination and four-point optimization" mechanism. Combined with a point coding system based on "port axis priority + functional surface view priority", measurement data from different sources have unified rule identification, solving the problem of poor data correlation.

[0063] 2. Innovative automated verification logic reduces reliance on manual labor. Breaking away from the constraints of traditional manual verification and professional software, it adopts a "group first, then judge" logic: automatically verifying the compliance of points based on geometric feature constraints, automatically judging the correctness of numbers through spatial projection and surrounding direction analysis, and outputting results in binary identifiers throughout the process, significantly improving verification efficiency.

[0064] 3. Innovate special member anomaly handling process to cover more scenarios. For "members with unequal heights at the front and rear", which are not covered by existing technology, a special handling process is constructed: first, the front and rear ends are distinguished by numbering rules and measurement points are extracted; then, the Z-axis distance difference is calculated to identify anomalies; finally, the coordinates are unified by the principle of "coordinate benchmark assimilation of the high end", ensuring that the measured value of the member length reflects the true size, filling the technical gap in the handling of special members.

[0065] 4. This invention has significant advantages in engineering practicality, technical adaptability, and industry promotion, and is expected to promote the standardization and upgrading of bolted truss girder measurement data preprocessing technology. Engineering practicality: The standardized benchmark system improves data correlation, reduces subsequent calculation deviations, and automated verification eliminates the need for full-time professional involvement. Technical adaptability: The specialized anomaly handling process can solve the parameter calculation problem for special-shaped members. Combined with a reproducible benchmark construction mode, it can adapt to dense scenarios with multiple members and bolt holes, breaking the scenario limitations of existing technologies. Industry promotion: Compared with existing technologies that rely on specific platforms, special equipment, or professional algorithms, this patented solution has a clear process and is easy to operate. On-site construction personnel do not need to master complex software or algorithm principles, making it easier to promote and apply within the industry and contributing to the popularization of digital processes in steel truss girder engineering.

[0066] 5. This invention addresses the problems of inconsistent benchmarks, difficulty in verifying data compliance, and deviations in measurement results caused by abnormal features in the preprocessing of measurement data for bolted steel truss members. This invention provides a standardized preprocessing scheme for measurement data of steel truss members, which ultimately eliminates data deviations, ensures that key measurement results accurately reflect the true values, and provides reliable data support for subsequent member-related calculations.

[0067] 6. The present invention unifies the measurement benchmark. The present invention establishes a reproducible geometric benchmark through the "three-point surface determination and four-point optimization" mechanism. Combined with the supporting point coding system, it realizes the regular identification of measurement data. This standardized benchmark system enables data obtained from different links and different instruments to have a unified reference, providing reliable data support for subsequent data processing and calculation of steel structure members, and effectively reducing the problem of poor data correlation in subsequent calculations caused by benchmark deviation.

[0068] 7. Automated Verification: This invention utilizes a "grouping first, then judgment" logic, based on geometric feature constraints and spatial projection analysis, to automatically determine the compliance and correctness of the numbering of each group of points. The entire process requires no manual intervention, and the results are output in binary format. This automated verification mode reduces reliance on personnel expertise and increases verification efficiency several times over, making it suitable for the construction needs of large-scale projects and avoiding delays in construction progress caused by manual verification.

[0069] 8. Precise handling of anomalies: This invention constructs a special processing flow for "uneven height members". First, a feature inspection system for the member ends is constructed according to the numbering rules. Then, anomalies are identified by calculating the Z-axis distance difference between corresponding points at the front and rear ends. Finally, the "high end coordinate benchmark assimilation" principle is used to eliminate deviations, ensuring that the measured length of the member reflects the true size. This solves the problem of slanted distance miscalculation caused by inconsistent heights at the member ends. Attached Figure Description

[0070] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0071] Figure 1 This is a schematic diagram showing the locations and numbers of each group of points on the front surface of the steel truss members;

[0072] Figure 2(a) is a schematic diagram of the front points and numbering of the front and rear ends of the steel truss members with unequal heights;

[0073] Figure 2(b) is a schematic diagram showing the location and numbering of the upper and lower points of the short wave of the steel truss members with unequal heights at the front and rear;

[0074] Figure 3 This is a flowchart of the preprocessing method for measurement data of bolted steel truss members according to the present invention. Detailed Implementation

[0075] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0076] (like Figure 3 (As shown) A method for preprocessing measurement data of bolted steel truss members, comprising the following steps:

[0077] S1. Define the measurement reference; S2. Check the data location; S3. Check the data number; S4. Check the port characteristics and processing.

[0078] S1. Define the measurement benchmark: This includes the measurement point extraction method and point coding system. First, define the measurement point extraction method, using the ports and nodes of steel structure members as the extraction carriers for bolt hole positions. Establish a local coordinate system based on the geometric parameters of the member cross-section. Define four reference holes that can form a rectangle on the functional surface of the port and node box openings. Use a laser scanner to collect the center of the hole to obtain the coordinates of the four measurement points. Establish a geometric benchmark using the "three-point surface determination, four-point optimization" mechanism. Then, construct a matching point numbering system. Within each functional surface, the points are numbered clockwise. Different functional surfaces are numbered according to view priority. Different ports or nodes are numbered according to the priority of the axis from left to right. For axes with the same orientation, the points are numbered according to the priority of the view orientation.

[0079] S2. Check data location: Divide the measurement data into groups of 4 points, and use the geometric feature constraint method to iteratively verify the compliance of each group of points:

[0080] First, overlap or nearest neighbor anomalies are identified by calculating the Euclidean distance between any two points.

[0081] Secondly, calculate the cosine of the angle between the vectors of opposite sides, compare it with the maximum error of the theoretical parallel condition, and verify the parallelism of the opposite side directions.

[0082] Finally, the absolute difference between the lengths of the opposite side and the diagonal is calculated. First, the opposite sides are verified to be of equal length to determine the parallelogram. Then, the diagonals are verified to be of equal length to filter the rectangles. Finally, the compliance of the rectangle composition of each group of points is output with binary identifiers.

[0083] S3. Check the data numbering: Each group of 4 points constitutes a functional surface. Based on spatial geometric projection and surrounding direction analysis, determine the correctness of the point numbering on each functional surface:

[0084] First, the normal vector of the plane is obtained by the cross product of the three-point vectors, and then the main projection plane is determined, realizing the orthogonal projection of the three-dimensional point set into two dimensions.

[0085] Secondly, the surrounding value is calculated for the four projected points, and the viewpoint influence is eliminated by combining the preset observation direction vector to determine the surrounding direction of the point number of each functional surface.

[0086] Finally, output the compliance of the numbering of each group of points in binary format.

[0087] S4. Check and process port characteristics: Establish a standardized port characteristic processing flow.

[0088] First, distinguish the front and rear ports according to the port numbering order of the axis, and extract the measurement points in combination with the functional surface numbering rules.

[0089] Secondly, calculate the distance difference in the Z-axis direction between the measurement points at corresponding positions on the front surfaces of the front and rear ports, and set a threshold to identify "uneven height rods".

[0090] Then, the principle of "high-side port coordinate assimilation" is adopted to unify the Z coordinates of the low-side port points with the Z coordinates of the high-side port points as the reference, thereby eliminating height deviation.

[0091] Finally, the Euclidean distance between the feature points at both ends is calculated along the axis to obtain the actual measured value of the rod length.

[0092] It should be noted that the precise definition of the measurement benchmark in this invention improves data reliability. By using the ports and nodes of steel structure members as the extraction carriers for bolt hole locations, and establishing a local coordinate system in conjunction with the geometric parameters of the member cross-sections, and employing a "three-point plane determination, four-point optimization" mechanism, the accumulation of measurement errors is effectively reduced. This avoids the systematic deviations caused by ambiguous coordinate system definitions in traditional methods, and is particularly suitable for steel truss members with complex geometries. The structured point numbering system, constructed according to rules such as functional surfaces, view priority, and axis direction, ensures that each measurement point has a unique identifier, and that the numbering logic is strongly correlated with the spatial structure of the member. This provides standardized input for subsequent data processing (such as deformation analysis and assembly verification), reducing the risk of errors from manual numbering.

[0093] Furthermore, efficient data location verification ensures geometric compliance. Multi-level geometric constraint checks use parameters such as Euclidean distance, cosine of the angle between opposite side vectors, and the difference in length between opposite sides and diagonals to iteratively verify whether each group of four points forms a compliant rectangle. It quickly identifies gross errors such as overlapping points and nearest-neighbor outliers; verifies the geometric conditions of parallelograms and rectangles to ensure the flatness of the port functional surface meets design requirements; and outputs results with binary identifiers for easy integration into automated processing workflows. Robust design employs a layered verification logic of "parallelograms first, then rectangles," progressively narrowing the error range. This avoids misjudgments caused by errors in single-condition judgments, improving verification accuracy.

[0094] Furthermore, intelligent data numbering checks eliminate spatial ambiguity. The orthogonal projection transformation from 3D to 2D uses the cross product of three-point vectors to obtain the plane normal vector, determining the principal projection plane and achieving dimensionality reduction from a 3D point set to a 2D plane. This simplifies subsequent orbital direction analysis and avoids misnumbering caused by changes in viewing angle in 3D space. Dynamic orbital direction determination combines the orbital values ​​of the projected points with the preset observation direction vector to eliminate the influence of viewing angle and automatically determine whether the point numbering order conforms to a clockwise rule. It adapts to measurement data under different observation angles; binary identifier output results support automatic comparison with the design model.

[0095] Furthermore, standardized port feature processing enables precise assembly control. The system identifies and corrects height differences between front and rear ports by calculating the Z-axis distance difference between corresponding measurement points on the front and rear ports, setting a threshold to identify "unequal height members," and unifying the coordinates of the lower port points using the coordinates of the higher port as a benchmark. This eliminates height deviations caused by manufacturing or installation errors, ensuring that the member axis aligns with the theoretical design and avoiding assembly stress concentration issues caused by height differences. Actual member length measurements are obtained by calculating the Euclidean distance between feature points on both ports along the axis, directly outputting the actual member length. This provides high-precision foundational data for structural health monitoring and deformation analysis, and supports updates to the digital twin model.

[0096] Finally, the potential for end-to-end automation enhances engineering efficiency. Modular design allows each step (baseline definition, location check, number verification, port processing) to be an independent module, which can be flexibly combined or replaced. It adapts to the measurement needs of different types of steel truss members, reducing algorithm development costs. Binary identifier output ensures all verification results are output in binary form (e.g., compliance identifiers, numbering direction identifiers), facilitating integration with automated assembly systems or BIM models. This achieves closed-loop control throughout the measurement-analysis-correction process, reducing manual intervention and improving construction efficiency.

[0097] In summary, this method significantly improves the preprocessing accuracy and efficiency of measurement data for bolted steel truss members through core technological innovations such as geometric benchmark optimization, multi-level verification, space-to-plane conversion, and standardized processing. It provides reliable technical support for the digital manufacturing and assembly of steel structures and is particularly suitable for engineering scenarios with extremely high accuracy requirements, such as large bridges and high-rise buildings.

[0098] In the above embodiments, further: in step S1, before data preprocessing, a benchmark for the measurement data structure must be established, namely, the measurement point extraction method and point coding system, to ensure that all measurement data are extracted according to specified rules. This includes the measurement point extraction method and the construction of a matching point numbering system;

[0099] Measurement point extraction method: First, the ports and nodes of the steel structure members are used as the carriers for extracting bolt hole positions; the center point of the member is determined by the geometric parameters of the member section as the origin of the local coordinate system, and a local coordinate system is established based on this: the X-axis is parallel to the length direction of the member, the Y-axis is perpendicular to the member section, and the Z-axis is parallel to the functional surface of the port; on each functional surface of the port or node box opening, four reference holes that can form a rectangle within the threshold range are defined, and the center of the reference hole is designated as the measurement point. The coordinates of the four measurement points are collected using a laser scanner. The four points provide a unique and reproducible geometric reference for hole position measurement through the "three-point plane determination and four-point optimization" method.

[0100] Construct a corresponding point numbering system: For a single functional surface of a port or node opening, take the upper left point of the functional surface as the starting point for numbering, and assign consecutive numbers to the four measurement points within the surface in a clockwise direction; For multiple functional surfaces contained in a single port or node opening, use the rod view as a spatial reference, and number the points on different functional surfaces of the opening sequentially according to the view orientation priority of the front surface, upper surface, rear surface, and lower surface; For multiple ports, openings, or node openings contained in a rod, use the rod's length axis as a reference, and number all ports, openings, and node openings sequentially according to the axis orientation priority from left to right. If the axis orientations are the same, then number them sequentially according to the view orientation priority of front, upper, rear, and lower.

[0101] It should be noted that the bolt hole positions are extracted from the ports and nodes of steel structural members. Key connection points in the design are directly used as measurement reference points. Bolt hole positions are the core control points for member assembly, and their positional accuracy directly affects connection strength and structural safety. Using these as a reference ensures that the measurement data is strongly correlated with actual assembly requirements. Bolt hole positions are typically precision-machined during manufacturing, resulting in high surface flatness and minimal error in the center coordinates acquired by the laser scanner, reducing systematic deviations caused by improper reference point selection. The local coordinate system is determined using the member's center point as the origin, combined with cross-sectional geometric parameters (such as cross-sectional shape and size): X-axis: parallel to the member's length direction (axial direction), ensuring length measurement is consistent with theoretical design; Y-axis: perpendicular to the member's cross-section, reflecting cross-sectional deformation or manufacturing errors; Z-axis: parallel to the port functional surface, strongly correlated with the assembly plane. Four reference holes are defined on the functional surface of the port or node opening. A geometric reference is established through a "three-point plane determination, four-point optimization" mechanism. The local coordinate system is automatically generated according to the member's geometric parameters, requiring no manual intervention and adapting to the measurement needs of members of different specifications. The "three-point plane determination" method quickly establishes the initial plane, reducing computational complexity. The "four-point optimization" method uses the least squares method to fit the optimal plane, eliminating the influence of reference hole machining errors or scanning noise and improving the accuracy of the reference plane. For the same member in different measurement scenarios, as long as the reference hole position remains consistent, the same local coordinate system can be reproduced, ensuring data consistency.

[0102] Furthermore, starting from the top left point of the functional surface, numbers are sequentially arranged clockwise (e.g., P1, P2, P3, P4). Each measurement point has a unique number within the functional surface, avoiding duplicate or omissions. Clockwise numbering aligns with the right-hand coordinate system in mathematics, facilitating subsequent calculations of geometric parameters such as the surrounding direction and normal vector. The numbering rules can be embedded with algorithms to automatically match points with numbers, reducing manual annotation errors. For multiple functional surfaces of a single port or node opening, numbering follows the view priority order of "front surface → top surface → rear surface → bottom surface." The numbering order aligns with the observation perspective during member assembly, enabling engineers to quickly locate problem points. The view priority is consistent with the standard view direction in Building Information Modeling (BIM), supporting direct import of measurement data into BIM for comparative analysis. Numbering rules for side surfaces (e.g., left surface, right surface) can be flexibly added to adapt to complex member structures. Along the axis of the member's length, ports or node openings are numbered in order from left to right; if the axis orientation is the same (such as in a symmetrical structure), the numbering is based on the view priority of "front → top → back → bottom"; all measurement points of the entire member are uniquely numbered to avoid numbering conflicts when crossing ports / nodes; the numbering order is consistent with the actual assembly order of the member, supporting step-by-step verification of assembly accuracy according to the numbering process; automated algorithms can process data in batches according to the numbering order, improving efficiency.

[0103] Finally, by using a high-precision datum surface ("three-point datum, four-point optimization") and a structured numbering system, systematic and random errors in the measurement data are reduced, providing a reliable foundation for subsequent deformation analysis and assembly deviation correction. Numbering rules and geometric constraints can be embedded in measurement software or robot control algorithms to achieve: automatic identification of measurement point carriers (e.g., bolt hole positions); automatic generation of local coordinate systems; automatic verification of point compliance (e.g., rectangular composition, numbering direction); and automatic output of measurement reports conforming to engineering standards. Datum definitions and numbering rules are directly linked to the design parameters of the members (e.g., cross-sectional shape, axial direction) and assembly processes (e.g., view priority, assembly sequence), ensuring that measurement data can be directly used to guide on-site construction and reducing data conversion errors. The structured data format (e.g., point numbers, coordinate values, datum surface parameters) conforms to international standards (e.g., ISO 10303 STEP) and can seamlessly integrate with BIM, digital twin, and other platforms, promoting the intelligent and digital development of steel structure engineering.

[0104] It is evident that the collaborative design of high-precision geometric benchmark construction and a structured point numbering system has achieved end-to-end optimization of steel structure measurement data from acquisition to management. Accuracy: Errors are reduced through a dynamic local coordinate system and optimization algorithms. Efficiency: Manual operations are reduced through automation rules. Engineering Value: By defining benchmarks and numbering logic that align with actual needs, assembly precision control and digital delivery are directly supported. This approach is particularly suitable for engineering scenarios with extremely high precision requirements for steel structures, such as large bridges and high-rise buildings, providing the industry with a replicable and scalable standardized solution.

[0105] In the above embodiment, further: in step S2, the measurement data is divided into groups of four points, with the first group of four points being A, B, C, and D. The position of each group of points is checked using a method based on geometric feature constraints.

[0106] S201. Calculate the Euclidean distance between any two points in a set to determine whether there are overlapping or nearest-neighbor outliers.

[0107]

[0108] S202. If the discreteness of the four points meets the requirements, calculate the cosine of the angle between the opposite side vectors AB and CD, and AD and BC.

[0109]

[0110] Set the cosine value of the opposite side to the theoretical parallel condition, and the maximum reference error with an absolute value of 1 for the cosine value, to determine whether the opposite side satisfies the directional parallelism constraint.

[0111] S203. If the four points satisfy the condition of parallel opposite sides, calculate the absolute differences between the lengths of opposite sides AB and CD, BC and DA, and the lengths of diagonals AC and BD. Set a maximum length error, first verify whether opposite sides are of equal length. If so, determine that the four points can form a parallelogram. Then verify whether the diagonals are of equal length. If so, further filter out rectangles from the parallelograms.

[0112] S204. Calculate each group of points in a loop, and finally output the preset requirements for whether each group of points can form a rectangle, with 1 for yes and 0 for no, using binary identifiers. This realizes the compliance check of the measurement data position and provides a geometric basis for subsequent data calculation.

[0113] It should be noted that step S2 uses a grouping inspection method based on geometric feature constraints to systematically verify whether every four points in the measurement data can form a rectangle. Its technical advantages are reflected in four aspects: accuracy control, anomaly handling, engineering adaptability, and automation support. It is especially suitable for compliance inspection of high-precision geometric benchmarks such as bolt hole positions in steel structures.

[0114] Among them, precision control: multi-level geometric constraints ensure the accuracy of rectangle determination. The geometric definition of a rectangle must simultaneously satisfy three conditions: parallel opposite sides, equal length opposite sides, and equal length diagonals. Step S2 verifies these conditions in stages, forming a closed-loop control chain of "distance filtering → direction constraint → length verification", significantly improving the determination accuracy. Euclidean distance pre-screening: eliminates overlapping or nearest-neighbor anomalies. Calculate the Euclidean distance between any two points in the group. If the distance is less than a threshold (such as twice the laser scanner's point error), it is determined to be an overlapping or nearest-neighbor anomaly. Quickly remove invalid data: overlapping points (such as scanning noise or labeling errors) will directly cause subsequent geometric calculations to fail (such as meaningless vector angles). Pre-screening can avoid invalid calculations. Reduce error propagation: nearest-neighbor points may be misjudged as rectangle vertices due to measurement errors. Pre-screening reduces the risk of misjudgment in subsequent steps. Parallelism constraint of opposite side directions: rigorous verification based on cosine values. Calculate the cosine of the angle between opposite vectors (e.g., AB and CD, AD and BC), and set theoretical parallelism conditions: the absolute value of the cosine is 1 (perfect parallelism), with a maximum allowable reference error (e.g., ±0.01) to accommodate measurement noise. The cosine value directly reflects the vector direction relationship and is more stable than traditional angle difference calculations (avoiding numerical errors from the inverse cosine function). By setting a preset error threshold, the tolerance range for parallelism is clearly defined, avoiding misjudging the rectangle's failure due to minor deviations. Verification of opposite side directions is independent of length; even if there are errors in side length (e.g., manufacturing tolerances), the direction can still be confirmed first. Verification of opposite side and diagonal lengths: Step-by-step screening of parallelograms and rectangles. First step: Verify that opposite sides are of equal length (AB=CD, BC=DA); if satisfied, it is determined to be a parallelogram. Second step: Based on the parallelogram, verify that the diagonals are of equal length (AC=BD); if satisfied, it is determined to be a rectangle. Set a maximum length error (e.g., ±0.5mm) to accommodate reasonable deviations in manufacturing or measurement. First, identify the parallelogram (basic geometric shape), then filter for rectangles (special parallelograms), following a progressive relationship defined by mathematical rules. An error threshold is used to distinguish between "theoretical rectangles" and "actually acceptable rectangles," avoiding the erroneous discarding of large amounts of data due to absolute idealization. If the diagonals are of unequal length but parallel, it can be identified as a parallelogram rather than a rectangle, providing a basis for subsequent structural deformation monitoring.

[0115] Among them, anomaly handling: systematically exclude non-rectangular data to ensure data quality. There may be anomalies such as noise points, misaligned points, and missing points in the measurement data. Step S2 achieves efficient anomaly detection through multi-condition joint verification. (1) Coincident / nearest neighbor point detection: avoid invalid calculation. If there are coincident points in the group (such as A and B coinciding), then the vector AB is a zero vector, and the subsequent angle and length calculations are meaningless. Pre-screening can directly mark the group as invalid, reducing resource waste. (2) Direction and length joint verification: exclude "pseudo rectangles". Point groups that only meet some conditions (such as opposite sides being parallel but not of equal length) will be excluded to avoid mistakenly judging non-rectangular shapes such as trapezoids and rhombuses as compliant. For example, if AB∥CD but AB≠CD, even if the diagonals are of equal length, it will still be judged as non-rectangular because it does not meet the parallelogram condition.

[0116] Among these aspects, engineering adaptability is crucial: It aligns with the measurement requirements of bolt holes in steel structures. The rectangular arrangement of bolt holes in steel structures is a core control element for assembly accuracy, and step S2's design directly serves this engineering scenario. It supports the reproducibility verification of the "four-point plane" datum. In step S1, the rectangular reference hole is used to establish a local coordinate system and geometric datum. Step S2 verifies whether the four points form a rectangle, thus checking the reproducibility of the datum plane in reverse: if the four points cannot form a rectangle, it indicates that the datum plane may have shifted due to member deformation or measurement errors, requiring recalibration. If the four points meet the rectangular condition, the datum plane is stable, and subsequent measurement data is reliable. It is also compatible with manufacturing tolerances and measurement errors. In steel structure manufacturing, bolt hole positions have allowable tolerances (e.g., ±1mm). Step S2 combines the theoretical rectangle with the actual tolerance range by setting length and direction error thresholds, avoiding the erroneous rejection of compliant data due to absolute idealization.

[0117] The system supports automated output of binary identifiers for seamless integration with subsequent calculations. Step S2 ultimately outputs the compliance of each group of points using binary identifiers (1 / 0), providing a standardized interface for automated processing. This efficient data labeling feature uses binary identifiers with minimal storage space (1 bit / group) and can be directly used for logical judgments (e.g., "IF identifier = 1 THEN retain data"), improving processing efficiency. Batch processing and decision-making are also supported. The automated algorithm can iteratively process all point groups, quickly filtering out compliant rectangular point groups, providing geometric foundation data for subsequent coordinate system transformations (e.g., using the rectangle's center as the origin of the local coordinate system); assembly deviation calculations (e.g., comparing the positional deviations of theoretical and actual rectangles); and digital model updates (e.g., importing compliant point groups into BIM for real-time calibration).

[0118] As can be seen, step S2, through phased geometric constraint verification, achieves high-precision and high-reliability checks on rectangular point groups in the measurement data. It features a closed-loop precision mechanism: multi-level verification from distance and direction to length ensures the mathematical rigor of rectangle determination; anomaly immunity: systematically excluding coincident points, pseudo-rectangles, and other abnormal data to guarantee data quality; engineering adaptability: directly serving the reproduction of bolt hole references and assembly precision control in steel structures; and automation friendliness: binary output supports efficient batch processing and seamless integration with digital engineering workflows. This method is particularly suitable for scenarios with extremely high geometric precision requirements (such as bolted connections of bridge steel box girders and node assembly of high-rise buildings), providing key technical support for digital construction of steel structures.

[0119] In the above embodiment, further: in step S3, a method based on spatial geometric projection and surrounding direction analysis is used to determine whether the numbering of each group of points is correct.

[0120] S301. Calculate any 3 points in a set of points (e.g., points A, B, and C in the first set, with their locations and numbers as follows). Figure 1 As shown in the figure, the cross product of the vectors is used to obtain the surface normal vector. The main projection plane is determined based on the maximum absolute value of each component of the normal vector, which is either the YZ, XZ, or XY plane, thus realizing the orthogonal projection transformation of the three-dimensional point set to the two-dimensional plane.

[0121] S302. Project a set of 4 points according to the main projection plane, and calculate the wrap value by adding the cross product of the coordinates of the adjacent points to the 4 projected points (P1(x1y1), P2(x2,y2), P3(x3,y3), P4(x4,y4)).

[0122]

[0123] S303. Preset observation direction vectors for different projection planes. Use the direction coefficient ±1, where 1 indicates observation from the positive direction and -1 indicates observation from the negative direction to unify the observation angle reference. Couple the orbit value with the observation direction coefficient to eliminate the influence of the observation angle on the orbit direction determination. If the calculation result is less than 0, determine that a group of 4 points is clockwise numbered; otherwise, determine that the 4 points are counterclockwise numbered.

[0124] S304. Calculate each group of points in a loop, and finally output whether the number of each group of points meets the preset requirements with binary identifiers, 1 for yes and 0 for no, so as to realize the automatic verification of the numbering order of points on each functional surface.

[0125] It should be noted that step S3 achieves automated verification of the numbering order of each group of points by using a method based on spatial geometric projection and surrounding direction analysis. Its technical advantages are reflected in four aspects: dimensionality reduction from three dimensions to two dimensions, unified observation perspective, rigorous mathematical verification, and engineering adaptability. It is especially suitable for scenarios that require strict numbering order, such as steel structure nodes and mechanical parts.

[0126] Among these features, the following steps are included: Dimensionality reduction from 3D to 2D: This simplifies complex geometric problems and improves computational efficiency. Directly determining the order of four points (clockwise / counterclockwise) in 3D space requires handling complex spatial directional relationships. Step S3 transforms this problem into a 2D planar problem through orthogonal projection, significantly reducing computational complexity. Unified observation perspective: This eliminates directional ambiguity and ensures consistent judgment. In 3D space, the numbering order of the same group of points may appear opposite depending on the observation perspective (e.g., counterclockwise from the front, clockwise from the back). Step S3 solves this problem by coupling observation direction coefficients. Mathematical rigor verification: This combines 3D information to avoid misjudgments in 2D projection. 2D projection may cause the calculation of wraparound values ​​to fail due to point overlap or collinearity (e.g., collinearity after four points are projected, resulting in a wraparound value of 0). Step S3 supplements the verification with 3D geometric constraints to ensure accuracy. Engineering adaptability: This directly serves the numbering verification of steel structures and mechanical parts. In scenarios such as bolt hole positions in steel structures and feature points of mechanical parts, the compliance of the numbering order directly affects assembly accuracy and automated processes. Step S3 is designed to closely align with these requirements. Points on each functional surface (e.g., a connection surface of a steel structure) need to be numbered in a specific order (e.g., counter-clockwise) to align with the assembly process or digital model. Step S3 directly verifies whether the point numbering meets the functional surface requirements by using preset observation direction coefficients. Seamless integration with engineering specifications: The numbering order requirements are flexibly configurable (e.g., adjusted via direction coefficients) to adapt to different standards (e.g., ISO, GB). Support for automated assembly: Compliant numbering sequences can be directly used for downstream tasks such as robot path planning and welding sequence optimization. A binary identifier (1=compliant, 0=non-compliant) is output for each group of points, forming structured data that can be directly used for logical judgments (e.g., "IF identifier = 1 THEN execute assembly") or visual marking (e.g., highlighting non-compliant point groups in the BIM model). High processing speed: Binary data occupies little space, suitable for real-time verification of large-scale point groups. Strong compatibility: Seamless integration with programming languages ​​such as Python and C++, supporting calls from industrial systems such as MES and ERP.

[0127] As can be seen, step S3 achieves efficient and accurate verification of the numbering order of 3D point groups through spatial geometric projection dimensionality reduction, unified observation perspective, rigorous mathematical verification, and engineering adaptability design. Its technical advantages can be summarized as follows: High computational efficiency: The 3D-to-2D projection simplifies the problem, making it suitable for batch processing. Viewpoint independence: Observational ambiguity is eliminated through directional coefficient coupling, ensuring consistency. Strong robustness: Combining 3D coplanarity checks and optional 3D verification avoids misjudgments. Engineering-friendly: Directly supports functional surface definition and automated assembly requirements, outputting structured data. This method is particularly suitable for scenarios sensitive to numbering order (such as steel structure node assembly and aircraft blade feature point marking), providing key technical support for geometric compliance checks in digital manufacturing.

[0128] In the above embodiments, further: In step S4, during the data preprocessing stage, a feature inspection system for the rod ends is constructed according to the numbering rules, which can determine whether there are any special cases in the rods, laying a standardized foundation for subsequent geometric calculations. The specific process is as follows:

[0129] S401. Based on the preset port numbering sequence of the member axis direction, the measurement point numbers of the front and rear ports of the member can be effectively distinguished; combined with the numbering sequence of a single port and a single functional surface, the rules for extracting measurement points can be clearly defined.

[0130] In the above embodiment, the measurement point extraction process in step S401 is further as follows:

[0131] S4011. First, ensure that the number of measurement points for each port is fixed and the same (e.g., 16 port points).

[0132] S4012. Secondly, when extracting the measurement points of the front port, take the upper left side of the port as the starting point and follow the numbering principle of different functional surfaces of a single port. Extract the continuous measurement points corresponding to the front port in ascending order (such as 1, 2, 3, 4, ..., 13, 14, 15, 16).

[0133] S4013. When extracting measurement points at the rear port, take the measurement point number corresponding to the total number of holes n of the rod as the starting point, and extract the continuous measurement point numbers corresponding to the rear port in reverse order (such as n, n-1, n-2, n-3, ..., n-12, n-13, n-14, n-15). The extraction range matches the total number of measurement points at a single port. The locations and numbers of the front and rear port points are shown in Figure 2(a) and Figure 2(b).

[0134] S402. For the identified front and rear ports, extract the measurement points at the corresponding positions on the front surface of the port. For example, the Z-axis coordinates of points numbered 1 and 4 correspond to points numbered n-14 and n-13. Calculate the normal distances D1 and D2 between the two points in the Z-axis direction (height direction), as shown in Figure 2(a).

[0135] S403. Set a distance difference judgment threshold (e.g., ±5mm). Compare the normal distances D1 and D2 between the front and rear port measurement points. When the measured distance difference d exceeds the threshold, it is determined that the current member is a "member with unequal heights at the front and rear". The coordinates of the port measurement points need to be standardized, as shown in Figure 2(a).

[0136] S404. Adopting the principle of "high port coordinate reference assimilation", the Z coordinate of the port measurement point on the side with the larger height value between the front and rear ports is extracted as the reference. The Z coordinate of the feature point on the side with the smaller height value is assigned to the reference value, that is, moved to the position of point J. This realizes the alignment of the coordinate reference of the two port measurement points in the height direction, eliminating the spatial coordinate deviation caused by the inconsistency of port height, as shown in Figure 2(a).

[0137] S405. After completing the port benchmark unification, calculate the Euclidean distance between the port feature points along the axis of the member, and use it as the final measured value of the member length L. This avoids the length calculation deviation caused by the height difference of the port and ensures that the member length calculation result reflects its true size, rather than the oblique distance affected by spatial height misalignment.

[0138] It should be noted that step S4, by constructing a feature inspection system for member ends and combining standardized measurement point extraction rules with a dynamic benchmark unification strategy, achieves accurate verification and standardized processing of member geometric parameters. Its technical advantages are reflected in four aspects: the standardization of measurement point extraction, adaptability to special working conditions, accuracy of geometric calculations, and directness in engineering applications. It is particularly suitable for scenarios with high spatial geometric accuracy requirements, such as steel structure members and mechanical shaft parts.

[0139] The standardization of measurement point extraction is crucial: standardized input ensures the reliability of geometric calculations. Steps S401-S4013 establish a standardized process for measurement point extraction by fixing the number of points, unifying the numbering order, and clarifying extraction rules, providing high-quality, unambiguous input data for subsequent geometric calculations. A fixed number of measurement points is preset for each port (e.g., 16), ensuring consistent data dimensions across all member ports. This avoids dimensional mismatches in subsequent calculations (e.g., distance, angle) caused by different numbers of points at different member ports. A fixed number of points facilitates the development of generalized processing programs (e.g., loop extraction, batch calculation), reducing manual intervention. Abnormal point counts (e.g., missing points) can be quickly identified as data errors, triggering warnings or retesting processes. Unified port numbering order clarifies spatial correspondence. The numbering order of front and rear ports is associated with the member axis direction, ensuring that "point 1" corresponds spatially in the front and rear ports (e.g., both located on the left or right). Combined with the numbering principle of individual functional surfaces (e.g., clockwise / counterclockwise), measurement points for specific functional surfaces (e.g., welding surfaces, assembly surfaces) can be quickly located. Even if the members rotate in space, the numbering order can still maintain logical consistency through the starting point positioning, avoiding confusion caused by changes in direction. Dynamic starting point design: adapts to different member structures. Supports unified processing of members with different numbers of holes (e.g., 16-hole and 32-hole members), eliminating the need to design separate extraction rules for each number of holes. When adding a new member type (e.g., more holes), only the value of n needs to be adjusted, without modifying the core logic.

[0140] Among them, the adaptability to special working conditions: dynamically handling members with unequal heights at the front and rear. Steps S402-S404 solve special working conditions where the heights of the front and rear ends are inconsistent (such as due to machining errors or assembly tilt) by calculating the normal distance, threshold judgment, and benchmark unification, avoiding geometric calculation distortion. Specific numerical values ​​(such as D1=3.2mm) rather than qualitative descriptions (such as "different heights") provide an objective basis for subsequent judgments. Calculating the distance between multiple point pairs (such as 1 and n-14, 4 and n-13) can identify local height differences (such as unequal heights on only one side) or global height differences. Threshold judgment: clearly defining the boundaries of special working conditions. Setting a distance difference judgment threshold (such as ±5mm), when the measured distance difference d exceeds the threshold, it is judged as "members with unequal heights at the front and rear". Fault-tolerant design: allowing a certain range of machining errors (such as ±5mm) to avoid triggering unnecessary benchmark unification processes due to small height differences. Standardized judgment: the threshold can be adjusted according to engineering needs (such as ±2mm for high-precision scenarios) to achieve flexible adaptation. High-port benchmark assimilation: Eliminates spatial coordinate deviation. The Z-coordinate of the measurement point on the port with the larger height value is selected as the benchmark, and the Z-coordinate of the other port is assigned to this benchmark value. Geometric consistency: After benchmark unification, the measurement points on both ports are aligned in the height direction, avoiding the calculation of the rod length as an oblique distance (not the true length) due to height differences. Minimal intervention principle: Only the coordinates on the side with the smaller height value are modified, preserving other dimensions of the original data (such as x and y coordinates) to reduce information loss. Visualization support: The benchmark-unified data can be directly used for 3D model rendering, intuitively displaying the true spatial posture of the rod.

[0141] The accuracy of geometric calculations is based on the actual size measurement using standardized data. Step S405 calculates the Euclidean distance along the axis to ensure that the length of the member reflects its true size, rather than the oblique distance affected by spatial height misalignment. This eliminates the interference of height difference on length calculation, ensuring that L is the projected length of the member on the horizontal plane (i.e., the design length). After benchmark unification, the z-component difference is 0, reducing one squaring operation and improving calculation efficiency. Anti-interference capability is also ensured by eliminating the influence of spatial posture. Even if the member is tilted in space (e.g., at an angle θ with the horizontal plane), after benchmark unification, the length calculation is still based on the horizontal projection, avoiding the error of oblique distance L′=L / cosθ. This meets the actual needs of length measurement in scenarios such as steel structures and mechanical parts (e.g., bolt hole spacing, effective length of shaft parts). It provides accurate basic data for subsequent assembly and strength analysis, avoiding cascading problems caused by length errors (e.g., bolts cannot be inserted, stress concentration).

[0142] The direct application in engineering seamlessly integrates assembly, inspection, and digital processes. Step S4 is designed to closely align with the actual needs of engineering scenarios such as steel structures and mechanical manufacturing. Its output data can be directly used for assembly guidance, quality inspection, and digital modeling. Assembly Guidance: Standardized benchmarks support automated integration. The coordinates of the member ports after benchmark standardization can be used for robot assembly path planning (e.g., welding robots moving along a standardized height trajectory). This avoids collisions of the robot's end effector or welding position deviations due to port height differences. Standardized data reduces the steps of manual parameter adjustment, accelerating production line changeover. Quality Inspection: Clearly determines member compliance. The measured length L is compared with the designed length L, and the member's compliance is determined by combining the height difference threshold. Both length and height difference are checked simultaneously to avoid loopholes in single-parameter testing (e.g., length is acceptable but height exceeds tolerance). Data before and after benchmark standardization is recorded for easy problem tracing (e.g., analyzing whether the height difference originates from processing or assembly). Digital Modeling: Supports BIM / MES system integration. Structured data (e.g., port point coordinates, length, height difference) is output and can be directly imported into BIM models or MES systems. It synchronizes the geometric state of physical components in real time to maintain the accuracy of the digital twin model. This provides foundational data for production scheduling, inventory management, and other tasks, optimizing resource allocation.

[0143] As can be seen, step S4 achieves high-precision and high-reliability processing of the geometric parameters of the members through standardized measurement point extraction, dynamic benchmark unification, Euclidean distance calculation, and engineering adaptation design. The input specifications are standardized: a fixed number of points and a unified numbering order ensure data quality; it adapts to special working conditions: dynamically handling members with unequal heights avoids calculation distortion; the calculation is accurate: after benchmark unification, the Euclidean distance directly reflects the true length; and it is engineering-friendly: the output data directly supports assembly, inspection, and digitalization processes. This method is particularly suitable for scenarios with high geometric accuracy requirements (such as steel structure bridge nodes and aero-engine shaft parts), providing key technical support for geometric compliance checks in intelligent manufacturing.

[0144] The present invention also claims protection for a system comprising a processor and a memory, the memory storing at least one instruction, at least one program, code set, or instruction set, the at least one instruction, the at least one program, the code set, or the instruction set being loaded and executed by the processor to implement the preprocessing method as described in any of the above.

[0145] The various embodiments in this specification are described in a related manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the system embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

[0146] The above description is merely a preferred embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention are included within the scope of protection of the present invention.

Claims

1. A method for preprocessing measurement data of bolted steel truss members, characterized in that, The steps include: S1, defining the measurement reference; S2, checking the data location; S3, checking the data number; S4, checking the port characteristics and processing. S1. Define the measurement benchmark: This includes the measurement point extraction method and point coding system. First, define the measurement point extraction method, using the ports and nodes of steel structure members as the extraction carriers for bolt hole positions. Establish a local coordinate system based on the geometric parameters of the member cross-section. Define four reference holes that can form a rectangle on the functional surface of the port and node box openings. Use a laser scanner to collect the center of the hole to obtain the coordinates of the four measurement points. Establish a geometric benchmark using the "three-point surface determination, four-point optimization" mechanism. Then, construct a matching point numbering system. Within each functional surface, the points are numbered clockwise. Different functional surfaces are numbered according to view priority. Different ports or nodes are numbered according to the priority of the axis from left to right. For axes with the same orientation, the points are numbered according to the priority of the view orientation. S2. Check data location: Divide the measurement data into groups of 4 points, and use the geometric feature constraint method to iteratively verify the compliance of each group of points: First, overlap or nearest neighbor anomalies are identified by calculating the Euclidean distance between any two points. Secondly, calculate the cosine of the angle between the vectors of opposite sides, compare it with the maximum error of the theoretical parallel condition, and verify the parallelism of the opposite side directions. Finally, the absolute difference between the lengths of the opposite side and the diagonal is calculated. First, the opposite sides are verified to be of equal length to determine the parallelogram. Then, the diagonals are verified to be of equal length to filter the rectangles. Finally, the compliance of the rectangle composition of each group of points is output with binary identifier. S3. Check data numbering: Each group of 4 points constitutes a functional surface. Based on spatial geometric projection and surrounding direction analysis, determine the correctness of the point numbering on each functional surface: First, the normal vector of the plane is obtained by the cross product of the three-point vectors, and then the main projection plane is determined, realizing the orthogonal projection of the three-dimensional point set into two dimensions; Secondly, the surrounding value is calculated for the four projected points, and the viewpoint influence is eliminated by combining the preset observation direction vector to determine the surrounding direction of the point number of each functional surface. Finally, output the compliance of the numbering of each group of points using binary identifiers; S4. Check and handle port characteristics: Establish a standardized port characteristic handling process; First, distinguish the front and rear ports according to the port numbering order of the axis, and extract the measurement points in combination with the functional surface numbering rules; Secondly, calculate the Z-axis distance difference between the measurement points at corresponding positions on the front surfaces of the front and rear ports, and set a threshold to identify "uneven height rods". Then, the principle of "high port coordinate assimilation" is adopted to unify the Z coordinates of the low port points with the Z coordinates of the high side port points as the reference, thereby eliminating height deviation; Finally, the Euclidean distance between the feature points at both ends is calculated along the axis to obtain the actual measured value of the rod length.

2. The pretreatment method according to claim 1, characterized in that: Step S1 includes the method for extracting measurement points and the construction of a corresponding point numbering system; Measurement point extraction method: First, the ports and nodes of the steel structure members are used as the carriers for extracting bolt hole positions; the center point of the member is determined by the geometric parameters of the member section as the origin of the local coordinate system, and a local coordinate system is established based on this: the X-axis is parallel to the length direction of the member, the Y-axis is perpendicular to the member section, and the Z-axis is parallel to the functional surface of the port; on each functional surface of the port or node opening, four reference holes that can form a rectangle within the threshold range are defined, and the center of the reference hole is designated as the measurement point. The coordinates of the four measurement points are collected using a laser scanner. The four points provide a unique and reproducible geometric reference for hole position measurement through the "three-point plane determination and four-point optimization" method. Construct a corresponding point numbering system: For a single functional surface of a port or node opening, take the upper left point of the functional surface as the starting point for numbering, and assign consecutive numbers to the four measurement points within the surface in a clockwise direction; For multiple functional surfaces contained in a single port or node opening, use the rod view as a spatial reference, and number the points on different functional surfaces of the opening sequentially according to the view orientation priority of the front surface, upper surface, rear surface, and lower surface; For multiple ports, openings, or node openings contained in a rod, use the rod's length axis as a reference, and number all ports, openings, and node openings sequentially according to the axis orientation priority from left to right. If the axis orientations are the same, then number them sequentially according to the view orientation priority of front, upper, rear, and lower.

3. The pretreatment method according to claim 1, characterized in that: In step S2, the measurement data is divided into groups of four points. Let the four points in the first group be A, B, C, and D. The position of each group of points is checked using a method based on geometric feature constraints. S201. Calculate the Euclidean distance between any two points in a set to determine whether there are overlapping or nearest-neighbor outliers. S202. If the discreteness of the four points meets the requirements, calculate the cosine of the angle between the opposite side vectors AB and CD, and AD and BC. Set the cosine value of opposite sides to the theoretical parallelism condition, and determine whether opposite sides satisfy the directional parallelism constraint by setting the maximum reference error of the absolute value of the cosine value as 1. S203. If the four points satisfy the condition that the opposite sides are parallel, calculate the absolute differences between the lengths of the opposite sides AB and CD, BC and DA, and the lengths of the diagonals AC and BD. Set the maximum length error, first verify whether the opposite sides are of equal length. If they are, determine that the four points can form a parallelogram. Then verify whether the diagonals are of equal length. If they are, further filter out rectangles from the parallelograms. S204. Calculate each group of points in a loop, and finally output the preset requirements for whether each group of points can form a rectangle, with 1 for yes and 0 for no, using binary identifiers. This realizes the compliance check of the measurement data position and provides a geometric basis for subsequent data calculation.

4. The pretreatment method according to claim 1, characterized in that: In step S3, a method based on spatial geometric projection and surrounding direction analysis is used to determine whether the numbering of each group of points is correct: S301. Calculate the cross product of any three points in a set of points to obtain the surface normal vector. Determine the main projection plane, YZ, XZ or XY plane, based on the maximum absolute value of each component of the normal vector, to realize the orthogonal projection transformation of the three-dimensional point set to the two-dimensional plane. S302. Project a set of 4 points according to the main projection plane, and calculate the wrap value by adding the cross product of the coordinates of adjacent points to the 4 projected points (P1(x1y1), P2(x2,y2), P3(x3,y3), P4(x4,y4)). S303. Preset observation direction vectors for different projection planes. Use the direction coefficient ±1, where 1 indicates observation from the positive direction and -1 indicates observation from the negative direction to unify the observation angle reference. Couple the circumference value with the observation direction coefficient to eliminate the influence of the observation angle on the circumference direction determination. If the calculation result is less than 0, determine that a group of 4 points is clockwise numbered; otherwise, determine that the 4 points are counterclockwise numbered. S304. Calculate each group of points in a loop, and finally output whether the number of each group of points meets the preset requirements with binary identifiers, 1 for yes and 0 for no, so as to realize the automatic verification of the numbering order of points on each functional surface.

5. The pretreatment method according to claim 1, characterized in that: In step S4, during the data preprocessing stage, a feature inspection system for the member ports is constructed based on the numbering rules. This system can determine whether there are any special cases in the members, laying a standardized foundation for subsequent geometric calculations. The specific process is as follows: S401. Based on the preset port numbering sequence of the member axis direction, the measurement point numbers of the front and rear ports of the member can be effectively distinguished; combined with the numbering sequence of a single port and a single functional surface, the rules for extracting measurement points can be clearly defined. S402. For the identified front and rear ports, extract the measurement points at the corresponding positions on the front surface of the port, and calculate the normal distances D1 and D2 between the two points in the Z-axis direction. S403. Set a distance difference judgment threshold, compare the normal distances D1 and D2 between the front and rear port measurement points, and when the measured distance difference d exceeds the threshold, it is determined that the current member is a "member with unequal height at the front and rear", and the coordinates of the port measurement points need to be standardized. S404. Adopting the principle of "high port coordinate reference assimilation", extract the Z coordinate of the port measurement point on the side with the larger height value between the front and rear ports as the reference, and assign the Z coordinate of the feature point on the side with the smaller height value to the reference value, that is, move it to the position of point J, so as to realize the coordinate reference of the two port measurement points in the height direction and eliminate the spatial coordinate deviation caused by the inconsistency of port height. S405. After completing the port benchmark unification, calculate the Euclidean distance between the port feature points along the axis of the member, and use it as the final measured value of the member length L. This avoids the length calculation deviation caused by the height difference of the port and ensures that the member length calculation result reflects its true size, rather than the oblique distance affected by spatial height misalignment.

6. The pretreatment method according to claim 5, characterized in that, The measurement point extraction process in step S401 is as follows: S4011. Ensure that the number of measurement points at each port is fixed and the same; S4012. When extracting measurement points at the front port, take the upper left side of the port as the starting point and follow the numbering principle of different functional surfaces of a single port to extract the continuous measurement points corresponding to the front port in ascending order. S4013. When extracting measurement points at the rear port, take the measurement point number corresponding to the total number of holes n of the rod as the starting point, and extract the continuous measurement point numbers corresponding to the rear port in reverse order. The extraction range matches the total number of measurement points at a single port.

7. A system, characterized in that: The system includes a processor and a memory, the memory storing at least one instruction, at least one program, a code set, or an instruction set, the at least one instruction, the at least one program, the code set, or the instruction set being loaded and executed by the processor to implement the preprocessing method as described in any one of claims 1 to 6.

Images