CN-122023280-A - Digital detection method and system for virtual pre-assembly of large-scale bending member

CN122023280ACN 122023280 ACN122023280 ACN 122023280ACN-122023280-A

Abstract

The invention discloses a digital detection method and a digital detection system for virtual pre-assembly of a large-scale bending member, wherein an initial geometric model is constructed by acquiring three-dimensional coordinates of surface feature points of the member, the initial geometric model is led into an intelligent pre-assembly analysis platform of a steel bridge, an axis curvature and torsion angle parameter is calculated by calling a bending member curvature inversion model, the spatial posture of the member is corrected by a special-shaped member posture correction model, the model is consistent with an actual member, a matching and positioning of virtual and actual mark points is completed by utilizing an image mark matching and recognition algorithm, a coordinate mapping relation is established, virtual pre-assembly is further executed in the platform, joint surface gaps, misalignment amount and axis alignment deviation are calculated, and a digital detection report comprising curvature distribution, posture parameters and joint deviation is finally output. According to the method, through cooperation of multiple models and full-flow digital operation, detection accuracy and efficiency are improved, and reliable technical support is provided for quality control of preassembled large-scale bending members.

Inventors

XIA HUA
YANG CHUNMING
NIE JIANGBAO
ZHANG FEI
JING SHIJIE

Assignees

安徽精工建设集团有限公司

Dates

Publication Date: 20260512
Application Date: 20260108

Claims (10)

1. The digital detection method for virtual pre-assembly of the large-scale bending member is characterized by comprising the following steps of: S1, acquiring three-dimensional coordinate data of discrete feature points on the surface of a large-scale bending member, constructing a member initial geometric model based on feature point space distribution, and extracting a member axis direction vector, a section contour boundary point set and relative distance parameters among feature points; S2, introducing the initial geometric model into an intelligent pre-splicing analysis platform of the steel bridge, and calling a bending component curvature inversion model to calculate the curvature distribution of the component axis to obtain curvature values and torsion angle parameters of each section of the axis; S3, carrying out posture adjustment on the initial geometric model by adopting a special-shaped component posture calibration model, and correcting component space posture parameters through three-dimensional coordinate deviation analysis of feature points to enable the model posture to be consistent with an actual component; S4, carrying out matching positioning on preset mark points on the surface of the component by using an image mark matching recognition algorithm, obtaining the accurate coordinates of the mark points in a virtual pre-spelling coordinate system, and establishing a coordinate mapping relation between the virtual component and the actual component; S5, based on the calibrated geometric model and coordinate mapping relation, virtual pre-assembly operation is executed in the intelligent pre-assembly analysis platform of the steel bridge, and gap values, offset and axis alignment deviation parameters of the assembly surfaces of the components are calculated; s6, outputting a member pre-assembly suitability detection result according to deviation parameters generated in the virtual pre-assembly process, and forming a digital detection report comprising curvature distribution, attitude parameters and splicing deviation.
2. The method for digitally testing virtual pre-assembly of a large-scale bending member according to claim 1, wherein the bending member curvature inversion model has the expression: , Wherein, the For the overall curvature of a point on the axis of the torque member, Is the three-dimensional coordinates of the feature points, As a parameter of the arc length of the axis, For the curvature correction factor(s), Is the torsion angle of the cross section of the component, As the second derivative in the x-direction of the axis, For the rate of change of the torsion angle along the axis, Respectively is The first derivative of direction along the axis.
3. The method for digitally detecting virtual pre-assembly of a large-scale bending member according to claim 1, wherein the expression of the profile member posture calibration model is: , Wherein, the For the amount of the attitude calibration, The direction vector component is adjusted for the pose, As the characteristic point coordinate deviation value, The weight coefficients are calibrated for the pose, Respectively the components are wound around Rotation deviation angle of the shaft.
4. The method for detecting the virtual pre-assembled digitization of the large-scale bending member according to claim 1, wherein the expression of the image mark matching recognition algorithm is as follows: , Wherein, the In order for the marker points to match the similarity, Respectively virtual and actual mark point sets, Is the feature vector of the i-th marker point, In order to match the coefficients of the coefficients, For the number of marker points, For the average spacing between virtual and actual marker points, Is a standard pitch parameter.
5. The method for detecting the virtual pre-assembly digitization of the large-scale bending member according to claim 1, wherein the pre-assembly deviation calculation model of the intelligent pre-assembly analysis platform of the steel bridge is as follows: , Wherein, the For the total pre-assembly deviation, As the weight coefficient of the deviation is used, For the deviation of the geometric model, For the purpose of calibrating the deviation of the posture, For the coordinate mapping deviation to be a function of the coordinate, For the local deviation of the kth splice face, The number of the splicing surfaces is the number.
6. The method for virtually pre-assembled digital detection of large-scale bending members according to claim 1, wherein the model for virtually pre-assembled digital detection parameters of large-scale bending members is: , Wherein, the In order to detect the parameters of the complex, As a result of the coupling coefficient, In order to integrate the curvature of the web, For the amount of the attitude calibration, In order to match the degree of similarity, For the total pre-assembly deviation, For the surface area of the component, Is the component volume.
7. The method for digitally testing virtual preassembly of large-scale torque components according to claim 1, wherein S3 comprises the following sub-steps: S31, extracting three-dimensional coordinate data of the marked feature points in the initial geometric model, comparing the three-dimensional coordinate data with actual measurement coordinates of the feature points corresponding to the actual components one by one, and calculating coordinate deviation values of each feature point in the x, y and z directions; S32, inputting coordinate deviation values into a special-shaped component gesture calibration model, and determining the optimal rotation angle and translation quantity of the component around x, y and z axes through iterative calculation, so that the deviation square sum of the model feature point coordinates and the actually measured coordinates is minimum; s33, according to the calculated rotation angle and translation amount, carrying out space attitude adjustment on the initial geometric model, updating three-dimensional coordinates of all feature points in the model, and forming a preliminarily calibrated geometric model; And S34, performing deviation verification on the model after preliminary calibration, and repeating S31 to S33 if the average deviation of the feature points exceeds a preset threshold value until the average deviation meets the detection precision requirement.
8. The method for digitally testing virtual preassembly of large-scale torque components according to claim 1, wherein S4 comprises the following sub-steps: S41, carrying out image acquisition on preset mark points on the surface of the component, extracting shape features, gray level features and spatial position features of each mark point, and constructing a mark point feature database; S42, generating virtual mark points corresponding to the actual mark points in the virtual pre-spelling model, and calculating the feature similarity of the virtual mark points and the actual mark points by adopting an image mark matching recognition algorithm; S43, screening mark point pairs with similarity higher than a set threshold value, establishing a one-to-one correspondence between virtual mark points and actual mark points, and determining a coordinate mapping matrix; S44, based on the coordinate mapping matrix, converting the three-dimensional coordinate data of the actual component into a virtual pre-spelling coordinate system, and completing the coordinate unification of the virtual component and the actual component.
9. The method for digitally testing virtual preassembly of large-scale torque components according to claim 1, wherein S5 comprises the following sub-steps: s51, importing the calibrated geometric model and the actual component data with unified coordinates into an intelligent pre-splicing analysis platform of the steel bridge, and setting pre-splicing constraint conditions including splicing surface fitting requirements, axis alignment standard and deviation allowable ranges; s52, performing virtual splicing operation of components in a platform according to a preset pre-splicing sequence, calculating the gap value and the offset of each splicing surface in real time, and recording the included angle deviation of the axis at the splicing position; S53, monitoring deviation data in the pre-assembly process in real time, and if the deviation of a certain splicing surface exceeds an allowable range, automatically adjusting the spatial position of the component, and carrying out splicing calculation again; and S54, after virtual pre-assembly of all the components is completed, summarizing deviation parameters and axis alignment conditions of all the splicing surfaces, and generating pre-assembly deviation detail data.
10. The digital detection system for virtual pre-assembly of large-scale bending members is characterized in that the system is applied to the digital detection method for virtual pre-assembly of large-scale bending members, which is disclosed in claim 1, and comprises the following steps: the three-dimensional coordinate data acquisition unit is used for acquiring three-dimensional coordinate data of the surface characteristic points of the large-scale bending member, establishing data transmission connection with the intelligent pre-splicing analysis platform of the steel bridge, and transmitting the acquired coordinate data to the platform in real time; The geometric model construction and curvature inversion unit is connected with coordinate data output by the three-dimensional coordinate data acquisition unit, a component initial geometric model is constructed, the curvature inversion model of the bending component is called to calculate the axis curvature distribution, and the model and curvature parameters are sent to the posture calibration unit; the special-shaped component posture calibration unit is used for receiving the model data transmitted by the geometric model construction and curvature inversion unit, correcting the spatial posture of the component through the special-shaped component posture calibration model, and transmitting the calibrated model to the mark point matching and positioning unit; The image mark matching and positioning unit is used for collecting the image information of the mark points on the surface of the component, performing matching and positioning of virtual and actual mark points by using an image mark matching and recognition algorithm, establishing a coordinate mapping relation and sending the coordinate mapping relation to the virtual pre-assembly unit; the virtual pre-assembly and deviation calculation unit is connected with the model, the coordinate mapping relation and the platform preset parameters after the gesture calibration, performs virtual pre-assembly operation, calculates splicing deviation parameters and transmits deviation data to the detection report generation unit; the digital detection report generating unit is used for receiving the deviation parameters, the curvature distribution and the attitude parameters output by the virtual pre-assembly and deviation calculating unit, integrating the data according to a preset format, generating a digital detection report comprising various detection indexes and outputting the visual detection result.

Description

Digital detection method and system for virtual pre-assembly of large-scale bending member Technical Field The invention relates to the technical field of large-scale bent member pre-assembly, in particular to a digital detection method and system for virtual pre-assembly of a large-scale bent member. Background The large-scale bending member is used as a steel bridge core stress part, and the manufacturing precision and the pre-assembly suitability of the large-scale bending member directly influence the overall bearing performance and the service safety of the bridge. Along with the development of bridge engineering to large-span and complex modeling directions, the curvature change, the space attitude and the splicing accuracy requirements of the bending member are continuously improved, and the traditional detection mode relying on manual measurement and entity pre-splicing is difficult to meet the requirements of efficient and accurate construction. The virtual pre-assembly technology is a key path for solving the pre-assembly problem of the large-scale bending member by virtue of the advantages of digitization and visualization, but how to realize the accurate mapping, posture calibration and deviation quantification of the geometric model and the actual member is still a technical pain point to be broken through in the industry, and a set of digital detection method and system which are combined with the multi-model cooperation and intelligent algorithm support are needed to be constructed, so that technical guarantee is provided for the management and control of the construction working medium of the steel bridge. The prior art has the two remarkable defects that on one hand, in the traditional virtual pre-assembly detection method, the geometric model construction and the posture calibration of an actual component lack of a mathematical model support with accurate adaptation, and the model posture and the actual component are subjected to multi-experience parameter adjustment, so that deviation exists between the model posture and the actual component, the real curvature distribution and the spatial position relation of the component cannot be accurately reflected, and the reliability of pre-assembly deviation calculation is affected, on the other hand, the marker point matching positioning and the coordinate mapping are not intelligent enough, a single feature matching mode is adopted, the influence of component surface environment and measurement errors is easy, the unified precision of the coordinates of the virtual component and the actual component is low, a multi-parameter coupling quantification system is not formed by pre-assembly deviation analysis, and key deviations such as joint surface gaps, misplacement, axis alignment and the like are difficult to comprehensively and accurately identify, so that the application value of a detection result is restricted. Disclosure of Invention In order to overcome the defects and shortcomings in the prior art, the invention provides a digital detection method and system for virtual pre-assembly of a large-scale bending member. The technical scheme includes that S1, three-dimensional coordinate data of discrete feature points on the surface of a large-sized bending member are obtained, an initial geometric model of the member is built based on spatial distribution of the feature points, a member axis direction vector, a cross section outline boundary point set and relative distance parameters among the feature points are extracted, S2, the initial geometric model is led into a steel bridge intelligent pre-splicing analysis platform, curvature distribution of the member axis is calculated by calling the bending member curvature inversion model to obtain curvature values and torsion angle parameters of each cross section of the axis, S3, the initial geometric model is subjected to posture adjustment by adopting the special-shaped member posture calibration model, the spatial posture parameters of the member are corrected through three-dimensional coordinate deviation analysis of the feature points to enable the model posture to be consistent with actual members, S4, accurate coordinates of the feature points in a virtual pre-splicing coordinate system are obtained, a coordinate mapping relation between the virtual pre-splicing coordinate system and the actual members is built, S5, virtual pre-operation face calculation is performed in the steel bridge intelligent pre-splicing analysis platform based on the calibrated geometric model and the coordinate mapping relation, the pre-splicing operation face is calculated, the pre-splicing parameter is calculated according to the calculated, and the gap deviation of the pre-splicing parameter is generated, and the pre-splicing parameter is generated according to the error of the position is generated, and the pre-splicing parameter is generated, and the matching parameter is formed by matching the matching and