Search

CN-121998957-A - Traceability diagnosis method and system for size errors of welding frame

CN121998957ACN 121998957 ACN121998957 ACN 121998957ACN-121998957-A

Abstract

The invention relates to the technical field of welding frame manufacturing and detection, in particular to a traceability diagnosis method and system for welding frame size errors, wherein the method comprises the steps of obtaining actual measurement three-dimensional coordinates of a series of preset measurement targets on a welding frame, wherein the measurement targets comprise a functional reference point group for establishing a coordinate reference and a diagnosis characteristic point group for deformation analysis; the method comprises the steps of calculating overall space transformation parameters between a frame design model and an actual measured frame based on actual measurement coordinates of a functional reference point group to eliminate overall pose deviation, compensating the actual measurement coordinates of a diagnosis feature point group on the basis to obtain residual deviation reflecting welding deformation, extracting a welding deformation mode through geometric analysis and mode recognition, further matching and reasoning the overall space transformation parameters, the welding deformation mode and a process error source mapping knowledge base, and outputting suspected process error sources and corresponding process optimization suggestions, so that accurate analysis and effective tracing of welding frame size errors are realized.

Inventors

  • JIN ENHUI
  • CHEN YANG

Assignees

  • 江苏小牛电动科技有限公司

Dates

Publication Date
20260508
Application Date
20260129

Claims (10)

  1. 1. A method for traceable diagnosis of dimensional errors of welded frames, the method comprising: obtaining actual measurement three-dimensional coordinates of a series of preset measurement targets on a welding frame, wherein the preset measurement targets comprise a functional reference point group for establishing a coordinate reference and a diagnosis characteristic point group for deformation analysis; Calculating integral space transformation parameters required for aligning the frame design model with the actual measurement frame through a space registration algorithm based on the actual measurement coordinates of the functional reference point group; Based on the integral space transformation parameters, compensating actual measurement coordinates of each diagnosis feature point in the diagnosis feature point group to obtain residual deviation reflecting welding deformation, and carrying out geometric analysis and pattern recognition on the residual deviation to extract at least one predefined welding deformation pattern from a pre-constructed deformation pattern library; and matching and reasoning the type and the magnitude of the integral space transformation parameters and the identified deformation modes with a pre-established process error source mapping knowledge base, and outputting one or more suspected process error sources and process optimization suggestions related to the suspected process error sources.
  2. 2. The traceable diagnostic method of welding carriage size errors of claim 1, wherein said process error source mapping knowledge base is constructed by: Collecting historical production data, including frame design parameters, welding process parameters, measured coordinate data, confirmed process error sources and rectifying and modifying effect records; based on the historical production data, establishing a mapping rule from the measured characteristics to a process error source through a data mining method; And merging an analysis result of a welding deformation physical mechanism model constructed based on the welding thermodynamics and the material mechanics principle into the mapping rule as priori knowledge.
  3. 3. The traceable diagnostic method of welding carriage size error according to claim 2, wherein said welding deformation physical mechanism model is a reduced order proxy model established by a finite element method, the construction of which comprises: performing a full parameterized finite element simulation of the welding process to generate a large-scale sample dataset covering a process parameter space; Extracting dominant deformation modes from the sample dataset by using an intrinsic orthogonal decomposition method; and establishing a rapid mapping relation from key process parameters to dominant modal coefficients through Kriging interpolation or a neural network to form a reduced-order proxy model.
  4. 4. The method for traceable diagnosis of dimensional errors of welded frames according to claim 1, wherein said matching and reasoning further comprises: calculating a comprehensive confidence coefficient for each outputted suspected process error source; the integrated confidence level is calculated based on at least two factors, namely the consistency of the results of different reasoning paths, the statistical frequency of the occurrence of the error source in the historical data, and the matching degree of the error source and the current frame characteristic combination.
  5. 5. The method for traceable diagnosis of weld frame dimensional errors according to claim 1, wherein performing geometric analysis and pattern recognition on the residual deviation comprises: grouping the diagnosis characteristic point groups according to the structural positions on the frame and the welding process units; Carrying out statistical analysis on the residual deviation in each group to extract at least one feature of spatial distribution, direction consistency and magnitude change trend of the residual deviation; the extracted features are matched with typical welding deformation patterns in the deformation pattern library to determine a deformation pattern corresponding to the current frame welding deformation state.
  6. 6. The method for traceable diagnosis of weld frame dimensional errors of claim 5, wherein statistically analyzing residual deviations within a group comprises: For point groups distributed on the tubular structure, calculating the tendency of bending or twisting through curve or curve fitting; for the point group distributed in the weld area, the deviation vectors of the symmetrical points located at both sides of the weld are analyzed to calculate the asymmetric shrinkage.
  7. 7. The traceable diagnosis method of welding frame size errors according to claim 5, wherein the matching is achieved by a twin neural network, specifically comprising: Inputting the extracted feature vector into one branch of the twin neural network, and inputting various typical mode features stored in the deformation mode library into the other branch; And identifying the mode with the highest similarity with the current characteristic as the welding deformation mode by calculating the similarity in the characteristic space, and outputting the similarity value as the mode confidence.
  8. 8. The traceable diagnostic method of welded frame dimensional errors of claim 1, wherein the spatial registration algorithm is a weighted robust registration algorithm comprising: assigning a weight to each point in the functional reference point group; aiming at minimizing the weighted residual error square sum, solving the optimal transformation parameters for aligning the design model with the actual measurement frame; wherein the weights are dynamically assigned based at least on functional importance of the corresponding fiducial points, or historical measurement stability information.
  9. 9. The traceable diagnostic method of welding carriage size error according to claim 1, wherein the position of the preset measurement target is determined by: Based on a digital design model and a welding process file of the welding frame, automatically identifying welding joints and assembly reference characteristics; calling a corresponding diagnosis target spot arrangement template for each welding joint according to the type, and generating the diagnosis characteristic point group; The functional fiducial point group is generated for each assembly fiducial feature.
  10. 10. A traceable diagnostic system for weld frame dimensional errors, the system comprising: The coordinate measuring and acquiring module is used for acquiring actual measurement three-dimensional coordinates of a series of preset measuring targets on the welding frame, wherein the preset measuring targets comprise a functional reference point group for establishing a coordinate reference and a diagnosis characteristic point group for deformation analysis; The space registration module is used for calculating integral space transformation parameters required for aligning the frame design model with the actual measurement frame through a space registration algorithm based on the actual measurement coordinates of the functional reference point group; The deformation mode recognition module is used for compensating the actual measurement coordinates of each diagnosis characteristic point in the diagnosis characteristic point group based on the integral space transformation parameters to obtain residual deviation reflecting welding deformation, and carrying out geometric analysis and mode recognition on the residual deviation so as to extract at least one predefined welding deformation mode from a pre-constructed deformation mode library; And the error tracing module is used for matching and reasoning the type and the magnitude of the integral space transformation parameter and the identified deformation mode serving as geometric behavior characteristics of the welding process with a pre-established process error source mapping knowledge base and outputting one or more suspected process error sources and process optimization suggestions related to the suspected process error sources.

Description

Traceability diagnosis method and system for size errors of welding frame Technical Field The invention relates to the technical field of welding frame manufacturing detection, in particular to a traceable diagnosis method and system for welding frame size errors. Background The welding frame is used as a core bearing component of products such as two-wheeled electric vehicles, the key size deviation design value can be caused by thermal deformation, restraint deformation and the like generated in the welding process, size errors are formed, the geometric quality of the welding frame is directly related to the assembly precision of the electric two-wheeled vehicle and the use safety of the products, the size errors of the welding frame not only comprise structural deformation caused by welding, but also can be superimposed to measure the integral pose deviation caused by factors such as the placement pose, the reference establishment mode and the like, so that different error sources in the measurement results are mutually coupled, and the difficulty of subsequent analysis and diagnosis is increased. At present, two types of technologies are generally adopted in the industry for controlling the size of a welding frame, namely, a special physical inspection tool designed based on a specific vehicle model is adopted, quick qualitative judgment is carried out on a key position through simulation assembly, but the mode is poor in universality, quantitative information reflecting a welding deformation mechanism is difficult to provide, clear guidance cannot be provided for process optimization, and a three-coordinate measuring machine is utilized to acquire high-precision point cloud data, alignment comparison is carried out by means of software and a CAD model to generate a detection report, however, the method does not effectively distinguish integral pose deviation from welding deformation, various error factors are mixed in the deviation data, the analysis result is highly dependent on experience judgment of engineers, representative deformation characteristics are difficult to systematically identify from the measured data, and the deformation characteristics are stably related to welding process parameters, so that the size control stays in a post inspection stage, and accurate traceability and feedforward regulation of a welding process are difficult to realize. The information disclosed in this background section is only for enhancement of understanding of the general background of the invention and should not be taken as an acknowledgement or any form of suggestion that this information forms the prior art already known to a person skilled in the art. Disclosure of Invention The invention provides a traceable diagnosis method and a traceable diagnosis system for welding frame size errors, and accordingly the problems in the background technology are effectively solved. In order to achieve the aim, the technical scheme adopted by the invention is that the traceability diagnosis method for the dimensional errors of the welding frame comprises the following steps: obtaining actual measurement three-dimensional coordinates of a series of preset measurement targets on a welding frame, wherein the preset measurement targets comprise a functional reference point group for establishing a coordinate reference and a diagnosis characteristic point group for deformation analysis; Calculating integral space transformation parameters required for aligning the frame design model with the actual measurement frame through a space registration algorithm based on the actual measurement coordinates of the functional reference point group; Based on the integral space transformation parameters, compensating actual measurement coordinates of each diagnosis feature point in the diagnosis feature point group to obtain residual deviation reflecting welding deformation, and carrying out geometric analysis and pattern recognition on the residual deviation to extract at least one predefined welding deformation pattern from a pre-constructed deformation pattern library; and matching and reasoning the type and the magnitude of the integral space transformation parameters and the identified deformation modes with a pre-established process error source mapping knowledge base, and outputting one or more suspected process error sources and process optimization suggestions related to the suspected process error sources. Further, the process error source mapping knowledge base is constructed by: Collecting historical production data, including frame design parameters, welding process parameters, measured coordinate data, confirmed process error sources and rectifying and modifying effect records; based on the historical production data, establishing a mapping rule from the measured characteristics to a process error source through a data mining method; And merging an analysis result of a welding deformation physical mechanism model constru