Search

CN-121982347-A - Intelligent processing and analyzing system for measurement data of arc hyperbolic aluminum single plate

CN121982347ACN 121982347 ACN121982347 ACN 121982347ACN-121982347-A

Abstract

The invention relates to the technical field of data processing, in particular to an intelligent processing and analyzing system for arc hyperbolic aluminum single plate measurement data, wherein real-point cloud data and corresponding standard design model data are obtained through a data acquisition module, and a characteristic analysis module is utilized to extract local Gaussian curvature variation indexes; the intelligent registration analysis module respectively endows the areas with differentiated first position constraint weight values and second position constraint weight values in a reverse curvature constraint objective function through a non-uniform weight matrix to realize preferential high-precision registration of key characteristic areas, and the final processing correction execution module respectively generates a rigid positioning instruction and a deformation compensation instruction to drive numerical control processing equipment based on the calculated weighted spatial deviation, so that the problem of precision imbalance caused by global equal constraint is effectively solved, and the manufacturing precision and the attaching quality of the components are remarkably improved.

Inventors

  • TANG YI

Assignees

  • 四川省恒屹达新材料科技有限公司

Dates

Publication Date
20260505
Application Date
20260123

Claims (10)

  1. 1. The intelligent processing and analyzing system for the measurement data of the arc hyperbolic aluminum single plate is characterized by comprising a data acquisition module (100), a characteristic analysis module (200), a region dividing module (300), an intelligent registration analysis module (400) and a processing correction execution module (500), wherein: the data acquisition module (100) acquires real-point cloud data of a target aluminum veneer and corresponding standard design model data; The feature analysis module (200) performs differential geometric operation on the real-time point cloud data and the standard design model data to extract a local Gaussian curvature variation index; The region dividing module (300) divides data points into a topological anchoring region and an elastic tolerance region according to the local Gaussian curvature variation index and the variation gradient thereof; The intelligent registration analysis module (400) establishes an iterative closest point registration process of spatial approximation of the real-point cloud data to the standard design model data and constructs a reverse curvature constraint objective function, calls the local Gaussian curvature variation index to generate a non-uniform weight matrix, and respectively endows the topological anchoring zone and the elastic tolerance zone with a differentiated first position constraint weight value and a differentiated second position constraint weight value in the reverse curvature constraint objective function through the non-uniform weight matrix; The processing correction execution module (500) generates a rigid positioning instruction aiming at the topological anchoring area and a deformation compensation instruction aiming at the elastic tolerance area based on the weighted space deviation calculated by the first position constraint weight value and the second position constraint weight value, and is used for driving numerical control processing equipment.
  2. 2. The intelligent processing and analyzing system for measuring data of an arc-shaped hyperbolic aluminum single plate according to claim 1, wherein the process of acquiring real-time point cloud data and corresponding standard design model data by the data acquisition module (100) specifically comprises: Controlling a three-dimensional laser scanning device to perform full-coverage continuous scanning on the surface of the target aluminum veneer, acquiring three-dimensional space coordinate information of each discrete point on the surface of the target aluminum veneer through a laser ranging principle, and packaging the acquired three-dimensional space coordinate information into real-time point cloud data after noise elimination and filtering pretreatment; And accessing an engineering database through a data interface, indexing and reading a theoretical three-dimensional geometric file generated by computer aided design software according to the unique number of the target aluminum veneer, and analyzing and loading the theoretical three-dimensional geometric file into corresponding standard design model data.
  3. 3. The intelligent processing and analyzing system for measuring data of an arc hyperbolic aluminum single plate according to claim 1, wherein the extracting process of the local gaussian curvature variation index specifically comprises: Traversing discrete space points contained in the real-time point cloud data and the standard design model data, and respectively calculating actual measurement Gaussian curvature and theoretical Gaussian curvature corresponding to each discrete space point; the absolute value of the difference between the gaussian curvature and the theoretical gaussian curvature is determined as the local gaussian curvature variation index of the discrete spatial point.
  4. 4. The intelligent processing and analyzing system for measurement data of an arc-shaped hyperbolic aluminum single plate according to claim 1, wherein the dividing of the topological anchoring area and the elastic tolerance area specifically comprises: calculating the variation gradient of the local Gaussian curvature variation index of each data point; Calculating the comprehensive characteristic value of the local Gaussian curvature variation index and the variation gradient through weighted summation; Defining a data point set with the comprehensive characteristic value strictly higher than a preset geometric sensitivity threshold value as a topology anchoring area, defining a data point set with the comprehensive characteristic value lower than or equal to the preset geometric sensitivity threshold value as an elastic tolerance area, and generating a binarization classification result.
  5. 5. The intelligent processing and analyzing system for measurement data of an arc-shaped hyperbolic aluminum single plate according to claim 1, wherein the intelligent registration and analyzing module (400) comprises a reverse curvature constraint registration construction unit (401), and the process of constructing a reverse curvature constraint objective function by the reverse curvature constraint registration construction unit (401) specifically comprises: Initializing a rigid body transformation matrix containing rotation parameters and translation parameters; applying the rigid body transformation matrix to transform actual point cloud data in iterative computation, and searching the nearest point on the surface of standard design model data to establish a point-to-point relationship; And constructing a minimized mathematical model containing the distance deviation term and the curvature deviation term as a reverse curvature constraint objective function.
  6. 6. The intelligent processing and analysis system for measurement data of an arc-shaped hyperbolic aluminum veneer according to claim 5, wherein the distance deviation term is defined as a square sum of spatial euclidean distances between all corresponding point pairs, and the curvature deviation term is defined as a square sum of gaussian curvature difference values between all corresponding point pairs.
  7. 7. The intelligent processing and analyzing system for measurement data of an arc-shaped hyperbolic aluminum single plate according to claim 4, wherein the intelligent registration analyzing module (400) comprises a region adaptive weighting solving unit (402), and the process of generating the non-uniform weight matrix by the region adaptive weighting solving unit (402) specifically comprises: converting each data point into a corresponding weight coefficient according to the binarization classification result, and constructing a non-uniform weight matrix taking the weight coefficients of all the data points as diagonal elements; Introducing the non-uniform weight matrix into a distance deviation term calculation formula of the inverse curvature constraint objective function, endowing a topology anchoring region with a high-magnitude first position constraint weight value, endowing a spring tolerance region with a low-magnitude second position constraint weight value, and solving an optimal transformation parameter by minimizing the weighted inverse curvature constraint objective function.
  8. 8. The intelligent processing and analyzing system for measurement data of an arc-shaped hyperbolic aluminum single plate according to claim 1, wherein the calculating process of the weighted spatial deviation specifically comprises: And unifying the real-point cloud data to a workpiece processing coordinate system by utilizing the optimal transformation parameters, calculating residual distance vectors of each point of the real-point cloud data relative to the design theoretical surface, and carrying out point-by-point adjustment calculation on the residual distance vectors by utilizing the first position constraint weight value and the second position constraint weight value to obtain weighted space deviation.
  9. 9. The intelligent processing and analyzing system for measuring data of an arc-shaped hyperbolic aluminum single plate according to claim 1, wherein the driving process of the numerical control machining equipment specifically comprises the following steps: Screening out weighted space deviation in the topological anchoring area, determining the optimal clamping position and posture parameters of the workpiece on the table top of the machine tool through six-degree-of-freedom pose calculation, and generating a rigid positioning instruction; meanwhile, extracting a residual distance vector in an elastic tolerance area, and converting the residual distance vector into a feeding depth correction amount of a forming cutter or a pressing head of a numerical control machine tool along the surface normal direction to generate a deformation compensation instruction; and finally, sequentially fusing and compiling the rigid positioning instruction and the deformation compensation instruction into a control code executable by the machine tool and transmitting the control code to the numerical control machining equipment.
  10. 10. The intelligent processing and analyzing system for measuring data of an arc hyperbolic aluminum veneer according to claim 3, wherein the calculation process of the actual measured gaussian curvature and the theoretical gaussian curvature specifically comprises: Traversing each discrete space point contained in the real-time point cloud data and the standard design model data, searching a neighborhood point set in a set range by taking each discrete space point as a center, constructing a local polynomial fitting curved surface based on the neighborhood point set by using a mobile least square method, and calculating corresponding actual measurement Gaussian curvature and theoretical Gaussian curvature by solving a first basic quantity and a second basic quantity of the local polynomial fitting curved surface at the current discrete space point.

Description

Intelligent processing and analyzing system for measurement data of arc hyperbolic aluminum single plate Technical Field The invention relates to the technical field of data processing, in particular to an intelligent processing and analyzing system for measuring data of an arc hyperbolic aluminum single plate. Background In the field of modern complex building curtain wall engineering, arc hyperbolic aluminum veneers are widely applied due to unique artistic expressive force and space molding capability, but the geometric accuracy control of the components in the manufacturing and assembling processes has extremely high difficulty. At present, a non-contact three-dimensional laser scanning technology is generally adopted in the industry to acquire surface point cloud data of an aluminum veneer, and the surface point cloud data is compared with a standard design model for analysis so as to evaluate the processing quality or generate a subsequent correction instruction. In the process, unifying measured data to a theoretical model coordinate system through a point cloud registration algorithm is a key link of data processing. However, the existing general point cloud registration technology is generally based on the principle of a least square method, so as to pursue that the global euclidean distance error is minimized as a single target, and this mathematical model defaults that the contribution degree of all discrete sampling points to the geometric shape is equal, that is, a calculation strategy of global uniform weights is adopted. The traditional processing mode ignores the non-uniformity characteristic of the geometric property of the surface of the arc hyperbolic aluminum veneer and the anisotropy of the material mechanics. In practice, the curvature characteristics of the different areas of the surface of the hyperbolic sheet have a significant difference in the definition of the spatial topology, namely the areas of high curvature or severe curvature change tend to be the "skeletons" that establish the spatial pose of the component, extremely sensitive to the positioning accuracy, while the areas of low curvature or tend to be flat belong to the "skin", allowing certain elastic deformations or manufacturing tolerances. When global alignment is performed by adopting equal constraint, accumulated deviation generated by gentle regions with larger occupied area and lower geometric importance often dominates the convergence direction of an objective function, so that an algorithm is forced to cater to the fitting degree of the non-key regions, the alignment precision of the high-curvature characteristic regions which are critical to positioning is sacrificed, the alignment result cannot truly reflect the key geometric characteristics of a component due to the algorithm defect of 'average sense', and further, the processing correction instruction generated based on the alignment result is coupled and confused when processing rigid positioning and deformation compensation, so that the industrial production requirement of high-precision hyperboloid is difficult to meet. Disclosure of Invention The invention aims to provide an intelligent processing and analyzing system for arc hyperbolic aluminum single board measurement data, which solves the problems in the prior art, and the specific technical problems include how to construct a relative space skeleton model reflecting physical position relation by analyzing port vectors and route hop information acquired by a physical connection interface, and performing intersection retrieval by combining a functional attribute distribution map and space and functional keywords extracted by voice instructions, so as to solve the technical problem that the existing modularized robot system is difficult to adaptively match a target module for instructions containing azimuth semantics due to lack of a space configuration sensing mechanism. In order to achieve the purpose, the invention provides the following technical scheme that the intelligent processing and analyzing system for the measurement data of the arc hyperbolic aluminum single plate comprises a data acquisition module, a characteristic analysis module, a region division module, an intelligent registration analysis module and a processing correction execution module, wherein: the data acquisition module acquires real-time point cloud data of a target aluminum veneer and corresponding standard design model data, and specifically comprises the following steps: Controlling a three-dimensional laser scanning device to perform full-coverage continuous scanning on the surface of the target aluminum veneer, acquiring three-dimensional space coordinate information of each discrete point on the surface of the target aluminum veneer through a laser ranging principle, and packaging the acquired three-dimensional space coordinate information into real-time point cloud data after noise elimination and filtering pretreatmen