Reverse modeling and quality control method of prefabricated components based on slicing method
Through laser scanning technology and slice method combined with chord height difference algorithm, the problem of low quality detection efficiency of prefabricated components in prefabricated buildings is solved, automated quality detection and precise processing are realized, and the production quality control efficiency of prefabricated components is improved.
Patent Information
- Application Number
- CN202210104963.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-01-28
- Publication Date
- 2024-11-08
- Estimated Expiration
- 2042-01-28
AI Technical Summary
In the prior art, the quality detection efficiency of prefabricated components of prefabricated buildings is inefficient and non-automated, resulting in the inability to effectively control the quality of prefabricated buildings.
The reverse modeling and quality control method of prefabricated components based on slice method is adopted, and three-dimensional point cloud data is obtained using laser scanning technology, feature points are extracted through slice method chord height difference algorithm, data preprocessing and reverse modeling are carried out to realize automated quality detection.
It improves the efficiency and accuracy of quality inspection of prefabricated components, realizes automated production quality control, reduces the difficulty of data processing, and intuitively evaluates component quality through data visualization.
Smart Images

Figure CN114529596B_ABST
Abstract
Description
Technical Field
[0001] The invention relates to a reverse modeling and quality control method for prefabricated components based on a slicing method, and belongs to the technical field of civil engineering and surveying and mapping. Background Art
[0002] Prefabricated buildings use a combination of prefabrication and cast-in-place construction, and the construction process is quite different from traditional building types. Prefabricated components of prefabricated buildings are produced on assembly lines in prefabrication factories and then transported to the construction site of prefabricated buildings. In most cases, prefabricated components need to be used in conjunction with cast-in-place buildings.
[0003] The production model of the assembly line has high economic benefits. However, since the production of prefabricated components is not on-site, the possibility of making changes if errors occur is relatively small. At the same time, the quality errors of prefabricated components will directly affect the quality of prefabricated buildings. Therefore, the quality requirements for prefabricated components are relatively high. Different from the quality requirements of cast-in-place concrete components, the production of prefabricated components is standardized, and the quality problems inside the components are often well controlled. In current practical applications, although prefabricated components are produced in a standardized manner in factories, the methods used for quality inspection mostly rely on manual measurement and measurement. This method is inefficient and has large errors, which is not conducive to mass production quality control of prefabricated components for prefabricated buildings. Summary of the invention
[0004] The technical problem to be solved by the present invention is to provide a method for reverse modeling and quality control of prefabricated components based on the slicing method, which utilizes laser scanning technology to collect surface quality information of prefabricated components, and performs reverse modeling through the slicing method to inspect the surface quality of prefabricated components, thereby overcoming the shortcomings of low efficiency and non-automation in the prior art of quality inspection of prefabricated components for assembled buildings.
[0005] The present invention adopts the following technical solutions to solve the above technical problems:
[0006] The reverse modeling and quality control method of prefabricated components based on the slicing method includes the following steps:
[0007] Step 1, obtaining three-dimensional point cloud data of the surface of prefabricated components of assembled buildings;
[0008] Step 2, preprocessing the three-dimensional point cloud data obtained in step 1, distinguishing plane points from boundary points, and removing discrete points and noise points in the three-dimensional point cloud data;
[0009] Step 3, setting a reference line according to the preprocessed point cloud data, slicing and segmenting the prefabricated building components along the reference line, extracting the height information of each slice, and projecting the slice onto a two-dimensional plane to obtain a two-dimensional slice plane;
[0010] Step 4, using the chord height difference algorithm, extracting feature points from points on the outer contour of the two-dimensional slice plane;
[0011] Step 5, connecting the feature points extracted in step 4 to obtain the two-dimensional contour of each slice, stretching the two-dimensional contour of each slice into a three-dimensional slice according to the height information of the slice, and combining all the three-dimensional slices according to the original slice positions to form a reverse reconstruction model;
[0012] Step 6: align the reverse reconstruction model obtained in step 5 with the BIM model of the prefabricated component forward design, analyze and synthesize them according to the quality control requirements, and use data visualization to evaluate whether the quality of the prefabricated components meets the requirements.
[0013] As a preferred solution of the present invention, in step 1, the three-dimensional point cloud data of the surface of the prefabricated component in .asc format is acquired by using a laser scanner.
[0014] As a preferred solution of the present invention, in step 2, the three-dimensional point cloud data of the surface of the prefabricated component in .asc format is converted into .obj format, and the three-dimensional point cloud data in .obj format is preprocessed using a K nearest neighbor classification algorithm.
[0015] As a preferred solution of the present invention, the specific process of step 3 is as follows:
[0016] Step 3.1, preset a reference line on the upper surface of the prefabricated component, establish a three-dimensional rectangular coordinate system to position the prefabricated component, make one of the side surfaces of the prefabricated component located in the YOZ plane, the side surface located in the YOZ plane is parallel to the cross section to be cut, the preset reference line direction is the X-axis direction, randomly sample 10% of the points on the surface where the reference line is located to fit the reference line, project the randomly sampled points onto the XOY plane, fit the points projected on the XOY plane, and obtain a unique quadratic polynomial by minimizing the square error;
[0017] The square error formula is:
[0018]
[0019] Where E is the square error, n is the number of projection points obtained by projecting onto the XOY plane, and y i is the y coordinate of the projection point projected onto the XOY plane, p(x i ) is the interpolation value of the k-th polynomial;
[0020] Let p(x i ) is expressed as the following linear equations:
[0021]
[0022] In the formula, x i is the x-coordinate of the point projected onto the XOY plane, a0, a1, …, a k is the coefficient;
[0023] make Multiply both sides of Y=XA by X T , that is, X T Y=X T XA, for the matrix equation X T Y=X T XA is inversely transformed to obtain:
[0024] A=(X T X) -1 X T Y, n>k
[0025] According to the obtained A, the interpolation points of two adjacent projection points among the n projection points are constructed, that is, n-1 interpolation points are obtained, the reference line is fitted according to the interpolation points, and the tangent line of each interpolation point on the reference line is calculated. The slope of the tangent line of each interpolation point is:
[0026] tant j =f(x)′
[0027] In the formula, tant j is the tangent slope corresponding to the jth interpolation point, and f(x) is the reference line fitted by the interpolation point;
[0028] The prefabricated component is sliced along the normal direction of each interpolation point, and the prefabricated component is divided into n slices;
[0029] Step 3.2, set each slice to be a straight line along its tangent direction, and keep the cross section of the slice unchanged, and rotate each slice around the Z axis:
[0030]
[0031] Where (x, y, z) is the coordinate of the point on the slice, (x′, y′, z′) is the coordinate of the point on the slice after rotation, is the rotation angle of the jth slice. The rotation angle is the angle between the normal direction of each slice and the Y axis of the three-dimensional rectangular coordinate system. The normal direction of each slice is calculated using the center point of the tangent line of each slice.
[0032] The height information of each slice is extracted using the x-coordinate of the point on the rotated slice, and the slice is projected onto a two-dimensional plane.
[0033] As a preferred solution of the present invention, the specific process of step 4 is as follows:
[0034] For the points on the outer contour of the two-dimensional slice plane, the chord height difference algorithm is used to determine whether each point is a feature point, as follows:
[0035] For the mth point P on the outer contour of the jth two-dimensional slice plane m , select two points P adjacent to it m-1 and P m+1 Construct line L: A m x+B m y+C m =0, calculate point P m The vertical distance d to the line L m , when d m When it is greater than or equal to the set threshold σ, point P m Feature points are extracted;
[0036]
[0037]
[0038] In the formula, (x m ,y m ) is point P m The coordinates of A m , B m , C m is the coefficient of the straight line L, M j is the number of points on the outer contour of the j-th two-dimensional slice plane, d m It is the chord height difference distance of the mth point on the outer contour of the jth two-dimensional slice plane.
[0039] Compared with the prior art, the present invention adopts the above technical solution and has the following technical effects:
[0040] 1. The present invention proposes an innovative data standardization processing method, which uses the KNN algorithm to clean the large amount of data generated during the quality inspection of prefabricated components of assembled buildings, compresses the data to 87.41% of the original number of data points while retaining the geometric characteristics of the data; changes the coordinates of the collected point cloud model so that its bottom surface is located in the XY plane of the three-dimensional rectangular coordinate system, which is convenient for the subsequent application of the slicing method. The sliced data is reduced to two-dimensional plane data, which greatly reduces the difficulty of subsequent work.
[0041] 2. The present invention proposes an automatic calculation algorithm for the production quality information of prefabricated components. The geometric data features of prefabricated components of assembled buildings are extracted through the slicing method and the chord height difference algorithm, and the automation of eigenvalue fitting is realized through Hough transformation. The traditional quality data collection and verification method of measurement is changed, and automatic and precise processing is realized.
[0042] 3. The present invention parses and stores the reverse modeled prefabricated component model in IFC format, making the reverse model more data compatible and providing data support for subsequent processing. The reverse model and the forward design model are aligned, and the surface quality of the prefabricated component is displayed by means of visualization to perform production quality management. BRIEF DESCRIPTION OF THE DRAWINGS
[0043] Figure 1 is a point cloud data conversion flow chart of the present invention;
[0044] Figure 2 It is a flow chart of the reverse modeling and quality control method of prefabricated components based on the slicing method of the present invention;
[0045] Figure 3 It is a characteristic data judgment diagram of the chord height difference algorithm of the present invention;
[0046] Figure 4 It is a visualization diagram for quality inspection of prefabricated components of the present invention. DETAILED DESCRIPTION
[0047] The embodiments of the present invention are described in detail below, and examples of the embodiments are shown in the accompanying drawings. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and cannot be interpreted as limiting the present invention.
[0048] The present invention proposes a method for collecting surface quality information of prefabricated components using laser scanning technology and performing reverse modeling through a slicing method to inspect the surface quality of prefabricated components, thereby replacing traditional manual visual observation and measurement methods to achieve stricter and more automated production quality control.
[0049] Based on laser scanning technology and point cloud data characteristics, the present invention proposes a prefabricated component quality detection method with point cloud data processing and calculation as the core. The present invention uses the KNN algorithm to first perform data cleaning, and then uses the slicing algorithm and the chord height difference algorithm to extract the characteristic values and the geometric parameters of the prefabricated components of the assembled building, and realizes the automatic fitting of the characteristic parameters through the Hough transform. The geometric quality of the prefabricated components is refined and automatically detected by comparing the inverse model and the forward model, thereby improving the efficiency of this process.
[0050] Figure 1 It is a point cloud data conversion flow chart of the present invention. Figure 2 It is a flow chart of the reverse modeling and quality control method of prefabricated building components based on the slicing method of the present invention, and the specific steps are as follows:
[0051] S1, using a laser scanner to obtain three-dimensional point cloud data on the surface of prefabricated components of assembled buildings; wherein the three-dimensional point cloud data of the prefabricated components in .asc format is obtained by using the laser scanner.
[0052] S2, preprocessing the obtained three-dimensional point cloud data, using K-Nearest Neighbor (KNN) classification algorithm to remove noise and reduce the amount of point cloud data;
[0053] The conversion format is .obj data format, and the K nearest neighbor classification algorithm is applied for data cleaning: The K nearest neighbor classification algorithm classifies the data points, sets the corresponding K value, selects the data from the given point cloud data set, and determines the point type with the most points of the same type among the K points with close distances, and assigns this type to the measured point. The proximity algorithm can ensure that the original geometric feature information can be retained during the point cloud data cleaning process, and effectively distinguish between plane points and boundary points. During the data cleaning process, the K value generates a distance threshold, that is, the judgment is made by type, and the threshold takes the minimum distance of points that are not of the same type. Points exceeding the threshold are judged as noise points and are removed from cleaning.
[0054] S3, sets a reference line for the preprocessed point cloud data, performs slice segmentation along the line, extracts the height information of each slice, and projects the slice onto a two-dimensional plane; the details are as follows:
[0055] S3.1 Reference line calculation: Assuming that the prefabricated component has a slight curvature, establish a three-dimensional rectangular coordinate system to position the prefabricated component so that one side of it is located in the YOZ plane (generally the outer surface parallel to the cross section to be cut). Assume that a curve (generally a second-order curve) is used as the reference line for the study, and the direction of the reference line is the X-axis direction. About 10% of the points in the plane where the reference line is located are randomly selected to fit the reference line. Project the randomly sampled point cluster onto the XOY plane, and then fit the unique quadratic polynomial to the projected n points (x i ,y i ), assuming the x-axis is the main direction:
[0056]
[0057] In the formula, p(x i ) is the interpolation value of the k-th polynomial, which can be expressed as a0, a1, ..., a k The linear equations are:
[0058]
[0059] Multiply both sides of Y=XA by X T , that is, X T Y=X T XA, for the matrix equation X r Y=X T XA is inversely transformed to obtain:
[0060] A=(XT X) -1 X T Y, n>k (3)
[0061] According to the obtained A, the interpolation points of two adjacent projection points among the n projection points are constructed, that is, n-1 interpolation points are obtained, the reference line is fitted according to the interpolation points, and the tangent line of each interpolation point on the reference line is calculated. The slope of the tangent line of each interpolation point is:
[0062] tant j =f(x)′ (4)
[0063] In the formula, tant j is the tangent slope corresponding to the jth interpolation point, and f(x) is the reference line fitted by the interpolation point;
[0064] The prefabricated component is cut along the normal direction of each interpolation point (Formula (5)), and the prefabricated component is divided into n slices;
[0065]
[0066] S3.2 Height information extraction: Assume that each slice is a straight line along its tangent direction and that the cross section of the slice remains unchanged. For each slice, use formula (6) to rotate the slice around the Z axis:
[0067]
[0068] The rotation angle is the angle between the normal direction of each slice j and the Y axis of the global rectangular coordinate. Since the thickness of the slice is small, the middle value of the slice can be used to calculate the normal direction of each slice.
[0069] S4, using a chord height difference algorithm to extract edge feature data of the slice, fitting the extracted data to obtain relevant slice two-dimensional contour parameters;
[0070] For points on the outer contour of the two-dimensional slice plane, the chord height difference algorithm (such as Figure 3 As shown in the figure, determine whether each point is a feature point, as follows:
[0071] The chord height distance can be obtained by formula (7):
[0072]
[0073]
[0074] For the mth point P on the outer contour of the jth two-dimensional slice plane m , select two points P adjacent to it m-1 and P m+1 Construct line L: Am x+B m y+C m =0, calculate point P m The vertical distance d to the line L m , when d m When it is greater than or equal to the set threshold σ, point P m As feature points, they are extracted; (x m ,y m ) is point P m The coordinates of A m , B m , C m is the coefficient of the straight line L, M j is the number of points on the outer contour of the j-th two-dimensional slice plane, d m is the chord height difference distance of the mth point on the outer contour of the jth two-dimensional slice plane. The characteristic points of the slice are finally determined by eigenvalue extraction, and the slice contour map is extracted.
[0075] S5, stretching the two-dimensional contour according to the slice height information to form a three-dimensional slice, and combining different slices according to the original slice positions to form a reverse reconstruction model;
[0076] S6, the obtained reverse modeling model is combined with the forward design BIM model, analyzed and synthesized according to the quality control requirements, and data visualization methods (such as Figure 4 As shown), it can intuitively evaluate whether the quality of prefabricated components meets the requirements.
[0077] The above embodiments are only for illustrating the technical idea of the present invention, and cannot be used to limit the protection scope of the present invention. Any changes made on the basis of the technical solution in accordance with the technical idea proposed by the present invention shall fall within the protection scope of the present invention.
Claims
1. A reverse modeling and quality control method for prefabricated components based on the slicing method, characterized in that: The steps include: Step 1, obtaining three-dimensional point cloud data of the surface of prefabricated components of assembled buildings; Step 2, preprocessing the three-dimensional point cloud data obtained in step 1, distinguishing plane points from boundary points, and removing discrete points and noise points in the three-dimensional point cloud data; Step 3, setting a reference line according to the preprocessed point cloud data, slicing and segmenting the prefabricated building components along the reference line, extracting the height information of each slice, and projecting the slice onto a two-dimensional plane to obtain a two-dimensional slice plane; Step 4, using the chord height difference algorithm, extracting feature points from points on the outer contour of the two-dimensional slice plane; Step 5, connecting the feature points extracted in step 4 to obtain the two-dimensional contour of each slice, stretching the two-dimensional contour of each slice into a three-dimensional slice according to the height information of the slice, and combining all the three-dimensional slices according to the original slice positions to form a reverse reconstruction model; Step 6: align the reverse reconstruction model obtained in step 5 with the BIM model of the prefabricated component forward design, analyze and synthesize them according to the quality control requirements, and use data visualization to evaluate whether the quality of the prefabricated components meets the requirements.
2. The method for reverse modeling and quality control of prefabricated components based on the slicing method according to claim 1 is characterized in that: In the step 1, the laser scanner is used to obtain the 3D point cloud data of the surface of the prefabricated component in .asc format.
3. The reverse modeling and quality control method for prefabricated components based on the slicing method according to claim 2 is characterized in that: In the step 2, the three-dimensional point cloud data of the surface of the prefabricated component in the .asc format is converted into the .obj format, and the three-dimensional point cloud data in the .obj format is preprocessed using a K nearest neighbor classification algorithm.
4. The method for reverse modeling and quality control of prefabricated components based on the slicing method according to claim 1 is characterized in that: The specific process of step 3 is as follows: Step 3.1, preset a reference line on the upper surface of the prefabricated component, establish a three-dimensional rectangular coordinate system to position the prefabricated component, make one of the side surfaces of the prefabricated component located in the YOZ plane, the side surface located in the YOZ plane is parallel to the cross section to be cut, the preset reference line direction is the X-axis direction, randomly sample 10% of the points on the surface where the reference line is located to fit the reference line, project the randomly sampled points onto the XOY plane, fit the points projected on the XOY plane, and obtain a unique quadratic polynomial by minimizing the square error; The square error formula is: Where E is the square error, n is the number of projection points obtained by projecting onto the XOY plane, and y i is the y coordinate of the projection point projected onto the XOY plane, p(x i ) is the interpolation value of the k-th polynomial; Let p(x i ) is expressed as the following linear equations: In the formula, x i is the x-coordinate of the point projected onto the XOY plane, a0, a1, …, a k is the coefficient; make Multiply both sides of Y=XA by X T , that is, X T Y=X T XA, for the matrix equation X T Y=X T XA is inversely transformed to obtain: A=(X T X) -1 X T Y,n>k According to the obtained A, the interpolation points of two adjacent projection points among the n projection points are constructed, that is, n-1 interpolation points are obtained, the reference line is fitted according to the interpolation points, and the tangent line of each interpolation point on the reference line is calculated. The slope of the tangent line of each interpolation point is: tan t j =f(x)′ Where tan t j is the tangent slope corresponding to the jth interpolation point, and f(x) is the reference line fitted by the interpolation point; The prefabricated component is sliced along the normal direction of each interpolation point, and the prefabricated component is divided into n slices; Step 3.2, set each slice to be a straight line along its tangent direction, and keep the cross section of the slice unchanged, and rotate each slice around the Z axis: Where (x, y, z) is the coordinate of the point on the slice, (x′, y′, z′) is the coordinate of the point on the slice after rotation, is the rotation angle of the jth slice. The rotation angle is the angle between the normal direction of each slice and the Y axis of the three-dimensional rectangular coordinate system. The normal direction of each slice is calculated using the center point of the tangent line of each slice. The height information of each slice is extracted using the x-coordinate of the point on the rotated slice, and the slice is projected onto a two-dimensional plane.
5. The method for reverse modeling and quality control of prefabricated components based on slicing method according to claim 1 is characterized in that: The specific process of step 4 is as follows: For the points on the outer contour of the two-dimensional slice plane, the chord height difference algorithm is used to determine whether each point is a feature point, as follows: For the mth point P on the outer contour of the jth two-dimensional slice plane m , select two points P adjacent to it m-1 and P m+1 Construct line L: A m x+B m y+C m =0, calculate point P m The vertical distance d to the line L m , when d m When it is greater than or equal to the set threshold σ, point P m Feature points are extracted; In the formula, (x m ,y m ) is point P m The coordinates of A m , B m , C m is the coefficient of the straight line L, M j is the number of points on the outer contour of the j-th two-dimensional slice plane, d m It is the chord height difference distance of the mth point on the outer contour of the jth two-dimensional slice plane.
Citation Information
Patent Citations
Automatic extraction method of fine-size information of ancient wooden building component
CN108846888A
Dynamo-based point cloud BIM automatic modeling method
CN111612911A
Method for the setpoint-actual value evaluation of a component against CAD data
DE102017002158A1
