Three-dimensional alignment method suitable for reverse modeling of compressor blade

Abstract

The application provides a three-dimensional alignment method suitable for reverse modeling of compressor blades, relates to the technical field of three-dimensional surveying and reverse modeling, and obtains an overall point cloud model and multiple local point cloud models by three-dimensional scanning of a compressor; a target blade to be subjected to three-dimensional alignment is determined in the local point cloud model, and a single-blade model is formed based on blade data corresponding to the target blade in the overall point cloud model; the single-blade model is aligned to the target blade in the local point cloud model; a first space coordinate system is established based on the single-blade model before alignment, and a second space coordinate system is established based on the single-blade model after alignment; the local point cloud model is transformed from the second space coordinate system to the first space coordinate system to complete three-dimensional alignment of the local point cloud model; and the above steps are repeated for other local point cloud models until three-dimensional alignment of all local point cloud models is completed. The application can reduce deviation in the three-dimensional alignment process and improve subsequent modeling accuracy.

Description

Technical Field

[0001] This application relates to the field of 3D mapping and reverse modeling technology, and in particular to a 3D alignment method suitable for reverse modeling of compressor blades. Background Technology

[0002] In the field of 3D mapping and reverse modeling of mechanical components, direct overall scanning of large-volume components is often difficult due to limitations in the coverage area of ​​optical surveying instruments and available space. There are generally two approaches to address this: one is to divide the large component to be scanned into several regions and perform optical scanning on different regions in batches, repeatedly scanning the overlapping areas between different regions to facilitate subsequent positioning, alignment, and assembly of the components; the other approach is to adjust the position of the surveying instrument to scan the overall outline of the component in one or two separate scans, ignoring fine features and focusing only on the overall shape. This scan data serves as the positioning and alignment standard for the local scanned parts obtained from subsequent fine scanning. Based on the obtained overall shape outline scan data, the local parts obtained from subsequent fine scanning can be aligned and assembled to obtain the overall 3D mapping result of the component.

[0003] For both of the aforementioned 3D mapping methods, regardless of which method is used, after obtaining the mapping results, it is necessary to further align and combine the mapping data according to the alignment and positioning standards between the scanned data. This ensures that the obtained 3D mapping results can eliminate further adjustment and alignment steps in the subsequent reverse modeling process. If the component data from the 3D mapping is not aligned, or if the deviation between components is large during alignment, the final reverse-modeled overall component will have a significant difference in geometry from the scanned component, leading to inaccurate subsequent machining or 3D simulation results. Therefore, component alignment of the 3D mapping results is an indispensable step in the reverse modeling process.

[0004] When performing 3D alignment of survey data, if the overall contour data of the component obtained from the scan is sufficient and the difference in coverage between it and the local data obtained from the subsequent fine scan is small, then aligning the local component to the overall component contour can achieve good results. However, when the coverage of the overall scan data and the local data is significantly different, the 3D alignment effect of the component will be poor. For example, when performing 3D mapping of compressor blades, the overall scan may only obtain local data of the suction or pressure surfaces of a certain stage of stationary or moving blades in the compressor. Because compressor blades are thin and have a large radial twist angle, when aligning the finely scanned blade data to the overall scan data, the blade pressure surface may be in close contact with the overall scanned blade suction surface, resulting in a very poor 3D alignment effect and insufficient accuracy in subsequent reverse modeling results. Summary of the Invention

[0005] This application aims to at least partially address one of the technical problems in the related art.

[0006] Therefore, one objective of this application is to propose a three-dimensional alignment method suitable for reverse modeling of compressor blades. Through S1, the compressor is scanned in three dimensions to obtain the overall point cloud model of the compressor, and multiple local three-dimensional scans are performed on the compressor to obtain multiple local point cloud models of the compressor.

[0007] S2, determine the target blade to be 3D aligned in any local point cloud model, obtain the blade data corresponding to the target blade in the overall point cloud model, and form a single blade model based on the blade data;

[0008] S3, align the single-blade model to the target blade in the local point cloud model;

[0009] S4. Select three first reference points from the single blade model before alignment and establish a first spatial coordinate system based on the three first reference points. Also, select three second reference points from the single blade model after alignment and establish a second spatial coordinate system based on the three second reference points.

[0010] S5 transforms the local point cloud model from the second spatial coordinate system to the first spatial coordinate system through coordinate transformation to complete the three-dimensional alignment of the local point cloud model;

[0011] S6. For each of the other local point cloud models, repeat steps S2 to S5 until the 3D alignment of all local point cloud models is completed to obtain the 3D aligned target overall point cloud model.

[0012] According to one embodiment of this application, determining the target blade to be 3D aligned in any local point cloud model includes: obtaining the number of point cloud data corresponding to each candidate blade included in the local point cloud model; and selecting the candidate blade with the largest number of point cloud data as the target blade.

[0013] According to one embodiment of this application, obtaining blade data corresponding to the target blade in the overall point cloud model and forming a single blade model based on the blade data includes: obtaining the target position corresponding to the target blade in the overall point cloud model; deleting point cloud data other than the target position in the overall point cloud model, obtaining the remaining point cloud data after deletion as blade data, and forming a single blade model based on the blade data.

[0014] According to one embodiment of this application, there is a one-to-one correspondence between the three first reference points and the three second reference points.

[0015] According to one embodiment of this application, aligning a single blade model to a target blade in a local point cloud model includes: performing coarse alignment and best-fit alignment on the single blade model and the local point cloud model.

[0016] According to one embodiment of this application, the first reference point and the second reference point are the boundary points of a single blade model.

[0017] According to one embodiment of this application, the three first reference points are not on a straight line and the three second reference points are not on a straight line.

[0018] According to one embodiment of this application, establishing a first spatial coordinate system based on three first reference points or establishing a second spatial coordinate system based on three second reference points includes: establishing a first plane through the three reference points; selecting two of the three reference points to establish a first straight line, and establishing a second plane through the first straight line and perpendicular to the first plane; taking any one of the three reference points as the origin of the spatial coordinate system, and taking the normals of the first plane and the second plane as the x-axis and y-axis of the spatial coordinate system, respectively, and establishing the spatial coordinate system according to the right-hand rule.

[0019] According to one embodiment of this application, before aligning the single-blade model to the target blade in the local point cloud model, the method further includes: preserving the data of the single-blade model to facilitate the establishment of a first spatial coordinate system.

[0020] According to one embodiment of this application, a three-dimensional alignment method applicable to reverse modeling of compressor blades further includes: preserving the data of the overall point cloud model and analyzing and comparing the overall point cloud model and the target overall point cloud model.

[0021] This application achieves at least the following beneficial effects: Considering the large data gap between the overall point cloud model obtained by the overall 3D scanning and the local point cloud model obtained by the fine scanning of the blade, as well as the problem that the pressure surface and suction surface scanning data are too close due to the thinness of the compressor blade, resulting in a large alignment deviation, this application enhances the 3D alignment effect between the overall point cloud model and the local point cloud model by anti-aligning the single blade model after deletion and adjustment to the local point cloud model of each level of fine scanning that needs to be aligned, reducing the deviation that occurs between the two in the 3D alignment process, and improving the accuracy of subsequent reverse modeling.

[0022] Furthermore, after anti-aligning the 3D mapping data, this application uses a spatial coordinate system and coordinate transformation to transform the local point cloud model to be aligned to the first spatial coordinate system of the overall point cloud model, achieving alignment between the finely scanned overall point cloud model and the local point cloud model. Compared to directly aligning the local point cloud model to the overall point cloud model, this reduces the deviation in 3D alignment of the blades and significantly improves the accuracy of 3D alignment. By anti-aligning the single blade model to the local point cloud model and then using coordinate transformation to re-align the local point cloud model to the original overall point cloud model, the 3D alignment effect of the blades is significantly improved, and the positioning relationship between blades at each level is guaranteed. Attached Figure Description

[0023] The above and / or additional aspects and advantages of this application will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein:

[0024] Figure 1 This is a schematic diagram illustrating an exemplary implementation of a three-dimensional alignment method for reverse modeling compressor blades, as shown in one embodiment of this application.

[0025] Figure 2 This is a schematic diagram of an overall point cloud model of a compressor, as shown in one embodiment of this application.

[0026] Figure 3 This is a schematic diagram of a local point cloud model of a compressor, as shown in one embodiment of this application.

[0027] Figure 4 This is a schematic diagram of a single-blade model shown in one embodiment of this application.

[0028] Figure 5 This is a schematic diagram illustrating an embodiment of the present application of aligning a single blade model to a target blade in a local point cloud model.

[0029] Figure 6(a) is a schematic diagram of selecting three first reference points from a single blade model before alignment and establishing a first spatial coordinate system based on the three first reference points, as shown in this application.

[0030] Figure 6(b) is a schematic diagram of selecting three second reference points from an aligned single-blade model and establishing a second spatial coordinate system based on the three second reference points, as shown in this application. Detailed Implementation

[0031] The embodiments of this application are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain this application, and should not be construed as limiting this application.

[0032] Figure 1 This is a schematic diagram of an exemplary embodiment of a three-dimensional alignment method for reverse modeling compressor blades, as shown in this application. Figure 1 As shown, this 3D alignment method for reverse modeling compressor blades includes the following steps:

[0033] S101, perform a 3D scan of the compressor to obtain the overall point cloud model of the compressor, and perform multiple 3D scans of the compressor to obtain multiple local point cloud models of the compressor.

[0034] A 3D scan of the overall profile of the compressor is performed to obtain a point cloud model of the compressor. Figure 2 This is a schematic diagram of an overall point cloud model of a compressor as shown in this application.

[0035] Multiple detailed 3D scans of the compressor were performed to obtain multiple local point cloud models of the compressor. Figure 3 This is a schematic diagram of a local point cloud model of a compressor as shown in this application.

[0036] S102, determine the target blade to be 3D aligned in any local point cloud model, obtain the blade data corresponding to the target blade in the overall point cloud model, and form a single blade model based on the blade data.

[0037] Identify the target blade to be 3D aligned in any local point cloud model. The target blade is a thin compressor blade with a large twist angle; however, it can also be a component of other shapes.

[0038] For example, when determining the target blade, the number of point cloud data corresponding to each candidate blade included in the local point cloud model can be obtained, and the candidate blade with the largest number of point cloud data can be taken as the target blade.

[0039] After determining the target blade to be 3D aligned in the local point cloud model, the target position corresponding to the target blade is obtained in the overall point cloud model. Point cloud data other than the target position in the overall point cloud model is deleted. For example, the redundant point cloud data of other stages of blades, casings, and hubs other than the target position are deleted. The remaining point cloud data after deletion is obtained as blade data, and a single blade model is formed based on the blade data. Figure 4 This is a schematic diagram of a single blade model shown in this application.

[0040] S103 aligns the single-blade model to the target blade in the local point cloud model.

[0041] The local point cloud model is used as an alignment reference, with its spatial position kept fixed. Several points on the local point cloud model are selected and matched with several points on the single-blade model, with a one-to-one correspondence between the selected spatial points. This multi-point correspondence is used to coarsely align the single-blade model to the local point cloud model. To improve the accuracy of subsequent fitting and alignment, the number of alignment points should be no less than three. After the coarse alignment of the single-blade model and the local point cloud model, the alignment deviation is reduced through best-fit alignment between the two models, ensuring a perfect match and meeting the accuracy requirements of subsequent reverse modeling. Figure 5 This is a schematic diagram illustrating how a single blade model is aligned to a target blade in a local point cloud model, as shown in this application.

[0042] S104. Select three first reference points from the single blade model before alignment and establish a first spatial coordinate system based on the three first reference points. Also, select three second reference points from the single blade model after alignment and establish a second spatial coordinate system based on the three second reference points.

[0043] Three first reference points are selected from the unaligned single-blade model, and three second reference points are selected from the aligned single-blade model. There is a one-to-one correspondence between the three first reference points and the three second reference points. The first and second reference points are the boundary points of the single-blade model; the three first reference points are not collinear, and the three second reference points are not collinear either.

[0044] Before aligning the single-blade model to the target blade in the local point cloud model, the process also includes: preserving the data of the single-blade model in order to establish a first spatial coordinate system.

[0045] Figure 6(a) is a schematic diagram of selecting three first reference points from a single blade model before alignment and establishing a first spatial coordinate system based on the three first reference points, as shown in this application.

[0046] As shown in Figure 6(a), the three first reference points selected on the single blade model before 3D alignment are N1(x1, y1, z1), N2(x2, y2, z2), and N3(x3, y3, z3). The three selected first reference points N1, N2, and N3 are all boundary points on the single blade model.

[0047] The process of establishing the first spatial coordinate system C1 based on the three selected first reference points N1, N2, and N3 is as follows: For the three first reference points, establish a first plane S1 through these three reference points; draw a first straight line L1 through points N1 and N3; rotate the first plane S1 by 90° around the first straight line L1 as an axis to form a second plane S2; draw normals n1 and n2 through point N1 to the first plane S1 and the second plane S2, respectively. Establish the first spatial coordinate system C1 with N1 as the origin and n1 and n2 as the first coordinate axis x and the second coordinate axis y. The first spatial coordinate system C1 conforms to the right-hand rule.

[0048] Figure 6(b) is a schematic diagram illustrating the selection of three second reference points from the aligned single-blade model and the establishment of a second spatial coordinate system based on these three second reference points. As shown in Figure 6(b), the three second reference points selected on the three-dimensional aligned single-blade model are: N1'(x1', y1', z1'), N2'(x2', y2', z2'), and N3'(x3', y3', z3'). The three selected second reference points N1', N2', and N3' are all boundary points on the single-blade model, so as to establish a correspondence with the three first reference points on the single-blade model before alignment: N1(x1, y1, z1), N2(x2, y2, z2), and N3(x3, y3, z3).

[0049] The process of establishing a second spatial coordinate system C1' based on the three selected second reference points N1', N2', and N3' is as follows: For the three second reference points, establish a first plane S1' through these three reference points; draw a first straight line L1' through points N1' and N3'; rotate the first plane S1' by 90° around the first straight line L1' to form a second plane S2'; draw normals n1' and n2' through point N1' to the first plane S1' and the second plane S2', respectively. Establish the second spatial coordinate system C1' with N1' as the origin and n1' and n2' as the first coordinate axis x and the second coordinate axis y, respectively. The second spatial coordinate system C1' conforms to the right-hand rule.

[0050] In this application, when establishing a spatial coordinate system based on three points on a single blade model, the three points selected are boundary points on the scan data of the single blade model. Using boundary points ensures that when re-selecting three points to establish a spatial coordinate system on the local point cloud model after alignment with the local point cloud model, the relative positions of the three points remain consistent with those before alignment, satisfying the correspondence between the points before and after alignment. This prevents deviations during subsequent coordinate transformations of the local point cloud model and improves the accuracy of the blade's 3D alignment.

[0051] S105 transforms the local point cloud model from the second spatial coordinate system to the first spatial coordinate system through coordinate transformation to complete the three-dimensional alignment of the local point cloud model.

[0052] By transforming coordinates, the local point cloud model is transformed from the second spatial coordinate system to the first spatial coordinate system to complete the 3D alignment of the local point cloud model.

[0053] During coordinate transformation, the origin N1' of the second spatial coordinate system coincides with the origin N1 of the first spatial coordinate system, and the xyz coordinate axes of the second spatial coordinate system are aligned with the xyz coordinate axes of the first spatial coordinate system.

[0054] S106. For each of the other local point cloud models, repeat steps S102 to S105 until the 3D alignment of all local point cloud models is completed to obtain the target overall point cloud model after 3D alignment.

[0055] For each of the other local point cloud models, repeat steps S102 to S105 above until the 3D alignment of all local point cloud models is completed to obtain the target overall point cloud model after 3D alignment.

[0056] This application proposes a 3D alignment method for reverse modeling compressor blades. The method involves: S1, performing a 3D scan of the compressor to obtain its overall point cloud model, and performing multiple local 3D scans to obtain multiple local point cloud models; S2, identifying the target blade to be aligned in any local point cloud model, obtaining the blade data corresponding to the target blade in the overall point cloud model, and assembling a single blade model based on the blade data; S3, aligning the single blade model to the target blade in the local point cloud model; and S4, retrieving the single blade model from its original state before alignment. In step S5, three first reference points are selected from the blade model, and a first spatial coordinate system is established based on these three reference points. Similarly, three second reference points are selected from the aligned single-blade model, and a second spatial coordinate system is established based on these three reference points. In step S6, the local point cloud model is transformed from the second spatial coordinate system to the first spatial coordinate system through coordinate transformation to complete the 3D alignment of the local point cloud model. In step S7, steps S2 to S5 are repeated for each other local point cloud model until the 3D alignment of all local point cloud models is completed, thus obtaining the 3D aligned target overall point cloud model. This application considers the large data difference between the overall point cloud model obtained from the overall 3D scan and the local point cloud model obtained from the fine-scan blade, as well as the problem of large alignment deviations caused by the thinness of the compressor blades leading to excessively close scanning data between the pressure and suction surfaces. By anti-aligning the adjusted single-blade model to the required level of fine-scan local point cloud models, the 3D alignment effect between the overall point cloud model and the local point cloud model can be enhanced, reducing the deviations that occur during the 3D alignment process and improving the accuracy of subsequent reverse modeling.

[0057] Furthermore, after anti-aligning the 3D mapping data, this application uses a spatial coordinate system and coordinate transformation to transform the local point cloud model to be aligned to the first spatial coordinate system of the overall point cloud model, achieving alignment between the finely scanned overall point cloud model and the local point cloud model. Compared to directly aligning the local point cloud model to the overall point cloud model, this reduces the deviation in 3D alignment of the blades and significantly improves the accuracy of 3D alignment. By anti-aligning the single blade model to the local point cloud model and then using coordinate transformation to re-align the local point cloud model to the original overall point cloud model, the 3D alignment effect of the blades is significantly improved, and the positioning relationship between blades at each level is guaranteed.

[0058] Furthermore, in this application, the overall point cloud model is preserved, and the overall point cloud model and the target overall point cloud model obtained after 3D alignment are analyzed and compared to obtain the changes in the point cloud before and after 3D alignment, so as to better perform data analysis.

[0059] In the description of this application, it should be understood that the terms "center", "longitudinal", "lateral", "length", "width", "thickness", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", "clockwise", "counterclockwise", "axial", "radial", "circumferential", etc., indicating the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, are only for the convenience of describing this application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of this application.

[0060] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this application, "multiple" means two or more, unless otherwise explicitly specified.

[0061] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., refer to specific features, structures, materials, or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0062] Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of this application.

Claims

1. A three-dimensional alignment method suitable for reverse modeling of compressor blades, characterized in that, include: S1. Perform a 3D scan of the compressor to obtain the overall point cloud model of the compressor, and perform a 3D scan of multiple local parts of the compressor to obtain multiple local point cloud models of the compressor. S2, determine the target blade to be 3D aligned in any of the local point cloud models, obtain the blade data corresponding to the target blade in the overall point cloud model, and form a single blade model based on the blade data; S3, Align the single-blade model to the target blade in the local point cloud model; S4, selecting three first reference points from the pre-aligned single-blade model and establishing a first spatial coordinate system based on the three first reference points; and selecting three second reference points from the aligned single-blade model and establishing a second spatial coordinate system based on the three second reference points, wherein the three first reference points and the three second reference points correspond one-to-one. Establishing a first spatial coordinate system based on the three first reference points or establishing a second spatial coordinate system based on the three second reference points includes: A first plane is established using the three first reference points or the three second reference points. Select two of the three reference points to establish a first straight line, and establish a second plane through the first straight line and perpendicular to the first plane; Take any one of the three reference points as the origin of the spatial coordinate system, and take the normals of the first plane and the second plane as the x-axis and y-axis of the spatial coordinate system, respectively, and establish the spatial coordinate system according to the right-hand rule. S5, by means of coordinate transformation, the local point cloud model is transformed from the second spatial coordinate system to the first spatial coordinate system to complete the three-dimensional alignment of the local point cloud model; S6. For each of the other local point cloud models, repeat steps S2 to S5 until the 3D alignment of all the local point cloud models is completed to obtain the 3D aligned target overall point cloud model.

2. The method according to claim 1, characterized in that, Determining the target blade to be 3D aligned in any of the local point cloud models includes: Obtain the number of point cloud data points corresponding to each candidate leaf included in the local point cloud model; The candidate leaf with the largest number of point cloud data points is selected as the target leaf.

3. The method according to claim 1 or 2, characterized in that, The step of acquiring the blade data corresponding to the target blade in the overall point cloud model and assembling a single blade model based on the blade data includes: Obtain the target position corresponding to the target blade in the overall point cloud model; The point cloud data in the overall point cloud model, except for the target location, is deleted. The remaining point cloud data after deletion is used as the blade data, and the single blade model is formed based on the blade data.

4. The method according to claim 1, characterized in that, Aligning the single-blade model to the target blade in the local point cloud model includes: The single-blade model and the local point cloud model are coarsely aligned and best-fit aligned.

5. The method according to claim 1, characterized in that, The first reference point and the second reference point are the boundary points of the single blade model.

6. The method according to claim 1, characterized in that, The three first reference points are not on a straight line and the three second reference points are not on a straight line.

7. The method according to claim 1, characterized in that, Before aligning the single-blade model to the target blade in the local point cloud model, the method further includes: The data of the single blade model is preserved in order to establish the first spatial coordinate system.

8. The method according to claim 1, characterized in that, The method further includes: The data of the overall point cloud model is retained, and the overall point cloud model and the target overall point cloud model are analyzed and compared.

Images