Complex prototype reconstruction model and registration evaluation method based on actual measurement data driving

Abstract

The invention discloses a complex prototype reconstruction model and registration evaluation method based on actual measurement data driving. The method comprises the following steps: step 1, constructing a measurement-registration unified coordinate system; 2, digital measurement of the actual profile and characteristics of the product is completed; 3, completing the solution of the concerned feature point location of the product; 4, completing a complex prototype reconstruction model based on lightweight characteristics; 5, completing alignment relation calculation based on local constraint matching; step 6, calculating the overall assembly offset degree of different reconstruction pattern machines; 7, completing registration evaluation according to the offset degree of the reconstruction pattern machine; and step 8, assembling sequence planning based on registration evaluation. Quantitative supporting guidance can be provided for assembly work, the reliability and efficiency of the assembly process can be effectively improved, in addition, a physical prototype of the reconstruction model can provide data input for digital integration and automatic assembly, and work such as traditional trial docking and manual mass trial assembly can be effectively solved.

Description

Technical Field

[0001] The present invention belongs to the field of digital manufacturing, and in particular relates to a complex prototype remodeling and registration evaluation method driven by measured data. Background Art

[0002] With the rapid development of aerospace technology, new aircraft products are constantly moving towards high-performance manufacturing and assembly. Digital manufacturing driven by actual data in physical scenarios has become an important development direction. A large and complex rotary product with complex structural forms, its manufacturing accuracy and performance after assembly registration are the key to product reliability. In the assembly process of this product, it is often necessary to complete multi-form docking between different parts based on key feature areas. The traditional assembly operation mode usually relies on dedicated large-scale process equipment to complete auxiliary product alignment and assembly operations. In the process, it is necessary to constantly rely on multi-person collaboration, and use manual vision, assembly and actuator coordination and other means to ensure the matching of product posture and corresponding features. This method has low operating efficiency, low degree of digitization, and risks of affecting product quality. At the same time, it relies on manual experience. In order to reduce risks, it is often necessary to conduct trial docking before formal assembly, and the increase in overall processes leads to further reduction in efficiency. With the continuous development of digital technology, laser trackers, photogrammetry and other methods are gradually used in related operation scenarios. By completing the construction of the measurement field, local feature detection is carried out, and workers are guided to adjust the assembly based on the measurement results. This method can obtain assembly information to a certain extent, but for the measurement of large and complex components, it is difficult to complete the rapid and accurate measurement of the entire area at one time. It is necessary to perform multiple station transfers or splicing, and set up relevant interface assistance. The measurement cycle is long and errors are prone to superposition. In addition, the product measurement state is difficult to cover the relationship between different assembly postures. The product is subject to gravity, tooling constraints and flipping operations, which are different from the final assembly state. Comprehensive data measurement of this type of product will generate a large amount of redundant data, and the processing time is long. It is difficult to effectively use the data for assembly assistance, which limits the actual application on site. On the other hand, the structure of this type of product is mostly large thin-walled weak rigid structure, which will inevitably cause irregular morphological changes in the assembly scene, and key docking features are prone to deviation. There are conditions such as half-moons and uneven docking that affect assembly. It is difficult to directly integrate the obtained measured data with the assembly operation equipment, further increasing the workload of manual on-site repair. Summary of the invention

[0003] In order to overcome the above-mentioned shortcomings of the prior art, the present invention proposes a complex prototype remodeling and registration evaluation method driven by measured data.

[0004] The technical solution adopted by the present invention to solve the technical problem is: a complex prototype remodeling and registration evaluation method driven by measured data, comprising the following steps:

[0005] Step 1: Construct a unified coordinate system for measurement and registration;

[0006] Step 2: Complete the digital measurement of the actual shape and features of the product;

[0007] Step 3: Complete the calculation of product focus feature points;

[0008] Step 4: Complete the remodeling of complex prototypes based on lightweight features;

[0009] Step 5: Complete the alignment relationship calculation based on local constraint matching;

[0010] Step 6, calculating the overall assembly deviation of different reconstructed prototypes;

[0011] Step 7: Complete the registration evaluation according to the offset of the reconstructed prototype;

[0012] Step 8: Assembly sequence planning based on registration evaluation.

[0013] Compared with the prior art, the present invention has the following positive effects:

[0014] The main technical problem solved by the present invention is to overcome the shortcomings of the existing assembly process. Taking a large and complex rotary product as the object, the present invention proposes a complex prototype remodeling and registration evaluation method based on measured data driven, aiming at a series of problems such as irregular geometric deformation, assembly feature deviation, difficulty in effective quantification and matching, and large workload of trial docking and manual repair work in the radial assembly and docking process. The method first completes the feature analysis and mapping relationship expression of the theoretical model according to the product structure design, and establishes the product measurement and assembly coordinate system; the digital measurement method is used to complete the rapid acquisition of the actual point cloud data of the product's focus features, completes the data processing under the assembly accuracy constraint, and carries out auxiliary assembly feature fitting and rapid reconstruction secondary generation at the same time, realizing the measured digital prototype remodeling based on lightweight features; a virtual registration characterization method based on assembly section features is established to meet the assembly evaluation of different reconstructed prototypes in the assembly coordinate system, realize the registration analysis of gap deviation and distance distribution in different quadrants, and guide the optimization of product selection and assembly methods. The method of the present invention realizes the measurement and reconstruction of the prototype model before product assembly, and at the same time uses the registration evaluation to provide quantitative support guidance for the assembly operation, which can effectively improve the reliability and efficiency of the assembly process. In addition, the remodeled physical prototype can provide data input for digital integration and automated assembly, and can effectively solve the traditional trial docking, manual large-scale trial assembly and other tasks. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] The present invention will now be described by way of example with reference to the accompanying drawings, in which:

[0016] Figure 1 Shown is a schematic diagram of the overall process of the present invention;

[0017] Figure 2 The large cylinder product A in the present invention is shown, which is used for radially docking with the large cylinder B;

[0018] Figure 3 The large cylinder product B in the present invention is shown, which is used for radially docking with the large cylinder A;

[0019] Figure 4 The figure shows the setting of the product measurement-registration coordinate system, taking a large cylindrical product A as an example;

[0020] Figure 5 The figure shows the distribution diagram and parameter representation of the prototype machine and theoretical model reconstructed with large cylinder A under the cross section of the docking hole;

[0021] Figure 6 The figure shows the distribution diagram and parameter representation of the first quadrant of the cross section of the docking hole of the large cylinder A reconstructed prototype and theoretical model;

[0022] Figure 7 The diagram shows the docking holes between the reconstructed prototype machine for cylinder part A and the reconstructed prototype machine for cylinder segment part B. DETAILED DESCRIPTION

[0023] The method of the present invention is specifically:

[0024] First, according to the product theoretical model, the analytical definition of product assembly features and auxiliary registration features is completed; on this basis, the coordinate system of product measurement-registration relationship is established; large cylindrical parts use horizontal support to complete the digital measurement of the focus features of the product assembly area and the acquisition of the focus feature point cloud, so as to achieve the acquisition of the assembly morphology features of the target product, and complete the data under the relevant accuracy constraints and key design values ​​such as product arc length according to the measured data, so as to achieve rapid processing and fitting; matching the theoretical model situation, the lightweight feature modeling method is adopted to complete the prototype remodeling of the measured product;

[0025] Secondly, based on the reconstructed prototype model, the alignment relationship calculation based on the associated constraint matching is carried out to realize the expression of the measurement-registration coordinate system in the reconstructed prototype. Based on this coordinate system, the segmentation based on the assembly section is carried out to complete the section-based registration evaluation. Finally, based on the registration evaluation data, optimized registration process guidance is provided to provide data input support for efficient digital integration and assembly operations, and to solve the bottlenecks of traditional trial docking and trial assembly.

[0026] The present invention is not only applicable to the large irregular cylindrical parts mentioned in the text, but also applicable to prototype remodeling and registration evaluation of other large and complex products.

[0027] The method of the present invention will be described in detail below with reference to the accompanying drawings.

[0028] like Figure 1 As shown in the figure, a complex prototype remodeling and registration evaluation method based on measured data driving includes the following contents:

[0029] The first aspect: provides a complex prototype remodeling method based on measured data

[0030] like Figure 2 The figure shows a horizontal placement diagram of a large cylindrical product A;

[0031] like Figure 3 The figure shows a schematic diagram of the horizontal placement of a large cylinder B;

[0032] Step 1: Construct a unified coordinate system for measurement and registration

[0033] For the registration of two large and complex rotating products, take cylinder A as an example, take the bottom surface of its butt end as the reference, take the center of the end circle as the center, and establish the spatial rectangular coordinate system XYZ in the way of horizontal support placement; Figure 4 The figure shows the setting of the product measurement-registration coordinate system, taking a large cylindrical product A as an example;

[0034] Step 2: Complete the theoretical model parameter analysis based on the registration features

[0035] Based on the spatial rectangular coordinate system, the cross-section is cut along the center of the circle of the key radial docking feature, and the intersection line with the product is defined as the registration section; Figure 5 Shown is a cross-sectional view of the registration of the butt joint of cylinder A.

[0036] Step 3: Complete the digital measurement of the actual shape and features of the product

[0037] The product supports horizontally, and non-contact scanning and measurement of product end faces, shapes, holes and other features of interest are carried out in the measurement-registration coordinate system. For continuous smooth curved surfaces, a fast scanning method is used to complete point cloud acquisition, and for features such as docking holes, a precise scanning method is used to complete detailed data acquisition of the edges of conical surfaces and circular holes; based on the point cloud data, data that has little impact on key features and deviates far away is removed to complete point cloud denoising and data streamlining processing;

[0038] Step 4: Complete the calculation of product feature points

[0039] In order to further improve the scanning accuracy and reduce the measurement and reconstruction errors, the product cone hole and internal circular hole are used for data processing respectively. First, according to the processed point cloud data, the feature fitting of the overall cone surface scanning data is performed for the hole position data, and the circular feature fitting of the outer and inner circles of the cone feature part is obtained. The two centers of the circles are extracted, and the concentricity deviation between the two is defined as e (the process deviation value of the radial hole position of the product, which has an offset of 0 to Xmm by installing a movable nut inside); when the on-site light source conditions affect the point cloud missing caused by the reflection of the metal inner hole, etc., firstly, the point cloud replacement processing of the local area is performed with reference to the point cloud around the hole position completed previously, and the hole position is replaced. The missing point cloud data is supplemented according to the nearby continuous point cloud. At the same time, the positioning holes in the four quadrants are used as target points, and the quadrant area method is used to reduce the error. The arc length of the 1 / 4 quadrant area of ​​the theoretical model is fitted along the normal direction of the point cloud, and the theoretical arc length of each hole position under the supplementary area is mapped to achieve accurate compensation of the docking hole features in the case of missing holes. The coordinate point solution of the assembly features is completed according to the actual assembly requirements, and the coordinate point extraction is completed at intervals of XXmm; the circle and center coordinates of the inner ring of the cone are extracted respectively; for other features such as continuous surfaces and planes, according to the corresponding relationship between the features and the theoretical model analysis, the continuous surface features and plane features are fitted respectively;

[0040] Step 5: Complete the complex prototype remodeling based on lightweight features

[0041] According to the coordinate points and feature relationships of each feature of the product, surface feature fitting and creation are carried out. By adopting the overall remodeling method of the prototype without solid features, rapid prototype remodeling is achieved to provide quantitative support for subsequent product registration analysis. Specifically, for the docking radial hole feature, the intermediate normal feature F of the hole is generated with the center of the circle through the previously extracted cone inner circle and the center coordinates Y (x, y, z). With the aid of the coordinate points obtained above, the extension of the non-thickness surface is completed in accordance with the relationship of continuity and no mutation of the surface, and the envelope fitting of the entire surface is achieved. Finally, the prototype remodeling of various lightweight features is completed.

[0042] Second aspect: A registration evaluation method based on the assembly section method is provided

[0043] Step 1: Complete the alignment relationship calculation based on local constraint matching

[0044] On the basis of remodeling the prototypes of cylinder A and cylinder B, a registration evaluation method based on the assembly section method was developed to calculate the pre-assembly and deformation deviation of the product in advance, to evaluate the assembly options of different cylinder sections A and B products, and to provide quantitative support for the sequential optimization strategy in the assembly process.

[0045] In the registration of cylinder A and cylinder B, the global feature matching method has the highest accuracy, but the efficiency is low, and it is easy to cause the deviation of the assembly features in the local area to increase. In response to the above problems, the present invention is based on the reconstruction of the prototype machine, takes the product assembly section as the basis for judgment, and uses the deviation between the docking hole and the assembly section as the main indicator for evaluation. First, the product assembly registration section is determined. On the docking end face, the theoretical model is divided along the center of the radial docking key feature, and the intersection line with the product is the theoretical registration section Mn. At the same time, the center coordinates Dn of the docking radial hole and the normal Fn of the center of the hole are extracted; in the same measurement-registration coordinate system, the measured product prototype barrel section A is imported, and the center coordinates and normal features of the docking radial hole of barrel section A are used to align with the theoretical model registration section model; as shown in FIG. Figure 7 Shown is the characterization method of various parameters of a product.

[0046] Step 2: Complete the offset definition and analysis based on the registration of the reconstructed prototype and the theoretical model;

[0047] Based on the alignment of the docking features between the barrel section A and the theoretical model, the offset analysis of the reconstructed prototype is carried out to provide guidance for product assembly, specifically:

[0048] In the section Mn, curve La is the curve formed by the coordinates of the center of the docking hole of the cylinder A; curve Lm is the curve formed by the coordinates of the center of the docking hole of the theoretical cylinder; △h is the distance deviation between the cylinder A and the center of the theoretical hole position, and due to product deformation, the deformation at different positions is different; when La deviates outward, it is at the top, and when it deviates inward, it is at the bottom; La(xan, zan), Lm(xmn, zmn), etc. are points on the corresponding curves; S(am-n) (n=1, 2, 3..) is the area of ​​the intersection of La and Lm, which represents the offset of the corresponding area, such as S(am-1) and S(am-2) are respectively the upper offset and the lower offset; Fan is the normal line of a docking hole on the cylinder A, and Fmn is the normal line of a docking hole on the theoretical model. The angle between the two is △θ, which represents the deviation of the tightening alignment of the two holes. For local deviations, movable docking nuts are used to meet the deviation correction of Emm. When this value is exceeded, the product cannot be docked and assembled.

[0049] f La (x) is the surface curve equation of the reconstructed prototype, f Lm (x) is the curve equation of the theoretical model surface. The product is processed by CNC and the surface is continuous. La (x)-f Lm (x)>0, the actual product is offset toward the outside, and when f La (x)-f Lm (x)<0, the blank is offset inward; when f La (x) = f Lm(x) is the point where the points coincide (x gn ,y gn ), divided by each common point area, the values ​​of the upper offset and the lower offset can be solved;

[0050]

[0051] Upper skewness f La (x)-f Lm (x)≥0; lower skewness f La (x)-f Lm (x)≤0; overall assembly offset When P tends to 0, the measured product has the smallest deformation from the theoretical model; when P>0, the cylinder deforms outward; when P<0, the cylinder deforms inward. The larger the value, the greater the deformation.

[0052] Step 3: Complete the offset analysis of different reconstructed prototypes;

[0053] The above analysis is based on the analysis of the theoretical docking assembly with cylinder A as an example, which is used to define the deformation of the overall assembly of the product. On this basis, taking cylinder A and cylinder B as an example, the offset analysis of the product docking assembly is completed to guide the planning of the selection and assembly path.

[0054] In the cross section Mn, curve La is the curve formed by the coordinates of the center of the butt hole of cylinder A; curve Lb is the curve formed by the coordinates of the center of the butt hole of cylinder B; △h' is the distance deviation between the center of cylinder A and cylinder B, which is affected by product deformation and has different deformations at different positions; when La deviates outward, it is at the top, and when it deviates inward, it is at the bottom; La (xan, zan), Lb (xbn, zbn), etc. are points on the corresponding curves; S (ab-n) (n = 1, 2,

[0055] 3..) is the intersection area of ​​La and Lb, representing the offset set, such as S(ab-1) and S(ab-2) are the upper offset and lower offset respectively;

[0056] f La (x) is the surface curve equation of the cylinder A reconstructed prototype, f Lb (x) is the curve equation of the reconstructed prototype of cylinder B. Both products are processed by CNC and have continuous shapes. La (x)-f Lb (x)>0, the cylinder A deflects outwards, and when f La (x)-f Lb (x)<0, the cylinder A deviates inward; when f La (x) = f Lb(x) is the point where the points coincide (x gn ,y gn ), divided by each common point area, the values ​​of the upper offset and the lower offset can be solved; Upper skewness f La (x)-f Lb (x)≥0; lower skewness f La (x)-f Lb (x)≤0; overall assembly offset When P tends to 0, the registration deviation is minimal; when P>0, the cylinder is deformed outward; when P<0, the cylinder is deformed inward;

[0057] Step 4: Complete the registration evaluation based on the deviation of the reconstructed prototype

[0058] (1) Take cylinder A and cylinder B as objects. During the docking process, the normal lines must be concentric. Both are through holes. After the final alignment, the two are fitted together without deviation. The inner wall of cylinder B adopts a movable docking nut structure, which can achieve a certain amount of activity compensation. Figure 7 As shown in the figure, Lan is the coordinate of a circle center of cylinder A, Lbn is the coordinate of a circle center of cylinder B, Fmn is the normal of the circle center of Lbn, and Lb'n is the initial coordinate of the center of the movable butt nut inside cylinder B. The center can move along the normal of Fmn, and the movable distance is △kn. The movable distance on both sides is 1 / 2△kn. When Lan, Lbn, and Lb'n meet the concentricity requirement, the initial assembly requirement can be met. Because the normals of the two begin to offset, they cannot be directly used. Fan and Fmn are angularly docked. First, according to the measured data processing results, the coordinates of Lan (Xan, Zan) and the coordinates of the point Lbn (Xbn, Zbn) are obtained. According to the relationship between the installation distance of the movable docking nut and the normal line of Fmn, the zero position coordinates of Lb'n (Xb'n, Zb'n) are calculated. Assuming that when Lb'n moves to the point Lb"n (Xb"n, Zb"n) where Lan and Lbn are collinear, the movable nut offset distance is:

[0059]

[0060] The slope k of the two points is:

[0061]

[0062] At the same time, Lan, Lbn, and Lb'n are collinear:

[0063] Its slope k' is:

[0064]

[0065] According to the normal relationship:

[0066] k*k'=-1

[0067]

[0068] According to the installation relationship, the distance △L(b'b) between Lbn and Lb'n is a known constant:

[0069]

[0070] According to the above relationship, the calculation of △k(bb') is finally completed. The movable nut is offset along the two ends of the normal line. Therefore, when △k(bb')≤1 / 2△kn, it indicates that the movable butt nut can cover the assembly requirements after the product is deformed.

[0071] Step 5: Assembly sequence planning based on registration evaluation

[0072] Based on the aforementioned assemblability analysis, the assemblability analysis of the two products can be completed, providing a basis for product selection; finally, in the actual assembly process, the matching relationship selection of the assembly is carried out according to the matching evaluation results, and finally sequence guidance is carried out to improve assembly reliability and efficiency.

[0073] Specifically: (1) According to the initial connection and positioning conditions, complete the three-dimensional scanning of the product's outer surface and hole position to obtain the situation of the pin hole after installation in the real assembly environment; (2) According to the above method, complete the product measurement of the batch of cylinder parts A and cylinder parts B to obtain the offset △k(bb') of the movable nut along the normal line and the overall assembly offset Prioritize products with smaller ∑Pab, and then complete the pairing of cylinder A and cylinder B according to the envelope range of △k(bb'). (3) Use the pin holes in the four quadrants to complete the preliminary connection and positioning of cylinder A and cylinder B; (4) According to the distribution results of P, follow the strategy of tightening from large to small and symmetrically to complete the path planning design of docking assembly; (5) Use the characteristics of graded assembly to solve the problem of product deformation, and complete the assembly of half the torque requirement value for the first time. On this basis, scan the outer surface again, and according to the scanning results, complete the calculation of the center coordinates, evaluate the change of P value, and again according to the distribution results of P, follow the strategy of tightening from large to small and symmetrically to complete the final assembly of the product.

[0074] The above-mentioned related scanning strategies, data and data processing calculation methods are all realized through the system integration and secondary development modules to realize automatic data recording and calculation analysis, and guide process personnel to implement operations.

[0075] The method provided by the present invention completes the complex prototype remodeling driven by measured data and the evaluation based on the registration section under the assembly docking of large-scale complex rotating products. It can effectively obtain the measured data of complex assembly products and generate prototype models. At the same time, it completes the registration evaluation based on the assembly section and process properties of the physical prototype, and can realize the functions of perception before assembly, execution during assembly, and evaluation after assembly. Relevant data can be collected and integrated through the work station system to provide accurate input for the automatic assembly actuator, and at the same time, it can complete the generation of physical prototypes of manufactured products. The technical system of the present invention can not only be used in the assembly docking of large-scale complex rotating bodies, but also is suitable for lightweight modeling and efficient and reliable assembly of prototypes of other types of complex products.

[0076] Implementation case description:

[0077] In the embodiment, a different large aluminum alloy barrel section A and barrel section B with a diameter of 2.5 meters and an overall wall thickness of the skin are 2 mm. There are different degrees of roundness deformation in the horizontal docking state of the two, and the maximum deformation reaches 5 mm, which is required to meet the overall efficient assembly of barrel section A and barrel section B.

[0078] Step 1: Construct a unified coordinate system for measurement and registration

[0079] For the cylinder A and cylinder B in the embodiment, take cylinder A as an example, take the bottom surface of its butt end as the reference, take the center of the end surface circle as the center, and establish a spatial rectangular coordinate system XYZ in a horizontal support placement manner; as shown in the attached figure Figure 4 shown.

[0080] Step 2: Complete the theoretical model parameter analysis based on the registration features

[0081] Based on the spatial rectangular coordinate system, the cross-section is cut along the center of the circle of the radial docking key feature to obtain the registration section of cylinder A and cylinder B; Figure 6 Shown is a cross-sectional view of the alignment of cylinder A and cylinder B.

[0082] Step 3: Complete the digital measurement of the actual shape and features of the product in the implementation example

[0083] The product supports horizontally, and non-contact scanning and measurement of product end faces, shapes, holes and other features of interest are carried out in the measurement-registration coordinate system. For continuous smooth curved surfaces, a fast scanning method is used to complete point cloud acquisition, and for features such as docking holes, a precise scanning method is used to complete detailed data acquisition of the edges of conical surfaces and circular holes; based on the point cloud data, data that has little impact on key features and deviates far away is removed to complete point cloud denoising and data streamlining processing;

[0084] Step 4: Complete the calculation of feature points of the product in the implementation example

[0085] The feature fitting of the overall conical surface scanning data is performed to obtain the circular feature fitting of the outer and inner circles of the conical feature part, and the two centers are extracted. The concentricity deviation between the two is defined as e (the process deviation value of the radial hole position of the product, which has an offset of 0 to 2 mm by installing a movable docking nut inside); at the same time, the point cloud around the hole completed in the third step is used as a reference to perform local area point cloud replacement processing, and the missing point cloud data of the hole is supplemented according to the nearby continuous point cloud. At the same time, the positioning holes in the four quadrants are used as target points, and the quadrant area method is used to reduce Error, according to the arc length of the 1 / 4 quadrant area of ​​the theoretical model, fit along the normal direction of the point cloud, and map along the theoretical arc length of each hole position under the supplementary area to achieve accurate compensation of the docking hole features in the case of missing holes. In the embodiment, there are 50 holes in the docking area, and the coordinate point extraction is completed according to the arc length of 2.5 meters in diameter and 157mm intervals; the circle and center coordinates of the inner circle of the cone are extracted respectively; for other features such as continuous surfaces and planes, according to the corresponding relationship between the features and the theoretical model analysis, the continuous surface features and plane features are fitted respectively;

[0086] Step 5: Complete the complex prototype remodeling based on lightweight features

[0087] According to the coordinate points and feature relationships of various product features, surface feature fitting and creation are carried out. By adopting the overall remodeling method of the prototype without solid features, rapid prototype remodeling is achieved to provide quantitative support for the product registration analysis of the embodiment; specifically, for the docking radial hole feature, the intermediate normal feature F of the hole is generated with the center of the circle through the previously extracted cone inner circle and the center coordinates Y (x, y, z); with the aid of the various coordinate points obtained above, the extension of the thickness-free surface is completed in accordance with the relationship of continuity and no mutation of the surface, and the envelope fitting of the entire surface is achieved; finally, the prototype remodeling of various lightweight features of the embodiment is completed.

[0088] Step 6: Complete the alignment relationship calculation based on local constraint matching

[0089] On the basis of remodeling the prototypes of cylinder A and cylinder B, a registration evaluation method based on the assembly section method was carried out to calculate the pre-assembly and deformation deviation of the product in advance.

[0090] First, determine the product assembly registration section. On the butt end face, the theoretical model is segmented along the center of the radial butt key feature, and the intersection with the product is the theoretical registration section Mn. At the same time, the center coordinates Dn of the butt radial hole and the normal line Fn of the hole center are extracted; in the same measurement-registration coordinate system, the measured product prototype barrel section A is imported, and the alignment with the theoretical model registration section model is performed according to the center coordinates and normal line features of the butt radial hole of barrel section A; Figure 7 Shown is a method for characterizing various parameters of an embodiment.

[0091] Step 7: Complete the offset analysis of different reconstructed prototypes;

[0092] Taking cylinder A and cylinder B as an example, the offset analysis of product docking assembly is completed to guide the planning of matching and assembly paths.

[0093] In the section Mn, curve La is the curve formed by the coordinates of the center of the docking hole of cylinder A; curve Lb is the curve formed by the coordinates of the center of the docking hole of cylinder B; △h' is the distance deviation between the centers of cylinder A and cylinder B, and due to product deformation, the deformation at different positions is different; when La deviates outward, it is at the top, and when it deviates inward, it is at the bottom; La (xan, zan), Lb (xbn, zbn), etc. are points on the corresponding curves; S (ab-n) (n = 1, 2, 3..) is the area of ​​the intersection of La and Lb, representing the offset set, such as S (ab-1) and S (ab-2) are schematic diagrams of the upper offset and the lower offset respectively;

[0094] f La (x) is the surface curve equation of the cylinder A reconstructed prototype, f Lb (x) is the curve equation of the reconstructed prototype of cylinder B. Both products are processed by CNC and have continuous shapes. La (x)-f Lb (x)>0, the cylinder A deflects outwards, and when f La (x)-f Lm (x)<0, the cylinder A deviates inward; when f La (x) = f Lb (x) is the point where the points coincide (x gn ,y gn ), divided by each common point area, the values ​​of the upper offset and the lower offset can be solved; Upper skewness f La (x)-f Lb (x)≥0; lower skewness f La (x)-f Lb (x)≤0; overall assembly offset When P tends to 0, the registration deviation is minimal; when P>0, the cylinder is deformed outward; when P<0, the cylinder is deformed inward;

[0095] Step 8: Complete the registration evaluation based on the deviation of the reconstructed prototype

[0096] (1) Take cylinder A and cylinder B as objects. During the docking process, the normal lines must be concentric. Both are through holes. After the final alignment, the two are fitted together without deviation. The inner wall of cylinder B adopts a movable docking nut structure, which can achieve a certain amount of activity compensation. Figure 7As shown in the figure, Lan is the coordinate of a circle center of cylinder A, Lbn is the coordinate of a circle center of cylinder B, Fmn is the normal of the circle center of Lbn, and Lb'n is the initial coordinate of the center of the movable butt nut inside cylinder B. The center can move along the normal of Fmn, and the movable distance is △kn. The movable distance on both sides is 1 / 2△kn. When Lan, Lbn, and Lb'n meet the concentricity requirement, the initial assembly requirement can be met. Because the normals of the two begin to offset, they cannot be directly used. Fan and Fmn are angularly docked. First, according to the measured data processing results, the coordinates of Lan (Xan, Zan) and the coordinates of the point Lbn (Xbn, Zbn) are obtained. According to the relationship between the installation distance of the movable docking nut and the normal line of Fmn, the zero position coordinates of Lb'n (Xb'n, Zb'n) are calculated. Assuming that when Lb'n moves to the point Lb"n (Xb"n, Zb"n) where Lan and Lbn are collinear, the movable nut offset distance is:

[0097]

[0098] The slope k of the two points is:

[0099]

[0100] At the same time, Lan, Lbn, and Lb'n are collinear:

[0101] Its slope k' is:

[0102]

[0103] According to the normal relationship:

[0104] k*k'=-1

[0105]

[0106] According to the installation relationship, the distance △L(b'b) between Lbn and Lb'n is a known constant of 5mm:

[0107]

[0108] According to the above relationship, the calculation of △k(bb') is finally completed, and the movable nut is offset along the two ends of the normal line. Therefore, when △k(bb')≤1 / 2△kn=1mm, the movable docking nuts of the cylinder A and B in the embodiment can cover the assembly requirements after the product is deformed.

[0109] Step 9: Assembly sequence planning based on registration evaluation

[0110] Based on the above analysis of assembly performance, the matching relationship selection of assembly is carried out according to the registration evaluation results, and sequence guidance is provided to improve assembly reliability and efficiency.

[0111] Specifically: (1) Complete the three-dimensional scanning of the product's outer surface and hole position based on the initial connection and positioning conditions to obtain the conditions of the pin hole after installation in the real assembly environment; (2) Complete the product measurement of the batch of cylinder parts A and cylinder parts B to obtain the offset △k(bb') of the movable nut along the normal line at both ends and the overall assembly offset Prioritize products with smaller ∑Pab, and then complete the pairing of cylinder A and cylinder B according to the envelope range of △k(bb'). (3) Use the pin holes in the four quadrants to complete the preliminary connection and positioning of cylinder A and cylinder B; (4) According to the distribution results of P, the path planning design of the docking assembly in the embodiment is completed according to the strategy of tightening from large to small and symmetrically; (5) The characteristics of graded assembly are used to solve the problem of product deformation, and the assembly of 5N (half the torque requirement value) is completed for the first time. On this basis, the outer surface is scanned again, and the center coordinate calculation is completed according to the scanning results. The change of P value is evaluated, and the final assembly of the product 10N is completed again according to the distribution results of P according to the strategy of tightening from large to small and symmetrically. According to the above method, the registration assembly of cylinder A and cylinder B in the embodiment is finally completed, which greatly reduces the time of manual coordination and trial docking of traditional trial assembly, and the assembly efficiency is improved by 40%.

Claims

1. A complex prototype remodeling and registration evaluation method based on measured data, characterized in that: The steps include: Step 1: Construct a unified coordinate system for measurement and registration; Step 2: Complete the digital measurement of the actual shape and features of the product; Step 3: Complete the calculation of product focus feature points; Step 4: Complete the remodeling of complex prototypes based on lightweight features; Step 5: Complete the alignment relationship calculation based on local constraint matching; Step 6, calculating the overall assembly deviation of different reconstructed prototypes; Step 7: Complete the registration evaluation according to the offset of the reconstructed prototype; Step 8: Assembly sequence planning based on registration evaluation.

2. The complex prototype remodeling and registration evaluation method based on measured data drive according to claim 1 is characterized by: The method for constructing a unified coordinate system for measurement and registration described in step 1 is: taking the bottom surface of the butt end of the product as the reference, taking the center of the end face circle as the center of the circle, and establishing a spatial rectangular coordinate system XYZ in a horizontal support placement manner.

3. The complex prototype remodeling and registration evaluation method based on measured data drive according to claim 1 is characterized by: The method for completing the digital measurement of the actual shape and features of the product described in step 2 is: in a unified coordinate system of measurement and registration, non-contact scanning measurement is performed on the focus features of the product, wherein: for continuous smooth curved surface parts, a fast scanning method is used to complete point cloud acquisition; for docking hole features, a precise scanning method is used to complete detailed data acquisition of the conical edge and the circular hole; then, based on the point cloud data, the data that has little impact on the key features and deviates far is removed to complete point cloud denoising and data simplification processing.

4. The complex prototype remodeling and registration evaluation method based on measured data drive according to claim 3 is characterized by: The method for completing the solution of the product's focus feature points described in step three is: perform feature fitting on the overall conical surface scanning data, obtain the circular feature fitting of the outer and inner circles of the cone feature part, and extract the two circle centers; supplement the point cloud data of the missing hole positions according to the nearby continuous point clouds, and at the same time use the positioning holes in the four quadrants as target points, and use the quadrant area method to fit along the normal direction of the point cloud according to the arc length of the 1 / 4 quadrant area of ​​the theoretical model, and map along the theoretical arc length of each hole position under the supplementary area to achieve accurate compensation of the docking hole features in the case of missing holes, and complete the coordinate point solution of the assembly features according to the actual assembly requirements, and complete the coordinate point extraction at the set interval; extract the circle and center coordinates of the inner circle of the cone; and complete the fitting of continuous surface features and plane features respectively.

5. The complex prototype remodeling and registration evaluation method based on measured data drive according to claim 4 is characterized by: The method for completing the remodeling of the complex prototype based on lightweight features described in step 4 is: for the docking radial hole feature, the middle normal feature of the hole is generated with the center of the circle through the inner circle of the cone and the center coordinates; With the aid of various coordinate points, the extension of the non-thickness surface is completed in accordance with the relationship of continuity and no mutation of the surface, and the envelope fitting of the entire surface is realized; finally, the prototype remodeling of the lightweight features of various products is completed.

6. The complex prototype remodeling and registration evaluation method based on measured data drive according to claim 1 is characterized by: The method for completing the alignment relationship calculation based on local constraint matching described in step five is: first, determine the product assembly registration section: on the docking end face, the theoretical model is divided along the center of the radial docking key feature, and the intersection with the product is the theoretical model registration section; at the same time, the center coordinates of the docking radial hole and the normal of the hole center are extracted; in the same measurement-registration coordinate system, the measured product prototype assembly segment is imported, and the docking radial hole center coordinates and normal features of the assembly segment are aligned with the theoretical model registration section.

7. The complex prototype remodeling and registration evaluation method based on measured data drive according to claim 1 is characterized by: The overall assembly deviation of different reconstructed prototypes is calculated according to the following formula: Where La and Lb represent the curves formed by the coordinates of the center of the docking holes of assembly segments A and B in the theoretical model registration section; S(ab-n) (n=1, 2, 3..) is the area of ​​the intersection of La and Lb; f La (x), f Lb (x) represents the surface curve equations of the reconstructed prototype of assembly segments A and B respectively.

8. The complex prototype remodeling and registration evaluation method based on measured data drive according to claim 7 is characterized by: The method for completing the registration evaluation according to the offset of the reconstructed prototype described in step 7 is: (1) First, according to the measured data processing results, the point coordinates Lan (Xan, Zan) and Lbn (Xbn, Zbn) of the assembly sections A and B are obtained, and according to the relationship between the installation distance of the movable butt nut and the normal line Fmn, the zero-position coordinates of Lb'n (Xb'n, Zb'n) are calculated; where Lan and Lbn are the coordinates of a circle center of the assembly sections A and B, respectively, Fmn is the normal line of the center of Lbn, and Lb'n is the initial coordinate of the center of the movable butt nut inside the assembly section B. The center of the circle can move along the normal line of Fmn, and the movable distance is △kn; (2) Assuming that Lb'n moves to the point where Lan and Lbn are collinear, it is point Lb"n (Xb"n, Zb"n). At this time, the offset distance of the movable docking nut is: The slope k of the two points is: At the same time, the slope k' of Lan, Lbn, and Lb'n after collinearity is: According to the normal relationship k*k'=-1, we get: According to the installation relationship, the distance △L(b'b) between Lbn and Lb'n is a known constant: (3) The offset distance △k(bb') of the movable butt nut is calculated based on the above relationship. When △k(bb')≤1 / 2△kn, it is determined that the movable butt nut can meet the assembly requirements after the product is deformed.

9. The complex prototype remodeling and registration evaluation method based on measured data drive according to claim 8, characterized in that: Step 8 The assembly sequence planning method based on registration evaluation includes the following steps: The first step is to complete the 3D scanning of the product's outer surface and hole positions based on the preliminary connection and positioning conditions, and obtain the conditions of the pin holes after installation in the real assembly environment; Step 2: Calculate the offset distance △k(bb') of the movable butt nut and the overall assembly offset Prioritize the product with smaller ∑Pab, and then complete the pairing of assembly segments A and B according to the envelope range of △k(bb'); Step 3: Use the pin holes in the four quadrants to complete the preliminary connection and positioning of assembly sections A and B; Step 4: According to the distribution results of the assembly offset P, the path planning design of the docking assembly is completed according to the strategy of tightening from large to small and symmetrically; Step 5: Assemble the products in different grades.

10. The complex prototype remodeling and registration evaluation method based on measured data drive according to claim 9, characterized in that: The method for graded assembly of products described in the fifth step is: first complete the assembly of half the torque requirement value, and on this basis, scan the outer surface again, complete the calculation of the center coordinates according to the scanning results, evaluate the change of P value, and then complete the final assembly of the product according to the distribution result of P in accordance with the strategy of symmetrical tightening from large to small.

Images