Blank machining surface allowance optimization method and system based on three-dimensional scanning measurement

Abstract

The invention discloses a workblank machining surface allowance optimization method and system based on three-dimensional scanning measurement, and the method comprises the steps: collecting workblank measurement point cloud data and design model point cloud data as original point cloud data, carrying out the point cloud simplification of the original point cloud data, obtaining a point cloud normal vector, and carrying out the redirection; after point cloud simplification, blank measurement point cloud data and design model point cloud are aligned; performing initial segmentation on the point cloud feature surface, then performing re-segmentation on the point cloud feature surface, and establishing a corresponding relation between the measurement point cloud feature surface and the design model point cloud feature surface; and establishing a mathematical model of the machining allowance based on the corresponding relationship between the feature surfaces of the measurement point cloud and the design model point cloud, and optimizing the constraint variance minimization matching allowance. According to the method, the feature surface extraction accuracy of the point cloud data is obviously improved, meanwhile, the technological requirements of allowance uniformity and specified machining surface allowance constraint in the allowance optimization process are met, machining surface allowance optimization can be obviously more convenient and accurate, and calculation convergence and accuracy are higher.

Description

Technical Field

[0001] The present invention belongs to the fields of computer graphics and three-dimensional point cloud registration, and particularly relates to a method and system for optimizing the machining allowance of a rough machining surface based on three-dimensional scanning measurement. Background Art

[0002] Aerospace integral structural parts and their blanks are large in size, complex in process features, and have many thin-walled structures. Before processing such workpieces, it is necessary to allocate machining allowances for the blanks. If the allowance distribution is uneven, it will affect the machining accuracy and surface quality, and even cause the part to be scrapped due to insufficient local allowance. Therefore, optimizing the machining allowance of the machining surface is an important requirement in the rough machining process of aerospace structural parts.

[0003] The problems in the allowance distribution process of aerospace integral structural part blanks are summarized as the following points: 1) High scrap rate of rough machining parts. On the one hand, the overall structural part blanks obtained by casting or forging have large production tolerances. Even for blanks of the same batch, their actual shapes are different, and it is difficult to directly utilize the self-positioning reference features of the blanks. Even for some parts, there are no geometric features that can be used as a reference. On the other hand, once a local material shortage occurs during the blank manufacturing process and is not discovered in advance, the allowance distribution at some positions is insufficient, resulting in part scrapping.

[0004] 2) It is impossible to specify the allowance distribution of each machining surface of the blank, and it is difficult to ensure uniform machining allowances on the machining surface. The traditional scribing machining positioning method only ensures sufficient local allowances for the blank and cannot evenly distribute the specified machining allowances according to process requirements. Uneven allowances will cause the back engagement of cut to change continuously during cutting, the cutting force to be unstable, resulting in an unstable cutting process, poor machining accuracy and surface quality.

[0005] 3) The large size of the blank increases the manufacturing cost. Due to the limitations of current machining methods and blank manufacturing processes, in the design and manufacturing process, in order to avoid the occurrence of material shortage as much as possible, a large allowance is left for the whole blank, which not only increases the blank manufacturing cost but also increases the cutting time.

[0006] Therefore, how to optimize the allowance distribution of each machining surface of the integral structural part blank before NC machining has become an urgent problem to be solved.

[0007] The prior art calculates the directed distance from the actual measurement point corresponding to any theoretical measurement point to the local theoretical plane of any theoretical measurement point, and uses this directed distance as the machining allowance of any theoretical measurement point and constrains and optimizes it. However, this method only realizes the optimization of the machining allowance of a single measurement point and cannot realize the overall constraint of the workpiece feature surface; and only uses the example of uniformly distributed machining allowances, and it is difficult to prove the optimization effect of the machining surface with different allowance requirements.

[0008] There is also a two-step registration method implemented by using the PCA algorithm and the Plane-ICP algorithm improved based on margin optimization to obtain the optimal transformation parameters. However, this method only realizes the machining margin optimization of several reference plane section lines.

[0009] The above methods have problems of uneven margin distribution and insufficient local machining margin on key surfaces, which may lead to part scrapping and low machining efficiency and reliability of workpieces. Summary of the Invention

[0010] The purpose of the present invention is to provide a method and system for optimizing the margin of the rough machining surface based on three-dimensional scanning measurement to solve the problem of part scrapping caused by uneven margin distribution and insufficient local machining margin on key surfaces.

[0011] To achieve the above object, the present invention adopts the following technical solutions: In the first aspect, the present invention provides a method for optimizing the margin of the rough machining surface based on three-dimensional scanning measurement, including: Collect the point cloud data of the rough measurement and the point cloud data of the design model as the original point cloud data, perform point cloud reduction on the original point cloud data, obtain the point cloud normal vector and perform reorientation; After point cloud reduction, align the point cloud data of the rough measurement and the point cloud of the design model; Perform initial segmentation on the point cloud feature surface, then perform re-segmentation on the point cloud feature surface, and establish the corresponding relationship between the measurement point cloud and the point cloud feature surface of the design model; Based on the corresponding relationship between the measurement point cloud and the point cloud feature surface of the design model, establish a mathematical model of the machining margin, and constrain the variance minimization to match the margin optimization.

[0012] Optionally, the performing point cloud reduction on the original point cloud data, obtaining the point cloud normal vector and performing reorientation includes: Adopt the voxel downsampling algorithm and the farthest point downsampling algorithm for point cloud reduction. The voxel downsampling algorithm realizes point cloud reduction by specifying the sampling grid size; the farthest point downsampling algorithm specifies the number of points after reduction and retains the contour before reduction; subsequently, use the point cloud local surface fitting normal vector calculation method to determine the straight line where the normal vector of a certain local point is located, and use the minimum spanning tree method to reorient the normal vector to point to the outside of the point cloud.

[0013] Optionally, after the point cloud reduction, the aligning the point cloud data of the rough measurement and the point cloud of the design model includes: The two point clouds are aligned using the sampling consistency algorithm for rough registration and the improved iterative closest point algorithm for fine registration. In the rough registration stage, the sampling consistency algorithm based on the fast point feature histogram feature is used. First, the fast point feature histogram feature of the point cloud is extracted, and then the corresponding relationship between the feature points is obtained through the sampling consistency rough registration algorithm, and the coordinate transformation between the two point clouds is solved. After the rough registration provides the initial value, the improved iterative closest point algorithm ICP is used to further align the two point clouds on the basis of the rough registration. The ICP algorithm takes the point in the blank point cloud data and the point in the designed model point cloud data that is the closest in distance as the corresponding point, and iteratively solves the optimal transformation relationship between the two sets of point clouds by minimizing the square of the distance between the corresponding point pairs to register the two point clouds.

[0014] Optionally, the objective function of the ICP algorithm is: (1) In formula (1), represents the rotation matrix; represents the translation vector; represents the point on the blank point cloud; represents the designed model the point on it that is the closest in distance to

[0015] Optionally, the initial segmentation of the point cloud feature surface is performed, and then the point cloud feature surface is re-segmented to establish the corresponding relationship between the measured point cloud and the designed model point cloud feature surface, including: The initial segmentation of the point cloud feature surface is performed using the random sample consensus algorithm, and then the region growing algorithm is used to re-segment the point cloud feature surface to establish the corresponding relationship between the blank and the designed model feature surfaces. The initial segmentation of the point cloud feature surface is performed using the random sample consensus algorithm based on octree cells, and the complete point cloud is segmented into multiple regular point clouds that have been extracted and the remaining unextracted point cloud, and then the region growing algorithm is used to re-segment the point cloud feature surface; Let the set of feature surfaces extracted from the blank measured point cloud be , and the set of feature surfaces extracted from the part designed model point cloud be , by selecting and a set of corresponding surfaces in to form a feature surface pair , and finally forming a set of feature surface pairs , and the points subordinate to the set of feature surface pairs are used as the input of the allowance distribution algorithm.

[0016] Optionally, based on the corresponding relationship between the measured point cloud and the designed model point cloud feature surface, a mathematical model of machining allowance is established, and the constraint variance is minimized to optimize the matching allowance, including: ​Establish a mathematical model of machining allowance, and use the positive or negative value of the allowance to indicate whether the allowance at a certain point is surplus or lacking; if the blank point cloud any point on the design model point cloud the nearest corresponding point is , and in the normal direction is the allowance value, that is to represent the allowance, and the expression is: (2) Use the plane points extracted in the feature surface extraction step , and through the established corresponding relationship of the feature surface, find in the design model point cloud the corresponding feature surface , and the points located on the feature surface ; Based on the variance minimization matching algorithm VMM, and based on the minimum nominal allowance constraint of different machining surfaces, a constrained variance minimization matching allowance optimization is proposed, and the optimization goal is to minimize the sum of the squares of the residuals of the measured point distance from its mean deviation distance.

[0017] Optionally, the VMM objective function expression is: (3) The VMM objective function solves the rotation matrix and the translation vector such that the allowance from the blank model point cloud to the design model point cloud has the minimum variance, and at the same time makes the allowance value of any point on the machining surface greater than or equal to the minimum nominal allowance as the objective function constraint; In summary, the mathematical model expression based on the constrained VMM algorithm is: (4).

[0018] In a second aspect, the present invention provides a blank machining surface allowance optimization system based on three-dimensional scanning measurement, including: A data acquisition module, which is used to collect blank measurement point cloud data and design model point cloud data as the original point cloud data, perform point cloud reduction on the original point cloud data, obtain the point cloud normal vector and perform redirection; A point cloud alignment module, which is used to align the blank measurement point cloud data and the design model point cloud after point cloud reduction; A feature segmentation module, which is used to perform an initial segmentation on the point cloud feature surface, then re-segment the point cloud feature surface, and establish the corresponding relationship between the measured point cloud and the point cloud feature surface of the design model; An optimization output module, which is used to establish a mathematical model of the machining allowance based on the corresponding relationship between the measured point cloud and the point cloud feature surface of the design model, and optimize the matching allowance by minimizing the constrained variance.

[0019] In a third aspect, the present invention provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of the method for optimizing the machining surface allowance based on three-dimensional scanning measurement are implemented.

[0020] In a fourth aspect, the present invention provides a computer-readable storage medium. The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the method for optimizing the machining surface allowance based on three-dimensional scanning measurement are implemented.

[0021] Compared with the prior art, the present invention has the following technical effects: With the rapid development of computer technology and three-dimensional digital detection technology, three-dimensional point cloud information of a blank can be quickly obtained by using a three-dimensional scanning device. After matching the blank measurement point cloud data and the design model point cloud data, the machining surface allowance is optimized, and finally, the method of quickly positioning the processed blank point cloud information on a numerically controlled machine tool begins to be gradually applied. Compared with the prior art, the present invention has the following advantages: 1) Improve the accuracy of machining positioning, prevent the problem of part scrapping caused by insufficient local allowance distribution or inaccurate positioning, and improve the first-pass qualification rate of part machining.

[0022] 2) Uniform allowance distribution. The uniform distribution of the allowance on important machining surfaces enables the tool to be in a stable state during the machining process, reduces cutting deformation, and ensures good machining accuracy and surface machining quality of the part.

[0023] 3) Improve the efficiency of machining positioning. By using the processed blank point cloud information, the position and posture of any blank clamped on the workbench can be quickly and efficiently determined, reducing the auxiliary time before machining.

[0024] 4) Since the possibility of scrapping caused by blank material shortage is reduced, it has a feedback effect on the casting or forging process of the blank, so that there is no need to leave too much allowance when designing the blank, reducing its volume and thus reducing the production cost of the blank. Description of the Drawings

[0025] Figure 1 It is a schematic diagram for calculating the local normal vector; Figure 2It is a flow chart for pre - alignment of point cloud registration; Figure 3 It is a flow chart for extraction of point cloud feature surfaces; Figure 4 It is a schematic diagram of the effect of allowance optimization; Figure 5 It is a flow chart of the present invention. Specific implementation manners

[0026] The present invention will be described in detail below with reference to the accompanying drawings and embodiments.

[0027] Embodiment 1. Please refer to Figure 5 , A method for optimizing the allowance of the rough - machined surface based on three - dimensional scanning measurement, including: Collect the point cloud data of the rough - machined measurement and the point cloud data of the design model as the original point cloud data, perform point cloud reduction on the original point cloud data, obtain the point cloud normal vector and perform redirection; After point cloud reduction, align the point cloud data of the rough - machined measurement and the point cloud of the design model; Perform an initial segmentation on the point cloud feature surfaces, then perform a re - segmentation on the point cloud feature surfaces, and establish the corresponding relationship between the measurement point cloud and the point cloud feature surfaces of the design model; Based on the corresponding relationship between the measurement point cloud and the point cloud feature surfaces of the design model, establish a mathematical model of the machining allowance, and optimize the matching allowance by minimizing the constrained variance.

[0028] By optimizing the allowance of the rough - machined surface, the present invention can significantly improve the problem that the allowance distribution of each rough - machined surface cannot be specified, and the problem that the overall allowance reserved for the rough - machined part is large, resulting in increased production costs. It solves the problem of uneven allowance distribution during rough - machining and the resulting rough - part shortage problem, providing a prerequisite for carrying out research on the allowance distribution of rough - machined surfaces and ensuring the uniformity of the quality of the machined surface.

[0029] Embodiment 2. Please refer to Figures 1 to 5 , The present invention provides a method for optimizing the allowance of the rough - machined surface, including the following steps: Step 1: Perform point cloud reduction on the original point cloud data, calculate the point cloud normal vector and perform redirection.

[0030] The present invention adopts a voxel down - sampling algorithm and a farthest - point down - sampling algorithm. Among them, the voxel down - sampling algorithm can quickly achieve point cloud reduction by specifying the sampling grid size; the farthest - point down - sampling can accurately specify the number of points after reduction and can better retain the contour before reduction.

[0031] The specific steps of the voxel down - sampling algorithm are as follows: Construct an axial AABB bounding box covering all measurement points. According to the point cloud scale and the requirement of reduction ratio, divide the bounding box into pieces with a side length of The small grids, called voxels, are finally calculated for the center of each small grid and used to replace all points within the grid.

[0032] The specific steps of the farthest point downsampling reduction algorithm are as follows: Let the input point cloud have points. First, calculate the centroid of the point cloud, and then select the point farthest from from the point cloud as the starting point to obtain the initial sampling point set . Gradually select the point farthest from the currently selected point set and add it to the current sampling point set; repeat the above steps until the number of elements in the sampling point set reaches the set sampling number .

[0033] The method of calculating the vector of the local surface fitting of the point cloud is used to determine the straight line where the local one-point normal vector is located, and the minimum spanning tree method is used to redirect the normal vector to point to the outside of the point cloud.

[0034] Step 2: Coarse registration using the sampling consistency algorithm and fine registration using the improved iterative closest point algorithm to align the two point clouds.

[0035] SAC-IA is a coarse registration method that applies the principle of random sampling consistency. The algorithm steps using FPFH features are as follows: (1) Calculate the FPFH feature points of the rough part measurement point cloud and the design model point cloud respectively. (2) Feature point matching: Match the FPFH feature points in and the two point clouds. The matching rule is that each feature point in the point cloud matches feature points in the point cloud .

[0036] (3) Matching point selection: Randomly sample feature points ( ) from the point cloud . Each of these sampling points has corresponding points in . Randomly select one point as the corresponding point of the sampling point, so that pairs of matching points are obtained, and they have similar FPFH features.

[0037] (4) Calculate the corresponding rotation and displacement according to the selected pairs of matching points, and solve the transformation matrix through the SVD method.

[0038] (5) Evaluate the registration effect of each group of transformation matrices using the Huber function under the current sampled feature points.

[0039] Verify the SAC-IA coarse registration method based on FPFH features using the measured point clouds of the blanks of three parts: connector, slide rail, and bearing block, and the point clouds of the design models respectively. The transformation matrices obtained for the point clouds of the connector blank, slide rail blank, and bearing block are 、 、 respectively. The results of the coarse registration verification are shown in the following table.

[0040]

[0041]

[0042] Table 1 Results of Coarse Registration Verification

[0043] Before and after coarse registration, the RMSE values of the point clouds of the three parts are reduced by 46.16%, 85.43%, and 98.40% respectively. It can be seen that the measured point clouds of the blanks and the point clouds of the design models are roughly aligned, providing a good initial value for subsequent fine registration.

[0044] The input of the ICP algorithm is two groups of point clouds, namely the measured data of the blank as the source point cloud and the point cloud data of the design model as the target point cloud The output is the coordinate transformation from the blank point cloud to the point cloud of the design model .

[0045] The algorithm takes the points with the shortest distance between the measured point cloud data of the blank and the point cloud data of the design model as corresponding points, and iteratively solves the optimal transformation relationship between the two groups of point clouds by minimizing the square of the distance between the corresponding point pairs, thereby registering the two point clouds. The objective function of the classic ICP algorithm is: (1) In Equation (1), represents the rotation matrix; represents the translation vector; represents the point on the blank point cloud; represents the design model the point on it that is the closest to . The objective function is a convex function and can be solved using the SVD linear optimization method.

[0046] In the actual situation, there is a certain allowance on each machined surface, and not every point has a correct corresponding point on the part point cloud. Therefore, truncated least squares is introduced in the ICP algorithm to prune the wrong point pairs. For each found its corresponding , calculate the square of the Euclidean distance between the point pairs , and select the first point pairs. Among them, is the total number of points in the blank point cloud, represents the minimum guaranteed rate of pairable data points. When , the algorithm is equivalent to the general ICP method.

[0047] The improved ICP fine registration method is verified by using the measured point clouds of the blanks of three parts, namely the joint, the slide rail, and the bearing block, and the point clouds of the part design models respectively. The transformation matrices obtained by applying the improved ICP fine registration algorithm to the joint blank point cloud, the slide rail blank point cloud, and the bearing block blank point cloud are , and respectively.

[0048]

[0049] The fine registration results of the improved ICP algorithm are shown in the table. By comparing the changes in the RMSE values before and after registration, the RMSE values of the point clouds of the three parts are reduced by 93.07%, 66.02%, and 7.48% respectively. After fine registration, the two point clouds are further aligned. Among them, the alignment of the bearing block point cloud is already good in the rough registration stage, so the change in the alignment degree of fine registration is not significant.

[0050] Table 2 Fine registration verification results

[0051] Step 3: Use the random sample consensus algorithm to perform an initial segmentation of the point cloud feature surface, and then use the region growing algorithm to re-segment the point cloud feature surface to establish the corresponding relationship between the feature surfaces of the blank and the design model.

[0052] Let the set of feature surfaces extracted from the measured point cloud of the blank be , and the set of feature surfaces extracted from the point cloud of the part design model be . By artificially selecting and in a set of corresponding surfaces to form a feature surface pair , and finally form a set of feature surface pairs . The points belonging to the set of feature surface pairs are used as the input of the allowance distribution algorithm.

[0053] Feature surfaces were extracted from the point clouds of four different parts, namely the docking piece, the sector plate, the S-piece, and the box seat piece. The extraction effects are compared as shown in the table below.

[0054] Table 3 Comparison of Feature Surface Extraction Effects for Four Different Parts

[0055] Step 4: Establish a mathematical model for machining allowance and propose an optimization method for matching allowance by minimizing the constraint variance.

[0056] If for any point on the blank point cloud the nearest corresponding point on the designed model point cloud is , the distance between and in the normal direction is the allowance value, that is is used to represent the allowance, and the expression is: (2) If is negative, it indicates insufficient allowance, meaning that the blank measurement point is within the designed model, lacking material at this point, and the allowance is negative, not meeting the requirements; conversely, the blank measurement point is outside the designed model, having sufficient machining allowance at this point, meeting the requirements.

[0057] Using the plane points extracted in the feature surface extraction step, through the established corresponding relationship of feature surfaces, find the corresponding feature surface in the designed model point cloud , and the point located on the feature surface . By obtaining the parameters of through plane least squares fitting, the distance from the point on the blank point cloud to the corresponding machining surface on the designed model can be calculated using the fitting parameters of , that is, the point-to-plane distance from to the point on the corresponding feature surface of the designed model is used to replace the two-point distance between the corresponding points to represent the machining allowance on the plane.

[0058] If the coordinates of a point on the blank are , the plane where is located corresponds to the machining plane , combined with the point-plane distance formula, the following is the mathematical model expression of the said allowance: (3) In formula (3), is equal to 0 or 1. When the point allowance is positive, there is machining allowance; When the point allowance is negative, there is a lack of material. Compared with calculating the distances between all corresponding points, this allowance mathematical model omits the search for the corresponding point relationships, so the calculation time can be significantly reduced.

[0059] Based on the variance minimization matching (VMM) algorithm, considering the minimum nominal allowance constraints of different machining surfaces, the present invention proposes a constrained variance minimization matching allowance optimization method. The optimization objective is to minimize the sum of the squares of the residuals of the deviation distances between the measured point distances and their mean. The VMM objective function expression is: (4) The VMM objective function solves the rotation matrix and the translation vector such that the variance of the allowance from the blank model point cloud to the design model point cloud is minimized. At the same time, considering the setting of the minimum nominal allowance for different machining surfaces, let the allowance value at any point on the machining surface be greater than or equal to as the constraint of the objective function.

[0060] In summary, the mathematical model expression based on the constrained VMM algorithm is: (5) For the constraints imposed on each key machining surface , the penalty function method is used to construct a parameterized augmented objective function. By adding penalty terms, the constrained optimization problem is transformed into an unconstrained optimization problem, and the genetic algorithm is used to solve it. To evaluate the uniformity of the allowance distribution on the machining surface after blank allowance optimization, the following two evaluation indicators are used: (6) In formula (6), represents the fluctuation of the allowance at each point after allowance optimization, measuring the uniformity of the allowance, which is the same as the optimization objective of VMM; represents the size of the overall allowance distribution after optimization.

[0061] The present invention provides a method for optimizing the stock allowance of a rough machining surface based on three-dimensional scanning measurement, which is used to solve the problem of part scrapping caused by uneven stock allowance distribution on critical surfaces and insufficient local machining allowance, and provides a prerequisite for meeting the process requirements of the stock allowance of each machining surface of the rough blank and improving the machining efficiency and reliability of the workpiece. As Figure 4 shown, it can effectively solve the problem of rough blank material shortage.

[0062] In another embodiment of the present invention, a system for optimizing the stock allowance of a rough machining surface based on three-dimensional scanning measurement is provided, which can be used to implement the above-mentioned method for optimizing the stock allowance of a rough machining surface based on three-dimensional scanning measurement. Specifically, the system includes: A data acquisition module, which is used to acquire the point cloud data of the rough blank measurement and the point cloud data of the design model as the original point cloud data, perform point cloud reduction on the original point cloud data, obtain the point cloud normal vector and perform redirection; A point cloud alignment module, which is used to align the point cloud data of the rough blank measurement and the design model point cloud after point cloud reduction; A feature segmentation module, which is used to perform primary segmentation on the point cloud feature surface, then perform secondary segmentation on the point cloud feature surface, and establish the corresponding relationship between the measurement point cloud and the design model point cloud feature surface; An optimization output module, which is used to establish a mathematical model of the machining allowance based on the corresponding relationship between the measurement point cloud and the design model point cloud feature surface, and constrain the variance minimization to match the allowance optimization.

[0063] The division of modules in the embodiments of the present invention is illustrative. It is only a logical function division. In actual implementation, there may be other division methods. In addition, in each embodiment of the present invention, each functional module can be integrated in a processor, or can exist separately physically, or two or more modules can be integrated in one module. The above integrated modules can be implemented in the form of hardware or in the form of software function modules.

[0064] In another embodiment of the present invention, a computer device is provided. The computer device includes a processor and a memory. The memory is used to store a computer program, and the computer program includes program instructions. The processor is used to execute the program instructions stored in the computer storage medium. The processor may be a Central Processing Unit (CPU), or may also be other general-purpose processors, Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field-Programmable Gate Arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. It is the computing core and control core of the terminal, and is suitable for implementing one or more instructions. Specifically, it is suitable for loading and executing one or more instructions in the computer storage medium to implement the corresponding method flow or corresponding function. The processor described in the embodiment of the present invention can be used for the operation of a method for optimizing the allowance of the rough machining surface based on three-dimensional scanning measurement.

[0065] In another embodiment of the present invention, a storage medium is also provided, specifically a computer-readable storage medium (Memory). The computer-readable storage medium is a memory device in the computer device and is used to store programs and data. It can be understood that the computer-readable storage medium here can include both the built-in storage medium in the computer device and, of course, the extended storage medium supported by the computer device. The computer-readable storage medium provides a storage space, and the operating system of the terminal is stored in this storage space. And, one or more instructions suitable for being loaded and executed by the processor are also stored in this storage space. These instructions can be one or more computer programs (including program codes). It should be noted that the computer-readable storage medium here can be a high-speed RAM memory or a non-volatile memory, such as at least one disk memory. One or more instructions stored in the computer-readable storage medium can be loaded and executed by the processor to implement the corresponding steps of the method for optimizing the allowance of the rough machining surface based on three-dimensional scanning measurement in the above embodiment.

[0066] Those skilled in the art should understand that the embodiments of the present invention can be provided as a method, a system, or a computer program product. Therefore, the present invention can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk memory, CD-ROM, optical memory, etc.) that contain computer-usable program code.

[0067] The present invention is described with reference to the flowcharts and / or block diagrams of methods, apparatuses (systems), and computer program products according to embodiments of the present invention. It should be understood that each flow and / or block in the flowchart and / or block diagram, as well as the combination of flows and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, such that the instructions executed by the processor of the computer or other programmable data processing devices generate means for implementing the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0068] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, such that the instructions stored in the computer-readable memory generate a manufactured article including instruction means that implement the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0069] These computer program instructions can also be loaded onto a computer or other programmable data processing device, such that a series of operation steps are performed on the computer or other programmable device to generate a computer-implemented process, and thus the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0070] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them. Although the present invention has been described in detail with reference to the above embodiments, those of ordinary skill in the art should understand that it is still possible to modify the specific implementation manners of the present invention or make equivalent substitutions. Any modification or equivalent substitution that does not depart from the spirit and scope of the present invention should be covered by the protection scope of the claims of the present invention.

Claims

1. A method for optimizing the machining surface allowance of a blank based on three-dimensional scanning measurement, characterized in that: include: Collect the blank measurement point cloud data and the design model point cloud data as the original point cloud data, simplify the original point cloud data, obtain the point cloud normal vector and redirect it; After the point cloud is simplified, the blank measurement point cloud data and the design model point cloud are aligned; Perform initial segmentation on the feature surface of the point cloud, and then segment the feature surface of the point cloud again to establish the corresponding relationship between the measured point cloud and the feature surface of the point cloud of the design model; Based on the correspondence between the measured point cloud and the characteristic surface of the design model point cloud, a mathematical model of the machining allowance is established, and the constraint variance is minimized to match the allowance optimization.

2. The method for optimizing the machining surface allowance of a blank based on three-dimensional scanning measurement according to claim 1, characterized in that: The point cloud simplification of the original point cloud data to obtain the point cloud normal vector and redirect it includes: The voxel downsampling algorithm and the farthest point downsampling algorithm are used to simplify the point cloud. The voxel downsampling algorithm realizes point cloud simplification by specifying the sampling grid size; the farthest point downsampling algorithm specifies the number of points after simplification and retains the outline before simplification; then the point cloud local surface fitting normal vector calculation method is used to determine the straight line where the normal vector of a local point is located, and the minimum spanning tree method is used to redirect the normal vector to point to the outside of the point cloud.

3. The method for optimizing the machining surface allowance of a blank based on three-dimensional scanning measurement according to claim 1, characterized in that: After the point cloud is simplified, the blank measurement point cloud data and the design model point cloud are aligned, including: The two point clouds are aligned using the sampling consistency algorithm for coarse registration and the improved iterative nearest corresponding point algorithm for fine registration. In the coarse registration stage, the sampling consistency algorithm based on the fast point feature histogram feature is used to first extract the fast point feature histogram feature of the point cloud, and then the corresponding relationship between the feature points is obtained through the sampling consistency coarse registration algorithm to solve the coordinate transformation between the two point clouds; after the coarse registration provides the initial value, the improved iterative nearest corresponding point algorithm ICP is used to further align the two point clouds on the basis of the coarse registration. The ICP algorithm converts the blank point cloud data into Design model point cloud data The points with the closest distance between them are taken as corresponding points, and the optimal transformation relationship between the two groups of point clouds is iteratively solved by minimizing the square of the distance between the corresponding point pairs, so as to align the two point clouds.

4. The method for optimizing the machining surface allowance of a blank based on three-dimensional scanning measurement according to claim 3, characterized in that: The objective function of the ICP algorithm is: (1) In formula (1), represents the rotation matrix; represents the translation vector; Represents a point on the blank point cloud; Representation design model Upper distance The nearest point.

5. The method for optimizing the machining surface allowance of a blank based on three-dimensional scanning measurement according to claim 1, characterized in that: The method of performing an initial segmentation on the point cloud feature surface, and then segmenting the point cloud feature surface again, and establishing a corresponding relationship between the measured point cloud and the design model point cloud feature surface, includes: The point cloud feature surface is initially segmented using the random sampling consensus algorithm, and then the point cloud feature surface is re-segmented using the regional growing algorithm to establish the corresponding relationship between the feature surface of the blank and the design model. The point cloud feature surface is initially segmented using the random sampling consensus algorithm based on the octree unit, and the complete point cloud is segmented into multiple extracted regular point clouds and the remaining unextracted point clouds. Then the point cloud feature surface is re-segmented using the regional growing algorithm; Assume that the blank measurement point cloud The feature surface set extracted from , part design model point cloud The feature face set extracted from , by selecting and A set of corresponding faces in the , and finally form a feature face set , the points to which the feature face set belongs are used as input to the margin allocation algorithm.

6. The method for optimizing the machining surface allowance of a blank based on three-dimensional scanning measurement according to claim 1, characterized in that: The mathematical model of machining allowance is established based on the correspondence between the measured point cloud and the characteristic surface of the design model point cloud, and the constraint variance minimization matching allowance optimization includes: Establish a mathematical model for machining allowance, and use the positive or negative value of the allowance to indicate whether the allowance at a certain point is surplus or lacking; if the blank point cloud Take office a little Designing model point cloud The closest corresponding point on is , and exist The distance in the normal direction is the margin value, that is, To express the margin, the expression is: (2) Use the plane points extracted in the feature surface extraction step , through the established feature face correspondence relationship, find Designing model point cloud The corresponding feature surface , and on the characteristic surface point ; Based on the variance minimization matching algorithm VMM and the minimum nominal margin constraints of different machining surfaces, a constrained variance minimization matching margin optimization is proposed. The optimization goal is to minimize the sum of squares of the residuals of the distance between the measuring point and its mean deviation.

7. The method for optimizing the machining surface allowance of a blank based on three-dimensional scanning measurement according to claim 6, characterized in that: The VMM objective function expression is: (3) VMM objective function solves the rotation matrix and the translation vector Make point cloud from blank model To design model point cloud The margin The variance is the smallest, and at the same time, the allowance value of any point on the machining surface is greater than or equal to the minimum nominal allowance As objective function constraints; In summary, the mathematical model expression based on the constrained VMM algorithm is: (4)。 8. A blank machining surface allowance optimization system based on three-dimensional scanning measurement, characterized in that: include: The data acquisition module is used to collect the blank measurement point cloud data and the design model point cloud data as the original point cloud data, simplify the original point cloud data, obtain the point cloud normal vector and redirect it; Point cloud alignment module, used to align the blank measurement point cloud data with the design model point cloud after point cloud simplification; The feature segmentation module is used to perform initial segmentation on the feature surface of the point cloud, and then segment the feature surface of the point cloud again to establish the corresponding relationship between the measured point cloud and the feature surface of the design model point cloud; The optimization output module is used to establish a mathematical model of machining allowance based on the correspondence between the measured point cloud and the characteristic surface of the design model point cloud, and to optimize the matching allowance by constraining variance minimization.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of a method for optimizing the machining surface allowance of a blank based on three-dimensional scanning measurement as described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of a method for optimizing the machining surface allowance of a blank based on three-dimensional scanning measurement as described in any one of claims 1 to 7 are implemented.

Images