Rapid Prototyping Method for Complex Structural Parts of RV Reducer Based on 3D Printing Technology

Abstract

The present invention relates to the field of computer-aided forming technology, and discloses a rapid prototyping method for complex structural parts of an RV reducer based on 3D printing technology. The method includes: receiving three-dimensional model data of the RV reducer and performing topology optimization; performing feature segmentation on the three-dimensional model to obtain differentiated sub-regions; filling each sub-region to generate an initial filling path; obtaining first printing deviation data and correcting the initial filling path; obtaining second printing deviation data and comparing it with a preset threshold to determine whether to trigger a path correction mechanism; if correction is required, dynamically adjusting the filling path. By processing three-dimensional model data by computer, optimizing the filling path, analyzing printing deviation and dynamically adjusting, this method improves the accuracy and efficiency of 3D printing of complex structural parts of the RV reducer.

Description

Technical Field

[0001] The present invention relates to the technical field of computer-aided forming, and more specifically, to a rapid prototyping method for complex structural parts of RV reducers based on 3D printing technology. Background Art

[0002] With the rapid development of 3D printing technology, 3D printing has been widely applied in various fields, especially having unique advantages in rapid prototyping. At present, 3D printing technology is increasingly used in the field of industrial manufacturing, especially in the design and manufacturing of complex structural parts in the field of mechanical transmission, such as RV reducers. 3D printing technology shows great application potential.

[0003] The patent application with publication number CN114536766A discloses a 3D printing rapid prototyping method. This method first obtains the three-dimensional data of the item to be printed and establishes a three-dimensional model, then prepares the raw materials and screens the raw materials, and then imports the raw materials into the 3D printer material tank, and finally obtains the 3D printed finished product. This method has a simple and reasonable process, does not require laser forming, does not require the raw materials to be limited, can also achieve rapid prototyping under normal temperature and pressure, shortens the prototyping cycle, saves costs, and the prototyping rate is controllable, and the use effect is good. However, this patent fails to propose a special optimization scheme for complex structural parts such as RV reducers, nor does it consider the deviation control problem during the printing process.

[0004] The patent application with authorization announcement number CN214788854U discloses a RV reduction device based on 3D printing. This device includes a primary reduction mechanism and a secondary reduction mechanism, and realizes two-stage reduction through specially designed planetary gear transmission and cycloid pinwheel transmission. Using 3D printing technology to manufacture key components can achieve the lightweight design and rapid manufacturing of RV reducers. However, this patent mainly focuses on the mechanism innovation of RV reducers, and lacks a systematic elaboration on how to use 3D printing technology to optimize the prototyping and quality control of complex structural parts.

[0005] The patent application with publication number CN112026166A discloses a preparation method for rapid prototyping 3D printed workpieces. This method uses a photosensitive resin and a binder fully kneaded as a composite printing material, and realizes 3D printing by using a fluid filling printing technology, which can effectively overcome the problems of long existing 3D printing time and unstable structure. However, this method focuses on the modification of printing materials and process optimization, and lacks consideration of how to optimize the three-dimensional model and path planning according to the structural characteristics of RV reducers.

[0006] In summary, the existing technology lacks a 3D printing optimization scheme for the complex structural characteristics of RV reducers, lacks effective control means for printing deviations, and it is difficult to ensure printing accuracy and efficiency. Summary of the Invention

[0007] To overcome the above-mentioned defects of the prior art, the present invention provides a rapid prototyping method for complex structural parts of an RV reducer based on 3D printing technology, which performs topological optimization and feature segmentation on a three-dimensional model, generates a differentiated sub-region filling scheme, and adaptively corrects and dynamically adjusts the filling path by obtaining historical and real-time printing deviation data, so as to improve the accuracy and efficiency of 3D printing of the RV reducer.

[0008] To achieve the above object, the present invention provides the following technical solutions:

[0009] Receive the three-dimensional model data of the RV reducer, perform topological optimization on the three-dimensional model to generate the first filling path data; perform feature segmentation on the three-dimensional model of the RV reducer to obtain differentiated sub-regions; fill each sub-region to generate the initial filling path of the complex structural parts of the RV reducer;

[0010] Obtain the first printing deviation data, and based on the first printing deviation data, correct the initial filling path data to generate the third filling path data; the first printing deviation data includes dimensional deviation data, shape deviation data, and surface quality deviation data;

[0011] Obtain the second printing deviation data, compare the second printing deviation data with a preset deviation threshold to determine whether it is necessary to trigger the path correction mechanism; if the path correction mechanism is triggered, dynamically adjust the third filling path data according to the second printing deviation data to generate the fourth filling path data; according to the fourth filling path data, perform the 3D printing process of the complex structural parts of the RV reducer.

[0012] Further, the differentiated sub-regions include sub-region A, sub-region B, sub-region C, and sub-region D;

[0013] The feature segmentation of the three-dimensional model of the RV reducer to obtain differentiated sub-regions includes:

[0014] Establish a three-dimensional finite element model of the RV reducer, perform structural mechanics analysis on the RV reducer under rated working conditions to obtain the stress nephogram and displacement nephogram;

[0015] Extract the high stress area from the stress nephogram and the deformation concentration area from the displacement nephogram;

[0016] Perform wall thickness analysis on the three-dimensional model of the RV reducer, and extract the area with a wall thickness greater than the preset wall thickness threshold T thick as the wall thickness area;

[0017] According to the high stress area, deformation concentration area, and wall thickness area, obtain sub-region A, sub-region B, sub-region C, and sub-region D.

[0018] Further, obtaining sub-regions A, B, C, and D based on the high-stress region, the deformation-concentrated region, and the wall-thickness region includes:

[0019] Subtracting the deformation-concentrated region and the thick-wall region from the high-stress region to obtain sub-region A;

[0020] Subtracting the high-stress region and the thick-wall region from the deformation-concentrated region to obtain sub-region B;

[0021] Subtracting the high-stress region and the deformation-concentrated region from the thick-wall region to obtain sub-region C;

[0022] Taking the region other than sub-regions A, B, and C as sub-region D.

[0023] Further, the initial filling path for generating the complex structural part of the RV reducer by filling each sub-region includes:

[0024] Adopting a high-strength filling pattern including but not limited to cross-cross and spiral, and a filling density ρ1 for sub-region A to generate the filling path of sub-region A and obtain the filling path data of sub-region A;

[0025] Adopting a high-stiffness filling pattern including but not limited to honeycomb and hexagonal grid, and a filling density ρ2 for sub-region B to generate the filling path of sub-region B and obtain the filling path data of sub-region B;

[0026] Adopting a low filling pattern including but not limited to triangle and square grid, and a filling density ρ3 for sub-region C to generate the filling path of sub-region C and obtain the filling path data of sub-region C;

[0027] Adopting a filling density ρ4 for sub-region D to generate the filling path of sub-region D and obtain the filling path data of sub-region D; where ρ2 > ρ1 > ρ3 > ρ4;

[0028] Combining the filling path data of each sub-region to generate the second filling path data;

[0029] Fusing the first filling path data with the second filling path data to generate the initial filling path of the complex structural part of the RV reducer.

[0030] Further, obtaining the first printing deviation data includes:

[0031] Comparing the 3D models of all historical printing tasks with the current 3D model of the RV reducer and calculating the shape feature similarity;

[0032] Extracting the shape feature similarity with the current 3D model of the RV reducer greater than the preset similarity threshold T simHistorical printing tasks are marked as similar historical printing tasks; the printing parameters of the similar historical printing tasks are extracted as the first historical printing data;

[0033] Extract the dimensional parameters of each similar historical printing task in the first historical printing data, compare the dimensional parameters with the designed dimensional parameters of the historical 3D model corresponding to the similar historical printing task, and calculate the dimensional deviation data;

[0034] Extract the shape parameters of each similar historical printing task in the first historical printing data, compare the shape parameters with the designed shape parameters of the historical 3D model corresponding to the similar historical printing task, and calculate the shape deviation data;

[0035] Extract the surface quality parameters of each similar historical printing task in the first historical printing data to obtain the surface quality deviation data.

[0036] Further, the obtaining of the second printing deviation data includes:

[0037] During the 3D printing process of the complex structural parts of the RV reducer, the printing environment parameters and material property parameters are collected in real time;

[0038] Input the printing environment parameters and material property parameters into the pre-trained printing quality prediction model to obtain the second printing deviation data;

[0039] The second printing deviation data includes second dimensional deviation data, second shape deviation data, and second surface quality deviation data.

[0040] Further, the deviation thresholds include a dimensional deviation threshold Td, a shape deviation threshold Ts, and a surface quality deviation threshold Tq;

[0041] The path correction mechanism includes dimensional deviation correction, shape deviation correction, and surface quality correction;

[0042] Comparing the second printing deviation data with the preset deviation threshold to determine whether to trigger the path correction mechanism includes:

[0043] If the second dimensional deviation data exceeds Td, trigger the dimensional deviation correction;

[0044] If the second shape deviation data exceeds Ts, trigger the shape deviation correction;

[0045] If the second surface quality deviation data exceeds Tq, trigger the surface quality correction.

[0046] A rapid prototyping system for complex structural parts of an RV reducer based on 3D printing technology, which is used to implement the above-mentioned rapid prototyping method for complex structural parts of an RV reducer based on 3D printing technology. The system includes:

[0047] Region segmentation module: It is used to receive the three-dimensional model data of the RV reducer, perform topological optimization on the three-dimensional model to generate the first filling path data; perform feature segmentation on the three-dimensional model of the RV reducer to obtain differentiated sub-regions;

[0048] Path filling module: It is used to fill each sub-region to generate the initial filling path of the complex structural parts of the RV reducer;

[0049] First correction module: It is used to obtain the first printing deviation data, and based on the first printing deviation data, correct the initial filling path data to generate the third filling path data; the first printing deviation data includes dimensional deviation data, shape deviation data and surface quality deviation data;

[0050] Second correction module: It is used to obtain the second printing deviation data, compare the second printing deviation data with the preset deviation threshold, and judge whether it is necessary to trigger the path correction mechanism; if the path correction mechanism is triggered, then according to the second printing deviation data, dynamically adjust the third filling path data to generate the fourth filling path data; according to the fourth filling path data, execute the 3D printing process of the complex structural parts of the RV reducer.

[0051] An electronic device includes a memory, a central processing unit, and a computer program stored on the memory and executable on the central processing unit. When the central processing unit executes the computer program, it implements the above-mentioned rapid prototyping method for complex structural parts of an RV reducer based on 3D printing technology.

[0052] A computer-readable storage medium stores a computer program, and when the computer program is executed, it implements the above-mentioned rapid prototyping method for complex structural parts of an RV reducer based on 3D printing technology.

[0053] Compared with the prior art, the beneficial effects of the present invention are:

[0054] By performing feature segmentation and differential filling on the 3D model of the RV reducer, and using historical printing data and real-time printing parameters for path correction, the 3D printing accuracy of complex structural parts has been greatly improved. In particular, a differential filling strategy is adopted for key areas such as high-stress areas and deformation concentration areas, effectively ensuring the mechanical properties and geometric accuracy of these areas; by using the methods of topology optimization and intelligent zoning filling, the printing path is reasonably planned, reducing unnecessary material consumption and printing time. At the same time, the mechanism of dynamically adjusting printing parameters also avoids repeated printing caused by printing deviations, further improving production efficiency; by real-time monitoring of environmental parameters and material properties during the printing process, and combining with a pre-trained quality prediction model, printing deviations during the printing process can be detected and corrected in a timely manner, effectively ensuring the dimensional accuracy, shape accuracy and surface quality of the final product; by optimizing the filling path and material use, raw material waste is reduced. At the same time, the one-time forming rate is increased, and the rework and scrap rates are reduced, thereby reducing production costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0056] Figure 1 It is the principle flowchart of the rapid prototyping method for complex structural parts of the RV reducer based on 3D printing technology in the present invention;

[0057] Figure 2 It is the method flowchart for performing feature segmentation on the 3D model of the RV reducer in the rapid prototyping method for complex structural parts of the RV reducer based on 3D printing technology in the present invention to obtain differential sub-regions;

[0058] Figure 3 It is the method flowchart for extracting high-stress areas from the stress nephogram in the rapid prototyping method for complex structural parts of the RV reducer based on 3D printing technology in the present invention;

[0059] Figure 4 It is the method flowchart for extracting deformation concentration areas from the displacement nephogram in the rapid prototyping method for complex structural parts of the RV reducer based on 3D printing technology in the present invention;

[0060] Figure 5 It is the method flowchart for obtaining sub-region A, sub-region B, sub-region C and sub-region D according to high-stress areas, deformation concentration areas and wall thickness areas in the rapid prototyping method for complex structural parts of the RV reducer based on 3D printing technology in the present invention;

[0061] Figure 6 It is the method flowchart for filling each sub-region and generating the second filling path data in the rapid prototyping method of the complex structural parts of the RV reducer based on 3D printing technology of the present invention;

[0062] Figure 7 It is the method flowchart for obtaining the first printing deviation data of each similar historical printing task in the rapid prototyping method of the complex structural parts of the RV reducer based on 3D printing technology of the present invention;

[0063] Figure 8 It is the method flowchart for determining whether to trigger the path correction mechanism in the rapid prototyping method of the complex structural parts of the RV reducer based on 3D printing technology of the present invention;

[0064] Figure 9 It is the functional module diagram of the rapid prototyping system for the complex structural parts of the RV reducer based on 3D printing technology in the present invention. Specific embodiments

[0065] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0066] Embodiment 1

[0067] Please refer to Figure 1 As shown, this embodiment provides a rapid prototyping method for the complex structural parts of the RV reducer based on 3D printing technology, including:

[0068] Step S1000, receiving the three-dimensional model data of the RV reducer, performing topology optimization on the three-dimensional model to generate the first filling path data; according to the three-dimensional model data of the RV reducer, establishing a three-dimensional finite element model of the RV reducer, performing feature segmentation on the three-dimensional model of the RV reducer to obtain differentiated sub-regions; filling each sub-region to generate the initial filling path of the complex structural parts of the RV reducer;

[0069] Further, step S1000 includes:

[0070] Step S1100, receiving the three-dimensional model data of the RV reducer, performing topology optimization on the three-dimensional model to generate the first filling path data;

[0071] Further, step S1100 includes:

[0072] Step S1110: Based on the 3D model data of the RV reducer, establish a topology optimization mathematical model;

[0073] Step S1120: Solve the topology optimization mathematical model to obtain the optimized filling material distribution information, and generate the first filling path data according to the filling material distribution information.

[0074] Specifically, first receive the 3D CAD model of the RV reducer, which contains detailed geometric information of the RV reducer, such as external dimensions, internal structure, wall thickness changes, etc. At the same time, it is also necessary to input the material property parameters of the RV reducer, such as material density, elastic modulus, Poisson's ratio, yield strength, etc. These parameters are important bases for structural analysis and optimization. In addition, it is also necessary to clarify the stress state of the RV reducer, including the magnitude, direction, and acting position of the load. Based on the above input information, build a topology optimization model in CAE software. The topology optimization model includes an objective function, constraint conditions, and design variables; the objective function is to minimize the filling material usage on the premise of meeting the strength and stiffness requirements, and the constraint conditions include stress constraints, displacement constraints, etc.; the purpose of topology optimization is to optimize the distribution of materials in the design domain on the premise of meeting the structural performance requirements, remove redundant or low-stress materials, and obtain a lightweight structure. By defining design variables (pseudo-density of each unit material), objective function (such as minimizing volume), constraint conditions (such as stress constraints, displacement constraints), etc., establish a topology optimization mathematical model.

[0075] Exemplarily, taking minimizing volume as an example, the design variable is the pseudo-density of each unit, the objective function is to minimize the sum of (pseudo-density × unit volume) of all units, and the constraint conditions generally include:

[0076] 1) Overall static equilibrium equation to ensure that the optimized structure can withstand external loads;

[0077] 2) Local stress constraint to ensure that the optimized structure does not undergo material yield failure under the action of the load. The commonly used Von-Mises stress is used as the criterion;

[0078] 3) Displacement constraint to ensure that the deformation at key positions after optimization meets the usage requirements and avoid excessive deformation affecting the transmission accuracy;

[0079] 4) Volume fraction constraint to control the volume of removed materials. Too much will lead to insufficient strength, and too little will not achieve the weight reduction effect. Generally, it is taken as 20% - 50%.

[0080] After the constraint definition is completed, the last step is to use an optimization algorithm for numerical solution. Topology optimization is a huge non-linear programming problem involving a large number of design variables and requires the use of mathematical programming methods. The existing topology optimization methods mainly include:

[0081] 1) The homogenization method first calculates the equivalent properties of porous materials using homogenization theory and then optimizes the material distribution using the optimality criterion method;

[0082] 2) The solid isotropic material with penalization (SIMP) method describes the non - linear relationship between the pseudo - density and material properties using a power function. The higher the power exponent, the closer it approaches the "0 - 1" solution;

[0083] 3) The level - set method implicitly describes the structural boundary using a level - set function and evolves the boundary shape through the Hamilton - Jacobi equation;

[0084] 4) The evolutionary structural optimization (ESO) method iteratively removes low - stress elements or adds high - stress elements based on heuristic criteria;

[0085] 5) The bi - directional evolutionary structural optimization (BESO) method introduces a material addition criterion based on ESO to accelerate the convergence rate.

[0086] These methods have their own advantages and disadvantages. The SIMP method is the most widely used, which can better ensure volume constraints and obtain a "black - and - white" solution. The algorithm iterates continuously to decrease the objective function value. When the change in the objective function value between two adjacent iterations is less than a certain threshold or the maximum number of iterations is reached, the optimization process ends.

[0087] Exemplarily, taking the SIMP method as an example, the basic steps are as follows:

[0088] 1) Given the initial design variable value \(x_e\), for example, for full filling \(x_e = 1\);

[0089] 2) Calculate the element stiffness matrix, stress, and sensitivity;

[0090] 3) Assemble the total stiffness matrix and solve for the displacement;

[0091] 4) Calculate the objective function value and its sensitivity with respect to \(x_e\), and construct an approximate optimization model;

[0092] 5) Solve the approximate optimization model using MMA and update \(x_e\);

[0093] 6) Repeat steps 2) - 5) until convergence.

[0094] Finally, the element pseudo - density distribution \(x_e\) is obtained. \(x_e = 1\) represents a material element, \(x_e=x\) min represents a void element, \(x\) min represents the minimum relative density value in the SIMP topology optimization method, which is a small positive number close to 0. \(0\lt x_e\lt1\) represents the intermediate transition region (gray - scale element). Gray - scale elements have no physical meaning. Set a pseudo - density threshold. Those greater than the threshold are defined as materials, and those less than the threshold are defined as voids. By introducing \(x\) min , the numerical problem of singularity of the total stiffness matrix during the optimization process can be avoided.

[0095] After processing the gray-scale elements, a relatively clear material distribution region is obtained, thus obtaining the optimal material distribution result. In the result, the regions with higher material pseudo-density (red) indicate that the materials there are essential and bear the main load; conversely, the regions with lower pseudo-density (blue) are redundant materials that can be removed. For a speed reducer, stress concentration areas such as the tooth roots and rims of gears usually cannot be optimized, while most materials in low-stress areas such as the centers of gears and the middle parts of shafts can be optimized away.

[0096] Based on this, a first filling path is generated. The path parameters include filling direction (such as X-direction, Y-direction), filling density (such as 80%), and filling speed (such as 60 mm / s), etc. The filling direction generally follows the principal stress direction and is consistent with the force; the filling density is proportional to the pseudo-density value at that place, and the higher the pseudo-density, the denser the filling; the filling speed takes into account the printing efficiency and forming quality. If it is too fast, the strength is poor, and if it is too slow, the efficiency is low. These parameters provide initial values for further optimization.

[0097] In step S1100, topology optimization is used to find the optimal material distribution of the RV speed reducer, which not only meets the strength and stiffness requirements of the structure but also can minimize the material usage to the greatest extent, effectively shortening the printing time and reducing the printing cost. The filling path parameters obtained by optimization are relatively macroscopic, defining a unified filling method for the entire structure and facilitating rapid prototyping. However, for parts like the RV speed reducer with complex structures and large spans, the stress characteristics of different regions vary significantly. If a single filling method is used, there may be problems with poor local printing quality. Therefore, on this basis, targeted filling optimization for different regions is still required.

[0098] In step S1200, according to the three-dimensional model data of the RV speed reducer, a three-dimensional finite element model of the RV speed reducer is established, and the three-dimensional model of the RV speed reducer is feature-segmented to obtain differentiated sub-regions;

[0099] Furthermore, as Figure 2 shown, step S1200 includes:

[0100] In step S1210, according to the three-dimensional model data of the RV speed reducer, a three-dimensional finite element model of the RV speed reducer is established, and structural mechanics analysis is performed on the speed reducer under rated conditions to obtain the stress nephogram and displacement nephogram; the three-dimensional finite element model contains n1 nodes;

[0101] Specifically, for an RV reducer, first, a finite element model needs to be created based on its 3D CAD model for finite element analysis. Finite element analysis means discretizing a continuous structure into several elements, establishing local coordinate systems on the elements, and then using nodes as the link between elements to establish the overall coordinate system. On this basis, the stiffness matrix, mass matrix, etc. of the elements are established and assembled into the total stiffness matrix and total mass matrix. Then, combined with boundary conditions, initial conditions, and load conditions, by solving the static equilibrium equation or dynamic equilibrium equation, the response quantities such as displacement, strain, and stress of the structure under the action of the load are obtained. Due to the complex structure of the RV reducer, mesh division needs to be carried out for different characteristic regions. For example, the meshes in stress concentration areas such as gear tooth surfaces and wheel rims should be relatively dense, and the general mesh size should not be greater than 0.5 mm; while the meshes in stress level areas such as the wheel body and the box can be slightly sparser, and the mesh size can be relaxed to 1 - 2 mm. The mesh element type can be tetrahedron or hexahedron elements. Mesh encryption treatment is also required at parts such as tooth surfaces and mating surfaces. After the mesh division is completed, material properties such as elastic modulus, Poisson's ratio, and density are defined. Then, boundary conditions are applied, such as fixing and constraining the bearing seat holes and simply supporting and constraining the output end; loads are applied, such as the torque on the input shaft and the reaction moment on the output shaft. The loads and boundary conditions should reflect the actual working conditions as much as possible. Finally, the solution is submitted, and the stress field and displacement field distributions of the reducer under the action of the load can be obtained and visually presented in the form of contour maps. The stress contour map reflects the stress level of each part, and the displacement contour map reflects the deformation degree of each part.

[0102] Through finite element analysis, the stress characteristics of the RV reducer can be simulated efficiently and at low cost, its mechanical behavior under actual working conditions can be predicted, and the weak links in the design can be detected early. Compared with physical tests, simulation analysis is not restricted by site, tooling, and external factors and can be carried out quickly, providing a reliable basis for subsequent structural optimization, reinforcement, etc. The stress contour map and displacement contour map visually present the complex stress state inside the RV reducer and are important inputs for carrying out filling optimization.

[0103] Step S1220, extract the high-stress area from the stress contour map;

[0104] Furthermore, as Figure 3 shown, step S1220 includes:

[0105] Step S1221, convert the stress contour map into a grayscale image, and the grayscale value of the grayscale image is inversely proportional to the stress value;

[0106] Step S1222, traverse the grayscale image and classify the pixel points with grayscale values less than the preset grayscale threshold T stress into the high-stress area.

[0107] Specifically, since cloud maps usually represent the magnitude of values by continuous color gradients, it is not convenient to directly identify and extract regions. Therefore, the color stress cloud map needs to be first converted into a grayscale map, that is, the color information is mapped to grayscale values. The magnitude of the grayscale value is inversely proportional to the stress magnitude at that location, and the greater the stress, the lower the grayscale value. Then, set the grayscale threshold T stress Traverse each pixel in the grayscale map, and classify the pixel points with grayscale values lower than the threshold T stress as high-stress regions. The selection of the threshold needs to be determined according to parameters such as the yield limit and fatigue limit of the material, and can be taken as 50% of the yield limit, or 1.5 times the fatigue limit, etc., to ensure the safety margin of the high-stress region. The stress cloud map contains a large amount of stress numerical information, which is not easy to directly extract regions. Through grayscale conversion and threshold segmentation, the continuous stress field can be transformed into discrete high- and low-stress regions, which is convenient for clustering and extraction.

[0108] Step S1230, extract the deformation concentration region from the displacement cloud map;

[0109] Furthermore, as Figure 4 shown, step S1230 includes:

[0110] Step S1231, calculate the displacement gradient of the nodes according to the displacement cloud map;

[0111] Step S1232, select the node with the largest displacement gradient as the seed point, and classify the pixel points with displacement gradients greater than the preset gradient threshold T disp among the adjacent nodes into the deformation region;

[0112] Step S1233, repeat S1232 until all nodes that meet the conditions are classified into the deformation region, and obtain the deformation concentration region.

[0113] Specifically, when extracting regions with large deformations, if only considering the absolute value of the displacement, regions with large rigid body displacements may also be extracted, but the deformations in these regions are not significant. Therefore, by calculating the displacement gradient of the nodes, the severity of the deformation is judged. The displacement gradient reflects the rapid change rate of the displacement. On the displacement cloud map, first select the node with the largest gradient as the seed point, and then search around it. Classify the points with gradients greater than the preset gradient threshold T disp among the adjacent nodes into the deformation region. Here, the region growing method is used for region segmentation, which can effectively gather points with similar deformation degrees. The threshold T dispIt can be set according to the gradient distribution characteristics of the displacement nephogram, such as taking twice the gradient mean value, etc. Repeat the expansion until there are no new nodes that meet the conditions and can be classified into the deformation area, and the concentrated deformation area is obtained. Excessive deformation will affect the transmission accuracy of the RV reducer, and the concentrated deformation area also needs to be filled emphatically to reduce the deformation. The displacement gradient combines the absolute value of the displacement and the relative change rate, and can better reflect the significant degree of deformation. Using the region growing algorithm to extract the deformation area avoids the subjectivity of selecting the gradient threshold, can adaptively extract the area where the deformation is concentrated according to the displacement distribution, and improves the pertinence of the filling optimization.

[0114] Step S1240: Perform wall thickness analysis on the 3D model of the RV reducer, and extract the area where the wall thickness is greater than the preset wall thickness threshold T thick as the wall thickness area;

[0115] Specifically, in addition to the stress level and deformation trend, the influence of the wall thickness also needs to be considered when selecting the filling method of the RV reducer. Generally speaking, the larger the wall thickness, the more inclined to use a low filling density to save materials; on the contrary, in the thinner wall thickness area, to ensure the surface quality, the filling density usually needs to be increased. Therefore, it is necessary to perform wall thickness analysis on the 3D model. Using the wall thickness analysis command in CAD software, the distance from the model surface to the inside can be quickly calculated. Set the wall thickness threshold T thick， and extract the area greater than this threshold as the wall thickness area. The value of Tthick needs to balance the filling density and the forming quality. If it is too large, the filling density is too low and the strength is insufficient; if it is too small, the purpose of reducing materials cannot be achieved. The average wall thickness of this type of part can be taken, or the recommended value in the 3D printing process guide can be referred to. Most of the existing filling methods are based on experience and rarely consider the influence of the wall thickness on the filling quality. Through wall thickness analysis, the thick wall area can be found, and suitable filling parameters can be matched for it, ensuring the local strength and surface quality while saving costs, and making the filling scheme more scientific and reasonable.

[0116] Step S1250: Obtain sub-region A, sub-region B, sub-region C, and sub-region D according to the high stress area, the concentrated deformation area, and the wall thickness area.

[0117] Furthermore, as Figure 5 shown, step S1250 includes:

[0118] Step S1251: Subtract the concentrated deformation area and the thick wall area from the high stress area to obtain sub-region A;

[0119] Step S1252: Subtract the high stress area and the thick wall area from the concentrated deformation area to obtain sub-region B;

[0120] Step S1253: Subtract the high stress area and the concentrated deformation area from the thick wall area to obtain sub-region C;

[0121] Step S1254: Define the area outside sub-regions A, B, and C as sub-region D.

[0122] Specifically, through the above analysis, the RV reducer is divided into high-stress regions, deformation-concentrated regions, and thick-wall regions. However, there are often overlaps between these regions. For example, the high-stress region may also be a region with large deformation, and the thick-wall region may also be a low-stress region. To simplify the region division, it is necessary to eliminate the coupling between regions.

[0123] Specifically:

[0124] Subtract the deformation-concentrated region and the thick-wall region from the high-stress region to obtain sub-region A caused solely by a high stress level.

[0125] Subtract the high-stress region and the thick-wall region from the deformation-concentrated region to obtain sub-region B caused solely by large deformation.

[0126] Subtract the high-stress region and the deformation-concentrated region from the thick-wall region to obtain sub-region C caused solely by a large wall thickness.

[0127] Define the area outside sub-regions A, B, and C as sub-region D, which has the lowest filling density.

[0128] In this way, decoupling is achieved between each sub-region, and the filling properties are more independent, facilitating subsequent targeted filling optimization.

[0129] Through the above steps, the overlap between different characteristic regions is eliminated, enabling each sub-region to have unique filling properties and avoiding filling redundancy. For complex parts, this method of dividing and conquering can greatly reduce the optimization difficulty, be carried out in parallel, and improve the optimization efficiency. Considering from different perspectives such as high-stress regions, deformation-concentrated regions, and thick-wall regions, the filling scheme is more comprehensive and it is not easy to miss key regions.

[0130] Step S1200 comprehensively considers factors such as the structural mechanics characteristics of the RV reducer and the requirements of the 3D printing process, and divides the complex overall model into multiple sub-regions with obvious differences. This is the prerequisite for realizing regional heterogeneous filling. The selection of the number of sub-regions needs to balance the filling fineness and the calculation cost. Although a larger number of sub-regions can better match the local structural characteristics, it will also significantly increase the difficulty of optimization and solution. This embodiment recommends dividing the RV reducer into 5 - 8 sub-regions, which not only meets the needs of differential filling but also does not bring an excessive calculation burden.

[0131] Step S1300: Fill each sub-region, generate the filling path data corresponding to each sub-region, and combine the filling path data of each sub-region to generate the second filling path data.

[0132] Furthermore, asFigure 6 As shown, step S1300 includes:

[0133] Step S1310, adopting a high-strength filling pattern including but not limited to cross-cross and spiral for sub-region A, and a filling density ρ1, generating a filling path for sub-region A, and obtaining filling path data for sub-region A;

[0134] Specifically, sub-region A is the high-stress area of the RV reducer. This area bears a large load and is prone to stress concentration, thus causing failures such as fatigue and yielding. Therefore, the filling of area A should focus on ensuring strength and stiffness. Commonly used high-strength filling patterns are: cross-cross, that is, alternating parallel lines are filled in two directions, and the crossing angle is usually 90°, which can provide relatively uniform mechanical properties; spiral filling, that is, continuous spiral filling from the inside out, with a smooth path, small stress concentration, and high strength. In addition, there are also high-strength patterns such as concentric circle filling and radial filling available for selection. The filling angle should be consistent with the principal stress direction at this place, so that the force on the material will be more smooth and stress mutation is not easy to occur. When the filling angle is consistent with the force direction and the filling spacing is controlled below 1 mm, a relatively high filling density can be obtained, generally reaching more than 80%, thus ensuring a relatively high local strength.

[0135] Adopting a high-strength filling pattern for sub-region A can maximize the mechanical properties of the 3D printing material, resist large external loads, and extend the service life of the high-stress area. Directional filling can make the force on the material more smooth, and dense filling can improve the local strength. This is particularly important for key stress areas such as tooth roots and journal necks. The strength of these areas directly determines the transmission performance and reliability of the RV reducer. Compared with traditional overall filling, the directional dense filling of additive manufacturing can increase the strength of the key area by more than 30%.

[0136] Step S1320, adopting a high-stiffness filling pattern including but not limited to honeycomb and hexagonal grid for sub-region B, and a filling density ρ2, generating a filling path for sub-region B, and obtaining filling path data for sub-region B;

[0137] Specifically, sub-region B is the deformation concentration area of the RV reducer. The deformation amount in this area is relatively large, and the deformation gradient is steep, which is likely to cause a decline in transmission accuracy. Therefore, when filling area B, emphasis should be placed on improving stiffness and controlling deformation. Commonly used high-stiffness filling patterns include: honeycomb filling, that is, nesting and filling a hexagonal honeycomb structure, which is widely used in thin-walled and plate-shell parts and has a relatively high flexural stiffness while achieving lightweight; hexagonal grid, that is, dividing the entire filling area into regular hexagonal grids and filling along the grids, similar to the triangular grid, used to improve the structural stiffness. The filling angle should be perpendicular to the deformation direction. For example, if the deformation is mainly along the radial direction, the filling lines should be mainly along the circumferential direction, so as to maximize the constraint of displacement in the deformation direction. The filling spacing should be small, generally less than 0.8 mm, and the corresponding filling density can reach more than 60% to obtain a higher deformation resistance.

[0138] Adopting a high-stiffness filling mode for sub-region B can effectively improve the anti-deformation ability of this area and ensure the operation accuracy of the RV reducer. Shape-selective filling can maximize the constraint effect of materials and inhibit the development of deformation. Dense filling can further strengthen the stiffness. Deformation problems are common quality hazards in RV reducers. Slight deformation can cause poor gear meshing, and severe deformation can even lead to jamming. By adopting zone-optimized filling, the stiffness of the deformation-sensitive area can be increased by more than 50%, significantly improving the transmission quality of the reducer.

[0139] Step S1330, adopt a low-filling mode including but not limited to triangular and square grids and a filling density ρ3 for sub-region C to generate the filling path of sub-region C and obtain the filling path data of sub-region C;

[0140] Specifically, sub-region C is the thick-wall area of the RV reducer. The wall thickness in this area is relatively large, the moment of inertia is high, and the force is relatively small. Therefore, the filling density of area C can be appropriately reduced to save printing materials while ensuring strength. Commonly used low-filling patterns include: triangular grid, that is, dividing the filling area into regular triangles and filling along the sides of the triangles to build a stable frame structure; square grid, that is, dividing the filling area into regular quadrilaterals and filling along the quadrilaterals, similar to the triangular grid, further reducing the filling density. For the thick-wall area, the filling density is generally controlled at 30% - 50%, which can not only meet the strength requirements but also effectively shorten the printing time. The filling spacing can be appropriately increased and is acceptable within 2 mm, further increasing the sparsity of the filling.

[0141] Sub-region C adopts a low-fill pattern, which saves printing materials to the greatest extent while meeting the strength requirements, reducing the manufacturing cost. The grid-like fill can form a skeleton support at key nodes, ensuring the minimum strength requirements of the thick-wall area. Reducing the ineffective fill can also shorten the printing time and improve production efficiency. The thick-wall area is usually the non-critical stress-bearing areas such as the wheel body and end cover of the reducer, allowing the fill rate to be moderately reduced. After the fill is optimized by region, the material reduction rate can reach more than 20%, with obvious cost reduction and efficiency improvement.

[0142] Step S1340, adopt a fill density ρ4 for sub-region D, generate the fill path of sub-region D, and obtain the fill path data of sub-region D; where ρ2 > ρ1 > ρ3 > ρ4;

[0143] Sub-region B is the deformation concentration area of the RV reducer, usually located at weak cross-sections such as the tooth root and transition fillet, with a large stress-strain gradient and prone to stress concentration. A high fill density means a higher volume fraction of the filling material inside this area, more fully stressed, better able to resist deformation, and avoid early failure caused by stress concentration. When the RV reducer transmits torque, the torsional force is transmitted along the normal direction of the key cross-section. The high fill density in sub-region B helps the material in this area to have better continuity and a smoother force transmission path, enabling the torsional force to be evenly and orderly transmitted from one side to the other, reducing the strain mismatch caused by stress mutation. Sub-region B may also involve the tooth surface contact area and needs to bear a large contact stress. The high fill density provides more solid material support, higher contact stiffness, can effectively relieve the peak contact stress, and improve the tooth surface fatigue life. A large amount of deformation energy accumulates in sub-region B and is more likely to cause thermal deformation during high-speed operation. A high fill density means a higher overall thermal conductivity of the material in the area, which helps the deformation energy to be dissipated in time, reduce the local temperature rise, and maintain dimensional stability. In contrast, although sub-region A is a high-stress area, the deformation is relatively uniform and the heat dissipation condition is better, so the fill density can be appropriately reduced. The C and D regions mainly play a supporting and connecting role, and the fill density can be further reduced to save the material consumption. Sub-region B comprehensively considers various factors such as strength, stiffness, deformation, and force transmission. It is the weak link of the RV reducer. Adopting the highest fill density is the result of a compromise optimization in terms of mechanics, thermology, and technology, which can maximize the flexible advantages of additive manufacturing and achieve lightweight design.

[0144] Sub-region D is a non-critical area of the RV reducer with low stress level, small deformation, and relatively thin wall thickness. The filling pattern and angle in this area have little impact on the performance of the reducer. To maximize the printing efficiency, the lowest filling density, generally between 15% and 20%, should be adopted in area D, and only the necessary skeleton needs to be formed. At the same time, the filling spacing can be widened to more than 3 mm to further reduce the printing time. After such treatment, the filling density and spacing in area D will be significantly different from those in areas A, B, and C, fully reflecting the optimization effect of zonal filling. The sparest filling in sub-region D can minimize the filling amount and the shortest printing time in the area with the lowest mechanical property requirements, reflecting the flexible production characteristics of additive manufacturing. If the lowest requirement filling area is filled homogenously in the traditional way, a large amount of redundancy will often be generated, wasting printing resources. However, zonal optimization filling can save this unnecessary consumption. The filling amount in area D can be reduced by more than 50%, and the printing efficiency can be increased by more than 30%. The existence of area D brings a greater optimization space for the filling scheme of the entire reducer.

[0145] Step S1350: Combine the filling path data of each sub-region to generate the second filling path data.

[0146] Specifically, after generating the filling paths for each sub-region as above, each sub-region of the RV reducer has a corresponding optimal filling scheme. At this time, it is necessary to assemble the filling path data of each region to generate a filling scheme representing the entire reducer. During the assembly, it should be noted that the filling paths in different regions should achieve a smooth transition at the interface, the filling lines should be as continuous as possible without gaps. The change in the filling angle in adjacent regions should be as small as possible to reduce stress mutation. The overall assembled filling path shows obvious zonal characteristics in terms of density and angle macroscopically, while achieving the continuity and transition of the paths in local details, reflecting the unity of integrity and locality.

[0147] Assembling each region to form the overall filling path of the reducer makes the zonal optimization scheme implemented as a numerical control program that can be used to guide printing, completing a key step from digital twin to physical manufacturing. The smooth transition of the filling path at the interface is achieved in the assembly scheme. While enjoying the benefits of zonal filling, the integrity and consistency of the formed part are ensured, avoiding the sudden change of physical properties between regions, and thus improving the comprehensive performance of the reducer. This is another advantage of zonal filling compared with overall filling. Through assembly, the optimization results of each region can be superimposed and amplified, greatly enhancing the applicability of additive manufacturing in the field of reducers.

[0148] In summary, step S1300 adopts a region - differentiated filling optimization strategy, making full use of the differences in force - bearing characteristics of different regions. For regions A, B, C, and D, filling methods with high strength, high stiffness, low filling rate, and the lowest filling rate are adopted respectively, meeting the requirements of strength, precision, lightweight, and high efficiency. This is difficult to achieve in overall filling and homogeneous filling. After assembling the solutions for each region, smooth transition of the path at the interface is realized, ensuring the integrity and consistency of the formed part. Region - based filling is an inevitable requirement for the flexible and intelligent development of additive manufacturing, conforming to the trend of the development of subtractive structures towards functional gradient and composite structures, realizing the integrated optimization design of "shape - material - force", and having broad application prospects.

[0149] In step S1400, the first filling path data and the second filling path data are fused to generate the initial filling path of the complex structural part of the RV reducer.

[0150] Specifically, after obtaining the first filling path and the second filling path respectively, the two need to be further fused to form the initial filling path parameters representing the entire additive manufacturing process. The first filling path is a macroscopic filling scheme obtained based on topology optimization, reflecting the optimal distribution of materials in the entire design domain and having the characteristics of global optimality; the second filling path is a refined filling scheme obtained based on stress - strain analysis, further refining the filling methods in different internal regions and having the characteristics of local optimality. Only by organically combining the results obtained at two scales and two stages, learning from each other's strengths and compensating for each other's weaknesses, can the best solution that takes into account both the overall and local aspects be obtained. The fusion process needs to consider the following factors: First, the difference in filling density between the two schemes. The weighted average of the two can be taken, and the weight is proportional to the degree of influence of the region on the performance of the reducer. Second, the difference in filling angle between the two schemes. Based on the results of force analysis, the printing direction is also taken into account. Third, the continuity of the fused filling path at the junction should have a smooth transition. In addition, the total filling rate of the fused scheme needs to be checked, which should not be too high or too low, generally controlled between 20% and 30%. After passing the check, parameters such as filling density, filling angle, and filling spacing are converted into numerical control codes to guide the actual filling process of the 3D printer, and finally the optimal filling configuration of the RV reducer that meets the usage requirements and can save printing materials is obtained.

[0151] The fusion of the two-stage filling scheme takes the advantages of global optimization and local optimization to form a unified final filling scheme with multiple scales and multiple objectives. On the one hand, the fusion filling scheme retains the optimization results of the macroscopic distribution of materials by topology optimization, which meets the overall requirements of load transmission; on the other hand, it incorporates the differentiated treatment of local structures by regional filling, which meets the local requirements of force. And it realizes the continuous transition of the two filling paths at the interface, avoiding the sudden change of density and angle. The filling rate verification further balances the gains and losses of material reduction and strength from the perspectives of cost and quality. The obtained NC code has fully considered the printing process requirements and can be directly used to guide printing, saving the manual programming link. Multi-stage fusion optimization makes up for the shortcomings of the current filling method that focuses on a single goal, and represents the mainstream direction of future additive manufacturing filling optimization technology.

[0152] The step of fusing the first filling path data with the second filling path data to generate an initial filling path for the complex structural part of the RV reducer includes:

[0153]

[0154] in:

[0155] : Initial filling path data after fusion;

[0156] : First fill path data;

[0157] : The second filling path data;

[0158] : The power index of the first filling path, used to adjust the weight of the first filling path in the fusion;

[0159] : The power index of the second filling path, used to adjust the weight of the second filling path in the fusion; and It is obtained based on experience or by fitting historical data. It is usually used to balance the needs of global and local optimization. It is also used to adjust the importance of global and local filling paths. When local needs are more important, Can be compared larger, to enhance the impact of local paths;

[0160] : Attenuation factor, reflecting the distance attenuation effect in the first filling path; Represents the distance from the path node to the center of the structure, is the attenuation parameter, which is the parameter that controls the influence range of the global path; Each node in the model can be Calculated by the Euclidean distance of the geometric center of the reducer; Determined through experiments or simulation analysis, usually reflecting the rate of path filling attenuation; Can effectively weaken the influence of paths far from the central area on the final filling path, ensuring the concentration of paths;

[0161] : Enhancement term for local paths, where is the stress intensity, is the enhancement coefficient, reflecting the importance of areas with high local stress for path correction; Extract local stress values through the stress nephogram in finite element analysis; Set according to the requirements of local material strength, used to control the weight of local stress on path enhancement; the higher the local stress, the larger the value, which can enhance the filling density and complexity of paths in this area and improve the local structural performance.

[0162] Path power exponent , attenuation parameter and enhancement coefficient are the main over-variables, which jointly determine the global and local optimization balance of the final filling path and the distribution characteristics of the paths. The finally generated initial filling path changes with the adjustment of these over-variables. When increases, paths farther away decay faster and the role of the global path decreases; when increases, areas with high local stress will obtain stronger path enhancement.

[0163] Through parameters and adjust the weights of global and local paths. This formula can automatically balance global optimization and local strengthening according to design requirements. This is especially important in complex structural parts, which can not only ensure the overall structural integrity but also strengthen areas that require high stress or high stiffness locally. The attenuation factor in the formula can effectively weaken the influence of paths in areas farther away, making the printing paths more concentrated and reducing material waste in areas far from the core area. This helps to improve printing efficiency and material utilization rate. The local path enhancement term enables high-stress areas to obtain additional path enhancement, ensuring the mechanical properties of these areas while avoiding material redundancy. Through this formula, the printing paths achieve a more refined dynamic balance between global and local optimization, adapting to different stress distributions and structural characteristics.

[0164] Step S2000, obtaining n2 similar historical printing tasks with similar structural features to the current RV reducer three-dimensional model, extracting printing parameters of n2 similar historical printing tasks as first historical printing data; analyzing the first historical printing data, obtaining first printing deviation data of each similar historical printing task, and based on the first printing deviation data, correcting the initial filling path data to generate third filling path data; the first printing deviation data includes size deviation data, shape deviation data and surface quality deviation data; n2 is a positive integer;

[0165] Furthermore, step S2000 includes:

[0166] Step S2100, obtaining n2 similar historical printing tasks with similar structural features to the current RV reducer three-dimensional model, and extracting printing parameters of the n2 similar historical printing tasks as the first historical printing data; n2 is a positive integer;

[0167] Furthermore, step S2100 includes:

[0168] Step S2110, comparing the 3D models of all historical printing tasks with the current RV reducer 3D model, and calculating the shape feature similarity;

[0169] Step S2120, extracting the shape feature similarity with the current RV reducer three-dimensional model is greater than a preset similarity threshold T sim The historical printing tasks are marked as similar historical printing tasks.

[0170] Specifically, historical printing tasks refer to all printing jobs completed by 3D printing equipment, including three-dimensional models of parts, printing parameter settings, printing effect evaluation and other contents. The three-dimensional model reflects the geometric shape, size and structural characteristics of the printed object; printing parameters include filling method, filling density, printing direction, layer thickness, printing speed, etc.; printing effect evaluation generally quantifies the quality characteristics of the printed parts through precision detection, mechanical testing, surface roughness measurement and other means. These data are usually stored in the database of the device or uploaded to the cloud server. In order to find the historical printing tasks that are most similar to the current RV reducer three-dimensional model, their shape features need to be compared first. The three-dimensional shape description algorithm can be used to extract the feature vectors of the model, such as volume, surface area, convex hull, Fourier descriptor, etc., and then calculate the similarity between the current model and the feature vectors of each historical model, such as Euclidean distance, Mahalanobis distance, angle cosine, etc. Set the similarity threshold T sim, extract the printing tasks corresponding to the historical printing models with similarity higher than this threshold. The selection of the threshold needs to consider both the number and the degree of similarity of the similar models. If it is too high, it may be impossible to find enough reference cases; if it is too low, the reference value is not great. Generally, the mean or median of the similarity of historical models can be taken. After extracting the similar historical printing tasks, use their printing parameters as the first historical printing data for subsequent path correction. Printing parameters are the key factors affecting the forming quality. Drawing on the experience of similar cases can avoid some known problems and provide better initial values for parameter optimization. The more similar models there are, the richer the historical data obtained, and the better the optimization effect.

[0171] The process of using historical data to guide the current printing task is essentially a case-based learning method. Compared with groping blindly, referring to the solutions of similar problems can greatly improve the optimization efficiency and reduce the trial-and-error cost. Especially for parts with complex structures and high process requirements, the selection of printing parameters requires more experience accumulation and inheritance. Through big data analysis and similarity matching, this method can automatically find relevant cases and avoid the repetitive labor of manual retrieval. Models with similar shape features often have similar stress and deformation characteristics, and the correlation of the forming process is very high. Therefore, using shape similarity as the retrieval condition can quickly lock in potential best practices. Although the printing parameters cannot be directly copied, as first-hand information, they can make the starting point of optimization higher and the direction clearer.

[0172] Step S2200, analyze the first historical printing data to obtain the first printing deviation data of each similar historical printing task, where the first printing deviation data includes dimensional deviation data, shape deviation data, and surface quality deviation data;

[0173] Furthermore, as Figure 7 shown, step S2200 includes:

[0174] Step S2210, extract the dimensional parameters of each similar historical printing task in the first historical printing data, compare the dimensional parameters with the designed dimensional parameters of the historical 3D model corresponding to the similar historical printing task, and calculate the dimensional deviation data;

[0175] Step S2220, extract the shape parameters of each similar historical printing task in the first historical printing data, compare the shape parameters with the designed shape parameters of the historical 3D model corresponding to the similar historical printing task, and calculate the shape deviation data;

[0176] Step S2230, extract the surface quality parameters of each similar historical printing task in the first historical printing data to obtain the surface quality deviation data.

[0177] Specifically, the first printing deviation data reflects the difference between the actual effect of historical prints and the design intent of the 3D model. A large deviation indicates that there is still room for improvement in the printing scheme, while a small deviation indicates that the scheme is relatively mature. To quantitatively evaluate the printing deviation, it is necessary to examine it from three aspects: dimensional accuracy, shape accuracy, and surface quality. Dimensional deviation refers to the difference between the actual processed size and the dimension marked on the drawing, and is commonly measured by indicators such as runout and coaxiality. Tools such as coordinate measuring machines, vernier calipers, and micrometers can be used to measure the key dimensions of the printed part, and then compare them with the design dimensions to calculate the deviation value. For mating surfaces with high dimensional accuracy requirements, such as bearing holes and spigot surfaces, the dimensional deviation should be controlled within 0.05 mm. Shape deviation refers to the difference between the actual processed contour and the theoretical design contour, and is used to evaluate the shape conformity of the printed part and the 3D model. Structured light scanners, laser trackers, etc. can be used to obtain the point cloud data of the printed part, and then compare it with the CAD model to calculate the shape deviation. Generally, it is expressed by indicators such as Hausdorff distance and surface reconstruction error. For RV reducers, the gear profile deviation requirement is relatively high, and exceeding 0.03 mm will affect the meshing performance. Surface quality deviation refers to the gap between the actual surface roughness, color, texture and other characteristics and the design requirements. A profilometer and a roughness meter can be used to measure the surface profile curve and roughness parameters, or a spectrophotometer can be used to measure the surface color, and visual inspection can be used to check for defects such as lamination marks, reticulations, and wrinkles. The proportion of the defective area and the surface roughness grade are commonly used quantitative indicators. Especially in areas with high appearance requirements such as tooth surfaces and spigot surfaces, surface quality is particularly important. The above three types of deviation data respectively reflect the advantages and disadvantages of historical printing schemes from different aspects. Dimensional accuracy is related to part assembly and interchangeability, and is a basic element for measuring printing level; shape accuracy affects the force deformation of parts and the fit of kinematic pairs, and is a guarantee of mechanical performance; surface quality determines the appearance and feel, and is the beauty of the printed part. Only when all aspects meet the standards can it be considered a successful practice. By comparing the deviation data of each historical scheme and finding high-quality schemes with small deviations and good stability, a reliable reference system can be provided for the path design of new models.

[0178] Generally speaking, the first printing deviation data objectively records the actual effect of historical printing and is a direct basis for judging the quality of printing parameters. Similar models use similar parameters, and the printing results will naturally be similar. By analyzing the deviation data, high-quality parameter combinations can be identified, inheriting excellent practices while also learning from failures and taking fewer detours. Those parameter values with small deviations will naturally become the preferred targets for path correction. The systematic mining of deviation data makes the selection of parameters traceable and no longer a matter of luck, and the printing success rate and stability will surely be greatly improved. In addition, the accumulation of deviation data also provides data support for the continuous improvement of the printing process, helping enterprises summarize rules, precipitate knowledge, and build their own core process know-how.

[0179] Step S2300: Based on the first printing deviation data, correct the initial filling path data to generate the third filling path data.

[0180]

[0181] Where:

[0182] : The third filling path data represents the corrected path value at the coordinate position ;

[0183] : The initial filling path data represents the initial path value at the coordinate position ;

[0184] : The number of similar historical printing tasks with similar three-dimensional model structure characteristics to the current reducer;

[0185] : The dimensional deviation influence coefficient corresponds to the dimensional deviation in the th similar historical printing task, reflecting the influence of dimensional deviation on path correction, and the coefficient value is determined by those skilled in the art through multiple experiments or optimizations; 1 ≤ i ≤ n2;

[0186] : The dimensional deviation data of the th similar historical printing task;

[0187] : The shape deviation influence coefficient corresponds to the shape deviation in the th similar historical printing task; reflecting the weight of shape deviation on filling path correction, obtained by fitting or optimizing the historical data of shape deviation, usually determined by multi-dimensional least squares method;

[0188] : The shape deviation data of the th similar historical printing task;

[0189] : The surface quality deviation influence coefficient corresponds to the surface quality deviation in the th similar historical printing task, reflecting the influence of surface quality deviation on the final path correction, determined by fitting the deviation between surface quality and design data or through empirical data;

[0190] : The surface quality deviation data of the th similar historical printing task;

[0191] : Weight factor, measuring the The similarity between a similar historical printing task and the current printing task; generally, the higher the similarity, the greater the weight, and the influence of each historical printing task on the third filling path data is adjusted by the weight factor.

[0192] With deviation data The larger the deviation is during the historical printing process, the more significant the fill path correction is. The increase in will increase the impact of similar historical tasks on path correction. Therefore, if a historical task is very similar to the current model, the corresponding deviation correction will have a greater impact.

[0193] The formula combines the deviation information (including size, shape, and surface quality) of multiple similar historical printing tasks with the current initial filling path data, and uses weight factors to accurately adjust the impact of different deviations on path correction, thereby improving the accuracy of the printing path. By introducing multiple deviation coefficients and weights, the formula can dynamically adapt to the data of different historical tasks, ensuring that the corrected path data is closer to the ideal printing result and reducing printing errors.

[0194] When correcting the initial filling path, it is necessary to fine-tune the paths of each area through an iterative algorithm or a model based on historical data to compensate for these deviations. For example, if the shape deviation of a certain area is large, the path of the area can be encrypted, that is, the density of the printing layer can be increased to improve the local shape accuracy. In addition, "filling path" refers to the movement trajectory of the nozzle or laser when the structure is built layer by layer in 3D printing. Different areas can adopt different filling modes (such as spiral, honeycomb, etc.) to adapt to material and mechanical requirements. By correcting the initial path, the defects of the finished product caused by errors during the printing process can be effectively reduced, and the geometric accuracy, mechanical properties and surface quality of the final RV reducer structural parts can be guaranteed. The logical reasoning of this correction step is based on the feedback of prior data. By dynamically adjusting the filling strategy, the material distribution and path planning are further optimized, thereby improving the consistency of the finished product. For example, if a certain area often has surface unevenness problems in historical tasks, the corrected path will reduce the occurrence of similar problems by increasing density or adjusting the printing order. In summary, the core of step S2300 is to use the feedback of historical data to optimize the printing path through the correction algorithm, so as to ensure the high-precision molding of complex structural parts of the RV reducer.

[0195] Step S3000, in real time, collect the printing environment parameters and material property parameters during the 3D printing process, input the printing environment parameters and material property parameters into the pre-trained printing quality prediction model to obtain the second printing deviation data; compare the size of the second printing deviation data with the preset deviation threshold to determine whether it is necessary to trigger the path correction mechanism; if the path correction mechanism is triggered, then according to the second printing deviation data, dynamically adjust the third filling path data to generate the fourth filling path data; according to the fourth filling path data, execute the 3D printing process of the RV reducer complex structural part.

[0196] Furthermore, step S3000 includes:

[0197] Step S3100, during the 3D printing process of the RV reducer complex structural part, in real time, collect the printing environment parameters and material property parameters;

[0198] Specifically, the printing environment parameters mainly include temperature, humidity and air flow velocity. Too high a temperature will cause the material viscosity to decrease, the fluidity to increase, and collapse deformation to occur; too low a temperature will cause the material viscosity to increase, the fluidity to become poor, and extrusion discontinuity to occur. Too high humidity is likely to cause the material to absorb moisture and generate bubbles, affecting the forming quality. Air flow disturbance will cause the molten material to cool rapidly and generate warping deformation. Therefore, it is necessary to strictly control the printing environment. By installing temperature and humidity sensors and wind speed sensors in the printing cavity, the environment parameters can be collected in real time. The material property parameters mainly include melting temperature, fluidity and shrinkage rate. The melting temperature directly determines the setting of the printing temperature. Too high a temperature will degrade the material, and too low a temperature will make it difficult to melt. Fluidity is an important index to evaluate the forming performance of the material. Materials with good fluidity are more likely to fill complex structures. The shrinkage rate affects the dimensional accuracy of the final part. The larger the shrinkage rate, the more likely it is to generate deformation. By installing an infrared thermometer at the print head, the melting temperature of the material can be monitored in real time, with an accuracy of up to ±1°C. Install a capillary rheometer at the feeding place to test the melt flow rate (MFR) of the material. The larger the MFR, the better the fluidity of the material. In addition, an dilatometer can measure the linear expansion coefficient of the material when heated, from which the shrinkage rate of the material can be estimated. These sensors are connected to the feeding system of the 3D printer and upload the measured material parameters to the control system in real time. Collecting printing parameters in real time is the basis for dynamically optimizing the path. Only by accurately grasping the changes in materials and the environment can it be judged whether the current filling method is reasonable and provide a reliable basis for subsequent path correction. The collection accuracy requirements for environmental parameters and material parameters are relatively high. Appropriate sensors need to be selected and well integrated with the printing equipment to minimize the impact on the normal printing process. At the same time, signal processing circuits should be equipped for the sensors to complete filtering, amplification, A / D conversion, etc., and convert the analog signals into digital signals for the control system to read and analyze.

[0199] Step S3200: Input the printing environment parameters and material property parameters into a pre-trained printing quality prediction model to obtain second printing deviation data; the second printing deviation data includes second dimensional deviation data, second shape deviation data, and second surface quality deviation data.

[0200] Specifically, the printing quality prediction model uses the Gradient Boosting Decision Tree (GBDT) algorithm. GBDT can construct a powerful integrated prediction model by combining the prediction results of multiple weak learners (decision trees). It can flexibly handle various types of features (such as continuous and discrete types), and has strong non-linear modeling ability and anti-noise ability. First, collect a large amount of known printing parameters (such as printing speed, printing temperature, material type, etc.) and corresponding multi-dimensional deviation data (such as dimensional deviation, shape deviation, surface quality deviation) as training samples. Then, through iterative training of decision trees for multiple rounds, each round learns the residuals of the previous round to continuously approximate the true value of the overall prediction result. During the generation of each decision tree, the greedy algorithm is used to select the optimal feature and splitting point for node splitting to maximize the information gain. After all decision trees are generated, they are linearly combined to obtain the final strong learner, which is used as the prediction model for printing deviation. When new printing environment parameters and material property parameters are input, the GBDT model can comprehensively consider various factors and predict the most likely dimensional deviation, shape deviation, and surface quality deviation values. The GBDT prediction model can obtain the printing parameters collected by sensors in real time, quickly estimate the deviation level of the current filling path, and provide an important basis for subsequent path optimization. Through big data-driven deviation prediction, combined with expert experience and machine learning algorithms, it is possible to continuously accumulate and optimize 3D printing process knowledge, providing strong support for realizing intelligent production.

[0201] Step S3300: Compare the second printing deviation data with a preset deviation threshold to determine whether to trigger a path correction mechanism; the deviation threshold includes a dimensional deviation threshold Td, a shape deviation threshold Ts, and a surface quality deviation threshold Tq; the path correction mechanism includes dimensional deviation correction, shape deviation correction, and surface quality correction.

[0202] Further, as Figure 8 shown, Step S3300 includes:

[0203] Step S3310: If the second dimensional deviation data exceeds Td, trigger dimensional deviation correction.

[0204] Step S3320: If the second shape deviation data exceeds Ts, trigger shape deviation correction.

[0205] Step S3330: If the second surface quality deviation data exceeds Tq, trigger surface quality correction.

[0206] Specifically, after the print deviation prediction value (second print deviation data) is obtained, it is necessary to evaluate the severity of the deviation. It is necessary to trigger the path correction only when it exceeds a certain range. Therefore, a series of deviation thresholds should be set in advance as a criterion. The dimensional deviation threshold Td can refer to the dimensional tolerance band marked in the RV reducer drawing, such as ±0.1mm. If the predicted dimensional deviation exceeds the tolerance band, it means that the current filling path cannot meet the design requirements and needs to be adjusted. The shape deviation threshold Ts can refer to the form and position tolerance requirements of the RV reducer, such as flatness ≤0.05mm and roundness ≤0.03mm. If the predicted shape deviation exceeds the tolerance range, it means that the filling path may introduce large deformation and needs to be optimized. The surface quality threshold Tq generally refers to the 3D printing process specifications. Different printing methods and materials will have different roughness requirements. For example, the surface roughness requirement for PA12 printed by the selective laser sintering (SLS) process is Ra1.6. If the predicted surface defect level exceeds the roughness limit, it means that the filling method will affect the surface performance and needs to be improved. The setting of the threshold needs to take into account both printing efficiency and quality. If it is too high, corrections will be triggered frequently, extending the printing cycle; if it is too low, the purpose of defect control cannot be achieved. A set of threshold standards commonly used in the industry can be formulated based on process test results and production experience. In actual applications, the system will automatically compare the predicted deviation with the threshold. Once it is found to exceed the standard, an alarm will be issued to prompt the operator to intervene. At the same time, the deviation data and threshold conditions will be recorded in detail for technicians to analyze the reasons and improve the process parameters. The printing deviation early warning mechanism can avoid the accumulation of defects and correct unreasonable filling paths in time, thereby ensuring the yield rate of RV reducers. On the one hand, the early warning is based on objective deviation thresholds, and on the other hand, it is inseparable from the guidance of expert experience. Printing companies should establish a complete deviation response plan, formulate corresponding correction strategies for deviations caused by different reasons, and be targeted. Deviation early warning makes path optimization more forward-looking, rather than a last-ditch effort, and is an important part of ensuring the quality of 3D printing.

[0207] Step S3400: if the path correction mechanism is triggered, the third filling path data is dynamically adjusted according to the second printing deviation data to generate fourth filling path data; and the 3D printing process of the complex structural part of the RV reducer is executed according to the fourth filling path data;

[0208] Specifically, for dimensional deviation, it is mainly achieved by adjusting the filling spacing and angle. If dimensional deviation correction is triggered, the filling spacing and filling angle in the third filling path data are adjusted to generate the fourth filling path data; the smaller the filling spacing, the closer the dimension is to the design value; the smaller the included angle between the filling angle and the long side, the less the dimension shrinks; the filling spacing refers to the distance between adjacent filling lines. The smaller the spacing, the higher the filling rate, and the closer the dimension is to the design value, but the printing efficiency will decrease. To find a balance between efficiency and accuracy, a variable spacing strategy can be adopted, with a small spacing near the outer contour and a large spacing in the internal area. The filling angle refers to the included angle between the filling line and the long side of the part. Vertical filling will exacerbate the thermal expansion and contraction effect, resulting in dimensional shrinkage. To reduce shrinkage, filling at a small angle with the long side can be selected, such as 15° or 30°. For shape deviation, it is mainly achieved by optimizing the filling direction and form. If shape deviation correction is triggered, the filling direction and filling form in the third filling path data are adjusted to generate the fourth filling path data; when the filling direction is consistent with the stress direction, deformation can be reduced; for the filling form, grid filling or honeycomb filling can be selected to improve the support stiffness; the filling direction should be as consistent as possible with the stress direction, so that the filling line bears the tensile stress and deformation can be prevented. When it is difficult to match the filling direction with the stress direction, cross filling can be adopted, and the mutually perpendicular filling lines form a grid-like support. The filling form also affects shape stability. Common ones include linear filling, grid filling, honeycomb filling, etc. Linear filling has a low density and is prone to deformation; grid filling has a high strength but stress concentration; honeycomb filling has a high strength and uniform stress distribution. The reasonable filling form can be selected in combination with the stress characteristics.

[0209] Regarding surface quality issues, it is mainly achieved by changing the filling density and speed. If surface quality correction is triggered, the filling density and filling speed in the third filling path data are adjusted to generate the fourth filling path data; the higher the filling density, the denser the surface; the slower the filling speed, the smoother the surface. The filling density is usually expressed by the filling rate, that is, the percentage of the filling line in the cross-section. The higher the filling rate, the denser the surface and the fewer defects, but the longer the printing time. A gradient filling strategy can be adopted, with a high filling rate on the surface and a low filling rate inside, taking into account both surface quality and cost. The filling speed also affects the surface quality. The faster the speed, the thicker the filling line and the rougher the surface. To obtain a smooth surface, the filling speed can be reduced near the outer surface to improve the resolution. However, if the speed is too slow, it will cause poor welding, so a trade-off is needed. The fourth filling path comprehensively considers the real-time deviation feedback and can better adapt to the changes in the printing state. Dynamically adjusting the filling parameters is a complex optimization problem that requires balancing multiple objectives, such as accuracy, efficiency, strength, cost, etc. It can be modeled as a multi-objective optimization and solved using an evolutionary algorithm to search for the optimal combination in the parameter space. The optimization process should make full use of expert knowledge to design reasonable optimization variables and constraints. For example, the angle should not be too large and the speed should not exceed the equipment limit, etc. In addition, the sensitivity of the parameters should be evaluated to identify the key parameters and reduce the optimization dimension. Dynamic path optimization is a closed-loop feedback control that continuously corrects the filling parameters by monitoring the deviation in real time and finally converges to a better printing plan. The optimized path can adaptively adjust to compensate for various interferences during the printing process, making the part quality more stable and reliable.

[0210] Step S3500, repeatedly execute Step S3100 to Step S3400 until the RV reducer printing is completed;

[0211] Specifically, since 3D printing is a progressive manufacturing process, dynamic path optimization needs to be implemented throughout the printing process. After printing starts, it enters the real-time monitoring mode to continuously collect printing parameters. Every certain period of time (such as 1 minute), a path evaluation is triggered to predict the current deviation. If the deviation exceeds the threshold, a path correction program is started, the optimization algorithm is called to generate a new path, and it is converted into process parameters and issued. During the printing process, path optimization and printing execution are carried out synchronously, and the two interact continuously to form a dynamic feedback. After multiple iterations, the printing parameters will be continuously corrected until the deviation converges within the allowable range. This real-time closed-loop control can correct deviations in a timely manner, dynamically compensate for process errors, and avoid the spread of defects. The frequency of repeated optimization needs to be determined according to specific circumstances. Too high a frequency will increase the computational burden, and too low a frequency will make it difficult to capture the printing dynamics. Dynamic path optimization runs through the entire printing cycle of the RV reducer, making the printing process more transparent and controllable. By real-time monitoring the printing status, evaluating the deviation risk, and automatically adjusting the process parameters, the quality of the parts is made more stable and uniform. Dynamic optimization must strictly control the time overhead to avoid affecting the printing progress. Efficient algorithms can be used, such as the differential algorithm for image processing, to quickly extract deviation information. The optimization search is preferably carried out using heuristic methods, such as genetic algorithms, to avoid falling into local optima. The adjustment of process parameters should be a smooth transition to avoid oscillations caused by step changes. Dynamic path optimization makes 3D printing an intelligent manufacturing process. Through real-time perception, dynamic decision-making, and online execution, high-quality part manufacturing is finally completed. This is the development trend of modern advanced manufacturing technology, representing a major leap of 3D printing technology from experience to science, from offline to online, and from open-loop to closed-loop.

[0212] Step S3600, after printing is completed, post-process the RV reducer and conduct quality inspection.

[0213] Specifically, after the printing task is completed, post-processing and quality inspection of the RV reducer are also required. Post-processing mainly includes substrate removal, surface polishing, heat treatment, etc., which can improve the mechanical properties and appearance quality of parts. Common substrate removal methods include solvent immersion, ultrasonic cleaning, etc., which can quickly dissolve the support material. Surface polishing can be carried out by sandpaper grinding, sandblasting, etc. to improve the surface finish. Heat treatment such as annealing or quenching can eliminate residual stress and stabilize the dimensions. The post-processing effect directly affects the use performance of parts, and the process parameters need to be strictly controlled and incorporated into the quality control system. Quality inspection mainly includes dimensional inspection, shape inspection and performance inspection. Dimensional inspection can use a coordinate measuring machine, vernier caliper, etc. to measure the key dimensions of parts and compare them with the drawing tolerances. Shape inspection can use an optical scanner, etc. to measure the surface profile, flatness, etc. and analyze the shape deviation. Performance inspection can test the strength, hardness, conductivity, etc. of parts to evaluate the use performance. Correlating and analyzing the dynamically optimized path parameters with the inspection results can evaluate the optimization effect and correct the optimization direction. Summarize successful optimization cases as best practices and apply them to subsequent tasks. Analyze the reasons for failed optimization cases, learn lessons, and avoid repeating mistakes. Post-processing and quality inspection are important links to ensure the performance of the RV reducer and also important means to evaluate the dynamic optimization effect. Through closed-loop optimization, data can be continuously accumulated, the model can be improved, and the intelligent level of path design can be enhanced. The dynamic path optimization combines online inspection and offline optimization, runs through the entire product life cycle, and finally realizes the high-quality and high-efficiency manufacturing of the RV reducer. This is an important milestone in the application of 3D printing in the industrial field, marking the leap of 3D printing from prototype manufacturing to large-scale production.

[0214] Embodiment 2

[0215] On the basis of Embodiment 1, this embodiment provides a rapid prototyping system for complex structural parts of an RV reducer based on 3D printing technology, as Figure 9 shown, including:

[0216] Region segmentation module: used to receive the three-dimensional model data of the RV reducer, perform topological optimization on the three-dimensional model to generate the first filling path data; perform feature segmentation on the three-dimensional model of the RV reducer to obtain differentiated sub-regions;

[0217] Path filling module: used to fill each sub-region to generate the initial filling path of the complex structural parts of the RV reducer;

[0218] First correction module: used to obtain the first printing deviation data, and based on the first printing deviation data, correct the initial filling path data to generate the third filling path data; the first printing deviation data includes dimensional deviation data, shape deviation data and surface quality deviation data;

[0219] Second correction module: used to obtain second printing deviation data, compare the second printing deviation data with a preset deviation threshold, and determine whether to trigger a path correction mechanism; if the path correction mechanism is triggered, the third filling path data is dynamically adjusted according to the second printing deviation data to generate fourth filling path data; the 3D printing process of the RV reducer complex structural part is executed according to the fourth filling path data.

[0220] In the region segmentation module, the topological optimization of the three-dimensional model to generate the first filling path data includes:

[0221] Step S1110, based on the three-dimensional model data of the RV reducer, establish a topological optimization mathematical model;

[0222] Step S1120, solve the topological optimization mathematical model to obtain the optimized filling material distribution information, and generate the first filling path data according to the filling material distribution information.

[0223] In the region segmentation module, the feature segmentation of the three-dimensional model of the RV reducer to obtain differentiated sub-regions includes:

[0224] Step S1210, based on the three-dimensional model data of the RV reducer, establish a three-dimensional finite element model of the RV reducer, perform structural mechanics analysis on the reducer under rated working conditions to obtain a stress nephogram and a displacement nephogram; the three-dimensional finite element model contains n1 nodes;

[0225] Step S1220, extract the high stress area from the stress nephogram;

[0226] Step S1230, extract the deformation concentration area from the displacement nephogram;

[0227] Step S1240, perform wall thickness analysis on the three-dimensional model of the RV reducer, and extract the area with a wall thickness greater than the preset wall thickness threshold T thick as the wall thickness area;

[0228] Step S1250, obtain sub-regions A, B, C, and D according to the high stress area, deformation concentration area, and wall thickness area.

[0229] In the path filling module, the filling of each sub-region to generate the initial filling path of the RV reducer complex structural part includes:

[0230] Step S1310, adopt a high-strength filling mode including but not limited to cross and spiral for sub-region A, and a filling density ρ1, generate the filling path of sub-region A, and obtain the filling path data of sub-region A;

[0231] Step S1320: Adopt a high-stiffness filling pattern including but not limited to honeycomb and hexagonal grids, and a filling density ρ2 for sub-region B to generate a filling path for sub-region B and obtain the filling path data of sub-region B;

[0232] Step S1330: Adopt a low filling pattern including but not limited to triangular and square grids, and a filling density ρ3 for sub-region C to generate a filling path for sub-region C and obtain the filling path data of sub-region C;

[0233] Step S1340: Adopt a filling density ρ4 for sub-region D to generate a filling path for sub-region D and obtain the filling path data of sub-region D; where ρ2 > ρ1 > ρ3 > ρ4;

[0234] Step S1350: Combine the filling path data of each sub-region to generate second filling path data;

[0235] Step S1400: Fuse the first filling path data and the second filling path data to generate an initial filling path for the complex structural part of the RV reducer.

[0236] The first correction module includes:

[0237] The first deviation acquisition unit: used to obtain n2 similar historical printing tasks with structural features similar to the current 3D model of the RV reducer, extract the printing parameters of the n2 similar historical printing tasks as the first historical printing data; n2 is a positive integer; analyze the first historical printing data to obtain the first printing deviation data of each similar historical printing task, and the first printing deviation data includes dimensional deviation data, shape deviation data, and surface quality deviation data;

[0238] The initial path correction unit: based on the first printing deviation data, correct the initial filling path data to generate third filling path data.

[0239] In the first deviation acquisition unit, the obtaining of n2 similar historical printing tasks with structural features similar to the current 3D model of the RV reducer includes:

[0240] Step S2110: Compare the 3D models of all historical printing tasks with the current 3D model of the RV reducer and calculate the shape feature similarity;

[0241] Step S2120: Extract the historical printing tasks with a shape feature similarity greater than the preset similarity threshold T sim to the current 3D model of the RV reducer, and mark them as similar historical printing tasks.

[0242] In the first deviation acquisition unit, the analyzing of the first historical printing data to obtain the first printing deviation data of each similar historical printing task includes:

[0243] Step S2210: Extract the dimensional parameters of each similar historical printing task in the first historical printing data, compare the dimensional parameters with the designed dimensional parameters of the corresponding historical 3D model of the similar historical printing task, and calculate the dimensional deviation data.

[0244] Step S2220: Extract the shape parameters of each similar historical printing task in the first historical printing data, compare the shape parameters with the designed shape parameters of the corresponding historical 3D model of the similar historical printing task, and calculate the shape deviation data.

[0245] Step S2230: Extract the surface quality parameters of each similar historical printing task in the first historical printing data to obtain the surface quality deviation data.

[0246] The second correction module includes:

[0247] The second deviation acquisition unit: During the 3D printing process of the complex structural parts of the RV reducer, it is used to collect the printing environment parameters and material property parameters in real time; input the printing environment parameters and material property parameters into the pre-trained printing quality prediction model to obtain the second printing deviation data; the second printing deviation data includes the second dimensional deviation data, the second shape deviation data, and the second surface quality deviation data.

[0248] The third path correction unit: It is used to compare the second printing deviation data with the preset deviation threshold to determine whether to trigger the path correction mechanism; the deviation threshold includes the dimensional deviation threshold Td, the shape deviation threshold Ts, and the surface quality deviation threshold Tq; the path correction mechanism includes dimensional deviation correction, shape deviation correction, and surface quality correction; if the path correction mechanism is triggered, the third filling path data is dynamically adjusted according to the second printing deviation data to generate the fourth filling path data; according to the fourth filling path data, the 3D printing process of the complex structural parts of the RV reducer is executed.

[0249] Embodiment 3

[0250] This embodiment discloses an electronic device, which may include one or more processors and one or more memories. Among them, computer-readable code is stored in the memory, and when the computer-readable code is run by one or more processors, it can execute the rapid prototyping method for the complex structural parts of the RV reducer based on 3D printing technology as described above.

[0251] The method or system according to the embodiments of the present application can also be implemented by means of the architecture of the electronic device disclosed in the present application. The electronic device may include a bus, one or more CPUs, a read-only memory (ROM), a random access memory (RAM), a communication port connected to a network, input / output components, a hard disk, etc. The storage device in the electronic device, such as the ROM or the hard disk, can store the method for rapid prototyping of complex structural parts of an RV reducer based on 3D printing technology provided by the present application. The method for rapid prototyping of complex structural parts of an RV reducer based on 3D printing technology may, for example, include: receiving three-dimensional model data of the RV reducer, performing topological optimization on the three-dimensional model to generate first filling path data; performing feature segmentation on the three-dimensional model of the RV reducer to obtain differentiated sub-regions; filling each sub-region to generate an initial filling path of the complex structural parts of the RV reducer; obtaining first printing deviation data, and based on the first printing deviation data, correcting the initial filling path data to generate third filling path data; the first printing deviation data includes dimension deviation data, shape deviation data, and surface quality deviation data; obtaining second printing deviation data, comparing the second printing deviation data with a preset deviation threshold to determine whether to trigger a path correction mechanism; if the path correction mechanism is triggered, then dynamically adjusting the third filling path data according to the second printing deviation data to generate fourth filling path data; performing a 3D printing process of the complex structural parts of the RV reducer according to the fourth filling path data.

[0252] Further, the electronic device may further include a user interface. Of course, the architecture disclosed in the present invention is only exemplary. When implementing different devices, one or more components of the electronic device disclosed in the present invention may be omitted according to actual needs.

[0253] Embodiment 4

[0254] This embodiment discloses a computer-readable storage medium, on which computer-readable instructions are stored. When the computer-readable instructions are run by a processor, the method for rapid prototyping of complex structural parts of an RV reducer based on 3D printing technology according to the embodiments of the present application described with reference to the above drawings can be executed. The storage medium includes, but is not limited to, for example, volatile memory and / or non-volatile memory. Volatile memory may, for example, include random access memory (RAM) and cache memory, etc. Non-volatile memory may, for example, include read-only memory (ROM), hard disk, flash memory, etc.

[0255] In addition, according to the embodiments of the present application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, the present application provides a non-transitory machine-readable storage medium storing machine-readable instructions that can be run by a processor to execute instructions corresponding to the method steps provided by the present application. For example: receiving three-dimensional model data of an RV reducer, performing topology optimization on the three-dimensional model to generate first filling path data; performing feature segmentation on the three-dimensional model of the RV reducer to obtain differentiated sub-regions; filling each sub-region to generate an initial filling path of the complex structural part of the RV reducer; obtaining first printing deviation data, and based on the first printing deviation data, correcting the initial filling path data to generate third filling path data; the first printing deviation data includes dimension deviation data, shape deviation data, and surface quality deviation data; obtaining second printing deviation data, comparing the second printing deviation data with a preset deviation threshold to determine whether to trigger a path correction mechanism; if the path correction mechanism is triggered, dynamically adjusting the third filling path data according to the second printing deviation data to generate fourth filling path data; performing a 3D printing process on the complex structural part of the RV reducer according to the fourth filling path data. When the computer program is executed by a central processing unit (CPU), the above functions defined in the method of the present application are executed.

[0256] The methods, systems, and devices of the present application can be implemented in many ways. For example, the methods, systems, and devices of the present application can be implemented by software, hardware, firmware, or any combination of software, hardware, and firmware. The above order of steps for the method is only for illustration, and the steps of the method of the present application are not limited to the specific order described above, unless otherwise specifically stated. In addition, in some embodiments, the present application can also be implemented as a program recorded in a recording medium, and these programs include machine-readable instructions for implementing the method according to the present application. Therefore, the present application also covers a recording medium storing a program for executing the method according to the present application.

[0257] In addition, parts of the above technical solutions provided in the embodiments of the present application that are consistent with the implementation principles of the corresponding technical solutions in the prior art are not described in detail to avoid excessive elaboration.

[0258] As described in the specific embodiments above, the purpose, technical solutions, and beneficial effects of the present invention have been further described in detail. It should be understood that the above is only the specific embodiments of the present invention and is not used to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included in the protection scope of the present invention.

Claims

1. A rapid prototyping method for complex structural parts of RV reducers based on 3D printing technology, characterized in that, The method includes: Receiving the three-dimensional model data of the RV reducer, performing topology optimization on the three-dimensional model to generate the first filling path data; performing feature segmentation on the three-dimensional model of the RV reducer to obtain differentiated sub-regions; filling each sub-region to generate the initial filling path of the complex structural parts of the RV reducer; the differentiated sub-regions include sub-region A, sub-region B, sub-region C, and sub-region D; The three-dimensional model of the RV reducer is segmented by features to obtain differentiated sub-regions, including: establishing a three-dimensional finite element model of the RV reducer, performing structural mechanics analysis on the RV reducer under rated working conditions to obtain stress nephograms and displacement nephograms; extracting high-stress regions from the stress nephograms and deformation concentration regions from the displacement nephograms; performing wall thickness analysis on the three-dimensional model of the RV reducer, and extracting the regions with wall thickness greater than the preset wall thickness threshold T thick as wall thickness regions; according to the high-stress regions, deformation concentration regions and wall thickness regions, obtaining sub-region A, sub-region B, sub-region C and sub-region D; Obtaining the first printing deviation data, and based on the first printing deviation data, correcting the initial filling path data to generate the third filling path data; the first printing deviation data includes dimensional deviation data, shape deviation data, and surface quality deviation data; The obtaining of the first printing deviation data includes: comparing the 3D models of all historical printing tasks with the current 3D model of the RV reducer, and calculating the shape feature similarity; extracting the historical printing tasks with the shape feature similarity to the current 3D model of the RV reducer greater than the preset similarity threshold T sim marking them as similar historical printing tasks; extracting the printing parameters of the similar historical printing tasks as the first historical printing data; extracting the dimension parameters of each similar historical printing task in the first historical printing data, comparing the dimension parameters with the designed dimension parameters of the historical 3D model corresponding to the similar historical printing task, and calculating the dimension deviation data; extracting the shape parameters of each similar historical printing task in the first historical printing data, comparing the shape parameters with the designed shape parameters of the historical 3D model corresponding to the similar historical printing task, and calculating the shape deviation data; extracting the surface quality parameters of each similar historical printing task in the first historical printing data to obtain the surface quality deviation data; Obtaining the second printing deviation data, comparing the size of the second printing deviation data with a preset deviation threshold to determine whether to trigger the path correction mechanism; if the path correction mechanism is triggered, dynamically adjust the third filling path data according to the second printing deviation data to generate the fourth filling path data; execute the 3D printing process of the complex structural parts of the RV reducer according to the fourth filling path data; The obtaining of the second printing deviation data includes: during the 3D printing process of the complex structural parts of the RV reducer, collecting the printing environment parameters and material property parameters in real time; inputting the printing environment parameters and material property parameters into a pre-trained printing quality prediction model to obtain the second printing deviation data; the second printing deviation data includes the second dimensional deviation data, the second shape deviation data, and the second surface quality deviation data.

2. The rapid prototyping method for complex structural parts of an RV reducer based on 3D printing technology according to claim 1, wherein The obtaining of sub-region A, sub-region B, sub-region C, and sub-region D according to the high stress area, deformation concentration area, and wall thickness area includes: Subtracting the deformation concentration area and the thick wall area from the high stress area to obtain sub-region A; Subtracting the high stress area and the thick wall area from the deformation concentration area to obtain sub-region B; Subtracting the high stress area and the deformation concentration area from the thick wall area to obtain sub-region C; Taking the area outside sub-region A, sub-region B, and sub-region C as sub-region D.

3. The rapid prototyping method for complex structural parts of an RV reducer based on 3D printing technology according to claim 2, wherein The filling of each sub-region to generate the initial filling path of the complex structural parts of the RV reducer includes: Adopting a high-strength filling mode including cross-cross and spiral lines and a filling density ρ1 for sub-region A to generate the filling path of sub-region A and obtain the filling path data of sub-region A; Adopting a high-stiffness filling mode including honeycomb and hexagonal grids and a filling density ρ2 for sub-region B to generate the filling path of sub-region B and obtain the filling path data of sub-region B; Adopting a low-filling mode including triangles and square grids and a filling density ρ3 for sub-region C to generate the filling path of sub-region C and obtain the filling path data of sub-region C; Adopting a filling density ρ4 for sub-region D to generate the filling path of sub-region D and obtain the filling path data of sub-region D; where ρ2 > ρ1 > ρ3 > ρ4; Combining the filling path data of each sub-region to generate the second filling path data; Fusing the first filling path data and the second filling path data to generate the initial filling path of the complex structural parts of the RV reducer.

4. The rapid prototyping method for complex structural parts of RV reducers based on 3D printing technology according to claim 1, wherein The deviation thresholds include a dimensional deviation threshold Td, a shape deviation threshold Ts, and a surface quality deviation threshold Tq; The path correction mechanism includes dimensional deviation correction, shape deviation correction, and surface quality correction; Comparing the second printing deviation data with the preset deviation thresholds to determine whether to trigger the path correction mechanism includes: If the second dimensional deviation data exceeds Td, trigger dimensional deviation correction; If the second shape deviation data exceeds Ts, trigger shape deviation correction; If the second surface quality deviation data exceeds Tq, trigger surface quality correction.

5. A rapid prototyping system for complex structural parts of an RV reducer based on 3D printing technology, which is used to implement the rapid prototyping method for complex structural parts of an RV reducer based on 3D printing technology according to any one of claims 1-4, characterized in that, The system includes: A region segmentation module: configured to receive the three-dimensional model data of the RV reducer, perform topological optimization on the three-dimensional model to generate first filling path data; perform feature segmentation on the three-dimensional model of the RV reducer to obtain differentiated sub-regions; A path filling module: configured to fill each sub-region to generate an initial filling path for the complex structural parts of the RV reducer; A first correction module: configured to obtain first printing deviation data, and based on the first printing deviation data, correct the initial filling path data to generate third filling path data; the first printing deviation data includes dimensional deviation data, shape deviation data, and surface quality deviation data; A second correction module: configured to obtain second printing deviation data, compare the second printing deviation data with the preset deviation thresholds to determine whether to trigger the path correction mechanism; if the path correction mechanism is triggered, dynamically adjust the third filling path data according to the second printing deviation data to generate fourth filling path data; perform the 3D printing process of the complex structural parts of the RV reducer according to the fourth filling path data.

6. An electronic device, comprising a memory, a central processing unit, and a computer program stored on the memory and executable on the central processing unit, characterized in that When the central processing unit executes the computer program, it implements the rapid prototyping method for complex structural parts of an RV reducer based on 3D printing technology according to any one of claims 1-4.

7. A computer-readable storage medium, characterized in that, A computer-readable storage medium stores a computer program, and when the computer program is executed, it implements the rapid prototyping method for complex structural parts of an RV reducer based on 3D printing technology according to any one of claims 1-4.

Images