A method and system for matching process parameters for additive manufacturing

Abstract

The application provides an additive manufacturing process parameter matching method and system, and relates to the technical field of metal additive manufacturing.The three-dimensional model of a part to be manufactured is introduced and converted into point cloud data, and standardized point cloud data is obtained through decentralization, scale normalization and uniform sampling; the local geometric features and global topological features of the data are extracted, and the entity structure of the three-dimensional model is partitioned; process parameters are matched in a mapping database based on the structure type, and when there are multiple groups of parameters, the optimal parameters are selected through a forming quality evaluation function, and the scanning path strategy is one-to-one corresponding to the structure type; the parameters are converted into control instructions to generate a process configuration file.The application realizes accurate matching of structure-level parameters, reduces manual intervention, improves forming quality and manufacturing stability, completes the automation of the closed loop from design to manufacturing, and is suitable for different additive manufacturing equipment.

Description

Technical Field

[0001] This invention relates to the field of metal additive manufacturing technology, and in particular to a method and system for matching additive manufacturing process parameters. Background Technology

[0002] Metal additive manufacturing, as a core technology of advanced intelligent manufacturing, has been widely used in the rapid prototyping of complex metal components for high-end equipment. Laser wire deposition and arc wire deposition, among other processes, have become mainstream technologies due to their high forming efficiency and material utilization. However, as part structures become increasingly complex, with thin walls, curved surfaces, and solids coexisting, the requirements for heat input, scanning strategies, energy density, and other process conditions vary significantly across different regions. Traditional uniform parameter settings struggle to balance overall forming quality and efficiency. Therefore, achieving precise matching between process parameters and structural characteristics has become a pressing issue for the industry.

[0003] Currently, the industry typically uses a combination of manual experience-based parameter adjustment and trial-and-error printing to address the forming requirements of complex structures. Technicians manually set parameters such as laser power, scanning speed, and wire feeding speed based on the part's structural type and their own engineering experience. They then verify and adjust the parameter combinations through multiple trial prints until the forming quality requirements are met, thereby reducing the impact of structural differences on the forming effect.

[0004] This type of manually-driven parameter setting method lacks a systematic mapping mechanism between structural partitions and process parameters. It cannot automatically identify geometric features and segment regions of the 3D model, cannot output adapted and differentiated process parameters for different structural regions, and cannot directly generate control commands that the equipment can execute. This results in long parameter tuning cycles, high trial and error costs, and a high degree of reliance on manual labor, making it difficult to achieve full automation of the additive manufacturing process from design to manufacturing. Summary of the Invention

[0005] This invention provides a method and system for matching additive manufacturing process parameters, which can output adapted and differentiated process parameters for different structural regions, thereby generating control commands that can be executed by the equipment, effectively reducing the parameter tuning cycle, lowering trial and error costs, and realizing full automation of the additive manufacturing process from design to manufacturing.

[0006] To achieve the above objectives, the embodiments of the present invention adopt the following technical solutions: In a first aspect, a method for matching additive manufacturing process parameters is provided, comprising: importing a three-dimensional model of a part to be manufactured and converting the three-dimensional model into point cloud data; preprocessing the point cloud data to obtain standardized point cloud data, wherein the standardized point cloud data has uniform size and a preset number of points; extracting geometric feature information based on the standardized point cloud data, and performing structural segmentation on the three-dimensional model according to the geometric feature information to obtain structural partitioning results, wherein the geometric feature information includes local geometric features and global topological features, and the structural partitioning results are used to characterize the structural types corresponding to different regions included in the three-dimensional model; determining the additive manufacturing process parameters corresponding to different regions included in the three-dimensional model based on the structural partitioning results; and converting the additive manufacturing process parameters corresponding to different regions included in the three-dimensional model into control instructions executable by the equipment to generate a process configuration file.

[0007] The method provided by this invention imports a 3D model of the part to be manufactured and converts it into point cloud data. After preprocessing, standardized point cloud data with uniform size and a preset number of points is obtained. Based on this data, local geometric features and global topological features are extracted and structural segmentation is performed. Finally, the additive manufacturing process parameters for each region are determined and control instruction files are generated. This solves the problem of poor compatibility between process parameters and model structure in traditional additive manufacturing, achieves accurate matching of process parameters, and effectively improves the forming quality and manufacturing efficiency of the parts.

[0008] In one possible implementation of the first aspect, the point cloud data is preprocessed to obtain standardized point cloud data, including: calculating the geometric centroid of the point cloud data and translating all points to coordinate positions with the centroid as the origin to obtain decentralized point cloud data; scaling the decentralized point cloud data to a unit sphere to obtain scale-normalized point cloud data; and performing uniform sampling on the scale-normalized point cloud data to obtain standardized point cloud data with a preset number of points.

[0009] The method provided by this invention achieves decentralization by calculating the geometric centroid of point cloud data and translating it to a coordinate position with the centroid as the origin. Then, it is scaled to a unit sphere to complete scale normalization. Finally, standardized point cloud data with a preset number of points is obtained through uniform sampling. This preprocessing process eliminates the influence of position offset, scale difference and data redundancy of the original point cloud data, ensuring the accuracy of subsequent feature extraction and structure segmentation, and providing a high-quality data foundation for subsequent process parameter matching.

[0010] In one possible implementation of the first aspect, the structure type includes ultrathin-walled structures, standard thin-walled structures, and thick-walled solid structures; the step of extracting geometric feature information based on standardized point cloud data and performing structural segmentation on the three-dimensional model according to the geometric feature information to obtain structural partitioning results includes: extracting local and global features from the standardized point cloud data to obtain geometric feature information including local geometric features and global topological features; and dividing the three-dimensional model into regions according to the geometric feature information to obtain structural partitioning results.

[0011] The method provided by this invention clearly defines the structural types, including ultra-thin-walled structures, standard thin-walled structures, and thick-walled solid structures. By extracting local and global features from standardized point cloud data, and then dividing the three-dimensional model into regions based on the extracted geometric feature information, the method achieves accurate identification and division of different structural types of the three-dimensional model. This avoids errors in process parameter adaptation caused by confusion between different structural regions and provides a precise structural basis for matching exclusive process parameters to each region.

[0012] In one possible implementation of the first aspect, the extraction of local and global features from standardized point cloud data includes: performing initial feature mapping on the standardized point cloud data using a multilayer perceptron to obtain initial features; processing the initial features using multi-level downsampling and K-nearest neighbor construction to extract multi-scale geometric features; introducing a self-attention mechanism to weight and enhance the multi-scale geometric features to obtain enhanced key manufacturing features; and performing feature fusion on the enhanced key manufacturing features through progressive upsampling to obtain geometric feature information including local geometric features and global topological features.

[0013] The method provided by this invention obtains initial features by performing initial feature mapping on standardized point cloud data using a multilayer perceptron. It then extracts multi-scale geometric features using multi-level downsampling and K-nearest neighbor construction. After weighted enhancement by a self-attention mechanism, it obtains geometric feature information containing local geometric features and global topological features through step-by-step upsampling and fusion. This feature extraction process preserves local detailed features while taking into account global structural features, and enhances the representation ability of key manufacturing features. It significantly improves the completeness and effectiveness of geometric feature information, providing strong support for the accuracy of structural segmentation.

[0014] In one possible implementation of the first aspect, the additive manufacturing process parameters include laser power, scanning speed, layer thickness, scanning path strategy, and wire feed speed; determining the additive manufacturing process parameters corresponding to different regions included in the three-dimensional model based on the structural partitioning results includes: matching the combination of process parameters corresponding to each region included in the three-dimensional model in a mapping database according to the structural type in the structural partitioning results, wherein the mapping database pre-stores additive manufacturing process parameters corresponding to each of multiple structural types.

[0015] The method provided by this invention specifies additive manufacturing process parameters including laser power, scanning speed, layer thickness, scanning path strategy, and wire feed speed. By matching the process parameter combination of the corresponding region in the structure type in the structure partitioning result with the structure type in the preset mapping database, and the mapping database pre-stores the process parameters corresponding to each structure type, it realizes the rapid and accurate matching of process parameters and structure type, avoids the subjectivity and error of manual parameter setting, and improves the efficiency and accuracy of process parameter determination.

[0016] In one possible implementation of the first aspect, based on the structural type in the structural partitioning results, the process parameter combination corresponding to each region included in the three-dimensional model is matched in the mapping database, including: when the structural type corresponds to only one set of process parameter combinations in the mapping database, the process parameter combination is determined as the target process parameter of the region; when the structural type corresponds to multiple sets of process parameter combinations in the mapping database, a comprehensive score of each set of process parameter combinations is calculated based on a preset forming quality evaluation function with forming surface finish, forming dimensional accuracy, melt pool stability and deposition efficiency as weighting factors, and the process parameter combination with the highest comprehensive score is selected as the target process parameter of the region.

[0017] The method provided by this invention directly determines the target process parameters when the structure type corresponds to only one set of process parameter combinations. When multiple sets of process parameter combinations are involved, the combination with the highest comprehensive score is selected based on a preset forming quality evaluation function that uses surface finish, dimensional accuracy, melt pool stability, and deposition efficiency as weighting factors. This approach ensures high efficiency in single-parameter combination scenarios and achieves optimal selection in multi-parameter combination scenarios, further improving the adaptability of process parameters to structure types and ensuring a balance between part forming quality and manufacturing efficiency.

[0018] In one possible implementation of the first aspect, the scanning path strategy corresponds one-to-one with the structure type, specifically including: when the structure type is an ultra-thin-walled structure, the scanning path strategy adopts single-track oscillating scanning; when the structure type is a standard thin-walled structure, the scanning path strategy adopts ring-shaped offset scanning; when the structure type is a thick-walled solid structure, the scanning path strategy adopts cross-fill scanning and contour scanning.

[0019] The method provided by this invention clearly defines the one-to-one correspondence between the scanning path strategy and the structure type. The ultra-thin wall structure adopts single-track swing scanning, the standard thin wall structure adopts ring offset scanning, and the thick wall solid structure adopts cross-fill scanning and contour scanning. This achieves precise adaptation of the scanning path strategy to different structure types, avoids the problem that a single scanning path cannot adapt to multiple structure types, effectively improves the forming accuracy and surface quality of each structural region, and reduces forming defects.

[0020] In one possible implementation of the first aspect, the step of converting the matched process parameters into control instructions executable by the equipment and generating a process configuration file includes: converting the matched process parameters into corresponding control instruction statements according to the control protocol of the additive manufacturing equipment; dynamically switching the control instructions of the corresponding regions according to the structural partitions and printing layer sequence of the three-dimensional model to form a continuous instruction sequence; and generating a process configuration file in a preset format based on the instruction sequence, wherein the preset format is a G-code file format or an XML format.

[0021] The method provided by this invention converts the matched process parameters into control instruction statements according to the control protocol of the additive manufacturing equipment. It dynamically switches the control instructions according to the structural partition and printing layer sequence to form a continuous instruction sequence, and finally generates a process configuration file in G code or XML format. This achieves seamless conversion of process parameters into executable instructions of the equipment. Moreover, the dynamic switching of instructions adapts to the differences in process parameters in different regions, ensuring the accurate execution of process parameters during the printing process and guaranteeing the consistency of part forming.

[0022] In one possible implementation of the first aspect, after structurally segmenting the 3D model based on the geometric feature information to obtain structural partitioning results, the method further includes: determining whether the structural partitioning results meet preset manufacturing constraints, wherein the preset manufacturing constraints are that the 3D model is structurally partitioned across the entire domain, the boundaries of each structural region are continuous, and the identification confidence is greater than a preset threshold; if the structural partitioning results do not meet the preset manufacturing constraints, the 3D model is re-segmented until structural partitioning results that meet the preset manufacturing constraints are obtained.

[0023] The method provided by this invention, after obtaining the structural partitioning results, determines whether they meet the constraints of completing the structural partitioning of the entire 3D model, ensuring the continuity of the boundaries of each structural region, and having an identification confidence level greater than a preset threshold. If the constraints are not met, the structural partitioning is re-performed. This verification and iterative optimization process avoids partitioning results with omissions, discontinuous boundaries, or low identification confidence levels from affecting subsequent process parameter matching, ensuring the reliability of the structural partitioning results and providing dual protection for subsequent process parameter matching and part forming quality.

[0024] Secondly, the present invention provides an additive manufacturing process parameter matching system, comprising: a point cloud conversion module for importing a three-dimensional model of a part to be manufactured and converting the three-dimensional model into point cloud data; a point cloud processing module for preprocessing the point cloud data to obtain standardized point cloud data, wherein the standardized point cloud data has a uniform size and a preset number of points; a structure segmentation module for extracting geometric feature information based on the standardized point cloud data and performing structural segmentation on the three-dimensional model according to the geometric feature information to obtain structural partitioning results, wherein the geometric feature information includes local geometric features and global topological features, and the structural partitioning results are used to characterize the structural types corresponding to different regions included in the three-dimensional model; a parameter determination module for determining the additive manufacturing process parameters corresponding to different regions included in the three-dimensional model based on the structural partitioning results; and a file generation module for converting the additive manufacturing process parameters corresponding to different regions included in the three-dimensional model into control instructions executable by the equipment and generating a process configuration file.

[0025] Thirdly, an electronic device is provided, the electronic device including a memory and one or more processors; the memory is coupled to the processors; wherein the memory stores computer program code, the computer program code including computer instructions, which, when executed by the processor, cause the electronic device to perform the method as described in any implementation of the first aspect.

[0026] Fourthly, a computer-readable storage medium is provided, including computer instructions that, when executed on an electronic device, cause the electronic device to perform a method as described in any implementation of the first aspect.

[0027] Fifthly, a computer program product is provided that, when run on a computer, causes the computer to perform the method in any implementation of the first aspect.

[0028] Understandably, the beneficial effects achieved by the system of the second aspect, the electronic device of the third aspect, the computer-readable storage medium of the fourth aspect, and the computer program product of the fifth aspect provided above can be referred to with reference to the beneficial effects of the first aspect and any of its possible design embodiments, which will not be repeated here. Attached Figure Description

[0029] Figure 1 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention; Figure 2 A flowchart of an additive manufacturing process parameter matching method provided in an embodiment of the present invention; Figure 3 A schematic diagram of a three-dimensional model of a part to be manufactured, provided for an embodiment of the present invention; Figure 4 A schematic diagram of a three-dimensional model of a pre-processed part to be manufactured, provided in an embodiment of the present invention; Figure 5 This is a schematic diagram of a three-dimensional model of a part to be manufactured after being segmented, as provided in an embodiment of the present invention. Figure 6 This is a schematic diagram of a geometric feature information extraction process provided in an embodiment of the present invention; Figure 7 A schematic diagram illustrating the matching process of additive manufacturing process parameters provided in an embodiment of the present invention; Figure 8 This is a schematic diagram of a matching system provided in an embodiment of the present invention. Detailed Implementation

[0030] The technical solutions of the embodiments of the present invention will be described below with reference to the accompanying drawings. In the description of the present invention, unless otherwise stated, " / " indicates that the objects before and after are in an "or" relationship. For example, A / B can represent A or B. The "or" in the present invention is merely a description of the relationship between the related objects, indicating that three relationships can exist. For example, A or B can represent: A alone, A and B simultaneously, and B alone. A and B can be singular or plural. Furthermore, in the description of the present invention, unless otherwise stated, "multiple" refers to two or more. "At least one of the following" or similar expressions refer to any combination of these items, including any combination of single or plural items.

[0031] Furthermore, to facilitate a clear description of the technical solutions of the embodiments of the present invention, the terms "first" and "second" are used in the embodiments of the present invention to distinguish identical or similar items with essentially the same function and effect. Those skilled in the art will understand that the terms "first" and "second" do not limit the quantity or execution order, and that the terms "first" and "second" are not necessarily different.

[0032] In this embodiment of the invention, the terms "exemplary" or "for example" are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" or "for example" in this embodiment of the invention should not be construed as superior or more advantageous than other embodiments or designs. Specifically, the use of terms such as "exemplary" or "for example" is intended to present the relevant concepts in a concrete manner for ease of understanding.

[0033] Metal additive manufacturing, as a core technology of advanced intelligent manufacturing, has been widely used in the rapid prototyping of complex metal components for high-end equipment. Laser wire deposition and arc wire deposition, among other processes, have become mainstream technologies due to their high forming efficiency and material utilization. However, as part structures become increasingly complex, with thin walls, curved surfaces, and solids coexisting, the requirements for heat input, scanning strategies, energy density, and other process conditions vary significantly across different regions. Traditional uniform parameter settings struggle to balance overall forming quality and efficiency. Therefore, achieving precise matching between process parameters and structural characteristics has become a pressing issue for the industry.

[0034] Currently, the industry typically uses a combination of manual experience-based parameter adjustment and trial-and-error printing to address the forming requirements of complex structures. Technicians manually set parameters such as laser power, scanning speed, and wire feeding speed based on the part's structural type and their own engineering experience. They then verify and adjust the parameter combinations through multiple trial prints until the forming quality requirements are met, thereby reducing the impact of structural differences on the forming effect.

[0035] This type of manually-driven parameter setting method lacks a systematic mapping mechanism between structural partitions and process parameters. It cannot automatically identify geometric features and segment regions of the 3D model, cannot output adapted and differentiated process parameters for different structural regions, and cannot directly generate control commands that the equipment can execute. This results in long parameter tuning cycles, high trial and error costs, and a high degree of reliance on manual labor, making it difficult to achieve full automation of the additive manufacturing process from design to manufacturing.

[0036] In view of this, embodiments of the present invention provide an additive manufacturing process parameter matching method and system. The method includes: importing a three-dimensional model of a part to be manufactured and converting the three-dimensional model into point cloud data; preprocessing the point cloud data to obtain standardized point cloud data, wherein the standardized point cloud data has uniform size and a preset number of points; extracting geometric feature information based on the standardized point cloud data, and performing structural segmentation on the three-dimensional model according to the geometric feature information to obtain structural partitioning results, wherein the geometric feature information includes local geometric features and global topological features, and the structural partitioning results are used to characterize the structural types corresponding to different regions included in the three-dimensional model; determining the additive manufacturing process parameters corresponding to different regions included in the three-dimensional model based on the structural partitioning results; and converting the additive manufacturing process parameters corresponding to different regions included in the three-dimensional model into control instructions executable by the equipment to generate a process configuration file.

[0037] The method provided by this invention imports a 3D model of the part to be manufactured and converts it into point cloud data. After preprocessing, standardized point cloud data with uniform size and a preset number of points is obtained. Based on this data, local geometric features and global topological features are extracted and structural segmentation is performed. Finally, the additive manufacturing process parameters for each region are determined and control instruction files are generated. This solves the problem of poor compatibility between process parameters and model structure in traditional additive manufacturing, achieves accurate matching of process parameters, and effectively improves the forming quality and manufacturing efficiency of the parts.

[0038] In some embodiments, the additive manufacturing process parameter matching method provided by the present invention can be executed by an additive manufacturing process parameter matching system 100 (hereinafter referred to as matching system 100).

[0039] As an example, the matching system 100 can be any electronic device 200 with data processing capabilities, such as a general-purpose computer, personal computer, laptop computer, switch, or tablet computer. The specific implementation of the matching system 100 is not limited here.

[0040] Figure 1 A schematic diagram of the hardware structure of an electronic device provided in an embodiment of the present invention is shown. The electronic device 200 includes a processor 210, a memory 220, and a communication interface 230.

[0041] Processor 210 may include one or more processing cores. Processor 210 connects to various parts within electronic device 200 using various interfaces and lines, and performs various functions and processes data of electronic device 200 by running or executing instructions, programs, code sets, or instruction sets stored in memory 220, and by calling data stored in memory 220. Optionally, processor 210 may be implemented using at least one of the following hardware forms: Central Processing Unit (CPU), Graphics Processing Unit (GPU), Digital Signal Processing (DSP), Field-Programmable Gate Array (FPGA), and Programmable Logic Array (PLA).

[0042] The memory 220 may include random access memory (RAM) or read-only memory (ROM). Optionally, the memory 220 may include a non-transitory computer-readable storage medium. The memory 220 may be used to store instructions, programs, code, code sets, or instruction sets. The memory 220 may include a program storage area. This program storage area may store instructions for implementing an operating system, instructions for implementing at least one function, instructions for implementing the various method embodiments described above, etc.

[0043] Communication interface 230 is used to communicate with other devices, equipment or communication networks, such as data storage devices, image processing devices or Ethernet, wireless access network (RAN), wireless local area network (WLAN), etc.

[0044] In terms of physical implementation, the aforementioned devices (such as processor 210, memory 220, and communication interface 230) can each be devices within the same device (such as a laptop computer). Alternatively, at least two of these devices can be located within the same device, i.e., as different devices within the same device, similar to the deployment of devices or components in a distributed system.

[0045] It is understood that the structure illustrated in this embodiment does not constitute a specific limitation on the electronic device 200. In other embodiments of the present invention, the electronic device 200 may include more or fewer components than illustrated, or combine some components, or split some components, or have different component arrangements. The illustrated components may be implemented in hardware, software, or a combination of software and hardware.

[0046] The following description, in conjunction with the accompanying drawings, illustrates an additive manufacturing process parameter matching method provided by an embodiment of the present invention.

[0047] Figure 2 This is a flowchart illustrating an additive manufacturing process parameter matching method provided in an embodiment of the present invention. Optionally, this method can be... Figure 1 The illustrated electronic device 200 performs this operation. The method may include the following steps: S1. Import the 3D model of the part to be manufactured and convert the 3D model into point cloud data.

[0048] For details, see Figure 3 , Figure 3 This is a schematic diagram of a three-dimensional model of a part to be manufactured, provided in an embodiment of the present invention. The three-dimensional model is typically imported in STL or OBJ format. After importation, the surface information of the model needs to be converted into a discrete set of three-dimensional coordinate points, i.e., point cloud data. Point cloud data can intuitively reflect the spatial geometry of the model, providing a stable data carrier for subsequent processing.

[0049] It should be noted that the above format of the three-dimensional model is only an example. The three-dimensional model can also adopt other formats. The embodiments of the present invention do not impose any special restrictions on the format of the three-dimensional model.

[0050] S2. Preprocess the point cloud data to obtain standardized point cloud data, wherein the standardized point cloud data has a uniform size and the number of points is a preset value.

[0051] It should be understood that the preprocessing of point cloud data can eliminate the differences in size and coordinates of the original model, so that part models of different specifications can enter a unified data processing space. Standardized point cloud data includes two features: uniform size and fixed number of points, which can ensure the stability of subsequent feature extraction and structural segmentation.

[0052] In one possible implementation, S2 includes: calculating the geometric centroid of the point cloud data and translating all points to coordinate positions with the centroid as the origin to obtain decentralized point cloud data; scaling the decentralized point cloud data to a unit sphere to obtain scale-normalized point cloud data; and performing uniform sampling on the scale-normalized point cloud data to obtain standardized point cloud data with a preset number of points.

[0053] Specifically, the geometric centroid is the center of all spatial coordinate points in the point cloud data. After calculating the geometric centroid, all point cloud data is translated so that the centroid coincides with the origin of the coordinate system. This eliminates the positional offset of the point cloud data in space, ensuring that the point cloud data of different models are all under a unified coordinate reference. After completing the decentralization process, the point cloud data is scaled up to the interior of a unit sphere. This eliminates the size differences between different part models, ensuring that all models are in the same scale space and avoiding interference from scale on feature recognition results.

[0054] Performing uniform sampling on scale-normalized point cloud data can control the number of points within a preset value, reducing the impact of redundant data on computational efficiency while preserving the complete geometric features of the point cloud data. After the three steps of decentralization, scale normalization, and uniform sampling, the resulting standardized point cloud data can stably adapt to subsequent feature extraction algorithms, ensuring the consistency of structural segmentation results.

[0055] In one example, combined Figure 3 See Figure 4 and Figure 5 , Figure 4 This is a schematic diagram of a pre-processed three-dimensional model of a part to be manufactured, provided in an embodiment of the present invention. Figure 5 This is a schematic diagram of a three-dimensional model of a part to be manufactured after segmentation, provided in an embodiment of the present invention. The preprocessing of the point cloud data described above to obtain standardized point cloud data specifically includes: First, input point cloud data (STL, OBJ, or point cloud format), containing N 3D coordinate points: ; Then, centroid translation (decentralization) is performed: the geometric centroid of the original point cloud is calculated. ; Translate all points to a point with the centroid as the origin: ; Eliminate the model's positional offset in space and align its center with the origin.

[0056] Next, scale normalization is performed: the maximum Euclidean distance from all points to the origin is calculated. ; Scaling the point cloud to within a unit sphere: ; Eliminate dimensional differences between different parts to unify the model to a standard scale space.

[0057] Finally, point cloud sampling is performed: uniform sampling is used, and the preset sampling quantity is M, then: .

[0058] This invention employs a three-step standardization process to ensure that model data from different sources and of different specifications can enter a unified processing flow. This reduces the interference of external factors on the algorithm results, improves the reliability of feature extraction and structural segmentation, provides stable and accurate data support for subsequent process parameter matching, and enables the overall solution to adapt to more types of part models.

[0059] S3. Extract geometric feature information based on standardized point cloud data, and perform structural segmentation on the 3D model according to the geometric feature information to obtain structural partitioning results.

[0060] The geometric feature information includes local geometric features and global topological features. Local geometric features are used to reflect the curvature and thickness changes of local regions of the model, while global topological features are used to characterize the overall spatial structure of the model.

[0061] Furthermore, the structural partitioning results are used to characterize the structural types corresponding to different regions included in the 3D model. Structural segmentation of the 3D model based on these two types of features can accurately identify different structural regions within the model, ensuring that each region corresponds to a specific structural type.

[0062] In some embodiments, the structure types include ultra-thin-walled structures, standard thin-walled structures, and thick-walled solid structures; wherein, ultra-thin-walled structures, standard thin-walled structures, and thick-walled solid structures are the three most common structural forms in metal additive manufacturing parts, and different structures have significant differences in their requirements for heat input, scanning method, and energy density during the forming process.

[0063] It should be noted that the above structural types are merely illustrative examples. Structural types may also include rod-shaped structures, block-shaped structures, and transition regions. The embodiments of the present invention do not impose any particular restrictions on the specific implementation of structural types.

[0064] Furthermore, S3 specifically includes: extracting local and global features from standardized point cloud data to obtain geometric feature information including local geometric features and global topological features; dividing the 3D model into regions based on the geometric feature information to obtain structural partitioning results.

[0065] Specifically, local geometric features are used to describe the curvature, wall thickness, and surface variations of the neighborhood of an individual point in the point cloud data, while global topological features are used to characterize the overall spatial connectivity and structural distribution of the model. Performing local and global feature extraction operations on standardized point cloud data allows for the complete acquisition of the model's fine geometric and overall structural information. Using the extracted geometric features as a basis for partitioning, the global space of the 3D model is divided into regions, ensuring that each partitioned region corresponds to a uniquely defined structural type. The resulting structural partitioning clearly labels the structural attributes of each region of the model.

[0066] This invention, by accurately distinguishing three typical structures and relying on complete geometric feature information to complete the region division, can effectively avoid parameter matching deviations caused by misidentification of structural regions, providing an accurate basis for determining subsequent differentiated process parameters, enabling different structural regions to obtain suitable process conditions, and improving the overall forming effect of parts.

[0067] Furthermore, the above-mentioned local and global feature extraction of standardized point cloud data includes: performing initial feature mapping on the standardized point cloud data through a multilayer perceptron to obtain initial features; processing the initial features using multi-level downsampling and K-nearest neighbor construction to extract multi-scale geometric features; introducing a self-attention mechanism to weight and enhance the multi-scale geometric features to obtain enhanced key manufacturing features; and performing feature fusion on the enhanced key manufacturing features through progressive upsampling to obtain geometric feature information including local geometric features and global topological features.

[0068] It should be noted that the multilayer perceptron can perform nonlinear transformations on the three-dimensional coordinate information of standardized point cloud data, mapping the original coordinate data into high-dimensional initial features, improving the feature representation capability of the data, and laying the foundation for subsequent deep feature extraction.

[0069] Multi-level downsampling can gradually reduce the amount of point cloud data and reduce the amount of computation, while preserving the core structural information. The K-nearest neighbor construction method can determine a fixed number of neighboring points around each point, and completely extract the geometric features of the local area. Through the combination of multi-level and multi-neighbor processing, multi-scale geometric features covering different levels of detail can be obtained.

[0070] The self-attention mechanism can assign higher weights to additive manufacturing-related structural features in multi-scale geometry, weaken the influence of irrelevant features, and significantly enhance key manufacturing features related to forming quality.

[0071] Upsampling can combine high-level abstract features with low-level detailed features layer by layer, achieving full fusion of features at different scales. The final output feature information includes both local geometric features and global topological features, which can comprehensively reflect the geometric structural properties of the model.

[0072] For example, see Figure 6 , Figure 6 This is a schematic diagram of a geometric feature information extraction process provided in an embodiment of the present invention. The standardized point cloud data obtained in S2 above is used as input, with 2048 input points, each containing three-dimensional coordinate information. First, an initial feature mapping is performed on the input points using a multilayer perceptron, increasing the feature dimension to 32 dimensions. Then, a multi-level downsampling structure is constructed, using the farthest point sampling method to progressively reduce the number of points from 2048 to 512, 128, 32, and 8 sampling points. In each level, a local neighborhood (K=32) is constructed based on the K-nearest neighbor method. Local geometric features are extracted through a multilayer perceptron with shared weights, and neighborhood features are fused to form a multi-scale representation. A self-attention mechanism is introduced after each feature extraction module. A query vector, key vector, and value vector are generated by linearly transforming the feature matrix. An attention weight matrix is ​​calculated, and the neighborhood features are weighted and summed to enhance the feature representation capability of key structural regions. After the encoding stage, feature interpolation and skip connections are used for stepwise upsampling, fusing high-level semantic features with low-level spatial features to restore the original point-level resolution. Finally, a multilayer perceptron outputs the structure category probability vector corresponding to each point. During network training, cross-entropy loss is used for supervised learning, and the Adam optimization algorithm is used for parameter updates. After training, the algorithm can achieve accurate segmentation of different geometric regions, providing structure category labels for subsequent process parameter matching steps.

[0073] The embodiments of the present invention employ a processing method that combines multi-level feature extraction and weighted enhancement, which can accurately capture the local details and overall distribution of the model, highlight the core features related to manufacturing, improve the recognition accuracy and stability of structural segmentation, and provide high-quality feature support for subsequent structural partitioning and parameter matching.

[0074] In other embodiments, after structural segmentation of the 3D model based on the geometric feature information to obtain structural partitioning results, the method further includes: determining whether the structural partitioning results meet preset manufacturing constraints, wherein the preset manufacturing constraints are that the 3D model is structurally partitioned across the entire domain, the boundaries of each structural region are continuous, and the recognition confidence is greater than a preset threshold; if the structural partitioning results do not meet the preset manufacturing constraints, the 3D model is re-segmented until structural partitioning results that meet the preset manufacturing constraints are obtained.

[0075] The preset manufacturing constraints include three core judgment criteria: the first is that the three-dimensional model completes the structural partitioning across the entire domain; the second is that the boundaries of each structural region remain continuous; and the third is that the confidence value of structural identification is greater than the set preset threshold.

[0076] The matching system 100 will verify the structural partitioning results item by item to ensure they meet the aforementioned constraints. If partitioning is omitted, boundaries are discontinuous, or the recognition confidence level is insufficient, the current result is determined not to meet the requirements. The matching system 100 will then re-invoke the structural segmentation process and perform feature extraction and region division operations again based on standardized point cloud data. This segmentation and verification process is repeated iteratively until the output structural partitioning results fully meet the preset manufacturing constraints before proceeding to the subsequent process parameter matching stage.

[0077] This invention improves the accuracy and reliability of structural partitioning results through constraint condition determination and iterative optimization, avoids parameter matching errors caused by abnormal partitioning results, provides a reliable guarantee for the accurate configuration of subsequent process parameters, and thus steadily improves the overall forming quality of parts.

[0078] S4. Based on the structural partitioning results, determine the additive manufacturing process parameters corresponding to the different regions included in the three-dimensional model.

[0079] Specifically, the additive manufacturing process parameters include laser power, scanning speed, layer thickness, scanning path strategy, and wire feed speed.

[0080] Among them, laser power is the core parameter for controlling the energy input during additive manufacturing, scanning speed is used to control the movement rate of fused wire deposition, layer thickness is used to set the thickness of a single deposition, scanning path strategy is used to plan the motion trajectory of the equipment, and wire feed speed is used to control the material delivery rate. The above five parameters together determine the forming quality and efficiency of additive manufacturing.

[0081] In some embodiments, determining the additive manufacturing process parameters corresponding to different regions included in the three-dimensional model based on the structural partitioning results includes: matching the combination of process parameters corresponding to each region included in the three-dimensional model in a mapping database according to the structural type in the structural partitioning results, wherein the mapping database pre-stores additive manufacturing process parameters corresponding to each of multiple structural types.

[0082] The mapping database pre-determines the corresponding process parameter combinations for ultra-thin-walled structures, standard thin-walled structures, and thick-walled solid structures based on experimental tests, historical process data, or numerical simulation results. Data records include the correspondence between structure types and various parameters. After reading the structure partitioning results, the system performs a search operation in the mapping database according to the structure type corresponding to each region, directly retrieving the matching parameter combinations to complete the process parameter determination process for each region.

[0083] In one example, see Figure 7 , Figure 7 This is a schematic diagram illustrating the matching process of additive manufacturing process parameters according to an embodiment of the present invention. The specific implementation is as follows: First, based on the structural partitioning results obtained in step S3 (which can also be understood as structural segmentation labels), different regions of the part are structurally classified and labeled. These structural segmentation labels include, but are not limited to, ultra-thin-walled structures, standard thin-walled structures, and thick-walled solid structures (multi-layer thick-walled structures). For different structural categories, corresponding core geometric feature parameters are extracted as the basis for judgment, such as wall thickness, radius of curvature, rate of change of cross section, or structural continuity index, and corresponding feature threshold ranges are set.

[0084] Subsequently, a mapping database between structural categories and process parameters is constructed. This mapping database includes at least the following fields: structural label, core geometric feature threshold range, scanning path strategy, laser power, scanning speed, and target energy density. Each structural category corresponds to one or more combinations of process parameters that satisfy forming quality constraints.

[0085] The parameter combinations can be derived from experimental test data, historical process data, or numerical simulation results, and are stored in a database in tabular form.

[0086] Taking a thin-walled ring structure as an example, when the wall thickness of a certain region is less than or equal to 2mm, it is determined to be an ultra-thin-walled structure. The corresponding single-channel swing scanning strategy is matched, the laser power is set to 800W, the scanning speed is set to 12mm / s, and the energy density is controlled to the low energy range. When the wall thickness is between 2mm and 5mm, it is determined to be a standard thin-walled structure, matched with a ring offset scanning strategy, laser power of 1200W, scanning speed of 10mm / s, and energy density in the medium range. When the wall thickness is greater than or equal to 5mm, it is determined to be a thick-walled solid structure, and a cross-filling plus contour scanning strategy is matched, with a laser power of 1800W, a scanning speed of 8mm / s, and the energy density controlled in the high energy range.

[0087] After the segmentation results are input into the database query module, the matching system 100 automatically retrieves the corresponding process parameter combinations based on the structural labels and geometric feature values, and outputs the matching results to obtain the additive manufacturing process parameters corresponding to different regions included in the three-dimensional model.

[0088] Optionally, if a certain structure category corresponds to multiple sets of optional parameters, the optimal parameter combination shall be selected according to the preset mass priority or energy stability evaluation function.

[0089] Through the above-mentioned structure category-parameter mapping mechanism, automatic matching of process parameters driven by geometry is realized, avoiding manual experience-based parameter adjustment and improving the forming consistency and manufacturing efficiency of different structural regions.

[0090] The embodiments of the present invention rely on a pre-built mapping database to achieve automatic parameter matching, abandoning the traditional method of manually setting parameters, reducing errors and fluctuations caused by human factors, improving the speed and consistency of parameter determination, and enabling different structural regions to quickly obtain suitable process conditions.

[0091] Optionally, based on the structure type in the structural partitioning results, the process parameter combinations corresponding to each region included in the 3D model are matched in the mapping database, including: When the structure type corresponds to only one set of process parameter combinations in the mapping database, the process parameter combination is determined as the target process parameter for the region; when the structure type corresponds to multiple sets of process parameter combinations in the mapping database, a comprehensive score for each set of process parameter combinations is calculated based on a preset forming quality evaluation function with forming surface finish, forming dimensional accuracy, melt pool stability, and deposition efficiency as weighting factors, and the process parameter combination with the highest comprehensive score is selected as the target process parameter for the region.

[0092] For example, the mapping database can store one or more sets of verified process parameter combinations for the same structure type. During matching, the system first determines the number of corresponding parameter combinations. When a structure type is associated with only one set of parameter combinations, the system directly determines that combination as the target process parameter for the current structure region, maintaining efficient parameter configuration. When a structure type is associated with multiple sets of optional parameter combinations, the system activates a preset forming quality evaluation function to perform calculations. This evaluation function uses surface finish, dimensional accuracy, melt pool stability, and deposition efficiency as weighting factors to quantify and score each parameter combination, then selects the parameter combination with the highest score as the target process parameter for the current region.

[0093] The embodiments of the present invention adopt parameter determination logic based on different situations. In the case of a single parameter combination, the processing flow is simplified and the overall operating efficiency is improved. In the case of a multi-parameter combination, the optimal selection is completed through quantitative evaluation, so that the degree of adaptation between process parameters and structural type reaches a higher level, taking into account both part forming quality and overall manufacturing efficiency, and making the final forming effect more stable and reliable.

[0094] In another possible implementation, the scanning path strategy corresponds one-to-one with the structure type, specifically including: when the structure type is an ultra-thin-walled structure, the scanning path strategy adopts single-track oscillating scanning; when the structure type is a standard thin-walled structure, the scanning path strategy adopts ring-shaped offset scanning; when the structure type is a thick-walled solid structure, the scanning path strategy adopts cross-fill scanning and contour scanning.

[0095] It should be understood that the scanning path strategy directly determines the motion trajectory of the additive manufacturing equipment, and its compatibility with the structure type significantly affects the final formed state of the part. Ultra-thin-walled structures have smaller wall thicknesses and are more sensitive to heat input; using single-pass oscillating scanning can control heat accumulation, ensuring dimensional accuracy and forming continuity in the thin-walled area. Standard thin-walled structures have a certain structural width; using ring-off offset scanning can ensure continuous and uniform trajectory, improving the surface finish and structural stability of the thin-walled portion. Thick-walled solid structures require sufficient infill density and bonding strength; using cross-fill scanning and contour scanning can achieve sufficient internal filling and precise external contour forming, balancing deposition efficiency and structural strength. Maintaining a one-to-one correspondence between various structures and scanning path strategies allows each type of structure to be processed and formed along the most suitable trajectory.

[0096] This invention configures a dedicated scanning path strategy for different structural characteristics, solving the problem that a single scanning path is difficult to adapt to multiple structures, improving the forming accuracy and surface quality of each structural region, reducing the probability of forming defects such as warping, holes and surface roughness, and optimizing the overall forming effect of complex structural parts.

[0097] S5. Convert the additive manufacturing process parameters corresponding to different regions included in the three-dimensional model into control instructions that can be executed by the equipment, and generate a process configuration file.

[0098] In one possible implementation, S5 specifically includes: converting the matched process parameters into corresponding control command statements according to the control protocol of the additive manufacturing equipment; dynamically switching the control commands of the corresponding regions according to the structural partitions and printing sequence of the three-dimensional model to form a continuous command sequence; and generating a process configuration file in a preset format based on the command sequence, wherein the preset format is a G-code file format or an XML format.

[0099] Specifically, additive manufacturing equipment typically operates according to fixed control protocols. Different process parameters need to be converted into instruction statements that the equipment can recognize according to the protocol rules. Parameters such as laser power, scanning speed, layer thickness, scanning path strategy, and wire feed speed can all be converted according to the corresponding protocol. After the instruction statement conversion is completed, the system dynamically switches the instruction statements for different areas according to the structural partitioning results of the 3D model and the set printing layer sequence. This allows the equipment to automatically execute the matching process parameters when processing the corresponding structural area, thus forming a continuous and complete instruction sequence. The instruction sequence is encapsulated according to a preset format to generate process configuration files in G-code or XML format. Both formats can be directly read and executed by mainstream additive manufacturing equipment.

[0100] This invention enables the direct conversion of process parameters into equipment execution instructions. By relying on dynamic switching logic to adapt to parameter differences in different structural regions, it ensures stable execution of the printing process, improves the automation level of equipment operation, makes the forming process of complex structural parts more coherent, and effectively guarantees forming consistency.

[0101] In one example, S5 specifically includes: First, the additive manufacturing process parameters corresponding to different regions of the 3D model output in step S4 are read. These process parameters include, but are not limited to, laser power, scanning speed, path planning strategy, layer thickness, and wire feed speed. The matching system 100 establishes parameter mapping rules according to the equipment control protocol, converting the abstract process parameters into equipment instruction statements. Specifically, the laser power parameter is mapped to a power control instruction, for example, using the format "M104 S1200" to indicate setting the laser output power; the scanning speed parameter is mapped to a feed speed instruction, for example, using the format "G1 F600" to indicate setting the scanning speed; and the path planning strategy is converted into a corresponding motion trajectory control statement, for example, "G1 X10.5 Y20.3 E0.1" to indicate executing movement along a specified coordinate path.

[0102] Subsequently, based on the part's partition structure labels, differentiated instruction sequences are generated for different regions and organized according to the printing layer order to form continuous manufacturing control codes. If the part contains multiple structural types, the system dynamically switches the corresponding process parameters during instruction generation based on the spatial region division, achieving adaptive parameter control at the structural region level.

[0103] Regarding the generated file format, this embodiment supports two output methods: one is to generate standard G-code files for direct reading and execution by additive manufacturing equipment; the other is to generate XML format process configuration files for parameter encapsulation and equipment interface calls. The generated files contain complete process parameter packages and path control information, achieving end-to-end automatic output from the 3D design model to the equipment execution instructions.

[0104] As can be seen from S1-S5, the method provided by the embodiments of the present invention constructs a complete technical process from model import to instruction output. By using automated data processing and parameter matching, it replaces the traditional parameter adjustment method that relies on manual experience, shortens the process debugging cycle, reduces trial and error costs, makes the forming quality of complex structural parts more stable, and improves the automation level of the overall manufacturing process.

[0105] This invention can also be understood as follows: Compared with existing technologies, this invention has the following beneficial effects: By extracting local geometric features and global topological features based on standardized point cloud data, it accurately partitions the ultra-thin-walled structure, standard thin-walled structure, and thick-walled solid structure of the 3D model and matches them with corresponding additive manufacturing process parameters, achieving parameter matching at the structural level and effectively improving the consistency of part forming quality; The entire process automatically matches process parameters through a mapping database, and in scenarios with multiple parameter combinations, it can also select the optimal solution through a forming quality evaluation function, reducing manual intervention and significantly reducing parameter tuning costs; It designs exclusive scanning path strategies for different structural types and ensures partition reliability through partition verification, which can adapt to the manufacturing needs of complex structural parts and improve the manufacturing stability of complex structural parts; This invention realizes an automated closed-loop process from design to manufacturing, greatly improving manufacturing efficiency; At the same time, the process parameter conversion process adapts to the control protocol of additive manufacturing equipment, and can generate process configuration files in G-code file format or XML format, which has good scalability and can adapt to different types of additive manufacturing equipment, making it more widely applicable.

[0106] The foregoing mainly describes the solutions of the embodiments of the present invention from a methodological perspective. It is understood that, in order to achieve the above-mentioned functions, the matching system 100 includes at least one of the hardware structures and software modules corresponding to the execution of each function. Those skilled in the art should readily recognize that, in conjunction with the units and algorithm steps of the various examples described in the embodiments disclosed herein, the embodiments of the present invention can be implemented in hardware or a combination of hardware and computer software. Whether a function is executed in hardware or by computer software driving hardware depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the embodiments of the present invention.

[0107] In this embodiment of the invention, the matching system 100 can be divided into functional units according to the above method example. For example, the matching system 100 can be divided into functional units corresponding to each function, or two or more functions can be integrated into one processing unit. The integrated unit can be implemented in hardware or as a software functional unit. It should be noted that the unit division in this embodiment is illustrative and only represents one logical functional division; other division methods may be used in actual implementation.

[0108] For example, Figure 8 A schematic diagram of the hardware structure of a matching system provided in an embodiment of the present invention is shown. The matching system 100 includes: a point cloud conversion module 110, used to import a three-dimensional model of a part to be manufactured and convert the three-dimensional model into point cloud data; a point cloud processing module 120, used to preprocess the point cloud data to obtain standardized point cloud data, wherein the standardized point cloud data has a uniform size and the number of points is a preset value; a structure segmentation module 130, used to extract geometric feature information based on the standardized point cloud data and perform structural segmentation on the three-dimensional model according to the geometric feature information to obtain structural partitioning results, wherein the geometric feature information includes local geometric features and global topological features, and the structural partitioning results are used to characterize the structural types corresponding to different regions included in the three-dimensional model; a parameter determination module 140, used to determine the additive manufacturing process parameters corresponding to different regions included in the three-dimensional model based on the structural partitioning results; and a file generation module 150, used to convert the additive manufacturing process parameters corresponding to different regions included in the three-dimensional model into control instructions executable by the equipment and generate a process configuration file.

[0109] It should be understood that specific descriptions of the above-mentioned optional methods can be found in the foregoing method embodiments, and will not be repeated here. Furthermore, explanations of any of the matching systems 100 provided above, as well as descriptions of their beneficial effects, can be found in the corresponding method embodiments described above, and will not be repeated here.

[0110] This invention also provides a computer-readable storage medium storing at least one computer instruction, which is loaded and executed by a processor to implement the methods of the various embodiments described above. Explanations of the relevant content and descriptions of the beneficial effects of any of the computer-readable storage media provided above can be found in the corresponding embodiments described above, and will not be repeated here.

[0111] This invention also provides a chip. This chip integrates a control circuit for implementing the functions of the matching system 100 described above, and one or more ports. Optionally, the functions supported by this chip are as described above, and will not be repeated here.

[0112] Those skilled in the art will understand that the program for implementing all or part of the steps of the above embodiments, which can be executed by a program instructing related hardware, can be stored in a computer-readable storage medium. The storage medium mentioned above can be a read-only memory, a random access memory, etc. The processing unit or processor mentioned above can be a central processing unit, a general-purpose processor, an application-specific integrated circuit (ASIC), a microprocessor (DSP), a field-programmable gate array (FPGA), or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof.

[0113] This invention also provides a computer program product containing instructions that, when executed on a computer, cause the computer to perform any of the methods described in the above embodiments. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the flow or function according to the embodiments of this invention is generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, computer instructions can be transmitted from one website, computer, server, or data center to another via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium accessible to a computer or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., SSD), etc.

[0114] It should be noted that the devices for storing computer instructions or computer programs provided in the embodiments of the present invention, such as, but not limited to, the aforementioned memory, computer-readable storage medium, and communication chip, are all non-transitory. Those skilled in the art should recognize that the functions described in the embodiments of the present invention in one or more of the above examples can be implemented using hardware, software, firmware, or any combination thereof. When implemented using software, these functions can be stored in a computer-readable storage medium or transmitted as one or more instructions or code on a computer-readable storage medium. Computer-readable storage media include computer storage media and communication media, wherein communication media include any medium that facilitates the transmission of computer programs from one place to another. Storage media can be any available medium accessible to general-purpose or special-purpose computers.

[0115] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention.

Claims

1. A method for matching additive manufacturing process parameters, characterized in that, include: Import the 3D model of the part to be manufactured, and convert the 3D model into point cloud data; The point cloud data is preprocessed to obtain standardized point cloud data, wherein the standardized point cloud data has a uniform size and the number of points is a preset value; Geometric feature information is extracted from standardized point cloud data, and the 3D model is structurally segmented based on the geometric feature information to obtain structural partitioning results. The geometric feature information includes local geometric features and global topological features. Local geometric features are used to describe the curvature, wall thickness, and surface changes of the neighborhood of a single point in the point cloud data. Global topological features are used to characterize the overall spatial connectivity and structural distribution of the model. The structural partitioning results are used to characterize the structural types corresponding to different regions included in the 3D model. The structural types include ultrathin-walled structures, standard thin-walled structures, and thick-walled solid structures. Based on the structural partitioning results, determine the additive manufacturing process parameters corresponding to the different regions included in the three-dimensional model; The additive manufacturing process parameters corresponding to different regions included in the three-dimensional model are converted into control instructions that can be executed by the equipment, and a process configuration file is generated. The additive manufacturing process parameters include laser power, scanning speed, layer thickness, scanning path strategy, and wire feed speed. The step of determining the additive manufacturing process parameters corresponding to different regions included in the 3D model based on the structural partitioning results includes: Based on the structural type in the structural partitioning results, the process parameter combination corresponding to each region included in the three-dimensional model is matched in the mapping database. The mapping database pre-stores additive manufacturing process parameters corresponding to each of the multiple structural types. The scanning path strategy corresponds one-to-one with the structure type, specifically including: When the structure type is an ultrathin-walled structure, the scanning path strategy adopts single-track swing scanning. When the structure type is a standard thin-walled structure, the scanning path strategy adopts a ring-shaped offset scanning. When the structure type is a thick-walled solid structure, the scanning path strategy adopts cross-fill scanning and contour scanning. The matched process parameters are converted into control commands executable by the equipment, and a process configuration file is generated, including: According to the control protocol of the additive manufacturing equipment, the matched process parameters are converted into corresponding control command statements; According to the structural partitioning and printing layer sequence of the three-dimensional model, the control commands for the corresponding regions are dynamically switched to form a continuous command sequence; A process configuration file in a preset format is generated based on the instruction sequence. The preset format is either a G-code file or an XML file.

2. The method according to claim 1, characterized in that, The point cloud data is preprocessed to obtain standardized point cloud data, including: Calculate the geometric centroid of the point cloud data, and translate all points to coordinate positions with the centroid as the origin to obtain decentralized point cloud data; The decentralized point cloud data is scaled down to a unit sphere to obtain scale-normalized point cloud data. Uniform sampling is performed on the scale-normalized point cloud data to obtain standardized point cloud data with a preset number of points.

3. The method according to claim 1, characterized in that, The process of extracting geometric feature information based on standardized point cloud data and performing structural segmentation on the 3D model based on the geometric feature information to obtain structural partitioning results includes: Local and global features are extracted from standardized point cloud data to obtain geometric feature information including local geometric features and global topological features; Based on the geometric feature information, the three-dimensional model is divided into regions to obtain structural partitioning results; The extraction of local and global features from standardized point cloud data includes: Initial features are obtained by performing initial feature mapping on standardized point cloud data using a multilayer perceptron; The initial features are processed using multi-level downsampling and K-nearest neighbor construction to extract multi-scale geometric features; A self-attention mechanism is introduced to weighted enhance multi-scale geometric features, resulting in enhanced key manufacturing features. By performing feature fusion on the enhanced key manufacturing features through stepwise upsampling, geometric feature information including local geometric features and global topological features is obtained.

4. The method according to claim 1, characterized in that, Based on the structural type in the structural partitioning results, the process parameter combinations corresponding to each region included in the 3D model are matched in the mapping database, including: When the structure type corresponds to only one set of process parameter combinations in the mapping database, the process parameter combination is determined as the target process parameter for the region. When the structure type corresponds to multiple sets of process parameter combinations in the mapping database, a comprehensive score for each set of process parameter combinations is calculated based on a preset forming quality evaluation function that uses the surface finish, dimensional accuracy, melt pool stability, and deposition efficiency as weighting factors. The process parameter combination with the highest comprehensive score is selected as the target process parameter for the region.

5. The method according to claim 1, characterized in that, After performing structural segmentation on the 3D model based on the geometric feature information to obtain structural partitioning results, the method further includes: Determine whether the structural partitioning result meets the preset manufacturing constraints. The preset manufacturing constraints are that the three-dimensional model completes structural partitioning, the boundaries of each structural region are continuous, and the recognition confidence is greater than a preset threshold. If the structural partitioning result does not meet the preset manufacturing constraints, the three-dimensional model is re-segmented until a structural partitioning result that meets the preset manufacturing constraints is obtained.

6. An additive manufacturing process parameter matching system, characterized in that, include: The point cloud conversion module is used to import the 3D model of the part to be manufactured and convert the 3D model into point cloud data; The point cloud processing module is used to preprocess the point cloud data to obtain standardized point cloud data, wherein the standardized point cloud data has a uniform size and the number of points is a preset value. The structure segmentation module is used to extract geometric feature information based on standardized point cloud data, and to perform structural segmentation on the 3D model according to the geometric feature information to obtain structural partitioning results. The geometric feature information includes local geometric features and global topological features. Local geometric features are used to describe the curvature, wall thickness and surface changes of the neighborhood of a single point in the point cloud data. Global topological features are used to characterize the overall spatial connectivity and structural distribution of the model. The structural partitioning results are used to characterize the structural types corresponding to different regions included in the 3D model. The structural types include ultrathin-walled structures, standard thin-walled structures and thick-walled solid structures. The parameter determination module is used to determine the additive manufacturing process parameters corresponding to different regions included in the three-dimensional model based on the structural partitioning results. The file generation module is used to convert the additive manufacturing process parameters corresponding to different regions included in the 3D model into control instructions that can be executed by the equipment, and generate a process configuration file. The additive manufacturing process parameters include laser power, scanning speed, layer thickness, scanning path strategy, and wire feed speed; the parameter determination module is used to: match the process parameter combination corresponding to each region included in the three-dimensional model in the mapping database according to the structure type in the structure partitioning result, wherein the mapping database pre-stores the additive manufacturing process parameters corresponding to each structure type among multiple structure types; The scanning path strategy corresponds one-to-one with the structure type, specifically including: when the structure type is an ultra-thin-walled structure, the scanning path strategy adopts single-track swing scanning; when the structure type is a standard thin-walled structure, the scanning path strategy adopts ring offset scanning; when the structure type is a thick-walled solid structure, the scanning path strategy adopts cross-fill scanning and contour scanning. The file generation module is specifically used for: converting the matched process parameters into corresponding control command statements according to the control protocol of the additive manufacturing equipment; dynamically switching the control commands of the corresponding areas according to the structural partitions and printing sequence of the three-dimensional model to form a continuous command sequence; and generating a process configuration file in a preset format based on the command sequence, wherein the preset format is a G-code file format or an XML format.

Images