A cloud-based model simulation control method and system based on pre-processing

Abstract

The present invention discloses a cloud-based model simulation control method and system based on pre-processing, which belongs to the field of model simulation technology. The simulation pre-processing scheme in the prior art has low simulation efficiency and low precision. The cloud-based model simulation control method based on pre-processing of the present invention processes the model object by constructing a model grouping unit, a model pre-processing unit, and a model assembly unit, which can effectively reduce the pre-processing difficulty of complex model objects, reduce the pre-processing workload, and reduce the simulation time; at the same time, it can ensure that all geometric features are captured and geometric feature information will not be lost, thereby effectively improving the simulation accuracy and efficiency. By constructing a cloud-based processing unit, the present invention enables online resources and offline operations to be effectively combined, without the need to frequently upload and download files, thereby significantly improving the access speed of simulation objects and effectively reducing the amount of synchronous data processing of the simulation model.

Description

Technical Field

[0001] The present invention relates to a pre-processing-based cloud model simulation control method and system, belonging to the technical field of model simulation. Background Art

[0002] Currently, Computer Aided Engineering (CAE) simulation is widely used in the manufacturing industry. The pre-processing stage is a key step in CAE, which lays the foundation for subsequent solutions and analysis. However, the existing pre-processing solutions have relatively low processing efficiency.

[0003] Furthermore, a Chinese patent application (publication number: CN107742042A) provides a CAE simulation pre-processing method, including the following steps: detecting user operations to enter the CAE finite element mesh generation operation interface; establishing a finite element model based on the user's operations; converting the finite element model into a digital file containing mesh node information; setting parameters for the digital file; and generating a computable file from the digital file to complete the pre-processing process.

[0004] Since the product structure to be simulated is generally composed of various parts of different shapes and the structural modeling is relatively complex, simply converting the finite element model directly into a digital file containing mesh node information will make the pre-processing work more complicated and time-consuming, and require higher computing resources. It may even lose some geometric feature information, affecting the simulation accuracy and efficiency.

[0005] Furthermore, the above-mentioned CAE simulation solutions and existing CAE simulation solutions usually adopt an offline method to perform simulation processing separately, and the file synchronization speed is slow. Simulation engineers need to upload and download files frequently, which makes the data sharing process very cumbersome and the workload of synchronizing and sharing simulation models very large.

[0006] The information disclosed in this Background Art is only for understanding the background of the present inventive concept and therefore it may include information that does not constitute prior art. Summary of the Invention

[0007] In response to the above problem or one of the above problems, an object of the present invention is to provide a cloud-based model simulation control method and system based on pre-processing, which fully considers the structural complexity of the simulation object, and groups a model object and performs geometric topology analysis by constructing a model grouping unit, a model pre-processing unit, and a model assembly unit; and performs geometric feature recognition on the part grouping data, and performs grid processing to obtain part mesh data; then identifies the assembly relationship between the parts, and assembles the part mesh data according to the assembly relationship to obtain a mesh simulation object, thereby effectively reducing the pre-processing difficulty of complex model objects, reducing the pre-processing workload, reducing simulation time, and reducing the demand for simulation computing hardware; at the same time, it can ensure that all geometric features are captured and geometric feature information will not be lost, thereby effectively improving simulation accuracy and efficiency.

[0008] In response to the above problem or one of the above problems, the second purpose of the present invention is to provide a cloud-based model simulation control method and system based on pre-processing, which can effectively combine online resources and offline operations, so that simulation engineers can access and edit simulation objects in the cloud on the local client like operating local files, without the need to frequently upload and download files. This can significantly improve the access speed of simulation objects, making the data sharing process simple and convenient, and effectively reducing the amount of synchronous data processing of the simulation model.

[0009] To achieve one of the above purposes, the first technical solution of the present invention is:

[0010] A cloud-based model simulation control method based on pre-processing includes the following contents:

[0011] Based on the pre-built model grouping unit, a certain model object is grouped and geometric topology analysis is performed to obtain part grouping data;

[0012] Using the pre-built model pre-processing unit, the geometric features of the part grouping data are recognized and meshed to obtain the component mesh data;

[0013] Use pre-built model assembly units to identify the assembly relationships between parts, and assemble the component mesh data according to the assembly relationships to obtain a mesh simulation object;

[0014] The pre-built cloud processing unit obtains the request information sent by the user; and according to the request information, the grid simulation object is processed and response feedback information is generated to realize the cloud model simulation control based on pre-processing.

[0015] The present invention fully considers the structural complexity of the simulation object, and groups a model object and performs geometric topology analysis by constructing a model grouping unit, a model pre-processing unit, and a model assembly unit; and performs geometric feature recognition on the part grouping data, and performs gridding processing to obtain part grid data; then, the assembly relationship between the parts is identified, and the part grid data is assembled according to the assembly relationship to obtain a grid simulation object, thereby effectively reducing the pre-processing difficulty of complex model objects, alleviating the pre-processing workload, reducing simulation time, and reducing the demand for simulation computing hardware; at the same time, it can ensure that all geometric features are captured and geometric feature information will not be lost, thereby effectively improving simulation accuracy and efficiency.

[0016] The present invention constructs a cloud processing unit to obtain request information sent by users; and processes the grid simulation object according to the request information and generates response feedback information, so that online resources and offline operations can be effectively combined, so that simulation engineers can access and edit cloud simulation objects on the local client like operating local files, without the need to frequently upload and download files, thereby significantly improving the access speed of simulation objects, making the data sharing process simple and convenient, and effectively reducing the amount of synchronous data processing of the simulation model.

[0017] As preferred technical measures:

[0018] Based on the pre-built model grouping unit, the method for grouping a model object and performing geometric topological analysis to obtain part grouping data is as follows:

[0019] Obtaining image information of a model object, including several projection views;

[0020] Adjust several projection views to a fixed size to obtain a model image in a unified format.

[0021] Use convolutional neural networks to extract features from the model image and generate feature maps;

[0022] According to the feature map, the geometric features of the model object are encoded and classified to obtain part grouping data.

[0023] As preferred technical measures:

[0024] The method for obtaining component mesh data by using the pre-built model pre-processing unit to perform geometric feature recognition and meshing on the component group data is as follows:

[0025] Perform geometric feature recognition on the part grouping data to obtain part face information; part face information includes the geometric elements of the parts and the topological information between the geometric elements; the geometric elements include vertex information, edge information and face information;

[0026] Process the facial information of the parts to obtain facial geometric features;

[0027] Based on the face geometry features, the mapping mesh algorithm, the sweeping mesh algorithm or the filling mesh algorithm is used to mesh the face information of the part to obtain mesh geometry information;

[0028] The mesh geometry information is repaired by the geometry optimization algorithm, and then the local area of ​​the mesh geometry information is optimized by the topology optimization algorithm to obtain the component mesh data.

[0029] As preferred technical measures:

[0030] The method for performing geometric feature recognition on part grouping data to obtain part facial information is as follows:

[0031] Perform geometric feature analysis on the part grouping data to obtain all vertex information;

[0032] According to the vertex information, adjacent vertices are connected to form edges to obtain edge information;

[0033] According to the edge information, adjacent edges are connected to form a ring to obtain edge combinations;

[0034] Based on edge combination, closed surface information is obtained;

[0035] Based on the geometric center of a model object, the closed surface information is processed to obtain the mid-surface position, that is, the surface information;

[0036] The face information, edge information and vertex information are coupled and summarized to obtain the part face information.

[0037] As preferred technical measures:

[0038] Use the pre-built model assembly unit to identify the assembly relationship between parts, and assemble the component mesh data according to the assembly relationship to obtain the mesh simulation object as follows:

[0039] Based on a certain model object, the assembly relationship between parts is identified, including bolt connection and / or welding and / or snap connection; and an assembly sequence is generated according to the assembly relationship;

[0040] Optimize the assembly sequence through dynamic programming method and / or genetic algorithm to generate the final assembly sequence information;

[0041] According to the final assembly sequence information, the component mesh data is assembled to obtain the mesh simulation object.

[0042] As preferred technical measures:

[0043] The method of obtaining the request information sent by the user through the pre-built cloud processing unit, processing the grid simulation object according to the request information, and generating response feedback information is as follows:

[0044] Get the attribute request information sent by the user;

[0045] Return attribute response information based on the requested attribute information, which includes attribute information and operation information supported by the server;

[0046] Based on the hypertext transfer protocol and according to the attribute information and operation information supported by the server, the operation request information sent by the user is obtained; the operation request information is a request to synchronously download a certain simulation object, a request to synchronously upload a certain simulation object, a request to delete a certain simulation object, or a request to lock a certain simulation object;

[0047] According to the operation request information, the grid simulation object is processed accordingly and response feedback information is generated.

[0048] As preferred technical measures:

[0049] The method for processing the grid simulation object accordingly and generating response feedback information according to the operation request information is as follows:

[0050] Step 11: The operation request information is a request to synchronously download a simulation object;

[0051] Parse the request for synchronous download of a simulation object to obtain authentication information, file path information requested for download, and request host name;

[0052] Step 12: Check the authentication information. If the authentication is successful, proceed to step 13.

[0053] When the authentication fails, a first response body is generated, the first response body including an authentication error status code; and step 14 is executed;

[0054] Step 13: searching for the specified mesh simulation object according to the file path information. If the mesh simulation object does not exist, generating a first response body including a file error status code; and executing step 14;

[0055] If the grid simulation object exists, generate a second response body, the second response body including a completion status code and the grid simulation object; and execute step 15;

[0056] Step 14: Process the first response body according to the Hypertext Transfer Protocol to obtain response feedback information, and transmit the response feedback information, i.e., the first response body, to the user terminal based on the request host name;

[0057] Step 15: Process the second response body according to the Hypertext Transfer Protocol to obtain response feedback information, and transmit the response feedback information, that is, the second response body, to the user end according to the request host name.

[0058] As preferred technical measures:

[0059] The method for processing the grid simulation object and generating response feedback information according to the operation request information is as follows:

[0060] Step 21: The operation request information is a request to synchronously upload a simulation object;

[0061] Parse the request to synchronously upload a simulation object to obtain authentication information, file path information requested for upload, and request host name;

[0062] Step 22: Check the authentication information. If the authentication is successful, proceed to step 23.

[0063] When the authentication fails, a third response body is generated, the third response body including a creation error status code; and step 24 is executed;

[0064] Step 23: Upload a simulation object according to the file path information and the Hypertext Transfer Protocol; adjust the grid simulation object according to the simulation object; or replace the original grid simulation object according to the simulation object; and generate a fourth response body, which includes a creation success status code; and execute step 25;

[0065] Step 24: Process the third response body according to the Hypertext Transfer Protocol to obtain response feedback information, and transmit the response feedback information, i.e., the third response body, to the user terminal based on the requesting host name;

[0066] Step 25: Process the fourth response body according to the Hypertext Transfer Protocol to obtain response feedback information, and transmit the response feedback information, that is, the fourth response body, to the user end according to the request host name.

[0067] As preferred technical measures:

[0068] The method for processing the grid simulation object and generating response feedback information according to the operation request information is as follows:

[0069] Step 31: The operation request information is a request to delete a simulation object or a request to lock a simulation object;

[0070] Parse the operation request information to obtain authentication information, file path information for request processing, and request host name;

[0071] Step 32: Check the authentication information. If the authentication is successful, proceed to step 33.

[0072] If the authentication fails, a failure status code is generated and step 34 is executed.

[0073] Step 33, performing a delete or lock operation on the mesh simulation object according to the file path information, and obtaining a success status code; then executing step 35;

[0074] Step 34: Process the failure status code according to the Hypertext Transfer Protocol to obtain response feedback information, and transmit the response feedback information to the user terminal based on the request host name;

[0075] Step 35: Process the success status code according to the Hypertext Transfer Protocol to obtain response feedback information, and transmit the response feedback information to the user end according to the request host name.

[0076] To achieve one of the above purposes, the second technical solution of the present invention is:

[0077] A pre-processing-based cloud model simulation control system, which applies the above-mentioned pre-processing-based cloud model simulation control method; it includes a pre-processing module, a cloud control module, an incremental synchronization module and a real-time file monitoring module;

[0078] The pre-processing module includes a model grouping unit, a model pre-processing unit, and a model assembly unit, which are used to group a certain model object and perform geometric topology analysis to obtain a mesh simulation object;

[0079] The cloud control module is provided with a cloud processing unit for obtaining the request information sent by the user; and according to the request information, processes the grid simulation object and generates response feedback information;

[0080] The incremental synchronization module is built using a block comparison algorithm and is used to transfer the changed parts of the file;

[0081] The real-time file monitoring module is built based on the file event monitoring mechanism. It is used to capture the save, modify, and delete operations of local files in real time and trigger instant synchronization. At the same time, it can cooperate with asynchronous queue processing to ensure that user operations are not blocked when multiple tasks are concurrent, and realize file editing and synchronization.

[0082] The present invention provides a pre-processing module, a cloud control module, an incremental synchronization module and a real-time file monitoring module, so that online resources and offline operations can be effectively combined, so that simulation engineers can access and edit simulation objects in the cloud on the local client just like operating local files, without the need to frequently upload and download files. This can significantly improve the access speed of simulation objects, make the data sharing process simple and convenient, and effectively reduce the amount of synchronous data processing of the simulation model.

[0083] At the same time, the present invention can ensure that local files and cloud files are always consistent by simulating the process to the cloud and combining it with a synchronization mechanism, thereby reducing the workload of manual synchronization.

[0084] Compared with the existing technical solutions, the present invention has the following beneficial effects:

[0085] The present invention fully considers the structural complexity of the simulation object, and groups a model object and performs geometric topology analysis by constructing a model grouping unit, a model pre-processing unit, and a model assembly unit; and performs geometric feature recognition on the part grouping data, and performs gridding processing to obtain part grid data; then, the assembly relationship between the parts is identified, and the part grid data is assembled according to the assembly relationship to obtain a grid simulation object, thereby effectively reducing the pre-processing difficulty of complex model objects, alleviating the pre-processing workload, reducing simulation time, and reducing the demand for simulation computing hardware; at the same time, it can ensure that all geometric features are captured and geometric feature information will not be lost, thereby effectively improving simulation accuracy and efficiency.

[0086] The present invention can effectively combine online resources and offline operations, so that simulation engineers can access and edit simulation objects in the cloud on the local client just like operating local files, without the need to frequently upload and download files. This can significantly improve the access speed of simulation objects, making the data sharing process simple and convenient, and effectively reducing the amount of synchronous data processing of the simulation model. BRIEF DESCRIPTION OF THE DRAWINGS

[0087] Figure 1 A schematic diagram of a flow chart of the cloud model simulation control method of the present invention;

[0088] Figure 2 A schematic diagram of grid optimization according to the present invention. DETAILED DESCRIPTION

[0089] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0090] Rather, the present invention encompasses any alternatives, modifications, equivalents, and solutions that fall within the spirit and scope of the present invention as defined by the claims. Furthermore, to facilitate a better understanding of the present invention, certain specific details are described in detail below in the detailed description of the present invention. Those skilled in the art will be able to fully understand the present invention without these details.

[0091] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which the present invention pertains. The terms used herein are for the purpose of describing specific embodiments only and are not intended to limit the present invention.

[0092] like Figure 1 As shown, the first specific embodiment of the cloud model simulation control method based on pre-processing of the present invention is:

[0093] A cloud-based model simulation control method based on pre-processing includes the following contents:

[0094] Based on the pre-built model grouping unit, a certain model object is grouped and geometric topology analysis is performed to obtain part grouping data;

[0095] Using the pre-built model pre-processing unit, the geometric features of the part grouping data are recognized and meshed to obtain the component mesh data;

[0096] Use pre-built model assembly units to identify the assembly relationships between parts, and assemble the component mesh data according to the assembly relationships to obtain a mesh simulation object;

[0097] The pre-built cloud processing unit obtains the request information sent by the user; and according to the request information, the grid simulation object is processed and response feedback information is generated to realize the cloud model simulation control based on pre-processing.

[0098] In this embodiment, a method for obtaining a request message sent by a user through a pre-built cloud processing unit, processing a grid simulation object according to the request message, and generating response feedback message is as follows:

[0099] Get the attribute request information sent by the user;

[0100] Return attribute response information based on the requested attribute information, which includes attribute information and operation information supported by the server;

[0101] Based on the hypertext transfer protocol and according to the attribute information and operation information supported by the server, the operation request information sent by the user is obtained; the operation request information is a request to synchronously download a certain simulation object, a request to synchronously upload a certain simulation object, a request to delete a certain simulation object, or a request to lock a certain simulation object;

[0102] According to the operation request information, the grid simulation object is processed and response feedback information is generated.

[0103] In this embodiment, the first method of performing corresponding processing on the grid simulation object according to the operation request information and generating response feedback information is as follows:

[0104] Step 11: The operation request information is a request to synchronously download a simulation object;

[0105] Parse the request for synchronous download of a simulation object to obtain authentication information, file path information requested for download, and request host name;

[0106] Step 12: Check the authentication information. If the authentication is successful, proceed to step 13.

[0107] When the authentication fails, a first response body is generated, the first response body including an authentication error status code; and step 14 is executed;

[0108] Step 13: searching for the specified mesh simulation object according to the file path information. If the mesh simulation object does not exist, generating a first response body including a file error status code; and executing step 14;

[0109] If the grid simulation object exists, generate a second response body, the second response body including a completion status code and the grid simulation object; and execute step 15;

[0110] Step 14: Process the first response body according to the Hypertext Transfer Protocol to obtain response feedback information, and transmit the response feedback information, i.e., the first response body, to the user terminal based on the request host name;

[0111] Step 15: Process the second response body according to the Hypertext Transfer Protocol to obtain response feedback information, and transmit the response feedback information, that is, the second response body, to the user end according to the request host name.

[0112] In this embodiment, the second method for processing the grid simulation object according to the operation request information and generating response feedback information is as follows:

[0113] Step 21: The operation request information is a request to synchronously upload a simulation object;

[0114] Parse the request to synchronously upload a simulation object to obtain authentication information, file path information requested for upload, and request host name;

[0115] Step 22: Check the authentication information. If the authentication is successful, proceed to step 23.

[0116] When the authentication fails, a third response body is generated, the third response body including a creation error status code; and step 24 is executed;

[0117] Step 23: Upload a simulation object according to the file path information and the Hypertext Transfer Protocol; adjust the grid simulation object according to the simulation object; or replace the original grid simulation object according to the simulation object; and generate a fourth response body, which includes a creation success status code; and execute step 25;

[0118] Step 24: Process the third response body according to the Hypertext Transfer Protocol to obtain response feedback information, and transmit the response feedback information, i.e., the third response body, to the user terminal based on the requesting host name;

[0119] Step 25: Process the fourth response body according to the Hypertext Transfer Protocol to obtain response feedback information, and transmit the response feedback information, that is, the fourth response body, to the user end according to the request host name.

[0120] In this embodiment, a third method for processing a grid simulation object according to the operation request information and generating response feedback information is as follows:

[0121] Step 31: The operation request information is a request to delete a simulation object or a request to lock a simulation object;

[0122] Parse the operation request information to obtain authentication information, file path information for request processing, and request host name;

[0123] Step 32: Check the authentication information. If the authentication is successful, proceed to step 33.

[0124] If the authentication fails, a failure status code is generated and step 34 is executed.

[0125] Step 33, performing a delete or lock operation on the mesh simulation object according to the file path information, and obtaining a success status code; then executing step 35;

[0126] Step 34: Process the failure status code according to the Hypertext Transfer Protocol to obtain response feedback information, and transmit the response feedback information to the user terminal based on the request host name;

[0127] Step 35: Process the success status code according to the Hypertext Transfer Protocol to obtain response feedback information, and transmit the response feedback information to the user end according to the request host name.

[0128] A specific embodiment of the cloud-based model simulation control system based on pre-processing of the present invention:

[0129] A pre-processing-based cloud model simulation control system, which applies the above-mentioned pre-processing-based cloud model simulation control method; the system includes a pre-processing module, a cloud control module, an incremental synchronization module, a real-time file monitoring module, a breakpoint resume module, a multi-user collaboration module, and a version control module;

[0130] The pre-processing module includes a model grouping unit, a model pre-processing unit, and a model assembly unit. It is used to group model objects and perform geometric topology analysis to obtain mesh simulation objects. It can be deployed locally or on a cloud server.

[0131] The cloud control module is provided with a cloud processing unit for obtaining the request information sent by the user; and according to the request information, processes the grid simulation object and generates response feedback information. It is connected to the pre-processing module through the server;

[0132] The incremental synchronization module is built using a block comparison algorithm and is used to transfer the changed parts of the file;

[0133] The real-time file monitoring module is built based on the file event monitoring mechanism. It is used to capture the save, modify, and delete operations of local files in real time and trigger instant synchronization. At the same time, it can cooperate with asynchronous queue processing to ensure that user operations are not blocked when multiple tasks are concurrent, and realize file synchronization while editing.

[0134] The breakpoint resume module is used to implement block upload or download, and can continue from the breakpoint after interruption; and dynamically compress the transmitted data;

[0135] The multi-user collaboration module is used to enable multiple users to access and edit files simultaneously, allowing multiple engineers to collaborate on the same project. It also supports a locking mechanism that allows only one user to modify the same file at a time, ensuring data integrity and consistency.

[0136] The version control module is used to implement file version control, allowing multiple versions of files to coexist and supporting difference comparison and recovery between versions.

[0137] The second specific embodiment of the cloud model simulation control method based on pre-processing of the present invention:

[0138] A cloud-based model simulation control method based on pre-processing includes a model object pre-processing stage and an online and offline combined cloud service interaction stage.

[0139] In this embodiment, the pre-processing stage of the model object can be completed locally or on the cloud platform, and includes the following steps:

[0140] S1: In the part grouping scenario, the model grouping unit uses the feature extraction capability of the convolutional neural network (CNN) to encode the geometric features of the parts and realize the automatic classification of geometric features. The following is a detailed explanation based on the specific data processing process and mathematical model:

[0141] Step 1. Input the image data of the 3D model of the part, which is a 2D projection view, including front view, top view, side view, etc., and the size is usually , the format is RGB.

[0142] Step 2. Resize the input image to a fixed size (such as ), unifying the input format for network processing. Then, a convolutional neural network, such as the open-source neural network framework YOLO, is used to extract features from the image and generate a feature map.

[0143] The method for extracting features from an image is as follows:

[0144] Based on the convolutional neural network CNN, taking the third-generation open source neural network framework YOLOv3 as an example, its feature extraction network Darknet-53 contains 53 convolutional layers. The specific steps are as follows:

[0145] Step 1. Input the original image.

[0146] Step 2. Convolve and downsample the original image, that is, use multiple convolutional layers and residual blocks to extract the features of the original image; and downsample through a convolutional layer or pooling layer with a stride of 2 to obtain sampled data;

[0147] The calculation formula of the convolution operation is as follows:

[0148]

[0149] in, is the output feature map, is the input feature map, is the convolution kernel, is the offset.

[0150] Step 3. Perform multi-scale feature fusion on the sampled data, that is, extract feature maps at different levels (such as 、 、 ), and utilize upsampling and skip connections to fuse features of different scales and obtain feature maps of multiple scales.

[0151] S4. Output feature maps of multiple scales for subsequent object detection.

[0152] The feature map is further divided into If the target center falls in a cell, then the cell is responsible for detecting the target. The target center refers to the coordinates of the center point of the object's ground truth bounding box (Box). For a bounding box, its position is usually represented by four values, including the coordinates of the upper left corner. and the lower right corner coordinates , or the center point coordinates Add width and height .

[0153] Each cell predicts B bounding boxes, which are rectangular regions used to accurately mark the location of objects in an image. In the open source neural network framework YOLO, each bounding box is represented by a four-dimensional vector Indicates that The coordinates of the center point of the bounding box are relative to the position of the cell, and its value range is 0~1; and The width and height of the bounding box are usually determined based on the aspect ratio of the entire image or the scaling factor of the prior box. Each bounding box outputs position information (x, y, w, h) and confidence. The position information is used to indicate the offset and width and height of the bounding box relative to the cell; the confidence It is used to reflect the probability that there is an object in the bounding box, that is, the degree of matching between the bounding box and the real object. The calculation formula is as follows:

[0154]

[0155] in, Indicates the probability of the existence of an object in the bounding box (two classifications), It represents the intersection-over-union ratio between the predicted box and the true box, which is used to measure the accuracy of the predicted box.

[0156] like and When , the final confidence level is .

[0157] Each cell also predicts C class probabilities, and the C class probabilities predicted for each bounding box are conditional probabilities: , which represents the probability that the target belongs to the i-th category when there is a target in the bounding box, where Indicates that the currently predicted target exists. is the i-th target category. Since each bounding box is responsible for detecting only one category, the sum of the C probabilities is 1.

[0158] S2: The model pre-processing unit utilizes an automatic mid-surface extraction module to perform in-depth geometric and topological analysis on the input model object. This includes geometric discretization, element identification, topology construction, Euler consistency checking, thickness analysis, identification of thin-walled areas, mid-surface generation, and mid-surface interpolation. Geometric discretization involves triangulating the mesh to generate a point cloud; element identification includes detecting faces (planes / curved surfaces), edges, and key points; topology construction includes establishing adjacency relationships; thickness analysis includes calculating local thickness; and mid-surface generation involves matching internal and external surfaces. This process then identifies geometric elements such as faces, edges, and points within the model object, as well as the topological relationships between them.

[0159] The model pre-processing unit first processes the input model using a boundary representation method to obtain vertex information, edge information, and face information. Vertex information includes all key vertices extracted from the model data. Edge information includes edges formed by connecting adjacent vertices and records the faces associated with each edge. Face information is composed of looped edges, forming the closed surface of the model. The position of the mid-surface is then determined based on preset rules or user-defined rules. Generally, the mid-surface is located near the geometric center of the entity. Then, a computer algorithm is used to generate a two-dimensional mid-surface based on the results of the geometric topology analysis and the definition of the mid-surface.

[0160] S3: Select meshing algorithms, including mapped mesh, structured mesh, swept mesh, filled mesh, quality control, and adaptive meshing. Each algorithm is suitable for different geometric features. At the same time, set up mesh quality control tools to ensure that the generated mesh quality meets the requirements of simulation analysis, and support adaptive meshing.

[0161] Mapped mesh is mainly used for regular quadrilaterals / hexahedrons to generate structured meshes; swept mesh can map cross-sectional meshes to volumes along a path, which is suitable for pipelines; filled mesh is used to automatically generate tetrahedrons / triangles for complex geometries; quality control is used to check indicators such as aspect ratio and Jacobian matrix; the adaptive partitioning algorithm automatically encrypts high-gradient areas based on the solution results.

[0162] S4: Through the geometric optimization algorithm, through small feature removal (such as chamfers and small holes), surface smoothing (B-spline fitting), gap filling (based on curvature matching) and topology simplification (edge / face merging), combined with curvature analysis to identify key areas, after optimization, the geometric continuity and mesh quality are guaranteed, which is suitable for complex model pre-processing. The mesh geometry information is repaired, and then the topology optimization algorithm is used to optimize the mesh in the local area, which can significantly improve the overall quality and computational efficiency of the mesh, meeting the needs of high-precision numerical simulation. The mesh optimization effect can be seen in Figure 2 .

[0163] S5: Construct an assembly recognition algorithm to identify the assembly relationship between parts, which includes the following:

[0164] The system identifies inter-part connection types through geometric feature extraction (such as bolt holes, welds, and snap-fit ​​structures) and topological relationship analysis. Leveraging the 3D information of the CAD model and combining it with a rule-based matching library to match connection features, it determines assembly relationships such as bolts, welds, and snap-fit ​​connections, providing a basis for generating an assembly sequence. Assembly relationships include bolted, welded, and snap-fit ​​connections, and a reasonable assembly sequence is generated based on these relationships. Assembly sequence optimization technology is also introduced. Using dynamic programming and genetic algorithms, it decomposes the assembly subproblems using dynamic programming and constructs an optimal path with the goal of minimizing conflicts. The genetic algorithm encodes and evolves the assembly sequence, optimizing the sequence through selection, crossover, and mutation. This allows for the consideration of multi-component collaborative assembly constraints, reducing interference and assembly conflicts, and improving assembly efficiency and stability. This ensures coordination and stability among components when assembling multiple parts simultaneously. Interactive user adjustment is also provided, allowing users to fine-tune assembly parameters according to actual needs, further improving assembly quality.

[0165] In this embodiment, during the online and offline cloud service interaction phase, a series of new methods and header fields are introduced via the Hypertext Transfer Protocol (HTTP). This allows users to conveniently manage and edit pre-processed data on remote servers through a web browser or other client (user end), just as they would local files. Related operations include mounting cloud storage paths, synchronously downloading and uploading simulation objects, deleting simulation objects, and locking simulation objects. The simulation objects are CAE models.

[0166] S1: The protocol interaction process for mounting a cloud storage path is as follows:

[0167] Step 1. The client sends a property information request to the server to obtain the properties and methods supported by the server. This request is similar to the client "detecting" the server's capabilities.

[0168] Step 2. After receiving the request for obtaining the resource's attribute information, the server returns an attribute information response containing a list of attributes and methods supported by the server. The client uses this information to determine which operations it can perform on the server.

[0169] Step 3. If the client needs to create a new directory on the server, it sends a request to create a new collection. This request is used to create a new collection, that is, a new directory.

[0170] Step 4. The client sends the file contents to the server via an upload request. An upload request is used to create a new resource or replace one.

[0171] S2: The protocol interaction process for synchronous downloading of simulation objects is as follows:

[0172] Step 1. The client sends a HTTP Get request to download a file at a specified path. The request includes necessary header information, such as the host name and authentication information.

[0173] Step 2. After receiving the request, the server searches for the specified file. If the file exists and the authentication succeeds, the server returns a completed status code, which can be represented by the number 200, and includes the file content in the response body.

[0174] Step 3. The client receives the file content returned by the server and saves it to the local file system.

[0175] S3: The protocol interaction process for synchronously uploading simulation objects is as follows:

[0176] Step 1. The client sends a HTTP upload request to upload a file in a specified path. The request includes necessary header information, such as the host name, authentication information, and file contents.

[0177] Step 2. After receiving the request, the server verifies the authentication information and processes the file upload. If the file upload is successful, the server returns a created status code, which can be represented by the number 201, indicating that the resource has been successfully created.

[0178] S4: The protocol interaction process for deleting a simulation object is as follows:

[0179] Step 1. The client sends a HTTP delete request to delete the file at the specified path. The request contains necessary header information, such as the host name and authentication information.

[0180] Step 2. After receiving the request, the server verifies the authentication information and processes the file deletion. If the file is successfully deleted, the server returns a status code with no relevant content, which can be represented by the number 204, indicating that the request was successful and there is no response body.

[0181] S5: The protocol interaction process for locking the simulation object is as follows:

[0182] Step 1. The client sends a HTTP lock request to lock the file in the specified path. The request contains necessary header information, such as the host name, authentication information, and lock information.

[0183] Step 2. After receiving the request, the server verifies the authentication information and processes the file lock. If the file is locked successfully, the server returns a locked status code, which can be represented by the number 200, and contains the lock token and other lock information in the response body.

[0184] At the same time, the present invention can ensure the high efficiency and real-time performance of file synchronization through an incremental synchronization mechanism, a real-time file monitoring mechanism, a breakpoint resume and compression mechanism, and a multi-user operation mechanism.

[0185] Incremental synchronization mechanism, including the following:

[0186] Using a block-by-block comparison algorithm, only the actual changes in the file (such as modified mesh data and parameter configurations) are transmitted, avoiding the need to transmit the entire file. For example, a 1GB CAE model only requires transmitting approximately 10% of the difference data, significantly reducing network load.

[0187] Real-time monitoring mechanism, including the following:

[0188] By monitoring file events at the operating system level (such as Windows' file change notification feature), local file save and modification operations are captured in real time, triggering instant synchronization. Combined with asynchronous queue processing, this ensures that user operations are not blocked during concurrent multi-tasking, achieving a seamless "edit while sync" experience.

[0189] Breakpoint resumable download and compression mechanism, including the following:

[0190] It supports block uploading and downloading, and can continue from the breakpoint after interruption; it dynamically compresses the transmitted data to further improve the transmission speed and remain stable in a weak network environment.

[0191] Multi-user operation mechanism, including the following:

[0192] It supports multiple users accessing and editing files simultaneously, allowing multiple engineers to collaborate on the same project and improving team work efficiency. It supports a locking mechanism to prevent multiple users from modifying the same file at the same time, ensuring data integrity and consistency. Data is transmitted via the Hypertext Transfer Protocol HTTPS to ensure the security of data transmission. It supports file version control, allowing multiple versions of files to coexist, and supports difference comparison and recovery between versions. The client can be configured with an automatic synchronization function. Through real-time monitoring and incremental transmission technology, it ensures that local CAE pre-processing data and cloud CAE pre-processing data are always consistent, reducing the workload of manual synchronization. Ensure that users can use it seamlessly on different platforms.

[0193] In summary, the present invention allows users to access and edit model files in the cloud on their local computers, just like operating local files, without the need to frequently upload and download files, thereby improving the efficiency of data access.

[0194] An embodiment of a device applying the method of the present invention:

[0195] An electronic device comprising:

[0196] one or more processors;

[0197] a storage device for storing one or more programs;

[0198] When the one or more programs are executed by the one or more processors, the one or more processors implement the above-mentioned cloud-based model simulation control method based on pre-processing.

[0199] A computer medium embodiment of the method of the present invention:

[0200] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the above-mentioned cloud-based model simulation control method based on pre-processing.

[0201] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware. Furthermore, the present application may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, optical storage, etc.) containing computer-usable program code.

[0202] The unit in this application is an object that objectively describes the morphological structure with the help of physical or virtual representation. The object is not equal to the physical body and is not limited to physical and virtual. It can be a data processing function, software program, processing mode, usage method, operation method, workflow, application process, electronic hardware, circuit module, processing system, system imitation or simulation object.

[0203] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the field can still modify or replace the specific implementation methods of the present invention. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be included in the scope of protection of the claims of the present invention.

Claims

1. A cloud-based model simulation control method based on pre-processing, characterized by: Includes the following: Based on the pre-built model grouping unit, a certain model object is grouped and geometric topology analysis is performed to obtain part grouping data; Using the pre-built model pre-processing unit, the geometric features of the part grouping data are recognized and meshed to obtain the component mesh data; Use pre-built model assembly units to identify the assembly relationships between parts, and assemble the component mesh data according to the assembly relationships to obtain a mesh simulation object; Obtain the request information sent by the user through the pre-built cloud processing unit; And according to the request information, the grid simulation object is processed and response feedback information is generated to realize cloud-based model simulation control based on pre-processing; It includes the following: Get the attribute request information sent by the user; Return attribute response information based on the requested attribute information, which includes attribute information and operation information supported by the server; Based on the hypertext transfer protocol and according to the attribute information and operation information supported by the server, the operation request information sent by the user is obtained; the operation request information is a request to synchronously download a certain simulation object, a request to synchronously upload a certain simulation object, a request to delete a certain simulation object, or a request to lock a certain simulation object; According to the operation request information, the grid simulation object is processed accordingly and response feedback information is generated.

2. The cloud-based model simulation control method based on pre-processing according to claim 1, characterized in that: Based on the pre-built model grouping unit, the method for grouping a model object and performing geometric topological analysis to obtain part grouping data is as follows: Obtaining image information of a model object, including several projection views; Adjust several projection views to a fixed size to obtain a model image in a unified format; Use convolutional neural networks to extract features from the model image and generate feature maps; According to the feature map, the geometric features of the model object are encoded and classified to obtain part grouping data.

3. The cloud-based model simulation control method based on pre-processing according to claim 1, characterized in that: The method for obtaining component mesh data by using the pre-built model pre-processing unit to perform geometric feature recognition and meshing on the component group data is as follows: Perform geometric feature recognition on the part grouping data to obtain part face information; part face information includes the geometric elements of the parts and the topological information between the geometric elements; the geometric elements include vertex information, edge information and face information; Process the facial information of the parts to obtain facial geometric features; Based on the face geometry features, a mapping mesh algorithm, a sweeping mesh algorithm, or a filling mesh algorithm is used to mesh the face information of the part to obtain mesh geometry information; The mesh geometry information is repaired by the geometry optimization algorithm, and then the local area of ​​the mesh geometry information is optimized by the topology optimization algorithm to obtain the component mesh data.

4. The cloud-based model simulation control method based on pre-processing according to claim 3, characterized in that: The method for performing geometric feature recognition on part grouping data to obtain part facial information is as follows: Perform geometric feature analysis on the part grouping data to obtain all vertex information; According to the vertex information, adjacent vertices are connected to form edges to obtain edge information; According to the edge information, adjacent edges are connected to form a ring to obtain edge combinations; Based on edge combination, closed surface information is obtained; Based on the geometric center of a model object, the closed surface information is processed to obtain the mid-surface position, that is, the surface information; The face information, edge information and vertex information are coupled and summarized to obtain the part face information.

5. The cloud-based model simulation control method based on pre-processing according to claim 1, characterized in that: Use pre-built model assembly units to identify the assembly relationships between parts, and assemble the component mesh data according to the assembly relationships to obtain the mesh simulation object as follows: Based on a certain model object, the assembly relationship between parts is identified, including bolt connection and / or welding and / or snap connection; and an assembly sequence is generated according to the assembly relationship; Optimize the assembly sequence through dynamic programming method and / or genetic algorithm to generate the final assembly sequence information; According to the final assembly sequence information, the component mesh data is assembled to obtain the mesh simulation object.

6. The cloud-based model simulation control method based on pre-processing according to claim 1, characterized in that: The method for processing the grid simulation object accordingly and generating response feedback information according to the operation request information is as follows: Step 11: The operation request information is a request to synchronously download a simulation object; Parse the request for synchronous download of a simulation object to obtain authentication information, file path information requested for download, and request host name; Step 12: Check the authentication information. If the authentication is successful, proceed to step 13. When the authentication fails, a first response body is generated, which includes an authentication error status code; And execute step 14; Step 13: searching for a specified mesh simulation object according to the file path information. If the mesh simulation object does not exist, generating a first response body, the first response body including a file error status code; And execute step 14; If the grid simulation object exists, generating a second response body, the second response body including a completion status code and the grid simulation object; And execute step 15; Step 14: Process the first response body according to the Hypertext Transfer Protocol to obtain response feedback information, and transmit the response feedback information, i.e., the first response body, to the user terminal based on the request host name; Step 15: Process the second response body according to the Hypertext Transfer Protocol to obtain response feedback information, and transmit the response feedback information, that is, the second response body, to the user end according to the request host name.

7. The cloud-based model simulation control method based on pre-processing according to claim 1, characterized in that: The method for processing the grid simulation object and generating response feedback information according to the operation request information is as follows: Step 21: The operation request information is a request to synchronously upload a simulation object; Parse the request to synchronously upload a simulation object to obtain authentication information, file path information requested for upload, and request host name; Step 22: Check the authentication information. If the authentication is successful, proceed to step 23. When the authentication fails, a third response body is generated, and the third response body includes a creation error status code; And execute step 24; Step 23, uploading a simulation object according to the file path information and the hypertext transfer protocol; and adjusting the grid simulation object according to a certain simulation object; or replacing the original grid simulation object according to a certain simulation object; and generating a fourth response body at the same time, the fourth response body including a creation success status code; And execute step 25; Step 24: Process the third response body according to the Hypertext Transfer Protocol to obtain response feedback information, and transmit the response feedback information, i.e., the third response body, to the user terminal based on the requesting host name; Step 25: Process the fourth response body according to the Hypertext Transfer Protocol to obtain response feedback information, and transmit the response feedback information, that is, the fourth response body, to the user end according to the request host name.

8. The cloud-based model simulation control method based on pre-processing according to claim 1, characterized in that: The method for processing the grid simulation object and generating response feedback information according to the operation request information is as follows: Step 31: The operation request information is a request to delete a simulation object or a request to lock a simulation object; Parse the operation request information to obtain authentication information, file path information for request processing, and request host name; Step 32: Check the authentication information. If the authentication is successful, proceed to step 33. When authentication fails, a failure status code is generated; And execute step 34; Step 33: Deleting or locking the mesh simulation object according to the file path information, and obtaining a success status code; Then execute step 35; Step 34: Process the failure status code according to the Hypertext Transfer Protocol to obtain response feedback information, and transmit the response feedback information to the user end according to the request host name; Step 35: Process the success status code according to the Hypertext Transfer Protocol to obtain response feedback information, and transmit the response feedback information to the user end according to the request host name.

9. A cloud-based model simulation control system based on pre-processing, characterized by: A cloud-based model simulation control method based on pre-processing as described in any one of claims 1 to 8 is applied; the method comprises a pre-processing module, a cloud-based control module, an incremental synchronization module, and a real-time file monitoring module; The pre-processing module includes a model grouping unit, a model pre-processing unit, and a model assembly unit, which are used to group a certain model object and perform geometric topology analysis to obtain a mesh simulation object; The cloud control module is provided with a cloud processing unit for obtaining the request information sent by the user; And according to the request information, the grid simulation object is processed and response feedback information is generated; The incremental synchronization module is built using a block comparison algorithm and is used to transfer the changed parts of the file; The real-time file monitoring module is built based on the file event monitoring mechanism. It is used to capture the save, modify, and delete operations of local files in real time and trigger instant synchronization. At the same time, it can cooperate with asynchronous queue processing to ensure that user operations are not blocked when multiple tasks are concurrent, and realize simultaneous editing and synchronization of files.

Images