A fast design model reconstruction method and system based on parameterized feature decoupling

Abstract

The application provides a fast design model reconstruction method and system based on parameterized feature decoupling, and relates to the technical field of data processing. The method comprises the following steps: selecting a product template from a parameterized product template library constructed according to a feature decoupling concept; obtaining a new design scheme through a save-as operation based on the selected product template; determining a process selection calculation book through a project requirement based on the new design scheme, inputting main body parameters of the design scheme according to the process selection calculation book and saving to obtain a parameterized design scheme; and inputting parameters of key parts in sequence and saving according to a pre-defined part type in the product template based on the parameterized design scheme to obtain a complete parameter configuration. The application realizes the standardization and normalization of product design, and finally achieves the goal of shortening the development cycle and reducing the product cost.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a rapid design model reconstruction method and system based on parametric feature decoupling. Background Technology

[0002] In the research and development of traditional heat exchange equipment, such as plate heat exchangers, in addition to problems such as a lack of standardized integration in the R&D process, difficulty in transforming personal experience into shared team assets, low module reuse rate, and low efficiency in collaborative design, there is also a core technical pain point: difficulty in rapidly reconstructing design models due to parametric feature coupling. For example, in the design of the top condenser of a crude propylene oxide tower, engineers need to rely on their past experience to start the design from scratch. There is neither a unified product template for reuse nor a standardized module library, such as core, tube-side manifold, and pressure plate module support. Furthermore, the parametric features of components in traditional designs are strongly coupled. The coupling relationship between features and parameters, such as the direct binding of the corrugated structure features of the core plate with the plate thickness parameters, the deep correlation between the flange sealing features of the pipe-side manifold and the box diameter parameters, and the inseparable link between the mounting hole features of the pressure plate module and the plate thickness parameters, leads to constraints on parameter adjustment and model reconstruction. Once a key parameter needs to be modified, such as adjusting the core plate thickness to adapt to working conditions, the features and parameters cannot be separated individually. The associated corrugated structure features must be redesigned simultaneously. It is not possible to quickly reconstruct the model based on the original model. The only option is to manually complete SolidWorks 3D modeling, 2D engineering drawing, and strength calculation, which increases the workload of repetitive tasks.

[0003] Meanwhile, although parts designed by different engineers may meet the dimensional requirements, differences in parametric feature coupling methods lead to parts interfaces that, despite being dimensionally matched, cannot be quickly replaced or reused due to different feature and parameter association logics. In addition, R&D data, such as models, drawings, and calculation sheets, are stored in a scattered manner, lacking a single data control mechanism similar to a TC system. Furthermore, because the coupling relationship between parametric features and parameters has not formed a unified management rule, the association data between features and parameters in the model is chaotic. This often results in inconsistent data in 2D engineering drawings after model modifications due to the failure of feature and parameter associations, and errors or omissions of feature-related parameters in the bill of materials. Summary of the Invention

[0004] The technical problem to be solved by this invention is to provide a rapid design model reconstruction method and system based on parametric feature decoupling, so as to realize the standardization and normalization of product design, and ultimately achieve the goal of shortening the development cycle and reducing product cost.

[0005] To solve the above-mentioned technical problems, the technical solution of the present invention is as follows:

[0006] Firstly, a rapid design model reconstruction method based on parametric feature decoupling, the method comprising:

[0007] Select a product template from the parametric product template library built based on the feature decoupling concept; based on the selected product template, obtain a new design scheme by saving it as a new one; based on the new design scheme, determine the process selection calculation sheet according to the project requirements, input the main parameters of the design scheme according to the process selection calculation sheet and save it to obtain the parametric design scheme;

[0008] Based on the parametric design scheme, according to the predefined component types in the product template, the parameters of key components are entered sequentially and saved to obtain the complete parameter configuration;

[0009] The complete parameter configuration is input into a pre-trained neural network model. By learning and analyzing the parameter-performance mapping relationship in historical successful design cases, the complete parameter configuration is optimized for multiple objectives to obtain the optimized parameter configuration.

[0010] Based on the optimized parameter configuration, perform strength calculations to obtain a strength calculation report; based on the strength calculation report and the optimized parameter configuration, open the associated product model in SolidWorks, perform cloning and borrowing operations on the parts to obtain the cloned model;

[0011] Based on the cloned model and optimized parameter configuration, check the parameters of the non-strength calculation parts. After confirming that they are correct, drive the SolidWorks model to update and obtain the updated 3D model.

[0012] Based on the updated 3D model, update the associated 2D engineering drawing to obtain the updated engineering drawing; store the updated 3D model and engineering drawing in the TC system.

[0013] Furthermore, a product template is selected from the parametric product template library constructed based on the feature decoupling concept; based on the selected product template, a new design scheme is obtained through a save operation; based on the new design scheme, a process selection calculation sheet is determined according to project requirements, the main parameters of the design scheme are input according to the process selection calculation sheet and saved, resulting in a parametric design scheme, including:

[0014] Receive login requests and verify identity information. Once verification is successful, grant access to the rapid design system to obtain the verification result.

[0015] Based on the verification results, the system receives a product template selection instruction and calls the selected product template data from the parameterized product template library built according to the feature decoupling concept. The product template achieves decoupling of product features through parameterization and modular design.

[0016] Based on the product template data retrieved, initiate a save operation to obtain a new design scheme instance;

[0017] Based on the new design scheme example, the main parameters of the design scheme are input based on the process selection calculation sheet, and the main parameters are associated with and saved with the newly generated design scheme example; based on the associated and saved main parameters, the parameterized design scheme is generated.

[0018] Furthermore, based on the parametric design scheme, according to the predefined component types in the product template, the parameters of key components are entered sequentially and saved to obtain a complete parameter configuration, including:

[0019] Based on the parametric design scheme, the predefined component types in the product template are parsed out based on the parametric design scheme.

[0020] For each predefined component type, the key component parameters are received sequentially.

[0021] The input key component parameters are associated with the current components and saved in real time, gradually building a complete parameter configuration corresponding to the design scheme.

[0022] Furthermore, the complete parameter configuration is input into a pre-trained neural network model. By learning and analyzing the parameter-performance mapping relationship in historical successful design cases, the complete parameter configuration is optimized through multi-objective optimization to obtain the optimized parameter configuration, including:

[0023] It receives the complete parameter configuration and sends the complete parameter configuration as input data to the pre-trained neural network model;

[0024] The neural network model establishes a multi-objective optimization function based on the parameter-performance mapping relationship in historical successful design cases. The optimization function includes at least performance indicators, cost indicators, and reliability indicators.

[0025] Based on the established multi-objective optimization function, a multi-objective genetic algorithm is used to optimize and solve the problem, resulting in an initial population containing multiple individual parameters. Based on the initial population, a new generation of population is iteratively generated by simulating selection, crossover, and mutation operations in the process of biological evolution. In the evolution of each generation of population, the Pareto final solution set is obtained by non-dominated sorting and crowding calculation.

[0026] Based on the obtained Pareto final solution set, the parameters are filtered according to the preset performance target constraints to obtain the final parameter configuration that meets the requirements.

[0027] The final parameter configuration is then compared with the historical experience knowledge base to verify its matching degree. After verification that it conforms to the experience patterns of historical successful cases, the optimized parameter configuration is obtained.

[0028] Furthermore, based on the optimized parameter configuration, strength calculations are performed to obtain a strength calculation report. Based on the strength calculation report and the optimized parameter configuration, the associated product model is opened in SolidWorks, and cloning and borrowing operations are performed on the parts to obtain the cloned model, including:

[0029] Receive the optimized parameter configuration, and automatically perform strength calculation based on the optimized parameter configuration to generate the corresponding strength calculation report;

[0030] Based on the strength calculation report and the optimized parameter configuration, SolidWorks is driven to open the associated 3D model of the product.

[0031] In the opened model environment, clone and borrow parts according to preset rules;

[0032] For the cloned parts, the system automatically reassigns drawing numbers and names to obtain the cloned product model.

[0033] Furthermore, based on the cloned model and optimized parameter configuration, the parameters of the non-strength calculation components are checked. After confirming that they are correct, the SolidWorks model is updated to obtain the updated 3D model, including:

[0034] Receive the cloned model and optimized parameter configuration; based on the optimized parameter configuration, automatically check the parameters of non-strength calculation components in the cloned model and perform consistency verification with the optimized parameter configuration;

[0035] Once all parameters of non-strength-calculated components have passed verification, the system automatically drives SolidWorks to parametrically update the model, resulting in an updated 3D model.

[0036] Furthermore, based on the updated 3D model, the associated 2D engineering drawings are updated to obtain updated engineering drawings; the updated 3D model and engineering drawings are stored in the TC system, including:

[0037] Based on the updated 3D model, the 3D model data is automatically extracted and SolidWorks is driven to update all related 2D engineering drawings.

[0038] Perform consistency checks on the two-dimensional engineering drawings to ensure that the two-dimensional engineering drawings are completely consistent with the three-dimensional model;

[0039] The updated 3D model and 2D engineering drawings that have passed verification are automatically stored in the TC system, and a storage record is generated in the TC system, thus completing the entire collaborative design process.

[0040] Secondly, a rapid design model reconstruction system based on parametric feature decoupling includes:

[0041] The acquisition module is used to select a product template from the parametric product template library built according to the feature decoupling concept; based on the selected product template, a new design scheme is obtained through a save operation; based on the new design scheme, the process selection calculation sheet is determined according to the project requirements, the main parameters of the design scheme are input according to the process selection calculation sheet and saved to obtain the parametric design scheme;

[0042] The calculation module is used to input the parameters of key parts sequentially according to the predefined part types in the product template based on the parameterized design scheme and save them to obtain a complete parameter configuration. The complete parameter configuration is then input into a pre-trained neural network model. By learning and analyzing the parameter-performance mapping relationship in historical successful design cases, the complete parameter configuration is optimized through multi-objective optimization to obtain the optimized parameter configuration.

[0043] The cloning module is used to perform strength calculations based on the optimized parameter configuration to obtain a strength calculation report; based on the strength calculation report and the optimized parameter configuration, it opens the associated product model in SolidWorks, performs cloning and borrowing operations on the parts, and obtains the cloned model; based on the cloned model and the optimized parameter configuration, it checks the parameters of the non-strength calculation parts, and after confirming that they are correct, it drives the SolidWorks model to update, and obtains the updated 3D model.

[0044] The processing module is used to update the associated 2D engineering drawing based on the updated 3D model to obtain the updated engineering drawing; and to store the updated 3D model and engineering drawing in the TC system.

[0045] Thirdly, a computing device includes:

[0046] One or more processors;

[0047] A storage device for storing one or more programs that, when executed by one or more processors, cause the one or more processors to implement the method.

[0048] Fourthly, a computer-readable storage medium storing a program that, when executed by a processor, implements the method.

[0049] The above-described solution of the present invention has at least the following beneficial effects:

[0050] By employing technical means such as product template library calls, parametric-driven design, multi-objective optimization of pre-trained neural network models, deep integration with SolidWorks, and single data control and storage in the TC system, and by standardizing the entire process through acquisition, design, calculation, cloning, and processing modules, and by building a product platform module library including core and tube-side manifold modules to improve reusability, this approach overcomes the technical problems in traditional heat exchanger R&D, such as lack of standardized integration of R&D processes, difficulty in transforming individual experience into shared team assets, low module reusability, poor consistency due to scattered R&D data, high cost of multi-role collaborative communication, and unstable design quality. This achieves standardization and normalization of product design, transforms individual experience into team assets, ensures consistency of R&D data throughout the entire process, shortens product development cycles, reduces redundant parts to lower product costs, improves product quality, and ultimately achieves the goal of using less time and less cost. Attached Figure Description

[0051] Figure 1 This is a flowchart illustrating a rapid design model reconstruction method based on parametric feature decoupling provided by an embodiment of the present invention.

[0052] Figure 2 This is a schematic diagram of a rapid design model reconstruction system based on parametric feature decoupling, provided by an embodiment of the present invention. Detailed Implementation

[0053] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

[0054] like Figure 1 As shown, an embodiment of the present invention proposes a rapid design model reconstruction method based on parametric feature decoupling, the method comprising the following steps:

[0055] Step 1: Select a product template from the parametric product template library built based on the feature decoupling concept; Based on the selected product template, obtain a new design scheme by saving it as a new template; Based on the new design scheme, determine the process selection calculation sheet according to the project requirements, input the main parameters of the design scheme according to the process selection calculation sheet and save it to obtain the parametric design scheme.

[0056] Step 2: Based on the parametric design scheme, input the parameters of key parts in sequence according to the predefined part types in the product template and save them to obtain the complete parameter configuration;

[0057] Step 3: Input the complete parameter configuration into the pre-trained neural network model. By learning and analyzing the parameter-performance mapping relationship in historical successful design cases, perform multi-objective optimization on the complete parameter configuration to obtain the optimized parameter configuration.

[0058] Step 4: Perform strength calculations based on the optimized parameter configuration to obtain a strength calculation report; based on the strength calculation report and the optimized parameter configuration, open the associated product model in SolidWorks, perform cloning and borrowing operations on the parts to obtain the cloned model;

[0059] Step 5: Based on the cloned model and optimized parameter configuration, check the parameters of the non-strength calculation parts. After confirming that they are correct, drive the SolidWorks model to update and obtain the updated 3D model.

[0060] Step 6: Update the associated 2D engineering drawing based on the updated 3D model to obtain the updated engineering drawing; store the updated 3D model and engineering drawing in the TC system.

[0061] In this embodiment of the invention, the technical means of selecting a corresponding product template from the product template library by logging into the rapid design system, generating a new design scheme based on the selected template through a save operation, inputting the main parameters in conjunction with the process selection calculation sheet, inputting key part parameters sequentially according to the predefined part types of the product template to form a complete parameter configuration, inputting the complete parameter configuration into a pre-trained neural network model and performing multi-objective optimization by combining the parameter and performance mapping relationship of historical successful design cases, automatically performing strength calculation based on the optimized parameters to generate a calculation sheet, and completing model cloning and borrowing and drawing number reassignment in SolidWorks, checking the non-strength calculation part parameters and driving the SolidWorks model update, linking the update of the two-dimensional engineering drawing, and storing the results in the TC system; thus, it overcomes the technical problems in traditional product development such as lack of standardized integration of R&D processes, difficulty in transforming personal experience into team shared assets, lack of optimization basis for parameter configuration, easy inconsistency between model and engineering drawing data, scattered and chaotic management of R&D data, and low module reuse rate; thereby achieving the standardization and normalization of the product design process, realizing the transformation of personal experience into team experience assets, improving the rationality of parameter configuration and module reuse rate, ensuring the consistency of R&D data throughout the process, shortening the product development cycle, and reducing the redundancy of part types to reduce product costs.

[0062] In a preferred embodiment of the present invention, step 1 above may include:

[0063] Step 1.1: Receive login requests and verify identity information. Upon successful verification, grant access to the rapid design system to obtain the verification result. Specifically, this includes: receiving login requests from engineers to the rapid design system; engineers entering their personal account and password on the login interface, such as the login account information for the Lanzhou Petrochemical Heat Exchanger Product Rapid Design Platform; verifying the authenticity and permissions of the entered identity information, including whether the account exists, whether the password matches, and whether the account has the necessary permissions to access the rapid design system; if the identity information verification passes, the system grants the engineer access to the rapid design system and generates a successful identity verification result; if the verification fails, the system prompts an error in the identity information and refuses to grant access, thus completing the login identity verification process.

[0064] Step 1.2: Based on the verification results, receive the product template selection instruction. Call the selected product template data from the parameterized product template library constructed according to the feature decoupling concept. The product template achieves decoupling of product features through parameterization and modular design. Specifically, this includes: receiving the product template selection instruction issued by the engineer based on the verification results, such as selecting a crude propylene oxide tower top condenser template suitable for specific operating conditions; the product template selection instruction triggers access to the preset parameterized product template library. The parameterized product template library is constructed based on the feature decoupling concept. Each type of product template stored in the library, such as different specifications of plate heat exchangers and tower top condenser templates, achieves decoupling of product features through parameterization and modular design. Each component in the template, such as the core, pipe-side manifold, and pressure plate module, is divided into independent parameterized modules. The geometric features of each module, such as the plate corrugation structure of the core and the flange sealing surface structure of the pipe-side manifold, are decoupled from key parameters, such as the plate thickness of the core and the diameter of the pipe-side manifold. That is, geometric features are used as... The basic structural framework exists independently, and key parameters are stored separately as adjustable variables. The two are flexibly connected through preset association rules rather than being directly bound in traditional designs. For example, the corrugated structure feature of the core module is a fixed basic form, and the plate thickness parameter is an independent variable. When modifying the thickness parameter, only the size adaptation of the corrugated structure needs to be driven by the association rules, without redesigning the corrugated feature itself. The flange bolt hole feature of the pipe-side manifold is an independent structure, associated with the box diameter parameter through positioning rules. When the diameter is adjusted, the bolt holes are automatically repositioned according to the rules, without reconstructing the hole feature. At the same time, the interfaces between modules also adopt standardized parametric design. Interface features, such as connection surfaces and positioning pin holes, are decoupled from interface parameters, such as mating dimensions and tolerances, to ensure that different modules can be assembled by matching interface parameters without adjusting the interface feature structure. According to the selection command, the corresponding product template data is retrieved from the template library. The product template data contains all decoupled parametric modules of parts, the association rules between features and parameters, and the interface standard information between modules.

[0065] Step 1.3: Based on the retrieved product template data, initiate a save as operation to obtain a new design scheme instance. Specifically, after the retrieved product template data, such as the crude propylene oxide tower top condenser template data, is loaded, the engineer initiates a save as operation for the crude propylene oxide tower top condenser template within the rapid design system. The system responds to the save as operation and creates a new design scheme instance based on the loaded template data. The new design scheme instance inherits the basic structural framework of the original template, such as the associated logic of the core, pipe-side manifold, and pressure plate components contained in the template. At the same time, it generates an independent scheme identifier to distinguish it from the original template data, avoiding any modification impact on the original template data, thereby obtaining a completely new design scheme instance.

[0066] Step 1.4: Based on the new design scheme example, input the main parameters of the design scheme based on the process selection calculation sheet, and associate and save the main parameters with the newly generated design scheme example; based on the associated and saved main parameters, drive the generation of parameterized design schemes, specifically including: based on the obtained new design scheme example, the engineer, according to the pre-compiled process selection calculation sheet, such as the process parameter calculation file of the crude propylene oxide tower top condenser, inputs the main parameters of the design scheme item by item in the corresponding parameter input interface of the rapid design system, such as plate thickness, number of plate segments, plate width, number of cores, and core stacking thickness; after the main parameters are input, the engineer clicks the save button in the system to associate and bind the input main parameters with the currently newly generated design scheme example, ensuring that the parameters correspond one-to-one with the scheme, and storing the associated parameters and scheme data; based on the associated and saved main parameters, the system automatically triggers the parameterization driving logic, constructs the basic parameter framework of the design scheme according to the main parameters, and obtains the parameterized design scheme basis.

[0067] In this embodiment of the invention, the technical means of receiving login requests and verifying identity information to grant system access permissions, receiving template selection instructions based on the verification results and calling corresponding template data from a remote product template library, generating a new design scheme instance through a save operation based on the called template data, and inputting main parameters in conjunction with the process selection calculation book and saving them in association with the new scheme instance to drive the generation of parameterized design basis are overcome. Therefore, the technical problems of insufficient security caused by the lack of identity verification in system access in traditional product design, low efficiency due to the lack of a standardized remote calling mechanism for product template acquisition, confusion due to the lack of a unified save operation for new design scheme generation, and easy loss or mismatch of main parameters and design scheme without association are achieved. Thus, the invention achieves the goals of ensuring the security of rapid design system access, improving the efficiency of product template calling, and standardizing the new design scheme generation process.

[0068] In a preferred embodiment of the present invention, step 2 above may include:

[0069] Step 2.1, based on the parametric design scheme, analyzes the predefined component types in the product template. Specifically, this includes: based on the parametric design scheme, activating the component type analysis function. This function reads the pre-set component structure data in the product template, such as the crude propylene oxide tower top condenser template. The data includes the classification definition and attribute information of all components to be designed within the template. By analyzing the component structure data, the predefined component types in the template are automatically identified and extracted, such as core, pipe-side flange collection box, pipe-side collection box, pressure plate module, pressure plate module, plate-side collection box, clamp module, sleeve module, support module, clamping stud, pipe-side gasket, and semi-circular sealing plate. At the same time, the functional positioning and relationship of each component type in the design scheme are clarified, so that engineers clearly know the scope of all components that the current design needs to cover.

[0070] Step 2.2: For each predefined component type, the system sequentially receives the input key component parameters. Specifically, after parsing the predefined component types, the parameters are presented in a pre-defined logical order, usually based on the importance of the components in the overall product structure or assembly logic, such as core components first and auxiliary components later. For the currently displayed component type, the engineer inputs the key component parameters item by item in the parameter input window according to process requirements and design standards. For example, for the core, the number of cores, core stacking thickness, and core spacing need to be input; for the tube-side baffle, the step height, number of teeth, adjacent spacing, width, thickness, and length need to be input. During the parameter input process, the system performs a preliminary check on the input parameter format and value range in real time. If the parameters do not meet the requirements, the engineer is promptly prompted to correct them to ensure the validity of the input parameters.

[0071] Step 2.3 involves associating the input key component parameters with the current component and saving them in real time, gradually building a complete parameter configuration corresponding to the design scheme. Specifically, this includes: whenever an engineer completes and confirms the input of key component parameters for a component type, automatically associating and binding all input parameters with the corresponding component type to establish a unique correspondence between parameters and components, avoiding confusion in parameter attribution; simultaneously, immediately saving the associated parameter data to the database storage area dedicated to this design scheme, recording the parameter input time and operator traceability information during the saving process; as engineers sequentially complete the input and association saving of parameters for all predefined component types, the system continuously integrates the parameter data of each component. Once all component parameters have been input, a complete and logically clear parameter configuration is automatically compiled, directly corresponding to the current design scheme.

[0072] In this embodiment of the invention, a technical approach is adopted to analyze the predefined component types in the product template based on a parametric design scheme, sequentially receive key component parameters for each predefined component type, associate the input key component parameters with the corresponding component and save them in real time to gradually build a complete parameter configuration. This overcomes the technical problems in traditional heat exchanger design, such as the lack of predefined component types leading to confusion in the scope of components during design, the lack of order in inputting key component parameters which can easily result in omissions or mismatches, the lack of association between parameters and components which can easily lead to loss and difficulty in traceability, and the inability to form a complete parameter configuration of the system. Thus, it achieves the goals of clarifying the design scope of components, standardizing the input process of key component parameters, ensuring that parameters and components correspond accurately and are traceable, and gradually building a complete parameter configuration that matches the design scheme.

[0073] In a preferred embodiment of the present invention, step 3 above may include:

[0074] Step 3.1: Receive the complete parameter configuration and send it as input data to the pre-trained neural network model. Specifically, this includes: after the engineer completes the input of key parameters for all predefined components, the system integrates and forms a complete parameter configuration corresponding to the current design scheme, such as the design scheme of the crude propylene oxide tower top condenser. The system automatically receives the complete parameter configuration corresponding to the design scheme of the crude propylene oxide tower top condenser, which includes the key parameters of all components such as the core, the tube-side flange collection box, and the pressure plate module.

[0075] First, the model training data is prepared by collecting historical successful design case data of similar heat exchange equipment, such as fully welded plate heat exchangers, detachable plate heat exchangers, and crude propylene oxide tower top condensers, from the company's accumulated R&D knowledge and historical design achievements. The data covers two core parts: one is the key parameters of all components of the product in each case, such as the number of cores, the stacked thickness of the cores, the size parameters of the flange manifold on the tube side, and the thickness parameters of the pressure plate module; the other is the actual operating performance, manufacturing cost, and long-term reliability of the product under the corresponding parameter combinations, such as heat exchange efficiency, material consumption cost, and structural fatigue strength. Then, the collected data is preprocessed to screen out valid cases with successful designs and complete data, and outliers and invalid data, such as cases with substandard performance due to design errors or incomplete parameter records, are removed. The parameter and indicator data are then standardized and organized to form a dataset that meets the requirements of model training.

[0076] Next, the model structure was designed. Considering the multi-dimensional mapping requirements of parameters, performance, cost, and reliability in product collaborative design, a suitable deep neural network structure was selected. The deep neural network structure includes an input layer, multiple hidden layers, and an output layer: The input layer is used to receive complete key parameters of components in the design scheme, such as the parameters of all predefined components such as the core, tube-side manifold, and pressure plate; The hidden layer learns the complex correlation between component parameters and product performance, cost, and reliability in historical data through nonlinear mapping of multiple neurons, such as the correlation between plate thickness parameters and heat exchange efficiency and material cost, and the correlation between core spacing parameters and equipment operating stability; The output layer outputs intermediate results related to multi-objective optimization, such as product performance scores, cost estimates, and reliability assessments predicted based on input parameters.

[0077] Next, model training and optimization were carried out. The preprocessed historical successful design case dataset was divided into training and validation sets. With the goal of learning the mapping relationship between parameters and performance, cost, and reliability in historical cases, the designed neural network model was trained using the training set data. A loss function was set to minimize the error between the model's predicted values ​​and the actual values ​​of historical cases. The model's weights and bias parameters were iteratively updated to continuously optimize the model's learning effect. After each iteration, the model's performance was tested using the validation set data to determine whether the model could accurately reproduce the correspondence between parameter inputs and indicator outputs in historical cases. For example, if historical parameters of a crude propylene oxide tower top condenser were input, would the model's output values ​​of heat exchange efficiency and cost estimates be consistent with the actual data of that case? If the model's prediction error exceeded the preset range, the number of hidden layers, the number of neurons, or the number of training iterations were adjusted until the model could stably and accurately learn the mapping rules between parameters and various indicators in historical successful design cases and had the ability to analyze based on new component parameter configurations.

[0078] Finally, the model is validated through experience matching. A new batch of successful design cases of similar heat exchange equipment that were not used in the training are called from the enterprise's historical experience knowledge base. The component parameters of these cases are input into the trained model, and the consistency between the model output results and the actual indicator data of the cases is compared. At the same time, it is checked whether the model's learning of parameter correlation rules conforms to historical experience. For example, can the model identify historical experience rules such as when the plate thickness exceeds a certain range, the cost increase will be greater than the performance improvement? After the model output is verified to conform to the experience logic of historical successful cases and the prediction accuracy meets the design requirements, the neural network model is completed.

[0079] Step 3.2: Based on the parameter-performance mapping relationship in historical successful design cases, the neural network model establishes a multi-objective optimization function. The optimization function includes at least performance indicators, cost indicators, and reliability indicators. Specifically, after receiving the complete parameter configuration, the neural network model calls the internally stored historical successful design case database and extracts the parameter-performance mapping relationship of similar products to the current design scheme, such as other specifications of crude propylene oxide tower top condensers and fully welded plate heat exchangers. The mapping relationship covers the actual operating performance, manufacturing cost, and long-term reliability data of the product under different parameter combinations. For example, the correspondence between plate thickness parameters and product heat exchange efficiency, material usage, and structural fatigue strength; the correspondence between core spacing parameters and medium flow resistance, equipment volume, and operational stability. Based on the mapping relationship, the model constructs a multi-objective optimization function, where the performance indicators mainly include the product's heat exchange efficiency, pressure loss, and medium adaptability; the cost indicators mainly include raw material procurement costs, component processing costs, and assembly costs; and the reliability indicators mainly include structural strength, sealing performance, and service life, ensuring that the optimization function can comprehensively cover the core requirements of the product design.

[0080] Step 3.3: Based on the established multi-objective optimization function, a multi-objective genetic algorithm is used to optimize and solve the problem, resulting in an initial population containing multiple individual parameters. Based on the initial population, a new generation of population is iteratively generated by simulating selection, crossover, and mutation operations in the biological evolution process. During the evolution of each generation of population, the Pareto final solution set is obtained through non-dominated sorting and crowding calculation. Specifically, this includes: starting the multi-objective genetic algorithm to optimize and solve the problem based on the established multi-objective optimization function; first, the algorithm uses each parameter in the complete parameter configuration as a variable to generate an initial parameter population, where each individual corresponds to a set of parameter combinations; then, the algorithm iteratively executes selection, crossover, and mutation operations to simulate the biological evolution process and calculates and evaluates the performance, cost, and reliability indicators corresponding to each set of parameter combinations.

[0081] For each parameter combination, such as the core plate thickness, core stack thickness, and pipe-side manifold diameter, the core parameters related to performance are first extracted: plate thickness affects the heat transfer coefficient; the smaller the thickness, the larger the heat transfer coefficient. Core stack thickness determines the total heat exchange area; the larger the stack thickness, the larger the heat exchange area. Pipe-side manifold diameter affects the medium flow velocity; the smaller the diameter, the higher the flow velocity, the stronger the turbulence, but the greater the pressure loss. Combining the logic of the heat transfer efficiency calculation formula, based on the correlation between heat transfer coefficient, heat exchange area, and logarithmic mean temperature difference, the core plate thickness, core stack thickness, and pipe-side manifold diameter are substituted to calculate the basic heat transfer efficiency. At the same time, according to the correlation between medium flow velocity and pipe resistance, the pressure loss value is calculated; the higher the flow velocity, the greater the pressure loss. The heat transfer efficiency and pressure loss are comprehensively scored according to a preset weight, such as heat transfer efficiency accounting for 70% and pressure loss accounting for 30%. The higher the heat transfer efficiency and the lower the pressure loss, the higher the performance target value, thus obtaining the performance target value corresponding to this parameter combination.

[0082] For each parameter combination, cost-related data is extracted from the parameters, including the thickness, quantity, and material parameters of the core plates, the diameter and length of the manifold on the pipe side, and the dimensions of the pressure plate module. The material weight of each component is calculated based on material density and component volume parameters, and then multiplied by the unit weight procurement cost of the corresponding material, such as the unit price of stainless steel, to obtain the total raw material cost. Next, based on the dimensional complexity of the components, such as the complexity of the corrugated structure of the plates and the processing difficulty of the manifold flanges, combined with historical processing cost data, the total processing cost is calculated according to the rule that larger sizes, more complex structures, longer processing times, and higher costs. Finally, the total raw material cost is added to the total processing cost to obtain the total cost corresponding to that parameter combination; the lower the cost, the higher the target cost value.

[0083] For each parameter combination, key reliability-related parameters are extracted: core plate thickness, pipe-side manifold wall thickness, and pressure plate module material. Based on plate thickness, material yield strength, and working pressure, the structural safety factor of the core is calculated. The safety factor equals the material yield strength divided by the actual working stress; stress is inversely proportional to plate thickness. Next, based on the pipe-side manifold wall thickness, diameter, and working pressure, its pressure-bearing safety factor is calculated; a thicker wall results in a higher safety factor. Finally, considering the material's corrosion resistance level (e.g., stainless steel has better corrosion resistance than ordinary carbon steel) and the equipment's design service life requirements, the expected service life is assessed; stronger corrosion resistance results in a longer expected service life. The structural safety factor, pressure-bearing safety factor, and expected service life are weighted according to preset criteria (e.g., each accounting for one-third in the overall score). A higher safety factor and a longer expected service life result in a higher reliability target value, thus obtaining the reliability target value corresponding to that parameter combination.

[0084] Through the above process, each parameter combination will obtain three target values: performance, cost, and reliability. In each iteration, the parameter combinations in the population are divided into different levels according to the superiority or inferiority of the indicators by a non-dominated sorting method, and inferior combinations that are completely surpassed by other combinations are eliminated. At the same time, the degree of difference between combinations of the same level is measured by crowding calculation, and combinations with large differences are retained to maintain population diversity and avoid the optimization results being limited to local optima. After multiple rounds of iteration, when the algorithm converges to a stable state, a set of parameter combinations that cannot improve one indicator without decreasing other indicators is selected, forming the Pareto final solution set. The Pareto final solution set contains multiple parameter configuration schemes that achieve different balance states between performance, cost, and reliability.

[0085] Step 3.4: Based on the obtained Pareto final solution set, the system filters according to preset performance target constraints to obtain the final parameter configuration that meets the requirements. Specifically, this includes: calling the preset performance target constraints, which are determined based on the customer needs, industry standards, and enterprise production capacity of the current design project. For example, for the top condenser of the crude propylene oxide tower, the preset constraints may include a heat exchange efficiency of not less than 92%, a unit heat exchange area manufacturing cost not exceeding a set threshold, structural strength meeting the requirements of GB / T151-2014 standard under design pressure and temperature conditions, and sealing performance ensuring no media leakage. Subsequently, each parameter configuration scheme in the Pareto final solution set is compared with these constraints one by one to filter out all parameter configuration schemes that meet all constraints. If there are multiple schemes that meet the conditions, the system further combines project priorities, such as prioritizing cost control or performance improvement, to select the parameter configuration that best meets the project requirements as the final parameter configuration that meets the requirements.

[0086] Step 3.5 involves verifying the matching degree of the final parameter configuration with the historical experience knowledge base. After verification that it conforms to the experience patterns of historical successful cases, the optimized parameter configuration is obtained. Specifically, this includes: calling upon the company's accumulated historical experience knowledge base, which stores a large amount of experience pattern data on the design of similar heat exchange equipment, including reasonable value ranges for key parameters of similar products, coordination and matching rules between parameters, and parameter taboos that are prone to design errors, such as the matching ratio range of plate thickness and core superposition thickness in the top condenser of crude propylene oxide tower, and the correspondence between the diameter of the pipe-side manifold and the medium flow rate; verifying the final parameter configuration against the experience patterns in the knowledge base item by item, checking whether each parameter in the parameter configuration is within the reasonable value range of historical successful cases, whether the combination of parameters conforms to the experience matching rules, and whether there are parameter settings similar to those in historical failure cases; if the verification results show that the parameter configuration fully conforms to the historical experience patterns, the system confirms that the parameter configuration is practically feasible and determines it as the optimized parameter configuration; if there are parameter items that do not conform to the experience patterns, the system prompts the engineer to review and adjust until the parameter configuration passes the experience matching verification.

[0087] In this embodiment of the invention, the following technical means are employed: receiving a complete parameter configuration and sending it to a pre-trained neural network model; establishing a multi-objective optimization function containing performance, cost, and reliability indicators based on the parameter-performance mapping relationship of historical successful design cases; using a multi-objective genetic algorithm to obtain the Pareto final solution set through non-dominated sorting and crowding calculation; filtering the solution set according to preset performance objective constraints to obtain the final parameter configuration; and verifying the matching degree of the final parameter configuration with a historical experience knowledge base. Therefore, this overcomes the technical problems in traditional heat exchanger design where parameter configuration relies solely on personal experience without scientific optimization basis, making it difficult to balance product performance, cost, and reliability indicators, and lacking historical experience verification of parameter configuration rationality, which easily leads to design errors. Thus, it achieves a multi-objective balance of product performance, cost, and reliability in the parameter configuration optimization process, ensuring that the optimized parameter configuration conforms to the empirical patterns of historical successful cases.

[0088] In a preferred embodiment of the present invention, step 4 above may include:

[0089] Step 4.1: Receive the optimized parameter configuration and automatically perform strength calculations based on it to generate a corresponding strength calculation report. This includes: receiving the optimized parameter configuration, which contains optimized key parameters for all components of the crude propylene oxide tower top condenser, such as the core, the tube-side flange manifold, the pressure plate module, and the plate-side manifold; then, the system starts a preset strength calculation program. This program pre-stores the strength calculation logic for similar heat exchanger components, such as specific calculation methods and industry standards for structural deformation resistance and pressure resistance, as well as relevant rules for structural strength verification and pressure resistance assessment. It automatically extracts core calculation data for each component based on the optimized parameters, such as extracting the core plate thickness and core stack thickness for calculating its structural deformation resistance, and extracting the dimensional parameters of the tube-side manifold for calculating its pressure resistance; based on the extracted core data, the strength calculation program automatically executes various strength calculation operations, ultimately generating detailed information including the strength calculation results for each component, the logic and standards used in the calculations, and whether they meet industry standard requirements, and outputting the corresponding strength calculation report in PDF format.

[0090] Step 4.2: Based on the strength calculation report and optimized parameter configuration, drive SolidWorks to open the associated 3D product model. Specifically, this includes: reading the generated strength calculation report and optimized parameter configuration, establishing a connection with SolidWorks software through the built-in interface program; based on the product type identifier in the optimized parameter configuration, such as the product model and component parameter characteristics of the crude propylene oxide tower top condenser, the system locates the product's 3D basic model matching the current design scheme in the pre-associated model library. The model is a standard 3D structural framework of the corresponding product template, containing the basic form of all predefined components; subsequently, the system automatically sends a command to SolidWorks, driving the SolidWorks software to start and open the associated product's 3D basic model, ensuring that the opened model is completely consistent with the design scheme corresponding to the current optimized parameter configuration and strength calculation report, eliminating the need for engineers to manually search and open it, and avoiding problems such as incorrect model selection or mismatch with the design scheme.

[0091] Step 4.3: In the opened model environment, perform cloning and borrowing operations on components according to preset rules. Specifically, after opening the associated 3D base model of the product in SolidWorks, the system performs cloning and borrowing operations according to preset rules. The borrowing rules are based on the universality of components and design requirements. General standard parts are borrowed first, while non-standard parts or parts that require parameter adjustment are cloned. The system categorizes and processes the components in the model. For general standard parts, such as clamping studs, pipe side gaskets, and standard-specification nuts and washers, the corresponding existing models are directly called from the enterprise component library for borrowing without recreating them. For non-standard parts or parts whose parameters have been optimized and adjusted, such as cores, pipe side flange assembly boxes, and pressure plate modules, whose parameters have been optimized in step 3 and differ from the base model parameters, a cloning operation is performed to copy the structural framework of the corresponding components in the base model, preserving interface compatibility and reserving space for subsequent parameter updates. The entire process is executed automatically according to preset rules, without the need for manual judgment and operation by engineers, ensuring the accuracy and standardization of cloning and borrowing.

[0092] Step 4.4: For the cloned parts, the system automatically reassigns drawing numbers and names to obtain the cloned product model. Specifically, this includes: After completing the part cloning operation, the drawing number and name allocation module is automatically started. This module has a pre-set unified drawing number coding rule for the enterprise. The drawing number includes information such as product model, part type identifier, and version number. For example, the drawing number of the cloned core of the crude propylene oxide tower top condenser can reflect the product number, core type, and optimized version. The part name naming rule is also included, and the name must clearly reflect the part type and the product to which it belongs. The system automatically generates a unique drawing number and corresponding name based on the type of the cloned part, such as the core, the tube-side flange collection box, the product crude propylene oxide tower top condenser, and the optimized parameter characteristics. The system then associates the drawing number and name with the cloned part model. After completing the drawing number and name allocation, the system integrates all the cloned part models into a complete 3D product model, thus obtaining the cloned product model, and records the correspondence between the drawing number and the part.

[0093] In this embodiment of the invention, the following technical means are employed: strength calculation is automatically performed based on optimized parameter configuration to generate a corresponding strength calculation report; SolidWorks is driven to open the associated 3D product model based on the strength calculation report and optimized parameter configuration; cloning and borrowing operations are performed on parts in the opened model environment according to preset rules; and the system automatically reassigns drawing numbers and names to the cloned parts. Therefore, this overcomes the technical problems of low efficiency and error-proneness in manually performing strength calculations in traditional heat exchanger design, the tendency to select the wrong SolidWorks model and mismatch with current design parameters when manually searching and opening the model, the lack of standardized part cloning and borrowing leading to repetitive design or interface incompatibility, and the difficulty in manually assigning drawing numbers due to confusion and lack of traceability. This achieves improved efficiency and accuracy of strength calculations, ensures that the opened SolidWorks model matches the current optimized parameters, reduces repetitive design and increases part reuse rate, and guarantees the uniqueness and traceability of the drawing numbers and names of cloned parts.

[0094] In a preferred embodiment of the present invention, step 5 above may include:

[0095] Step 5.1: Receive the cloned model and optimized parameter configuration; based on the optimized parameter configuration, automatically check the parameters of non-strength calculation components in the cloned model and perform consistency verification with the optimized parameter configuration. Specifically, this includes: receiving the cloned product model, such as the cloned model of the crude propylene oxide tower top condenser, including all component models such as the core, tube-side flange collection box, clamp module, sleeve module, support module, clamping stud, tube-side gasket, and semi-circular sealing plate, as well as the optimized parameter configuration; subsequently, the system automatically identifies non-strength calculation components in the cloned model. Components refer to parts that do not need to participate in the initial strength calculation but are crucial to product assembly and function, such as clamp modules, sleeve modules, support modules, clamping studs, tube-side gaskets, and semi-circular sealing plates; further improving... The current parameters of non-strength calculation components in the cloned model are retrieved, such as the size parameters of the clamp module, the length parameters of the sleeve module, the specifications of the clamping stud, and the thickness parameters of the pipe side gasket. Simultaneously, the target parameters of the corresponding non-strength calculation components are retrieved from the optimized parameter configuration. The extracted current model parameters and the optimized target parameters are then checked for consistency, verifying whether they match perfectly. For example, whether the length parameter of the sleeve module in the model matches the value in the optimized configuration, and whether the specifications of the clamping stud meet the optimization requirements. If the check finds a parameter mismatch in a non-strength calculation component, the location and specific value of the parameter difference are automatically indicated. The engineer corrects the discrepancy based on the prompts, and the check is re-executed. If the parameters of all non-strength calculation components are consistent with the optimized parameter configuration, the consistency check is complete.

[0096] Step 5.2: After all parameters of non-strength calculation components have passed verification, the system automatically drives SolidWorks to perform parametric updates on the model, obtaining the updated 3D model. Specifically, after all parameters of non-strength calculation components have passed consistency verification, the system automatically sends a parametric update command to SolidWorks through the integration interface with the SolidWorks software. The parametric update command carries the target parameters of all components in the optimized parameter configuration, including both strength calculation and non-strength calculation components. Examples include the core stack thickness and plate length parameters of the core, the size parameters of the flange assembly box on the pipe side, the specification parameters of the clip module, and the length parameters of the sleeve module. Upon receiving the update command and target parameters, SolidWorks automatically... The cloned product model is adjusted based on the parameters. For components used in strength calculations, their structural dimensions are modified according to the optimized parameters. For example, the overall thickness of the core is adjusted to match the optimized core stacking thickness, and the diameter and height of the pipe-side flange assembly box are adjusted according to the optimized dimensions. For components not used in strength calculations, their specifications and assembly positions are updated according to the optimized parameters. For example, the dimensions of the sleeve module are adjusted to match the optimized length, and the corresponding model parameters of the clamping stud are replaced according to the optimized specifications. During the model adjustment process, SolidWorks automatically maintains the assembly relationships between components to ensure that each component can still be assembled normally after the parameters are updated. For example, after adjusting the thickness of the pipe-side gasket, the assembly gap of the relevant components is automatically adapted. After all component parameters are updated, a 3D model that perfectly matches the optimized parameter configuration is formed, which is the updated 3D model.

[0097] In this embodiment of the invention, the system receives a cloned model and an optimized parameter configuration. Based on the optimized parameter configuration, it automatically checks the parameters of non-strength calculation components in the cloned model and performs consistency verification between these parameters and the optimized parameter configuration. After all parameters of non-strength calculation components pass the verification, the system automatically drives SolidWorks to perform parameterized updates on the model. This overcomes the technical problems in traditional heat exchanger design where manual checking of non-strength calculation component parameters is prone to omissions or errors, deviations between parameters and optimized configurations are difficult to detect in a timely manner, and manual SolidWorks model updates are inefficient and prone to model-parameter mismatch due to improper operation. Thus, it achieves the technical effects of controlling the accuracy of non-strength calculation component parameters, ensuring complete consistency with the optimized parameter configuration, reducing errors caused by manual intervention, efficiently completing model parameterized updates, and obtaining an updated 3D model that meets the requirements of optimized parameters.

[0098] In a preferred embodiment of the present invention, step 6 above may include:

[0099] Step 6.1: Based on the updated 3D model, automatically extract the 3D model data and drive SolidWorks to update all related 2D engineering drawings. Specifically, this includes: based on the obtained updated 3D model, such as the 3D model of the crude propylene oxide tower top condenser, including the latest structure and parameters of all components such as the core, tube-side flange assembly box, pressure plate module, and plate-side assembly box, automatically starting data extraction to extract key data from the 3D model. Key data includes the actual dimensions of each component, such as the plate thickness and core stacking thickness of the core, the diameter and height of the tube-side flange assembly box, the assembly relationship between components, the overall product structure view information, and related list data, such as the quantity and specifications of each component; subsequently, through interaction with SolidWorks... The lidWorks integration interface transfers the extracted 3D model data to the SolidWorks software and sends a linkage update command, driving SolidWorks to automatically locate all 2D engineering drawings associated with the 3D model, such as core engineering drawings, pipe-side flange assembly box engineering drawings, plate-side assembly box engineering drawings, and complete machine assembly engineering drawings. Based on the received 3D model data, SolidWorks automatically updates the 2D engineering drawings synchronously, including adjusting the dimension annotations to ensure consistency with the actual dimensions of the 3D model, updating the structural views to reflect the latest assembly status of the 3D model, and correcting the list to match the quantity and specifications of parts in the 3D model. The entire process requires no manual operation by engineers, realizing the linkage update of 2D engineering drawings and 3D models.

[0100] Step 6.2: Perform consistency verification on the 2D engineering drawings to ensure complete consistency between the 2D engineering drawings and the 3D model. Specifically, this includes: after the 2D engineering drawings are updated, an automatic consistency verification is initiated to verify all updated 2D engineering drawings. The verification mainly includes three aspects: first, dimensional consistency verification, comparing the dimensions of each component marked on the 2D engineering drawings, such as the width of the side assembly box and the length of the clamping studs, with the actual dimensions of the corresponding components in the 3D model to check for any dimensional deviations; second, visual consistency verification... Figure 1 The consistency check involves verifying whether the structural views in the 2D engineering drawings, such as the front view, top view, and sectional views, fully reflect the latest structure of the 3D model, ensuring no views are missing or the structure is displayed incorrectly, such as not showing the assembly and adjustment of parts in the 3D model. The third check is the list of components consistency check, which compares the list of components attached to the 2D engineering drawings with the list extracted from the 3D model, checking whether the quantity, specifications, and names of the components are completely matched, such as the quantity of pipe-side gaskets and the specifications of semi-circular sealing plates. If inconsistencies are found during the check, such as a dimension in the 2D engineering drawings not matching the dimension in the 3D model, the difference is automatically marked and the specific difference is indicated. After the engineer corrects the 2D engineering drawings according to the prompts, the consistency check is re-executed. This process continues until all 2D engineering drawings are consistent with the 3D model in terms of dimensions, views, and list of components, at which point the check is passed.

[0101] Step 6.3 involves automatically storing the updated 3D model and 2D engineering drawings, which have passed verification, into the TC system and generating a storage record within the TC system, thus completing the entire collaborative design process. Specifically, this includes: after the 2D engineering drawings pass consistency verification, the system automatically establishes a data connection with the TC system, packages the updated 3D model file containing the latest 3D structural data of all components and all verified 2D engineering drawing files, such as core engineering drawings and pipe-side flange assembly box engineering drawings, and sends them to the TC system according to a preset transmission protocol; during file transmission, the system monitors the transmission status in real time to ensure that the files are not damaged or lost; after the files are successfully transmitted to the TC system, the system automatically generates a storage record in the TC system, including the drawing number of the design scheme (e.g., the drawing number HS972.00 for the crude propylene oxide tower top condenser), the storage path of the 3D model and 2D engineering drawings, file version, storage time, and operator information; simultaneously, the TC system categorizes and archives the received files, storing them in association with the preliminary data of the design scheme, such as process selection calculation sheets and strength calculation sheets, forming a complete design data archive.

[0102] In this embodiment of the invention, the technical means of automatically extracting 3D model data based on the updated 3D model and driving SolidWorks to update all related 2D engineering drawings, performing consistency verification on the 2D engineering drawings to check whether the dimensions, annotations, BOM information in the drawings match the corresponding data in the 3D model, and automatically storing the updated 3D model and 2D engineering drawings that pass the verification to the TC system and generating storage records containing information such as storage path, drawing number, and version in the TC system, overcomes the technical problems of low efficiency and easy error in dimensioning due to human operation in the traditional heat exchanger design, inconsistent 3D model and 2D engineering drawing data leading to chaotic production basis, scattered storage of R&D data without a unified management mechanism making it difficult to trace, and inability to efficiently complete the collaborative design process loop. Thus, it achieves the goals of improving the efficiency and accuracy of 2D engineering drawing updates, ensuring complete consistency between 3D model and 2D engineering drawing data to avoid production errors, and realizing unified storage management and full-process traceability of R&D data.

[0103] like Figure 2 As shown, embodiments of the present invention also provide a rapid design model reconstruction system based on parametric feature decoupling, comprising:

[0104] The acquisition module is used to select a product template from the parametric product template library built according to the feature decoupling concept; based on the selected product template, a new design scheme is obtained through a save operation; based on the new design scheme, the process selection calculation sheet is determined according to the project requirements, the main parameters of the design scheme are input according to the process selection calculation sheet and saved to obtain the parametric design scheme;

[0105] The calculation module is used to input the parameters of key parts sequentially according to the predefined part types in the product template based on the parameterized design scheme and save them to obtain a complete parameter configuration. The complete parameter configuration is then input into a pre-trained neural network model. By learning and analyzing the parameter-performance mapping relationship in historical successful design cases, the complete parameter configuration is optimized through multi-objective optimization to obtain the optimized parameter configuration.

[0106] The cloning module is used to perform strength calculations based on the optimized parameter configuration to obtain a strength calculation report; based on the strength calculation report and the optimized parameter configuration, it opens the associated product model in SolidWorks, performs cloning and borrowing operations on the parts, and obtains the cloned model; based on the cloned model and the optimized parameter configuration, it checks the parameters of the non-strength calculation parts, and after confirming that they are correct, it drives the SolidWorks model to update, and obtains the updated 3D model.

[0107] The processing module is used to update the associated 2D engineering drawing based on the updated 3D model to obtain the updated engineering drawing; and to store the updated 3D model and engineering drawing in the TC system.

[0108] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A fast design model reconstruction method based on parameterized feature decoupling, characterized in that, The method comprises: selecting a product template from a parameterized product template library constructed according to a feature decoupling concept, the product template being decoupled from product features through parameterization and modular design, and the product template being suitable for a crude propylene oxide column overhead condenser template under a specific working condition; obtaining a new design scheme through a save-as operation based on the selected product template; determining a process selection calculation sheet based on the new design scheme, inputting main body parameters of the design scheme according to the process selection calculation sheet, and saving to obtain a parameterized design scheme; based on the parameterized design scheme, inputting parameters of key parts in sequence according to pre-defined part types in the product template and saving to obtain a complete parameter configuration; placing the complete parameter configuration into a pre-trained neural network model, performing multi-objective optimization on the complete parameter configuration by learning and analyzing a parameter-performance mapping relationship in historical successful design cases to obtain an optimized parameter configuration; performing strength calculation according to the optimized parameter configuration to obtain a strength calculation sheet; opening an associated product model in SolidWorks based on the strength calculation sheet and the optimized parameter configuration, performing cloning and borrowing operations on parts to obtain a cloned model, including: receiving the optimized parameter configuration and automatically performing strength calculation based on the optimized parameter configuration to generate a corresponding strength calculation sheet; opening an associated product three-dimensional model in SolidWorks according to the strength calculation sheet and the optimized parameter configuration; performing cloning and borrowing operations on parts in the opened model environment according to a preset rule; allocating a part number and a name to the cloned parts automatically to obtain a cloned product model; checking parameters of non-strength calculation parts according to the cloned model and the optimized parameter configuration, and driving a SolidWorks model to update after confirming that the parameters are correct to obtain an updated three-dimensional model, including: receiving the cloned product model, the cloned model of the crude propylene oxide column overhead condenser including all part models of a core, a tube-side method collection box, a clip module, a sleeve module, a support module, a clamping stud, a tube-side gasket, and a semicircular sealing plate, and the optimized parameter configuration; identifying non-strength calculation parts in the cloned model, the parts being parts that do not need to participate in early strength calculation but are essential to product assembly and function, including the clip module, the sleeve module, the support module, the clamping stud, the tube-side gasket, and the semicircular sealing plate; further extracting current parameters of the non-strength calculation parts in the cloned model, including size parameters of the clip module, length parameters of the sleeve module, specification parameters of the clamping stud, and thickness parameters of the tube-side gasket, and calling target parameters of the non-strength calculation parts from the optimized parameter configuration; The consistency of the extracted model current parameters and the optimized target parameters is checked one by one, whether they are completely matched, whether the length parameters of the casing module in the model are consistent with the values in the optimized configuration, and whether the specifications of the clamping studs meet the optimization requirements; if the check finds that the parameters of a certain non-strength calculation component do not match, the parameter difference position and specific value are automatically prompted, and the check is re-executed after correction; if the parameters of all non-strength calculation components are consistent with the optimized parameter configuration, the consistency check is completed; After all the parameters of the non-strength calculation components are checked, the system automatically drives SolidWorks to update the model, and obtains an updated three-dimensional model; According to the updated three-dimensional model, update the associated two-dimensional engineering drawings to obtain updated engineering drawings; store the updated three-dimensional model and engineering drawings to the TC system, including: Based on the updated three-dimensional model, automatically extract three-dimensional model data and drive SolidWorks to update all associated two-dimensional engineering drawings; Conduct consistency check on the two-dimensional engineering drawings to ensure that the two-dimensional engineering drawings are completely consistent with the three-dimensional model; Store the updated three-dimensional model and two-dimensional engineering drawings that pass the check to the TC system and generate a storage record in the TC system, completing the entire collaborative design process.

2. The fast design model reconstruction method based on parameterized feature decoupling according to claim 1, characterized in that, Select a product template from the parameterized product template library constructed based on the feature decoupling concept; based on the selected product template, obtain a new design scheme through the save operation; based on the new design scheme, determine the process selection calculation book through the project requirement, input the main parameters of the design scheme according to the process selection calculation book and save, to obtain a parameterized design scheme, including: Receive a login request and verify the identity information, and grant access to the rapid design system after verification, to obtain a verification result; Based on the verification result, receive a product template selection instruction, call the selected product template data from the parameterized product template library constructed based on the feature decoupling concept, and realize the decoupling of product features through parameterization and modular design of the product template; According to the called product template data, initiate a save operation to obtain a new design scheme instance; According to the new design scheme instance, input the main parameters of the design scheme based on the process selection calculation book, and associate and save the main parameters with the newly generated design scheme instance; generate a parameterized design scheme based on the associated and saved main parameters.

3. The fast design model reconstruction method based on parameterized feature decoupling according to claim 2, characterized in that, Based on the parameterized design scheme, input the parameters of the key parts according to the pre-defined component types in the product template and save, to obtain complete parameter configuration, including: Based on the parameterized design scheme, and based on the parameterized design scheme, parse the pre-defined component types in the product template; For each pre-defined component type, sequentially receive the input key part parameters; Associate the input key part parameters with the current component and save in real time, and gradually build a complete parameter configuration corresponding to the design scheme.

4. The fast design model reconstruction method based on parameterized feature decoupling according to claim 3, characterized in that, The complete parameter configuration is input into the pre-trained neural network model, the parameter and performance mapping relationship in the historical successful design case is learned and analyzed, the complete parameter configuration is multi-objective optimized, and the optimized parameter configuration is obtained, including: Receiving the complete parameter configuration, and sending the complete parameter configuration as input data to the pre-trained neural network model; The neural network model establishes a multi-objective optimization function based on the parameter and performance mapping relationship in the historical successful design case, and the optimization function at least includes performance indicators, cost indicators and reliability indicators; According to the established multi-objective optimization function, a multi-objective genetic algorithm is used for optimization and solution, and an initial population containing multiple individual parameters is obtained; Based on the initial population, a new generation population is iteratively generated through selection, crossover and mutation operations in the biological evolution process; In each generation population evolution process, the non-dominated sorting and congestion degree calculation are used to obtain the final Pareto solution set; Based on the obtained Pareto final solution set, the final parameter configuration meeting the requirements is obtained according to the preset performance target constraint condition; The final parameter configuration is verified by matching degree with the historical experience knowledge base, and the optimized parameter configuration is obtained after verifying that it meets the experience law of the historical successful case.

5. A fast design model reconstruction system based on parameterized feature decoupling, the system implements the method of any one of claims 1 to 4, characterized in that, Including: An acquisition module is configured to select a product template from a parameterized product template library constructed according to a feature decoupling concept; Based on the selected product template, a new design scheme is obtained through a save operation; Based on the new design scheme, a process selection calculation book is determined according to the project requirements, the main body parameters of the design scheme are input according to the process selection calculation book and saved, and a parameterized design scheme is obtained; A calculation module is configured to input the parameters of key parts in sequence according to the pre-defined part types in the product template based on the parameterized design scheme, and save the parameters to obtain a complete parameter configuration; The complete parameter configuration is input into the pre-trained neural network model, the parameter and performance mapping relationship in the historical successful design case is learned and analyzed, the complete parameter configuration is multi-objective optimized, and the optimized parameter configuration is obtained; A cloning module is configured to perform strength calculation according to the optimized parameter configuration to obtain a strength calculation book; based on the strength calculation book and the optimized parameter configuration, an associated product model is opened in SolidWorks, and cloning and borrowing operations are performed on the parts to obtain a cloned model; according to the cloned model and the optimized parameter configuration, the parameters of non-strength calculation parts are checked, and the SolidWorks model is updated after the parameters are confirmed to be correct to obtain an updated three-dimensional model; A processing module is configured to update the associated two-dimensional engineering drawing according to the updated three-dimensional model to obtain an updated engineering drawing; and store the updated three-dimensional model and the engineering drawing to a TC system.

6. A computing device, comprising: Including: One or more processors; A storage device for storing one or more programs, when the one or more programs are executed by the one or more processors, so that the one or more processors implement the method of any one of claims 1 to 4.

7. A computer readable storage medium characterized in that, The computer readable storage medium stores a program, and the program is executed by the processor to implement the method in any one of claims 1 to 4.

Images