Precise mold design optimization method and system based on data mining

Abstract

The invention discloses a precision mold design optimization method and system based on data mining, and the method comprises the steps: obtaining historical mold data in a mold case database, carrying out the analysis of unstructured data in the historical mold data, and extracting key historical data; obtaining current mold data to obtain target material data, target process parameters and target structure data, and obtaining first verification data; screening out a historical mold corresponding to the target material data and the material characteristic data from the key historical data, and optimizing the target process parameters and the target structure data according to the process parameters and the structure design data in the historical mold to obtain an optimized mold; and performing simulation testing on the optimized mold to obtain second verification data, determining an optimization target according to errors between the first verification data and a preset requirement and between the second verification data and a preset requirement, and performing secondary optimization on the current mold or the optimized mold based on the optimization target to obtain a final optimized precision mold.

Description

Technical Field

[0001] The present invention relates to the technical field of mold design and manufacturing, and in particular to a precision mold design optimization method and system based on data mining. Background Art

[0002] Currently, mold design relies heavily on individual engineers' experience, requiring repeated trial and error to adjust design parameters, resulting in low efficiency and high costs. Furthermore, the multi-source, heterogeneous information involved in the design process, including process parameters, material properties, and simulation data, is stored in a decentralized manner and lacks unified integration. However, data mining is currently being used to assist in optimizing precision mold design, thereby reducing its complexity and improving its efficiency.

[0003] However, existing mold design optimization methods can only focus on auxiliary design and optimization of existing or historical molds. For molds with completely new structures or performance requirements, they are prone to rely on data from historical molds for optimization, resulting in molds with completely new structures or performance requirements failing to meet design goals and requirements, and failing to accurately and precisely implement auxiliary design of molds. At the same time, molds that do not meet the standards after optimization using existing methods also increase the complexity of optimizing the design of new molds and reduce the efficiency of design optimization. Summary of the Invention

[0004] The present invention provides a precision mold design optimization method and system based on data mining to solve the technical problem in the existing technology that it is easy to rely on historical mold data for optimization, resulting in molds with completely new structures or performance requirements being unable to meet design goals and requirements.

[0005] In order to solve the above technical problems, an embodiment of the present invention provides a precision mold design optimization method based on data mining, comprising: Obtain historical mold data from a mold case database, and parse unstructured data in the historical mold data to extract key historical data; wherein each historical mold corresponds to a key historical data, and the key historical data includes: material characteristic data, process parameters, structural design data, and verification data; Acquiring current mold data, parsing the current mold data to obtain target material data, target process parameters, and target structure data, and performing simulation testing on the current mold to obtain first verification data; Screening out historical molds corresponding to target material data and material characteristic data from the key historical data, and optimizing the target process parameters and target structure data based on the process parameters and structure design data in the historical molds to obtain an optimized mold; A simulation test is performed on the optimized mold to obtain second verification data, and an optimization target is determined based on the errors between the first verification data and the second verification data and the preset requirements, and the current mold or the optimized mold is optimized again based on the optimization target to obtain the final optimized precision mold.

[0006] As a preferred solution, the acquisition of historical mold data in the mold case database and parsing of unstructured data in the historical mold data to extract key historical data specifically includes: Obtain historical molds and their corresponding historical mold data in the mold case database; Organize the historical mold data corresponding to each historical mold and integrate the unstructured data in the historical mold data; The unstructured data in the historical mold data is subjected to deep semantic analysis to extract material feature data, process parameters, structural design data and verification data of each historical mold.

[0007] As a preferred solution, the step of acquiring current mold data, parsing the current mold data to obtain target material data, target process parameters, and target structure data, and performing simulation testing on the current mold data to obtain first verification data specifically includes: Obtain current mold data, organize the current mold data, and integrate the unstructured data in the current mold data; Performing deep semantic analysis on the unstructured data in the current mold data to extract target material data, target process parameters, and target structure data in the current mold; By presetting simulation test conditions, a simulation test is performed on the current mold, and test data of the current mold is collected during the simulation test as first verification data of the current mold.

[0008] As a preferred solution, the method of screening out historical molds corresponding to target material data and material characteristic data from the key historical data, and optimizing the target process parameters and target structure data according to the process parameters and structure design data in the historical molds to obtain the optimized mold specifically includes: Comparing the material characteristic data in the key historical data with the target material data, thereby screening out historical molds corresponding to the target material data and the material characteristic data; generating mold constraints according to the process parameters and structural design data of the historical mold; determining an optimization weight of a target mold according to the first verification data; According to the mold constraints and optimization weights, an optimization parameter combination is generated by presetting the LSTM prediction optimization model; The target process parameters and target structural data are optimized according to the optimization parameter combination, and an optimized mold is generated based on the optimized target process parameters and target structural data.

[0009] As a preferred solution, the method for constructing the preset LSTM prediction optimization model includes: Build the structure of the initial LSTM model; Preprocessing historical mold data in the mold case database to obtain training data; Designing an optimization algorithm and constructing a loss function based on verification data in the historical mold data; Based on the optimization algorithm, combined with the loss function, and using the training data, iteratively train the initial LSTM model; When the preset number of iterations is reached, the preset LSTM prediction optimization model is obtained.

[0010] As a preferred solution, the optimized mold is simulated and tested to obtain second verification data, and an optimization target is determined based on the errors between the first verification data and the second verification data and the preset requirements, and the current mold or the optimized mold is optimized for a second time based on the optimization target to obtain the final optimized precision mold, specifically including: Performing a simulation test on the optimized mold using preset simulation test conditions, and collecting test data of the optimized mold during the simulation test as second verification data; wherein both the first verification data and the second verification data include a mold trial defect record and a mold life indicator; determining a first error between the first verification data and the second verification data; When a first error between the first verification data and the second verification data is less than a preset range, generating a first optimization target using a preset mold optimization model based on the difference between the mold trial defect record and the mold life index in the first verification data and the preset requirement; generating a second optimization target by using a preset mold optimization model according to the difference between the mold trial defect record and the mold life index in the second verification data and the preset requirement; The first optimization target and the second optimization target are integrated to obtain a final optimization target, and the optimized mold is optimized for a second time based on the final optimization target to obtain a final optimized precision mold.

[0011] As a preferred solution, it also includes: When a first error between the first verification data and the second verification data is not less than a preset range, calculating a second error between the mold trial defect record and the mold life index in the first verification data and the preset requirement, and calculating a third error between the mold trial defect record and the mold life index in the second verification data and the preset requirement; When the second error is less than the third error, a first optimization target is generated using a preset mold optimization model based on the difference between the mold trial defect record and the mold life index in the first verification data and the preset requirement, and the target mold is secondary optimized based on the first optimization target to obtain a final optimized precision mold; When the third error is less than the second error, a second optimization target is generated through a preset mold optimization model based on the difference between the mold trial defect record and mold life index in the second verification data and the preset requirements, and the optimized mold is optimized for a second time based on the second optimization target to obtain the final optimized precision mold.

[0012] Accordingly, the present invention also provides a precision mold design optimization system based on data mining, comprising: A history module is used to obtain historical mold data from the mold case database and parse the unstructured data in the historical mold data to extract key historical data. Each historical mold corresponds to a key historical data, and the key historical data includes: material characteristic data, process parameters, structural design data, and verification data; a parsing module, configured to obtain current mold data, parse the current mold data to obtain target material data, target process parameters, and target structure data, and perform simulation testing on the current mold to obtain first verification data; An optimization module is used to screen out historical molds corresponding to target material data and material characteristic data from key historical data, and optimize the target process parameters and target structure data based on the process parameters and structure design data in the historical molds to obtain an optimized mold; The secondary module is used to perform simulation testing on the optimized mold to obtain second verification data, and determine the optimization target based on the errors between the first verification data and the second verification data and the preset requirements, and perform secondary optimization on the current mold or the optimized mold based on the optimization target to obtain the final optimized precision mold.

[0013] Accordingly, the present invention also provides a terminal device, comprising a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein when the processor executes the computer program, it implements the precision mold design optimization method based on data mining as described in any one of the above.

[0014] Accordingly, the present invention also provides a computer-readable storage medium, which includes a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute any one of the data mining-based precision mold design optimization methods described above.

[0015] Compared with the prior art, the embodiments of the present invention have the following beneficial effects: The technical solution of the present invention obtains historical mold data from a mold case database to extract corresponding key historical data, and obtains current mold data to screen out historical molds corresponding to target material data and material characteristic data, and optimizes based on the historical mold to obtain an optimized mold, and then simulates and tests the optimized mold to obtain second verification data to determine the error between the first verification data and the preset requirements, and finally determines the optimization target, avoiding optimizing a brand new current mold based solely on the data of the optimized mold, and can also be optimized in combination with historical mold data when designing a non-brand new current mold, so as to perform secondary optimization on the current mold or optimized mold based on the optimization target to obtain the final optimized precision mold, thereby reducing the complexity of optimizing the design of a brand new mold and improving the design optimization efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 : A flow chart of a precision mold design optimization method based on data mining provided by an embodiment of the present invention; Figure 2 : A structural diagram of a precision mold design optimization system based on data mining provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0017] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention. Example 1

[0018] Please refer to Figure 1 , a precision mold design optimization method based on data mining provided by an embodiment of the present invention includes the following steps S101-S104: Step S101: Obtain historical mold data from a mold case database, parse the unstructured data in the historical mold data, and extract key historical data; wherein each historical mold corresponds to a key historical data, and the key historical data includes: material characteristic data, process parameters, structural design data, and verification data.

[0019] As a preferred solution of this embodiment, the acquisition of historical mold data in the mold case database and parsing of unstructured data in the historical mold data to extract key historical data specifically includes: Obtain historical molds and their corresponding historical mold data in the mold case database; Organize the historical mold data corresponding to each historical mold and integrate the unstructured data in the historical mold data; The unstructured data in the historical mold data is subjected to deep semantic analysis to extract material feature data, process parameters, structural design data and verification data of each historical mold.

[0020] In this embodiment, the key historical data of the historical mold extracted include but are not limited to material characteristic data, process parameters, structural design data and verification data, wherein the material characteristic data include material mechanical properties and forming performance parameters. The material mechanical properties include shear strength (τ), yield strength (σs), tensile strength (σb), elastic modulus, Poisson's ratio, material heat treatment curve and hardness change, etc. The forming performance parameters include minimum bending radius, elongation of plastic material, rebound angle (Δα), neutral layer coefficient and plastic shrinkage rate, etc. Process parameters include stamping and injection molding parameters. Stamping parameters include blanking clearance (single-side clearance ratio (%t), depending on the material, e.g., 7-10%t for mild steel), drawing coefficient (m value, e.g., initial drawing coefficient of 0.5-0.6 for flangeless cylindrical parts), and flanging coefficient (K value, e.g., the ultimate flanging coefficient of 0.68-0.72 for mild steel). Injection molding parameters include flash value (a key parameter for controlling flash defects, e.g., 0.03-0.05mm for ABS), mold and melt temperatures (affecting the molding shrinkage of crystalline plastics), holding time and cooling rate (needed to match the material's phase change characteristics). Structural design data includes mold cavity dimensions, parting surface location, gating system layout, ejector stroke, stamping part dimensional tolerances, and mold component clearances. Verification data includes mold trial defect records (e.g., sink mark area, weld line location) and mold life indicators (preferably, a lifespan of ≥500,000 cycles for stamping molds and ≥1 million cycles for injection molds).

[0021] In this example, the mold case database covers different types of molds (such as injection molds and stamping molds). Each case includes detailed mold design drawings, process documentation, material specifications, performance test reports, and other information. In addition to data provided by the internal design department, actual manufacturing process data, such as processing parameters and quality inspection results, is collected from the production workshop. Furthermore, external data sources, such as material performance data provided by suppliers and customer feedback on user experiences, are integrated to enrich the data dimensionality.

[0022] In this embodiment, data conversion technology is used to transform unstructured data within historical mold data, such as design drawings (CAD files) and technical documents (PDF, Word, etc.), into structured data. For example, geometric dimensions, tolerances, material annotations, and other information from the design drawings are extracted through the CAD software API and stored in corresponding fields of a relational database. Text mining techniques are then used to extract key parameters and descriptions from the technical documents for structured storage. Furthermore, data from different data sources is fused to resolve issues such as inconsistent data formats and data redundancy, integrating the unstructured data within the historical mold data.

[0023] In this embodiment, natural language processing (NLP) and knowledge graph technology are used to perform deep semantic analysis of text information in unstructured data. Preferably, for descriptive text in mold design documents, NLP methods such as part-of-speech tagging, named entity recognition, and semantic role labeling are used to understand key concepts and relationships in the text, such as material type, process steps, design requirements, etc. Based on the semantic analysis results, a feature extraction algorithm is designed to accurately extract the material feature data, process parameters, structural design data, and verification data of the mold. That is, for material feature data, key features such as strength, toughness, and thermal stability of the material are extracted by parsing the parameter range and performance indicators in the material specification manual; for process parameters, specific values ​​and control ranges such as processing temperature, pressure, and time are extracted from the process file.

[0024] Step S102: Acquire current mold data, and analyze the current mold data to obtain target material data, target process parameters and target structure data, and perform simulation testing on the current mold to obtain first verification data.

[0025] As a preferred solution, the step of acquiring current mold data, parsing the current mold data to obtain target material data, target process parameters, and target structure data, and performing simulation testing on the current mold data to obtain first verification data specifically includes: Obtain current mold data, organize the current mold data, and integrate the unstructured data in the current mold data; Performing deep semantic analysis on the unstructured data in the current mold data to extract target material data, target process parameters, and target structure data in the current mold; By presetting simulation test conditions, a simulation test is performed on the current mold, and test data of the current mold is collected during the simulation test as first verification data of the current mold.

[0026] In this embodiment, relevant data of the current mold is collected, including design drawings, process documents, material specifications, etc. For the unstructured data in the current mold data, such as CAD design drawings (.dwg, .igs, etc. formats) and technical documents (.pdf, .doc, etc. formats), data conversion technology is used to convert them into structured data formats, that is, geometric dimensions, tolerances, material annotations and other information in the design drawings are extracted through the API of the CAD software and stored in the corresponding fields of the relational database; text mining technology is used to extract key parameters and descriptions from the technical documents and perform structured storage, and then natural language processing (NLP) and knowledge graph technology are used to perform deep semantic analysis of the text information in the unstructured data to extract the target material data, target process parameters and target structure data of the current mold.

[0027] In this embodiment, the conditions and parameters of the simulation test are set according to the design requirements and application scenarios of the mold. For example, for injection molds, the simulation test conditions may include injection pressure, temperature, holding time, etc.; for stamping molds, they may include stamping speed, material thickness, mold gap, etc. A simulation test is then performed on the current mold. During the simulation process, the actual working state of the mold is simulated, and mold performance data such as stress distribution, deformation, and filling status are collected as the first verification data of the current mold. At the same time, the verification data generated during the simulation test is collected and stored, and together with other mold data, it constitutes a complete data set, providing a basis for subsequent analysis and optimization.

[0028] Step S103: Filter out historical molds corresponding to target material data and material feature data from the key historical data, and optimize the target process parameters and target structure data according to the process parameters and structure design data in the historical molds to obtain an optimized mold.

[0029] As a preferred solution, the method of screening out historical molds corresponding to target material data and material characteristic data from the key historical data, and optimizing the target process parameters and target structure data according to the process parameters and structure design data in the historical molds to obtain the optimized mold specifically includes: Comparing the material characteristic data in the key historical data with the target material data, thereby screening out historical molds corresponding to the target material data and the material characteristic data; generating mold constraints according to the process parameters and structural design data of the historical mold; determining an optimization weight of a target mold according to the first verification data; According to the mold constraints and optimization weights, an optimization parameter combination is generated by presetting the LSTM prediction optimization model; The target process parameters and target structure data are optimized according to the optimization parameter combination, and an optimized mold is generated based on the optimized target process parameters and target structure data.

[0030] In this embodiment, the target material data of the current mold is compared with the material characteristic data in key historical data. This involves comparing parameters such as material composition, mechanical properties, and thermal properties. Specifically, by comparing key indicators such as yield strength, thermal conductivity, and linear expansion coefficient, the historical mold most similar to the target material data is identified. Based on the comparison results, the historical mold that matches the target material data is selected. Historical molds with similar material properties to the current mold can therefore be directly optimized using their design and process parameters. These historical molds' design and process parameters provide valuable reference and optimization value for the current mold.

[0031] In this embodiment, process parameters (such as injection temperature, holding pressure, cooling time, etc.) and structural design data (such as mold geometry, size, assembly relationship, etc.) are extracted from the screened historical molds. The design requirements of the current mold and the data extracted from the historical molds are combined to generate mold constraints. The constraints include the range of process parameters, structural design restrictions, etc., which are used to guide the optimal design of the current mold.

[0032] In this embodiment, the first validation data of the current mold (e.g., stress distribution, deformation, and filling conditions from simulation tests) is used to evaluate the mold's performance across various performance indicators. Based on the performance evaluation results, optimization weights for the target mold are determined. These weights reflect the relative importance of different performance indicators in the mold optimization process.

[0033] In this embodiment, a predictive optimization model based on a long short-term memory (LSTM) network is established. The LSTM model is trained using mold constraints and optimization weights as input. Model prediction generates multiple optimization parameter combinations. These optimization parameter combinations include different process parameter settings and structural design adjustment schemes. Ultimately, the optimized target process parameters and target structural data are obtained based on these optimization parameter combinations, thereby generating an optimized mold.

[0034] As a preferred solution, the method for constructing the preset LSTM prediction optimization model includes: Build the structure of the initial LSTM model; Preprocessing historical mold data in the mold case database to obtain training data; Designing an optimization algorithm and constructing a loss function based on verification data in the historical mold data; Based on the optimization algorithm, combined with the loss function, and using the training data, iteratively train the initial LSTM model; When the preset number of iterations is reached, the preset LSTM prediction optimization model is obtained.

[0035] In this embodiment, the input and output dimensions of the LSTM model are determined based on the characteristics of the mold design optimization problem. The input may include mold material characteristic data, process parameters, structural design data, etc., and the output is the target optimized mold performance indicator, that is, the optimized parameter combination. The appropriate number of hidden layers and neurons per layer is selected, along with a suitable activation function. The model weight parameters are initialized, either using random initialization or empirically based pre-trained model parameters. Furthermore, the historical mold data in the mold case database is cleaned to remove outliers, missing values, and duplicate data to ensure data accuracy and completeness.

[0036] In this embodiment, a suitable optimization algorithm, such as Adam or RMSprop, is selected based on the characteristics and training requirements of the LSTM model. This algorithm can effectively adjust the weight parameters of the model and minimize the loss function. Then, based on the goal of mold design optimization, a loss function is designed. The mean square error (MSE) loss function can be used to measure the difference between the model prediction value and the actual value, or a composite loss function can be constructed by combining multiple performance indicators to comprehensively consider various aspects of the mold performance.

[0037] In this embodiment, training hyperparameters such as the learning rate, batch size, and number of iterations are determined. The learning rate determines the step size for model parameter updates, the batch size affects the amount of data used for each parameter update, and the number of iterations determines the total duration of model training. The initial LSTM model is iteratively trained using training data. In each iteration, the training data is fed into the model in batches. Predicted values ​​and loss function values ​​are calculated through forward propagation. Then, an optimization algorithm is used for backward propagation to update the model's weight parameters and gradually reduce the loss function value. Ultimately, after reaching the preset number of iterations, the preset LSTM prediction optimization model is obtained.

[0038] Step S104: Perform simulation testing on the optimized mold to obtain second verification data, and determine the optimization target based on the errors between the first verification data and the second verification data and the preset requirements, and perform secondary optimization on the current mold or the optimized mold based on the optimization target to obtain the final optimized precision mold.

[0039] As a preferred solution, the optimized mold is simulated and tested to obtain second verification data, and an optimization target is determined based on the errors between the first verification data and the second verification data and the preset requirements, and the current mold or the optimized mold is optimized for a second time based on the optimization target to obtain the final optimized precision mold, specifically including: Performing a simulation test on the optimized mold using preset simulation test conditions, and collecting test data of the optimized mold during the simulation test as second verification data; wherein both the first verification data and the second verification data include a mold trial defect record and a mold life indicator; determining a first error between the first verification data and the second verification data; When a first error between the first verification data and the second verification data is less than a preset range, generating a first optimization target using a preset mold optimization model based on the difference between the mold trial defect record and the mold life index in the first verification data and the preset requirement; generating a second optimization target by using a preset mold optimization model according to the difference between the mold trial defect record and the mold life index in the second verification data and the preset requirement; The first optimization target and the second optimization target are integrated to obtain a final optimization target, and the optimized mold is optimized for a second time based on the final optimization target to obtain a final optimized precision mold.

[0040] In this embodiment, the preset simulation test conditions can be independently set based on actual conditions and needs. By performing a simulation test on the optimized mold and collecting test data from the optimized mold during the simulation test as second verification data, a first error between the first verification data and the second verification data can be calculated. By determining whether the first error exceeds a preset range, the difference in mold test defect records and mold life indicators between the current mold and the optimized mold can be determined. Ultimately, the final optimization target for the secondary optimization of the optimized mold can be determined, thereby obtaining the final optimized precision mold.

[0041] In this embodiment, a first optimization target is generated using a preset mold optimization model based on the differences between the mold trial defect records and mold life indicators in the first verification data and the preset requirements. The first optimization target primarily addresses the problems and deficiencies of the mold before optimization and proposes improvement directions and goals. A second optimization target is generated using the preset mold optimization model based on the differences between the mold trial defect records and mold life indicators in the second verification data and the preset requirements. The second optimization target primarily addresses the performance of the optimized mold and proposes further optimization directions and goals.

[0042] In this embodiment, the first and second optimization objectives are fused, comprehensively considering the optimization directions and requirements of both, to obtain the final optimization objective. The fusion method can use weighted averaging, multi-objective optimization, and other techniques to ensure that the final optimization objective takes into account both the improvement of the problem before optimization and the performance improvement after optimization.

[0043] In this example, the final optimization goal is applied to the mold optimization. Secondary optimization is performed by adjusting the mold's process parameters and structural design. For example, process parameters such as injection temperature and holding pressure are adjusted, or the mold's geometry, dimensions, and other structural designs are modified to meet the final optimization goal. After secondary optimization, the final optimized precision mold design is obtained.

[0044] It can be understood that when the first error between the first verification data and the second verification data is less than the preset range, it means that the trial mold defect records and mold life indicators between the current mold and the optimized mold are very small, that is, the performance gap between the two is also very small, which makes the process parameters and structural design data between the current mold and the optimized mold very small. Therefore, when performing secondary optimization, the first optimization target corresponding to the first verification data and the second optimization target corresponding to the second verification data can be integrated to ensure that the optimized targets come from the current mold and the optimized mold, avoiding the situation where inaccurate optimization is caused by optimizing data from only a single dimension, and obtaining the final optimization target, so that the optimized mold is optimized for the second time based on the final optimization target to obtain the final optimized precision mold.

[0045] As a preferred solution, it also includes: When a first error between the first verification data and the second verification data is not less than a preset range, calculating a second error between the mold trial defect record and the mold life index in the first verification data and the preset requirement, and calculating a third error between the mold trial defect record and the mold life index in the second verification data and the preset requirement; When the second error is less than the third error, a first optimization target is generated using a preset mold optimization model based on the difference between the mold trial defect record and the mold life index in the first verification data and the preset requirement, and the target mold is secondary optimized based on the first optimization target to obtain a final optimized precision mold; When the third error is less than the second error, a second optimization target is generated through a preset mold optimization model based on the difference between the mold trial defect record and mold life index in the second verification data and the preset requirements, and the optimized mold is optimized for a second time based on the second optimization target to obtain the final optimized precision mold.

[0046] In this embodiment, when the first error between the first and second verification data is no less than a preset range, a second error between the mold test defect records and mold life indicator in the first verification data and the preset requirement is calculated, as is a third error between the mold test defect records and mold life indicator in the second verification data and the preset requirement. Error calculation can employ methods such as mean square error (MSE) and mean absolute error (MAE), with the specific choice determined based on actual needs.

[0047] In this embodiment, the second and third errors are compared to determine which set of verification data is closer to the preset requirements. When the second error is smaller than the third error, a first optimization target is generated using a preset mold optimization model based on the difference between the mold trial defect records and mold life indicators in the first verification data and the preset requirements. Based on the first optimization target, a secondary optimization is performed on the target mold to obtain the final optimized precision mold.

[0048] Furthermore, when the third error is less than the second error, a second optimization target is generated using the preset mold optimization model based on the mold test defect records and the difference between the mold life indicator and the preset requirements in the second verification data. Based on the second optimization target, the optimized mold is optimized again to obtain the final optimized precision mold.

[0049] It is understandable that when the first error between the first verification data and the second verification data is greater than the preset range, it means that the trial mold defect record and mold life index between the current mold and the optimized mold are quite different, that is, the process parameters, structural design data, etc. of the current mold are optimized and changed significantly during the first optimization process. Therefore, it is necessary to further determine the difference between the current mold and the optimized mold and the preset requirements, and then select the one with the smaller error as the target of the secondary optimization, calculate the error between the trial mold defect record and the mold life index in the first verification data and the preset requirements, and finally generate the corresponding optimization target through the preset mold optimization model to achieve the secondary optimization of the target mold or the optimized mold. This avoids the optimization deviation caused by the difference between the newly designed mold and the historical mold data in the database, and at the same time uses the error between the preset requirements to determine the final optimization target, ensuring that the secondary optimization can be directly guided by the preset requirement target result, so that the final optimized precision mold can meet the preset requirements, ensuring the accuracy and efficiency of the precision mold design optimization.

[0050] The implementation of the above embodiment has the following effects: The technical solution of the present invention obtains historical mold data from a mold case database to extract corresponding key historical data, and obtains current mold data to screen out historical molds corresponding to target material data and material characteristic data, and optimizes based on the historical mold to obtain an optimized mold, and then simulates and tests the optimized mold to obtain second verification data to determine the error between the first verification data and the preset requirements, and finally determines the optimization target, avoiding optimizing a brand new current mold based solely on the data of the optimized mold, and can also be optimized in combination with historical mold data when designing a non-brand new current mold, so as to perform secondary optimization on the current mold or optimized mold based on the optimization target to obtain the final optimized precision mold, thereby reducing the complexity of optimizing the design of a brand new mold and improving the design optimization efficiency. Example 2

[0051] See also Figure 2 , which is a precision mold design optimization system based on data mining provided by the present invention, including: The history module 201 is used to obtain historical mold data from the mold case database and parse the unstructured data in the historical mold data to extract key historical data. Each historical mold corresponds to a key historical data, and the key historical data includes: material characteristic data, process parameters, structural design data, and verification data; The parsing module 202 is used to obtain current mold data, parse the current mold data to obtain target material data, target process parameters, and target structure data, and perform simulation testing on the current mold to obtain first verification data; The optimization module 203 is used to select historical molds corresponding to target material data and material characteristic data from the key historical data, and optimize the target process parameters and target structure data based on the process parameters and structure design data in the historical molds to obtain an optimized mold; The secondary module 204 is used to perform simulation testing on the optimized mold to obtain second verification data, and determine the optimization target based on the errors between the first verification data and the second verification data and the preset requirements, and perform secondary optimization on the current mold or the optimized mold based on the optimization target to obtain the final optimized precision mold.

[0052] As a preferred solution, the acquisition of historical mold data in the mold case database and parsing of unstructured data in the historical mold data to extract key historical data specifically includes: Obtain historical molds and their corresponding historical mold data in the mold case database; Organize the historical mold data corresponding to each historical mold and integrate the unstructured data in the historical mold data; The unstructured data in the historical mold data is subjected to deep semantic analysis to extract material feature data, process parameters, structural design data and verification data of each historical mold.

[0053] As a preferred solution, the step of acquiring current mold data, parsing the current mold data to obtain target material data, target process parameters, and target structure data, and performing simulation testing on the current mold data to obtain first verification data specifically includes: Obtain current mold data, organize the current mold data, and integrate the unstructured data in the current mold data; Performing deep semantic analysis on the unstructured data in the current mold data to extract target material data, target process parameters, and target structure data in the current mold; By presetting simulation test conditions, a simulation test is performed on the current mold, and test data of the current mold is collected during the simulation test as first verification data of the current mold.

[0054] As a preferred solution, the method of screening out historical molds corresponding to target material data and material characteristic data from the key historical data, and optimizing the target process parameters and target structure data according to the process parameters and structure design data in the historical molds to obtain the optimized mold specifically includes: Comparing the material characteristic data in the key historical data with the target material data, thereby screening out historical molds corresponding to the target material data and the material characteristic data; generating mold constraints according to the process parameters and structural design data of the historical mold; determining an optimization weight of a target mold according to the first verification data; According to the mold constraints and optimization weights, an optimization parameter combination is generated by presetting the LSTM prediction optimization model; The target process parameters and target structural data are optimized according to the optimization parameter combination, and an optimized mold is generated based on the optimized target process parameters and target structural data.

[0055] As a preferred solution, the method for constructing the preset LSTM prediction optimization model includes: Build the structure of the initial LSTM model; Preprocessing historical mold data in the mold case database to obtain training data; Designing an optimization algorithm and constructing a loss function based on verification data in the historical mold data; Based on the optimization algorithm, combined with the loss function, and using the training data, iteratively train the initial LSTM model; When the preset number of iterations is reached, the preset LSTM prediction optimization model is obtained.

[0056] As a preferred solution, the optimized mold is simulated and tested to obtain second verification data, and an optimization target is determined based on the errors between the first verification data and the second verification data and the preset requirements, and the current mold or the optimized mold is optimized for a second time based on the optimization target to obtain the final optimized precision mold, specifically including: Performing a simulation test on the optimized mold using preset simulation test conditions, and collecting test data of the optimized mold during the simulation test as second verification data; wherein both the first verification data and the second verification data include a mold trial defect record and a mold life indicator; determining a first error between the first verification data and the second verification data; When a first error between the first verification data and the second verification data is less than a preset range, generating a first optimization target using a preset mold optimization model based on the difference between the mold trial defect record and the mold life index in the first verification data and the preset requirement; generating a second optimization target by using a preset mold optimization model according to the difference between the mold trial defect record and the mold life index in the second verification data and the preset requirement; The first optimization target and the second optimization target are integrated to obtain a final optimization target, and the optimized mold is optimized for a second time based on the final optimization target to obtain a final optimized precision mold.

[0057] As a preferred solution, it also includes: When a first error between the first verification data and the second verification data is not less than a preset range, calculating a second error between the mold trial defect record and the mold life index in the first verification data and the preset requirement, and calculating a third error between the mold trial defect record and the mold life index in the second verification data and the preset requirement; When the second error is less than the third error, a first optimization target is generated using a preset mold optimization model based on the difference between the mold trial defect record and the mold life index in the first verification data and the preset requirement, and the target mold is secondary optimized based on the first optimization target to obtain a final optimized precision mold; When the third error is less than the second error, a second optimization target is generated through a preset mold optimization model based on the difference between the mold trial defect record and mold life index in the second verification data and the preset requirements, and the optimized mold is optimized for a second time based on the second optimization target to obtain the final optimized precision mold.

[0058] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working process of the device described above can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.

[0059] The implementation of the above embodiment has the following effects: The technical solution of the present invention obtains historical mold data from a mold case database to extract corresponding key historical data, and obtains current mold data to screen out historical molds corresponding to target material data and material characteristic data, and optimizes based on the historical mold to obtain an optimized mold, and then simulates and tests the optimized mold to obtain second verification data to determine the error between the first verification data and the preset requirements, and finally determines the optimization target, avoiding optimizing a brand new current mold based solely on the data of the optimized mold, and can also be optimized in combination with historical mold data when designing a non-brand new current mold, so as to perform secondary optimization on the current mold or optimized mold based on the optimization target to obtain the final optimized precision mold, thereby reducing the complexity of optimizing the design of a brand new mold and improving the design optimization efficiency. Example 3

[0060] Accordingly, the present invention also provides a terminal device, comprising: a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein when the processor executes the computer program, it implements the precision mold design optimization method based on data mining as described in any one of the above embodiments.

[0061] The terminal device of this embodiment includes: a processor, a memory, and a computer program and computer instructions stored in the memory and capable of running on the processor. When the processor executes the computer program, each step in the above embodiment 1 is implemented, such as Figure 1 Alternatively, when the processor executes the computer program, the functions of each module / unit in the above-mentioned device embodiment, such as the optimization module 203, are implemented.

[0062] Exemplarily, the computer program can be divided into one or more modules / units, and the one or more modules / units are stored in the memory and executed by the processor to complete the present invention. The one or more modules / units can be a series of computer program instruction segments that can perform specific functions, and the instruction segments are used to describe the execution process of the computer program in the terminal device. For example, the optimization module 203 is used to screen out historical molds corresponding to target material data and material feature data from key historical data, and optimize the target process parameters and target structure data based on the process parameters and structural design data in the historical mold, thereby obtaining an optimized mold;

[0063] The terminal device may be a computing device such as a desktop computer, laptop, PDA, or cloud server. The terminal device may include, but is not limited to, a processor and memory. Those skilled in the art will appreciate that the schematic diagram is merely an example of a terminal device and does not limit the terminal device. The terminal device may include more or fewer components than shown, or a combination of certain components, or different components. For example, the terminal device may also include input / output devices, network access devices, buses, and the like.

[0064] The processor may be a central processing unit (CPU), other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor. The processor is the control center of the terminal device and connects various parts of the entire terminal device using various interfaces and lines.

[0065] The memory can be used to store the computer programs and / or modules. The processor implements various functions of the terminal device by running or executing the computer programs and / or modules stored in the memory and calling the data stored in the memory. The memory may mainly include a program storage area and a data storage area. The program storage area may store an operating system, at least one application required for a function, etc.; the data storage area may store data created based on the use of the mobile terminal, etc. In addition, the memory may include high-speed random access memory and non-volatile memory, such as a hard disk, internal memory, a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, at least one disk storage device, a flash memory device, or other volatile solid-state storage device.

[0066] If the module / unit integrated into the terminal device is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present invention can implement all or part of the process steps in the above-mentioned method embodiments by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium. When executed by a processor, the computer program can implement the steps of each of the above-mentioned method embodiments. The computer program includes computer program code, which can be in source code form, object code form, executable file, or some intermediate form. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard drive, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electric carrier signal, telecommunication signal, and software distribution medium. It should be noted that the content of the computer-readable medium can be appropriately increased or decreased based on the requirements of legislation and patent practice in a jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer-readable media does not include electric carrier signals and telecommunication signals. Example 4

[0067] Accordingly, the present invention also provides a computer-readable storage medium, which includes a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute the precision mold design optimization method based on data mining as described in any one of the above embodiments.

[0068] The specific embodiments described above further illustrate the objectives, technical solutions, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the scope of protection of the present invention. In particular, it should be noted that any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included within the scope of protection of the present invention for those skilled in the art.

Claims

1. A precision mold design optimization method based on data mining, characterized in that: include: Obtaining historical mold data from a mold case database, parsing unstructured data in the historical mold data, and extracting key historical data; wherein each historical mold corresponds to a key historical data, and the key historical data includes: material characteristic data, process parameters, and structural design data; Acquire current mold data, analyze the current mold data to obtain target material data, target process parameters, and target structure data, and perform simulation testing on the current mold based on the analyzed data to obtain first verification data; Screening out historical molds whose material characteristic data corresponds to the target material data of the current mold from the key historical data, and optimizing the target process parameters and target structure data based on the process parameters and structure design data in the historical molds to obtain an optimized mold; A simulation test is performed on the optimized mold to obtain second verification data, and an optimization target is determined based on the errors between the first verification data and the second verification data and the preset requirements, and the current mold or the optimized mold is optimized for a second time based on the optimization target to obtain the final optimized precision mold.

2. The method for optimizing precision mold design based on data mining according to claim 1, characterized in that: The process of obtaining historical mold data from a mold case database and parsing unstructured data in the historical mold data to extract key historical data specifically includes: Obtain historical molds and their corresponding historical mold data in the mold case database; Organize the historical mold data corresponding to each historical mold and integrate the unstructured data in the historical mold data; The unstructured data in the historical mold data is subjected to deep semantic analysis to extract material feature data, process parameters, structural design data and verification data of each historical mold.

3. The method for optimizing precision mold design based on data mining according to claim 2, characterized in that: The acquiring of current mold data, parsing the current mold data to obtain target material data, target process parameters, and target structure data, and performing simulation testing on the current mold data to obtain first verification data specifically includes: Obtain current mold data, organize the current mold data, and integrate the unstructured data in the current mold data; Performing deep semantic analysis on the unstructured data in the current mold data to extract target material data, target process parameters, and target structure data in the current mold; By presetting simulation test conditions, a simulation test is performed on the current mold, and test data of the current mold is collected during the simulation test as first verification data of the current mold.

4. A precision mold design optimization method based on data mining as claimed in claim 3, characterized in that: The step of selecting a historical mold corresponding to target material data and material characteristic data from the key historical data, and optimizing the target process parameters and target structure data according to the process parameters and structure design data in the historical mold to obtain an optimized mold specifically includes: Comparing the material characteristic data in the key historical data with the target material data, thereby screening out historical molds corresponding to the target material data and the material characteristic data; generating mold constraints according to the process parameters and structural design data of the historical mold; determining an optimization weight of a target mold according to the first verification data; According to the mold constraints and optimization weights, an optimization parameter combination is generated by presetting the LSTM prediction optimization model; The target process parameters and target structural data are optimized according to the optimization parameter combination, and an optimized mold is generated based on the optimized target process parameters and target structural data.

5. The method for optimizing precision mold design based on data mining according to claim 4, characterized in that: The method for constructing the preset LSTM prediction optimization model includes: Build the structure of the initial LSTM model; Preprocessing historical mold data in the mold case database to obtain training data; Designing an optimization algorithm and constructing a loss function based on verification data in the historical mold data; Based on the optimization algorithm, combined with the loss function, and using the training data, iteratively train the initial LSTM model; When the preset number of iterations is reached, the preset LSTM prediction optimization model is obtained.

6. The method for optimizing precision mold design based on data mining according to claim 1, characterized in that: The simulation test of the optimized mold is performed to obtain second verification data, and an optimization target is determined according to the errors between the first verification data and the second verification data and the preset requirements, and a secondary optimization is performed on the current mold or the optimized mold based on the optimization target to obtain a final optimized precision mold, specifically including: Performing a simulation test on the optimized mold using preset simulation test conditions, and collecting test data of the optimized mold during the simulation test as second verification data; wherein both the first verification data and the second verification data include a mold trial defect record and a mold life indicator; determining a first error between the first verification data and the second verification data; When a first error between the first verification data and the second verification data is less than a preset range, generating a first optimization target using a preset mold optimization model based on the difference between the mold trial defect record and the mold life index in the first verification data and the preset requirement; generating a second optimization target by using a preset mold optimization model according to the difference between the mold trial defect record and the mold life index in the second verification data and the preset requirement; The first optimization target and the second optimization target are integrated to obtain a final optimization target, and the optimized mold is optimized for a second time based on the final optimization target to obtain a final optimized precision mold.

7. The method for optimizing precision mold design based on data mining according to claim 6, characterized in that: Also includes: When a first error between the first verification data and the second verification data is not less than a preset range, calculating a second error between the mold trial defect record and the mold life index in the first verification data and the preset requirement, and calculating a third error between the mold trial defect record and the mold life index in the second verification data and the preset requirement; When the second error is less than the third error, a first optimization target is generated using a preset mold optimization model based on the difference between the mold trial defect record and the mold life index in the first verification data and the preset requirement, and the target mold is secondary optimized based on the first optimization target to obtain a final optimized precision mold; When the third error is less than the second error, a second optimization target is generated through a preset mold optimization model based on the difference between the mold trial defect record and mold life index in the second verification data and the preset requirements, and the optimized mold is optimized for a second time based on the second optimization target to obtain the final optimized precision mold.

8. A precision mold design optimization system based on data mining, characterized in that: include: A history module is used to obtain historical mold data from the mold case database and parse the unstructured data in the historical mold data to extract key historical data. Each historical mold corresponds to a key historical data, and the key historical data includes: material characteristic data, process parameters, structural design data, and verification data; a parsing module, configured to obtain current mold data, parse the current mold data to obtain target material data, target process parameters, and target structure data, and perform simulation testing on the current mold to obtain first verification data; An optimization module is used to screen out historical molds corresponding to target material data and material characteristic data from key historical data, and optimize the target process parameters and target structure data based on the process parameters and structure design data in the historical molds to obtain an optimized mold; The secondary module is used to perform simulation testing on the optimized mold to obtain second verification data, and determine the optimization target based on the errors between the first verification data and the second verification data and the preset requirements, and perform secondary optimization on the current mold or the optimized mold based on the optimization target to obtain the final optimized precision mold.

9. A terminal device, characterized in that: It includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, and when the processor executes the computer program, it implements the precision mold design optimization method based on data mining as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium includes a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute the precision mold design optimization method based on data mining according to any one of claims 1 to 7.

Images