Garment process modular design method and system based on parameter adjustability

Abstract

The invention relates to the technical field of garment processes, in particular to a garment process modular design method and system based on adjustable parameters. The method comprises the following steps: acquiring a basic process parameter set; defining a process association relationship according to the basic process parameter set to obtain a process association diagram; performing meta-model construction according to the process association diagram to obtain a process meta-model; obtaining a target style parameter set; mapping parameters in the target style parameter set into the process meta-model, and performing process parameter adjustment to obtain an adjusted process parameter set; performing process flow optimization according to the adjusted process parameter set to obtain a parameterized process model; performing process flow analysis on the parameterized process model to obtain a process flow diagram; and performing module boundary division on the process flow diagram to obtain a module boundary. According to the method, the defects of the existing method in the aspects of efficiency and flexibility are overcome, and a more intelligent, efficient and flexible process design system is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of clothing technology, and in particular to a clothing technology modular design method and system based on adjustable parameters. Background Art

[0002] The clothing manufacturing industry is undergoing a transformation from mass production to personalized customization. Consumers' growing demand for diversified and personalized clothing styles requires companies to respond quickly to market changes and shorten product development cycles. Traditional clothing process design methods mainly rely on manual experience and fixed processes, which are difficult to meet such rapidly changing needs. In order to improve efficiency and flexibility, the modular design method of clothing process came into being. Its core idea is to decompose the complex clothing process into several independent and reusable process modules, and then recombine these modules according to different styles and design requirements to form a new process flow. Some early modular design methods were mainly based on the simple splitting and reorganization of the process flow, lacking parametric control and intelligent support. Later, parametric modeling and CAD / CAM technology began to be applied to clothing process design. Some rule-based expert systems are also used to guide the formulation of process flows.

[0003] Although the modular design method of clothing technology has made some progress, there are still the following disadvantages: 1. The process design efficiency is low; many existing methods still require a lot of manual intervention, such as the division, sorting and connection of modules. Especially for complex styles, the design efficiency is still low.

[0004] 2. Insufficient flexibility in process adjustment: Some methods are not modular enough, and the dependencies between modules are complex, making it difficult to flexibly adjust according to new requirements. For example, modifying a parameter requires redesigning the entire process. Summary of the invention

[0005] Based on this, it is necessary to provide a parameter-adjustable clothing process modular design method and system to solve at least one of the above technical problems.

[0006] To achieve the above purpose, a parameter-adjustable clothing process modular design method is provided, comprising the following steps: Step S1: obtaining a basic process parameter set; defining a process association relationship according to the basic process parameter set to obtain a process association diagram; constructing a metamodel according to the process association diagram to obtain a process metamodel; Step S2: obtaining a target style parameter set; mapping the parameters in the target style parameter set to the process meta-model, adjusting the process parameters, and obtaining an adjusted process parameter set; optimizing the process flow according to the adjusted process parameter set, and obtaining a parameterized process model; Step S3: performing process flow analysis on the parameterized process model to obtain a process flow chart; performing module boundary division on the process flow chart to obtain a module boundary set; performing module information extraction on the parameterized process model according to the module boundary set, and constructing a process module library to obtain a process module library; Step S4: obtaining a new design requirement parameter set; performing module matching on the process module library according to the new design requirement parameter set to obtain a candidate module set; performing process reorganization optimization according to the process meta-model and the candidate module set to obtain a reorganized process flow; Step S5: Performing model process simulation on the reorganized process flow according to the process element model to obtain original simulation data; performing process verification on the original simulation data to obtain simulation results to achieve the clothing process modular design task.

[0007] The present invention lays the foundation for the modular design of the entire clothing process by constructing a process meta-model containing all process information, associations and parametric expressions. It not only realizes the structuring and standardization of process knowledge, but also supports flexible configuration based on parameters, providing a solid foundation for subsequent process flow customization and optimization. Automatic customization of process parameters and processes based on style design. By mapping style parameters to the process meta-model, adjusting parameters and optimizing processes, the system can automatically generate a process flow that matches a specific style design, significantly improving the efficiency and accuracy of process design. The modular decomposition of the process flow and the construction of a process module library are realized. By decomposing the process flow into independent and reusable modules and storing them in a module library, the system can support rapid retrieval, reorganization and reuse of modules, improve the flexibility and efficiency of process design, and promote the accumulation and precipitation of process knowledge. Rapid reconstruction and optimization of process flow based on module library. By intelligently matching, sorting and connecting process modules, the system can quickly generate new process flows according to new design requirements and optimize them to meet different design and production requirements, greatly shortening the process design cycle. Virtual simulation verification and optimization of process flows are realized. By simulating the actual production process in a virtual environment and analyzing and evaluating the simulation results, the system can verify and optimize the process flow, improve the reliability and efficiency of process design, and reduce the risks and costs in actual production. Therefore, the present invention provides a modular design method for clothing process based on adjustable parameters, which effectively overcomes the shortcomings of the existing modular design method for clothing process based on adjustable parameters in terms of efficiency and flexibility through innovations in key links such as module matching, parametric modeling, module division, parameter mapping and process verification, and realizes a more intelligent, efficient and flexible process design system for clothing enterprises.

[0008] Preferably, step S1 comprises the following steps: Step S11: collecting basic process parameters according to common clothing processes to obtain a basic process parameter set; Step S12: defining the process association relationship according to the basic process parameter set to obtain a process association diagram; Step S13: abstracting the process module of the process association diagram to obtain an abstract process module library; Step S14: parametrically expressing the abstract process module library to obtain a parameterized process module library; Step S15: constructing a meta-model for the parameterized process module library, the process association diagram, and the basic process parameter set to obtain a process meta-model.

[0009] The present invention systematically collects the basic parameters of common clothing processes, establishes a comprehensive parameter database, provides a data basis for the subsequent definition of process associations, construction of process modules and parameterized expressions, ensures the integrity and accuracy of process information, and avoids process design problems caused by missing or incorrect information. By defining the associations between processes and presenting them in the form of a graph, the execution order, dependency and conditional parameters between different processes are clearly expressed, providing a logical framework for subsequent process module abstraction and process optimization, making the process clearer, easier to understand and manage, and supporting more complex process logic, such as conditional branches and parallel execution. The process nodes in the process association graph are abstracted into process modules, which realizes the modularization of the process flow, decomposes the complex process flow into independent and reusable units, provides a basis for subsequent module reorganization and process customization, improves the flexibility and efficiency of process design, and facilitates the management and reuse of modules. By expressing the process module in a parameterized form, the flexible configuration and adjustment of the process parameters are realized, so that the process module can adapt to different styles, fabrics and design requirements, enhances the adaptability and customizability of the process design, and provides a basis for subsequent parameter optimization and automatic adjustment. The parameterized process module library, process association diagram and basic process parameters are combined into a unified process meta-model to build a complete and parameterizable process knowledge base, which provides a unified data platform for subsequent process flow generation, module reorganization and simulation verification, realizes the centralized management and efficient utilization of process knowledge, and supports parameter-based process flow customization and optimization.

[0010] Preferably, step S12 comprises the following steps: Step S121: construct a knowledge graph according to the basic process parameter set to obtain a clothing process knowledge graph; Step S122: creating a process node for the basic process parameter set to obtain a process node set; Step S123: performing relationship reasoning on the process node set according to the clothing process knowledge graph to obtain a reasoning relationship set; Step S124: constructing association relationships for the process node set according to the inference relationship set to obtain an initial process association graph; Step S125: Manually review and modify the initial process association diagram to obtain a process association diagram.

[0011] The present invention relates process parameters, material properties and process knowledge in a structured form by constructing a clothing process knowledge graph, providing a rich knowledge base for subsequent relational reasoning and process optimization, enabling the system to more intelligently understand the mutual influence and constraint relationship between process parameters, and supporting more complex process logic reasoning. Each process is created as an independent node, laying the foundation for the subsequent construction of a process association graph and relational reasoning, enabling the system to more clearly express and process the relationship between different processes, and supporting more sophisticated process control and management. By using the knowledge graph for relational reasoning, the implicit association between process nodes can be discovered, such as inferring the required process steps based on fabric characteristics, expanding the scope of process association, enabling the system to automatically infer the complex relationship between processes, not limited to pre-defined rules, thereby improving the automation and intelligence level of the process. The initial process association graph is constructed according to the reasoning relationship set, and the process relationship obtained by reasoning is visualized, which provides convenience for subsequent manual review and correction, enabling experts to more intuitively understand and verify the results of the system's automatic reasoning, and make necessary corrections and supplements, thereby ensuring the accuracy and completeness of the process association graph. The initial process association diagram is reviewed and revised by experts, combining expert experience and domain knowledge, which makes up for the shortcomings of automated reasoning, ensures the accuracy and practicality of the process association diagram, and provides a reliable foundation for subsequent modular decomposition and reorganization.

[0012] Preferably, step S2 comprises the following steps: Step S21: acquiring design parameters for a target style to obtain a target style parameter set; Step S22: Mapping the parameters in the target style parameter set to the process meta-model to obtain a mapping parameter set; Step S23: adjusting the process parameters of the process element model according to the mapping parameter set to obtain an adjusted process parameter set; Step S24: selecting a process module for the process element model according to the adjusted process parameter set, and generating a process flow to obtain an initial process flow; Step S25: Optimize the initial process flow to obtain a parameterized process model.

[0013] The present invention obtains the design parameters of the target style through multiple ways, including obtaining size data through three-dimensional human body scanning, selecting style parameters from a style template library, and querying fabric characteristics from a material database, thereby ensuring the comprehensiveness and accuracy of the target style information and providing a reliable data basis for subsequent parameter mapping and process parameter adjustment. The target style parameters are mapped to the relevant parameters in the process meta-model, a bridge between the style design and the process parameters is established, parameter-driven process flow customization is realized, and the system can automatically adjust the process parameters according to different style designs, thereby improving the efficiency and automation of process design. The parameters in the process meta-model are adjusted according to the mapping parameter set, and the process parameters based on the style design are customized, so that the process parameters and the style design are accurately matched, thereby ensuring that the quality and style of the final product meet the design requirements. According to the adjusted process parameter set, a suitable process module is selected from the process meta-model and an initial process flow is generated, thereby realizing the automatic generation of the process flow, greatly improving the efficiency of the process design, and ensuring that the generated process flow matches the target style parameters. By optimizing the initial process flow, such as parallelizing the parallelizable steps, removing redundant steps, etc., the production efficiency is further improved, the production cost is reduced, and the process flow is made more streamlined and efficient, and finally a parameterized process model optimized for a specific style is obtained.

[0014] Preferably, step S3 comprises the following steps: Step S31: performing process flow analysis on the parameterized process model to obtain a process flow chart; Step S32: identifying key nodes of the process flow chart to obtain a key node set; Step S33: dividing the process flow chart into module boundaries according to the key node set to obtain a module boundary set; Step S34: extracting module information from the parameterized process model according to the module boundary set to obtain a module information set; Step S35: construct a process module library for the module information set to obtain a process module library.

[0015] The present invention converts the parameterized process model into a visualized process flow chart, so that the process flow is more intuitive, easy to understand and analyze, provides a visualized basis for subsequent key node identification and module boundary division, and facilitates manual intervention and adjustment. By identifying key nodes in the process flow chart, such as branch, merge and loop nodes, a basis is provided for module boundary division, the rationality and effectiveness of module division are ensured, and the generated modules are more independent and reusable. The module boundaries are divided according to the key node set, and the process flow is decomposed into a series of independent modules, so that the modularization of the process flow is realized, the foundation is laid for the subsequent module reorganization and reuse, and the flexibility and efficiency of process design are improved. The detailed information of each module is extracted, including process steps, input and output parameters, execution time, resource requirements, and dependencies and constraints between modules, so that complete data is provided for building a process module library, and the integrity and accuracy of module information are ensured. The extracted module information is constructed into a structured process module library, so that the storage, management and retrieval of process modules are realized, and rich module resources are provided for subsequent module matching and reorganization, which is convenient for the reuse and sharing of process modules and supports the accumulation and precipitation of process knowledge.

[0016] Preferably, step S33 includes the following steps: Step S331: performing initial module segmentation on the process flow chart according to the key node set to obtain an initial module segment set; Step S332: performing module complexity evaluation on the initial module segment set to obtain a module complexity index set; Step S333: using a preset complexity threshold to identify high-complexity modules in the module complexity index set, and obtaining a high-complexity module segment set; Step S334: identifying split points of the high-complexity module segment set to obtain split point data; Step S335: redrawing the boundary of the high-complexity module segment set according to the split point data to obtain boundary redrawing data; Step S336: adjusting the module boundaries of the high-complexity module segment set according to the boundary redrawing data to obtain an adjusted module segment set; Step S337: Determine the module boundary according to the adjusted module segment set and the initial module segment set to obtain a module boundary set.

[0017] The present invention performs preliminary segmentation of the process flow chart through key nodes, forms an initial module division scheme, provides a basis for subsequent module adjustment and optimization based on complexity, and avoids the subjectivity and inefficiency of completely manual module division. The complexity of each initial module segment is evaluated, and the complexity of the module is quantified from multiple dimensions such as the number of steps, the number of parameters, the number of parameter types, and the density of dependencies, providing an objective basis for the subsequent identification of high-complexity modules and avoiding the limitations of empirical judgment alone. Through the preset complexity threshold, the module with too high complexity is quickly identified, providing a clear goal for the subsequent module splitting and optimization, avoiding unnecessary splitting of all modules, and improving the efficiency of module division. The split point of the high-complexity module is identified, and the appropriate split position inside the module is found, which provides guidance for the subsequent boundary redrawing, ensures that the complexity of the module after splitting is reduced, and maintains the logical coherence inside the module. The boundary of the high-complexity module is redrawn according to the split point data, and a new module division scheme is generated, which reduces the complexity of the module and improves the reusability and maintainability of the module. The module boundaries are adjusted according to the boundary redrawing data, and the new modules after splitting are integrated into the module set, forming a better module partitioning scheme and improving the granularity and reusability of the modules. The module segments after adjustment are integrated with the initial module segments that have not been split, and the module boundaries are finally determined, forming a complete and optimized module partitioning scheme, which provides accurate module boundary information for subsequent module information extraction and process module library construction.

[0018] Preferably, step S4 comprises the following steps: Step S41: obtaining a new design requirement parameter set; performing module matching on the process module library according to the new design requirement parameter set to obtain a candidate module set; Step S42: Sort the candidate module set according to the process meta-model to obtain a module sequence; Step S43: Connect the module sequence according to the process element model to obtain a module network; Step S44: verify the module network process, and perform process reorganization and optimization to obtain a reorganized process flow.

[0019] The present invention obtains a new design requirement parameter set and performs module matching, and selects candidate modules that meet the new design requirements from the process module library, thereby providing a basis for subsequent module sorting and connection, realizing rapid reconstruction of the process flow based on the module library, and reducing the workload of designing the process flow from scratch. The candidate modules are sorted according to the module dependencies and constraints defined in the process metamodel, ensuring that the order of module connection conforms to the process logic, avoiding unreasonable process flow caused by module connection errors, and ensuring the correctness and feasibility of the generated process flow. The sorted modules are connected according to the connection mode defined in the process metamodel to form a module network, construct a complete process flow structure, and support multiple connection modes such as parallel, branch, and loop, realizing flexible customization of the process flow, and providing a basis for subsequent process verification and optimization. By performing process verification and optimization on the module network, such as checking the correctness of module connection, eliminating circular dependencies, optimizing process parameters, etc., the quality and efficiency of the reorganized process flow are further improved, ensuring that the generated process flow meets the design requirements and process constraints, and achieving optimal performance.

[0020] Preferably, step S41 includes the following steps: Step S411: performing text parsing and keyword extraction on the new design requirement parameter set to obtain the key requirement parameters; Step S412: performing semantic role labeling on the key parameters of the requirements to obtain semantic role labeling data; Step S413: querying the pre-built clothing craft domain ontology knowledge base according to the required key parameters and the semantic role annotation data to obtain ontology knowledge data; Step S414: Structuring the demand intent of the ontology knowledge data to obtain structured demand intent; Step S415: constructing module feature vectors according to the process module library to obtain a module feature vector set; Step S416: Calculate the semantic similarity between the requirements and the modules according to the structured requirements intention and the module feature vector set to obtain a module semantic similarity list; Step S417: Perform module scoring based on machine learning according to the structured demand intention and the module feature vector set to obtain a module matching score list; Step S418: Screening and sorting candidate modules in the module semantic similarity list and the module matching score list to obtain a candidate module set.

[0021] The present invention converts unstructured text information into structured key parameters by performing text parsing and keyword extraction on the new design requirement parameter set, providing a basis for subsequent semantic analysis and module matching, so that the system can understand the core content of the design requirements. Semantic role annotation is performed on the key parameters of the requirements, and the semantic roles of the keywords in the requirement description, such as action, location, object, etc., are identified, so as to more deeply understand the semantic information of the design requirements, and provide more accurate semantic information for subsequent ontology knowledge base query and requirement intention structuring. By querying the clothing technology field ontology knowledge base, the key parameters of the requirements are associated with the domain knowledge, and richer semantic information is obtained, such as the hierarchical relationship and attributes of the concepts, so as to expand the semantic scope of the requirement information and provide more comprehensive knowledge support for more accurate module matching. The ontology knowledge data is combined with the semantic role annotation data, the requirement intention is structured, and a clear and explicit requirement expression is formed, which provides structured data input for subsequent module matching and improves the accuracy and efficiency of module matching. By constructing the module feature vector, the module's text description, input and output parameters and other information are converted into numerical vector representations, providing a data basis for subsequent semantic similarity calculations and machine learning scoring, enabling the system to quantitatively compare and analyze modules. By calculating the semantic similarity between requirements and modules, the correlation between requirements and modules is quantified, providing an important basis for module screening, enabling the system to give priority to modules that are closer to the semantics of requirements. By using machine learning models to score modules, the complex associations between requirements and modules in historical data can be learned, so as to more accurately predict the degree of matching between modules and current requirements, and improve the intelligence level and accuracy of module matching. By comprehensively considering multiple aspects of information such as semantic similarity, machine learning scoring, and parameter constraints, candidate modules are screened and sorted, ensuring that the final selected module not only meets the semantic requirements, but also conforms to the new requirements in terms of parameters, improving the accuracy and reliability of module matching.

[0022] Preferably, step S5 comprises the following steps: Step S51: constructing a simulation environment according to the process element model to obtain a virtual simulation environment; Step S52: importing the reorganized process flow into the virtual simulation environment, performing correlation on the relevant processes, and obtaining the imported process flow; Step S53: simulate and run the imported process flow to obtain original simulation data; Step S54: Analyze and evaluate the original simulation data to obtain a simulation evaluation report; Step S55: Optimize and adjust parameters of the reorganization process flow according to the simulation evaluation report to obtain simulation results.

[0023] The present invention builds a virtual simulation environment according to the process element model, simulates the real clothing production process and equipment in the virtual environment, provides a realistic simulation platform for subsequent process verification and optimization, and avoids the cost and risk of testing in actual production. The reorganized process flow is imported into the virtual simulation environment and the relevant processes are associated, so that the virtual environment can perform simulation according to the reorganized process flow, and the abstract process flow is converted into a specific simulation operation, which provides a basis for subsequent simulation operation and data collection. By running the imported process flow in the virtual simulation environment, the actual clothing production process is simulated, and various key data such as time, material consumption and quality parameters are collected, which provides data support for the subsequent simulation result analysis and evaluation. By analyzing and evaluating the original simulation data, such as calculating the total production time, total material consumption and various product quality indicators, and comparing them with the preset standards, the performance of the reorganized process flow can be comprehensively evaluated, which provides a direction and basis for the subsequent process flow optimization. According to the simulation evaluation report, the parameters of the reorganization process were optimized and adjusted, and the optimization effect was verified through repeated simulation. Finally, a verified and optimized process and the best combination of process parameters were obtained, which maximized production efficiency, reduced costs, and ensured product quality.

[0024] Preferably, the present invention also provides a clothing process modular design system based on adjustable parameters, which is used to execute the clothing process modular design method based on adjustable parameters as described above, and the clothing process modular design system based on adjustable parameters includes: A metamodel construction unit is used to obtain a basic process parameter set; define a process association relationship according to the basic process parameter set to obtain a process association diagram; and construct a metamodel according to the process association diagram to obtain a process metamodel; The parametric design unit is used to obtain a target style parameter set; map the parameters in the target style parameter set to the process meta-model, adjust the process parameters, and obtain an adjusted process parameter set; optimize the process flow according to the adjusted process parameter set to obtain a parametric process model; The modular decomposition unit is used to perform process analysis on the parameterized process model to obtain a process flow chart; divide the process flow chart into module boundaries to obtain a module boundary set; extract module information from the parameterized process model according to the module boundary set, and construct a process module library to obtain a process module library; The module reorganization unit is used to obtain a new design requirement parameter set; perform module matching on the process module library according to the new design requirement parameter set to obtain a candidate module set; perform process reorganization optimization according to the process meta-model and the candidate module set to obtain a reorganized process flow; The process verification unit is used to perform model process simulation on the reorganized process flow according to the process meta-model to obtain original simulation data; perform process verification on the original simulation data to obtain simulation results to realize the modular design task of clothing process. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 A schematic diagram of the steps of a modular design method for garment craftsmanship based on adjustable parameters; Figure 2 Detailed implementation flow chart of step S3 in the present invention; Figure 3 It is a schematic diagram of the detailed implementation steps of step S4 in the present invention.

[0026] The purpose, features and advantages of the present invention will be further described with reference to the accompanying drawings in conjunction with the embodiments. DETAILED DESCRIPTION

[0027] The technical method of the present invention is described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by technicians in this field without creative work are within the scope of protection of the present invention.

[0028] In addition, the accompanying drawings are only schematic illustrations of the present invention and are not necessarily drawn to scale. The same reference numerals in the figures represent the same or similar parts, and their repeated description will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities and do not necessarily correspond to physically or logically independent entities. The functional entities can be implemented in software form, or implemented in one or more hardware modules or integrated circuits, or implemented in different networks and / or processor methods and / or microcontroller methods.

[0029] It should be understood that, although the terms "first", "second", etc. may be used herein to describe various units, these units should not be limited by these terms. These terms are used only to distinguish one unit from another unit. For example, without departing from the scope of the exemplary embodiments, the first unit may be referred to as the second unit, and similarly the second unit may be referred to as the first unit. The term "and / or" used herein includes any and all combinations of one or more of the listed associated items.

[0030] To achieve this, please refer to Figures 1 to 3 , a parameter-adjustable clothing process modular design method, comprising the following steps: Step S1: obtaining a basic process parameter set; defining a process association relationship according to the basic process parameter set to obtain a process association diagram; constructing a metamodel according to the process association diagram to obtain a process metamodel; Step S2: obtaining a target style parameter set; mapping the parameters in the target style parameter set to the process meta-model, adjusting the process parameters, and obtaining an adjusted process parameter set; optimizing the process flow according to the adjusted process parameter set, and obtaining a parameterized process model; Step S3: performing process flow analysis on the parameterized process model to obtain a process flow chart; performing module boundary division on the process flow chart to obtain a module boundary set; performing module information extraction on the parameterized process model according to the module boundary set, and constructing a process module library to obtain a process module library; Step S4: obtaining a new design requirement parameter set; performing module matching on the process module library according to the new design requirement parameter set to obtain a candidate module set; performing process reorganization optimization according to the process meta-model and the candidate module set to obtain a reorganized process flow; Step S5: Performing model process simulation on the reorganized process flow according to the process element model to obtain original simulation data; performing process verification on the original simulation data to obtain simulation results to achieve the clothing process modular design task.

[0031] In the embodiment of the present invention, reference Figure 1 As shown, it is a schematic diagram of the step flow of the clothing process modular design method based on adjustable parameters of the present invention. In this example, the clothing process modular design method based on adjustable parameters includes the following steps: Step S1: obtaining a basic process parameter set; defining a process association relationship according to the basic process parameter set to obtain a process association diagram; constructing a metamodel according to the process association diagram to obtain a process metamodel; In the embodiment of the present invention, first, the basic parameters of commonly used clothing processes are collected, such as sewing line type, stitch length, speed, cutting method, precision, ironing temperature, pressure, etc., to form a basic process parameter set. Then, using the graph database Neo4j, a process association graph is constructed according to the logical relationship and conditional parameters between processes. The nodes represent the processes, and the edges represent the sequential, parallel or conditional relationships between processes. Finally, the process association graph, the basic process parameter set, and the abstracted process modules and their parametric expressions are integrated into Neo4j to form the final process metamodel, which contains standardized descriptions of all process modules and possible connections and combinations between them.

[0032] Step S2: obtaining a target style parameter set; mapping the parameters in the target style parameter set to the process meta-model, adjusting the process parameters, and obtaining an adjusted process parameter set; optimizing the process flow according to the adjusted process parameter set, and obtaining a parameterized process model; In the embodiment of the present invention, the size, style and fabric parameters of the target style are obtained through three-dimensional human body scanning, style template library and material database to form a target style parameter set. Then, these parameters are mapped to the corresponding process parameters in the process meta-model, for example, the chest circumference is mapped to the body width, and the parameter value is adjusted according to the mapping relationship. Finally, according to the adjusted parameter set, a suitable process module is selected from the process meta-model, and the generated initial process flow is optimized using optimization algorithms such as genetic algorithms, such as parallelizing parallelizable steps, removing redundant steps, etc., and finally a parameterized process model for a specific style is obtained.

[0033] Step S3: performing process flow analysis on the parameterized process model to obtain a process flow chart; performing module boundary division on the process flow chart to obtain a module boundary set; performing module information extraction on the parameterized process model according to the module boundary set, and constructing a process module library to obtain a process module library; In an embodiment of the present invention, the process flow of the parameterized process model is visualized as a process flow chart using the Graphviz tool, where nodes represent process steps and edges represent dependencies between steps. Then, the key nodes (branch, merge, and loop nodes) in the flow chart are identified by a depth-first search algorithm, and the module boundaries are divided based on the key nodes to form a module boundary set. Finally, module information is extracted from the parameterized process model according to the module boundary set, including process steps, parameters, dependencies, etc., and this information is stored in a MongoDB database to build a process module library.

[0034] Step S4: obtaining a new design requirement parameter set; performing module matching on the process module library according to the new design requirement parameter set to obtain a candidate module set; performing process reorganization optimization according to the process meta-model and the candidate module set to obtain a reorganized process flow; In an embodiment of the present invention, a new design requirement parameter set input by a user is obtained, such as size modification, style change, etc. Then, NLP technology (word segmentation, semantic role labeling), ontology knowledge base and machine learning model are used to analyze and understand the requirements, and the most suitable candidate module set is matched from the process module library. Finally, according to the module dependencies and constraints defined in the process metamodel, a topological sorting algorithm is used to sort the candidate modules and connect them to form a module network. The module network is then verified and optimized, such as removing redundant modules, adjusting the module order, etc., and finally a reorganized process flow that meets the new design requirements is obtained.

[0035] Step S5: Performing model process simulation on the reorganized process flow according to the process element model to obtain original simulation data; performing process verification on the original simulation data to obtain simulation results to achieve the clothing process modular design task; In an embodiment of the present invention, a virtual simulation environment is built in Clo3D software based on the parameters in the process meta-model, including a virtual human model, clothing styles, fabrics and equipment. Then, the reorganized process flow is converted into an instruction sequence recognizable by Clo3D and run in the simulation environment to collect original simulation data, such as time, material consumption, quality parameters, etc. Next, the simulation data is analyzed to evaluate the efficiency, cost and quality of the process flow, and a simulation evaluation report is generated. Finally, according to the evaluation report, the process flow parameters are adjusted and optimized using optimization algorithms such as genetic algorithms, such as modifying the sewing speed, adjusting the order of process steps, etc., and finally the optimized simulation results are obtained, including the final process parameters and performance indicators.

[0036] Preferably, step S1 comprises the following steps: Step S11: collecting basic process parameters according to common clothing processes to obtain a basic process parameter set; Step S12: defining the process association relationship according to the basic process parameter set to obtain a process association diagram; Step S13: abstracting the process module of the process association diagram to obtain an abstract process module library; Step S14: parametrically expressing the abstract process module library to obtain a parameterized process module library; Step S15: constructing a meta-model for the parameterized process module library, the process association diagram, and the basic process parameter set to obtain a process meta-model.

[0037] In an embodiment of the present invention, first, a database containing all commonly used clothing processes is established, such as sewing, cutting, ironing, bonding, embroidery, etc. Then, for each process, the relevant basic parameters are recorded in detail. Taking the sewing process as an example, the recorded parameters include line type (for example: polyester thread, cotton thread), stitch length (for example: 2mm, 3mm), sewing speed (for example: 800 stitches / minute, 1000 stitches / minute), sewing stitch type (for example: straight line, lock edge), sewing force size (for example: 0.5N, 1N), etc. For the cutting process, the recorded parameters include cutting method (for example: manual cutting, laser cutting), cutting accuracy (for example: ±1mm, ±0.5mm), piece size (for example: length, width), etc. Similarly, record the temperature, pressure, time and other parameters of the ironing process, the adhesive type, glue coating amount, curing time and other parameters of the bonding process, and the embroidery thread type, needle density, embroidery pattern and other parameters of the embroidery process. All collected parameter information is stored in the database to form a structured basic process parameter set, and the value range and measurement unit are set for each parameter.

[0038] The process association graph is created using the graph database Neo4j. Each process collected in step S11 is created as a node in the graph database, and the node attributes include the process name and description. Then, based on the logical relationship between the processes, directed edges are created between the nodes to represent the execution order and dependency relationship of the processes. For example, the "cutting pieces" node points to the "stitching pieces" node, indicating that stitching can only be performed after cutting is completed. The attributes of the edge include the relationship type (for example: sequential relationship, parallel relationship, conditional relationship) and conditional parameters. For example, if the fabric thickness is greater than 3mm, pre-shrinkage treatment is required, then a conditional relationship edge is created between the "fabric detection" node and the "pre-shrinkage treatment" node, and the conditional parameter is "fabric thickness>3mm". For parallel relationships, such as "stitching sleeves" and "stitching body" can be performed at the same time, then a bidirectional edge is created between the two nodes. The final generated process association graph clearly shows the dependency relationship and execution order between different processes, as well as the conditional parameters that trigger different process paths.

[0039] Analyze the process association graph generated in step S12, and aggregate process nodes with similar functions or operation steps into abstract process modules. For example, aggregate nodes such as "cutting the front piece", "cutting the back piece", and "cutting the sleeve piece" into a "cutting module". Each module represents an independent process unit and defines its input and output. For example, the input of the "cutting module" is "fabric", and the output is "cut garment pieces". The input of the "sewing sleeves" module is "cut sleeve pieces" and "cut garment body", and the output is "sewn sleeves". All abstracted process modules and their input and output information are stored in the MongoDB database to form an abstract process module library.

[0040] From the basic process parameter set obtained in step S11, extract the parameters related to each abstract process module and associate them with the module. For example, associate the parameters such as "line type", "needle length", "sewing speed" with "sewing module". In addition, define the parameters of the combination relationship between control modules, such as module connection mode (for example: flat seam, overlock), module sequence, conditional parameters, etc. For example, in addition to sewing parameters, the parameters of the "sew sleeve" module also include connection mode parameters (for example: flat seam, overlock), and connection sequence parameters with the "sew body" module. Add these parameter information to the MongoDB database to form a parameterized process module library, and each module contains the required process parameters and control module combination parameters.

[0041] Import the parameterized process module library generated in step S14 into the Neo4j graph database, and create each module as a node. The node attributes include the module name, description, input and output interfaces, and related process parameters. Then, according to the process association graph generated in step S12, establish connections between the module nodes to represent the dependencies and execution order between the modules. The attributes of the edge include the relationship type and conditional parameters. At the same time, the basic process parameter set in step S11 is also imported into Neo4j and associated with the corresponding module nodes. Finally, the information in the Neo4j database constitutes a complete process metamodel, which includes standardized descriptions of all process modules, possible connections and combinations between them, and related process parameters, and all information is expressed in a parameterized form.

[0042] Preferably, step S12 comprises the following steps: Step S121: construct a knowledge graph according to the basic process parameter set to obtain a clothing process knowledge graph; Step S122: creating a process node for the basic process parameter set to obtain a process node set; Step S123: performing relationship reasoning on the process node set according to the clothing process knowledge graph to obtain a reasoning relationship set; Step S124: constructing association relationships for the process node set according to the inference relationship set to obtain an initial process association graph; Step S125: Manually review and modify the initial process association diagram to obtain a process association diagram.

[0043] In an embodiment of the present invention, a graph database Neo4j is used to construct a clothing process knowledge graph. Each process parameter in the basic process parameter set obtained in step S11 is used as a node in the knowledge graph, such as "line type", "needle length", "fabric thickness", etc. At the same time, related processes and materials are also created as nodes, such as "sewing", "cutting", "cotton", "wool", etc. Then, relationships are established between nodes based on process specifications, material properties and expert experience. For example, relationships (cotton, shrinkage, high), (wool, elasticity, medium), (sewing, line type, polyester thread), (needle length, influence, stitching strength), etc. are created. Relationship types include "material properties", "process requirements", "parameter influence", etc. In this way, a clothing process knowledge graph containing the relationship between fabrics, processes and parameters is constructed and stored in the Neo4j database.

[0044] From the basic process parameter set obtained in step S11, extract each process, such as sewing, cutting, ironing, etc., and create corresponding process nodes in the Neo4j graph database. The attributes of each node contain the name, description and related parameter information of the process. For example, create a node named "sewing", which is described as "the process of connecting two or more pieces of cloth together with sewing thread", and the related parameters include "line type", "needle length", "sewing speed", etc. These parameters are connected to the corresponding parameter nodes through relationships. All created process nodes constitute a process node set.

[0045] Using the clothing process knowledge graph constructed in step S121 and the graph algorithm provided by the APOC library, the process node set created in step S122 is subjected to relational reasoning. For example, it is known that there are relations (wool, shrinkage rate, high) and (high shrinkage rate, pre-shrinkage treatment, necessary) in the knowledge graph, as well as process nodes "cutting" and "pre-shrinkage treatment". Through the path finding algorithm `apoc.path.expandConfig` in the APOC library, it can be inferred that if the fabric is wool, the "pre-shrinkage treatment" node needs to be executed after the "cutting" node, and this reasoning relationship (cutting, conditional relationship; fabric = wool, pre-shrinkage treatment) is added to the reasoning relationship set. The reasoning relationship set stores the relationship between all process nodes obtained through knowledge graph reasoning.

[0046] In the Neo4j graph database, according to the inference relationship set obtained in step S123, connections are established between the process nodes created in step S122 to form an initial process association graph. Different types of relationships are used to represent different relationship types. For example, a solid arrow is used to represent a sequential relationship, a dotted arrow is used to represent a conditional relationship, and a bidirectional arrow is used to represent a parallel relationship. For a conditional relationship, the parameters of the triggering condition and the corresponding parameter values ​​are stored as attributes of the relationship. For example, for the relationship (cutting, conditional relationship; fabric = wool, pre-shrunk treatment), "fabric = wool" is used as an attribute of the relationship.

[0047] The domain experts review the initial process association diagram generated in step S124, for example, to check whether the connections between process nodes are correct, whether the condition parameters are reasonable, whether there are missing or redundant relationships, etc. The experts can directly modify the process association diagram through the interface of the Neo4j graph database, such as adding, deleting or modifying nodes and edges, and adjusting condition parameters. The reviewed and revised process association diagram serves as the final process association diagram, providing a basis for subsequent modular decomposition and reorganization.

[0048] Preferably, step S2 comprises the following steps: Step S21: acquiring design parameters for a target style to obtain a target style parameter set; Step S22: Mapping the parameters in the target style parameter set to the process meta-model to obtain a mapping parameter set; Step S23: adjusting the process parameters of the process element model according to the mapping parameter set to obtain an adjusted process parameter set; Step S24: selecting a process module for the process element model according to the adjusted process parameter set, and generating a process flow to obtain an initial process flow; Step S25: Optimize the initial process flow to obtain a parameterized process model.

[0049] In the embodiment of the present invention, the size parameters of the target style, such as chest circumference, waist circumference, hip circumference, length of clothing, sleeve length, etc., are obtained through a three-dimensional body scanner. At the same time, the style of the target style is selected through a pre-defined style template library, such as slim fit, loose fit, A-line, etc. In addition, the characteristic parameters of the fabric used for the target style, such as thickness, elasticity, shrinkage, etc., are queried through the material database. The obtained size parameters, style parameters, and material parameters are stored in a file in JSON format to form a target style parameter set.

[0050] Read the target style parameter set in JSON format generated in step S21, and map the parameters to the process metamodel generated in step S15. For example, map the "chest circumference" parameter in the target style parameter set to the "body width" parameter of the "cutting module" in the process metamodel, and map the "length" parameter to the "body length" parameter of the "cutting module". For style parameters, such as the "slim fit" style, map it to the "stitch length" parameter of the "sewing module" and set it to a smaller stitch length value. Store all mapping relationships and corresponding parameter values ​​in another JSON file to form a mapping parameter set. The mapping relationship is based on a pre-defined rule base and machine learning model.

[0051] Read the mapping parameter set generated in step S22, and adjust the parameters in the process meta-model generated in step S15 according to the mapping relationship and parameter values ​​therein. For example, according to the rule of "body width = chest circumference / 2+5cm" in the mapping parameter set, calculate the parameter value of "body width" of "cutting module". According to the rule of "stitch length = 2mm" corresponding to the "slim fit" style, set the parameter value of "stitch length" of "sewing module" to 2mm. Store all adjusted process parameter values ​​in a new JSON file to form an adjusted process parameter set.

[0052] Read the adjusted process parameter set generated in step S23, and select a suitable process module from the process metamodel generated in step S15 according to the parameter values ​​therein. For example, according to the parameter values ​​of "body width" and "body length", select "cutting module"; according to the parameter values ​​of "line type", "needle length", "sewing speed", etc., select "sewing module". Then, according to the dependency relationship and execution order between the modules defined in the process metamodel, connect the selected modules to form an initial process flow. The generated initial process flow is stored in the Neo4j database in the form of a directed acyclic graph, where the nodes in the graph represent the process modules and the edges represent the execution order between the modules.

[0053] The initial process flow generated in step S24 is optimized. For example, if there are two or more modules that can be executed in parallel, the process is adjusted so that these modules are executed in parallel to shorten the production time. If there are redundant process steps, they are deleted to improve production efficiency. The optimization process is based on pre-defined optimization rules and genetic algorithms. The optimized process flow together with the adjusted process parameter set generated in step S23 constitutes a parameterized process model, and the parameterized process model is stored in the Neo4j database, which contains the process parameters for a specific style and the optimized process flow.

[0054] Preferably, step S3 comprises the following steps: Step S31: performing process flow analysis on the parameterized process model to obtain a process flow chart; Step S32: identifying key nodes of the process flow chart to obtain a key node set; Step S33: dividing the process flow chart into module boundaries according to the key node set to obtain a module boundary set; Step S34: extracting module information from the parameterized process model according to the module boundary set to obtain a module information set; Step S35: construct a process module library for the module information set to obtain a process module library.

[0055] As an example of the present invention, refer to Figure 2 As shown, in this example, step S3 includes: Step S31: performing process flow analysis on the parameterized process model to obtain a process flow chart; In the embodiment of the present invention, the parameterized process model generated in step S25 is read from the Neo4j database, including the process flow and process parameters. The process flow in the parameterized process model is converted into a visual process flow chart using the Graphviz tool. Each node in the figure represents a process step (for example, cutting the front piece, sewing the shoulder, and ironing the collar), and the node attributes include information such as the step name, execution time, and required resources. The directed edges between the nodes represent the sequence and dependency between the steps, and the dependency type (for example, sequential, parallel, conditional) is marked on the edge.

[0056] Step S32: identifying key nodes of the process flow chart to obtain a key node set; In the embodiment of the present invention, the process flow chart generated in step S31 is analyzed, and the flow chart is traversed using a depth-first search algorithm to identify key nodes. Key nodes include branch nodes (nodes with multiple subsequent steps), merge nodes (nodes with multiple predecessor steps), and loop nodes (nodes that constitute a loop). The identified key nodes are stored in a list to form a key node set, and the list elements include node IDs and node types. For example, if the step of "sewing cuffs" can be followed by "sewing side seams" or "sewing collars", then "sewing cuffs" is identified as a branch node and added to the key node set.

[0057] Step S33: dividing the process flow chart into module boundaries according to the key node set to obtain a module boundary set; In an embodiment of the present invention, the process flow chart is divided into several modules according to the key node set in step S32. The module boundary is bounded by the key node. The process step sequence between two adjacent key nodes is regarded as a module, the step sequence between the starting node and the first key node, and the step sequence between the last key node and the ending node are also regarded as modules respectively. The starting node and the ending node ID of each module are recorded, and this information is stored in a list to form a module boundary set. For example, if "cutting garment pieces" is the starting node and "stitching shoulders" is the key node, all steps between "cutting garment pieces" and "stitching shoulders" constitute a module, and the node IDs of "cutting garment pieces" and "stitching shoulders" are recorded in the module boundary set.

[0058] Step S34: extracting module information from the parameterized process model according to the module boundary set to obtain a module information set; In an embodiment of the present invention, the information of each module is extracted from the parameterized process model generated in step S25 according to the module boundary set generated in step S33. For example, the process steps contained in the module, the input and output parameters of each step, the execution time, the resource requirements, etc. are extracted. In addition, the dependencies and constraints between the modules are extracted. For example, module A must be executed before module B, and module C needs to meet specific conditions before it can be executed. The information of each module is stored as a JSON object, including module ID, module name, process step list, input and output parameter list, dependency list, etc. The JSON objects of all modules are stored in a list to form a module information set.

[0059] Step S35: constructing a process module library for the module information set to obtain a process module library; In the embodiment of the present invention, the module information set generated in step S34 is imported into the MongoDB database to construct a process module library. Each module is stored in the database as a document, and the document fields include module ID, module name, process step list, input and output parameter list, dependency list, etc., which are consistent with the JSON object structure in step S34. Through the indexing function of the MongoDB database, module information can be quickly retrieved according to module ID, name, etc.

[0060] Preferably, step S33 includes the following steps: Step S331: performing initial module segmentation on the process flow chart according to the key node set to obtain an initial module segment set; Step S332: performing module complexity evaluation on the initial module segment set to obtain a module complexity index set; Step S333: using a preset complexity threshold to identify high-complexity modules in the module complexity index set, and obtaining a high-complexity module segment set; Step S334: identifying split points of the high-complexity module segment set to obtain split point data; Step S335: redrawing the boundary of the high-complexity module segment set according to the split point data to obtain boundary redrawing data; Step S336: adjusting the module boundaries of the high-complexity module segment set according to the boundary redrawing data to obtain an adjusted module segment set; Step S337: Determine the module boundary according to the adjusted module segment set and the initial module segment set to obtain a module boundary set.

[0061] In the embodiment of the present invention, the process flow chart generated in step S31 is initially segmented using the key node set obtained in step S32. The step sequence between the starting node and the first key node, the step sequence between adjacent key nodes, and the step sequence between the last key node and the ending node are defined as an initial module segment. A unique ID is assigned to each initial module segment, and the process step sequence, the starting key node ID, and the ending key node ID contained therein are recorded. The information of all initial module segments is stored in a list to form an initial module segment set.

[0062] Traverse each module segment in the initial module segment set generated in step S331. For each module segment, count the number of process steps it contains, the number of process parameters involved, and the number of parameter types. In addition, analyze the dependencies between the process steps within the module segment and calculate the dependency density, for example, using the number of dependencies between steps divided by the square of the number of steps. The number of steps, number of parameters, number of parameter types, and dependency density of each module segment are used as module complexity indicators and stored in a JSON object. The complexity indicator JSON objects of all module segments are stored in a list to form a module complexity indicator set.

[0063] The complexity threshold is preset, for example, the step number threshold is 10, the parameter number threshold is 20, and the dependency density threshold is 0.5. The module complexity index set generated in step S332 is traversed, and if any complexity index of a module segment exceeds the preset threshold, the module segment is marked as a high-complexity module segment, and its information is added to the high-complexity module segment set.

[0064] Traverse each module segment in the high-complexity module segment set identified in step S333. Analyze the process steps and dependencies within the module segment to find a suitable split point. Prioritize the steps with the least information transfer or parameter dependency with other steps as split points to reduce the coupling between modules. Record the split point step ID of each high-complexity module segment to form split point data.

[0065] According to the split point data identified in step S334, the boundaries of the high-complexity module segment are redrawn. The original high-complexity module segment is split into two or more new module segments at the split point. A unique ID is assigned to each newly generated module segment, and the process step sequence, the starting key node ID and the ending key node ID (the split point is used as a new key node) contained therein are recorded. The information of all new module segments and the ID and split point information of the original module segments are stored in a list to form boundary redrawing data.

[0066] According to the boundary redrawing data generated in step S335, the initial module segment set is updated, and the high-complexity module segment is replaced with the new module segment generated after the splitting to form an adjusted module segment set.

[0067] The adjusted module segment set generated in step S336 is traversed, the start node ID and the end node ID of each module segment are extracted, and the information is stored in a list to form a final module boundary set. The information of the initial module segment that has not been split is also included in the final module boundary set.

[0068] Preferably, step S4 comprises the following steps: Step S41: obtaining a new design requirement parameter set; performing module matching on the process module library according to the new design requirement parameter set to obtain a candidate module set; Step S42: Sort the candidate module set according to the process meta-model to obtain a module sequence; Step S43: Connect the module sequence according to the process element model to obtain a module network; Step S44: verify the module network process, and perform process reorganization and optimization to obtain a reorganized process flow.

[0069] As an example of the present invention, refer to Figure 3 As shown, in this example, step S4 includes: Step S41: obtaining a new design requirement parameter set; performing module matching on the process module library according to the new design requirement parameter set to obtain a candidate module set; In the embodiment of the present invention, new design requirement parameters are input through the user interface, such as modifying the size of the garment (for example, increasing the length of the garment by 2 cm), changing the style (for example, changing from slim to loose), or adding a new design element (for example, adding a zipper pocket). These new design requirement parameters are stored as a new design requirement parameter set in JSON format. Then, using the module matching method described in S418, according to the new design requirement parameter set, a matching module is retrieved from the process module library constructed in step S35, and the matching module information is stored in a list to form a candidate module set.

[0070] Step S42: Sort the candidate module set according to the process meta-model to obtain a module sequence; In an embodiment of the present invention, the process metamodel constructed in step S15 is read from the Neo4j database to obtain the dependencies and constraints between the modules. Then, the candidate module set obtained in step S41 is sorted using a topological sorting algorithm to generate a module sequence that satisfies the dependencies and constraints defined in the process metamodel. For example, if the process metamodel stipulates that the "cutting module" must be executed before the "sewing module", then in the sorted module sequence, the "cutting module" is located in front of the "sewing module". The sorted module sequence is stored in a list, and the list elements are module IDs.

[0071] Step S43: Connect the module sequence according to the process element model to obtain a module network; In an embodiment of the present invention, the modules in the module sequence generated in step S42 are connected according to the module connection mode defined in the process meta-model constructed in step S15 to form a module network. The module network is a directed graph, in which the nodes represent modules and the edges represent the connection relationship between modules. The attributes of the edge include the connection type (e.g., sequential connection, parallel connection, conditional connection) and conditional parameters. For example, if the "sew cuffs" module and the "sew body" module can be executed in parallel, a parallel connection relationship is established between the two modules. The generated module network is stored in a Neo4j database.

[0072] Step S44: verifying the process of the module network and reorganizing and optimizing the process to obtain a reorganized process flow; In the embodiment of the present invention, a rule-based reasoning engine is used to verify the module network generated in step S43, for example, to check whether the connection between modules complies with the rules and constraints defined in the process metamodel, whether there are circular dependencies, etc. If problems are found, the module network is adjusted and optimized according to predefined optimization rules and heuristic algorithms, such as modifying module parameters, adjusting module order, adding or deleting modules, etc. The optimization goals include minimizing production time, minimizing material consumption, maximizing product quality, etc. The optimized module network is the reorganized process flow, and it is stored in the Neo4j database.

[0073] Preferably, step S41 includes the following steps: Step S411: performing text parsing and keyword extraction on the new design requirement parameter set to obtain the key requirement parameters; Step S412: performing semantic role labeling on the key parameters of the requirements to obtain semantic role labeling data; Step S413: querying the pre-built clothing craft domain ontology knowledge base according to the required key parameters and the semantic role annotation data to obtain ontology knowledge data; Step S414: Structuring the demand intent of the ontology knowledge data to obtain structured demand intent; Step S415: constructing module feature vectors according to the process module library to obtain a module feature vector set; Step S416: Calculate the semantic similarity between the requirements and the modules according to the structured requirements intention and the module feature vector set to obtain a module semantic similarity list; Step S417: Perform module scoring based on machine learning according to the structured demand intention and the module feature vector set to obtain a module matching score list; Step S418: Screening and sorting candidate modules in the module semantic similarity list and the module matching score list to obtain a candidate module set.

[0074] In the embodiment of the present invention, Jieba word segmentation tool is used to perform word segmentation and part-of-speech tagging on the text description in the new design requirement parameter set in JSON format obtained in step S41. For example, for the requirement description "add a zipper pocket on the left chest", the word segmentation result is "add / v a / m zipper / n pocket / n on / p left chest / n". Then, nouns and verbs are extracted as the key parameters of the requirements, such as "left chest", "add", "zipper", and "pocket". The extracted key parameters of the requirements are stored in a list.

[0075] Use the LTP semantic role labeling tool to perform semantic role labeling on the key parameters of the requirements extracted in step S411. For example, for "add a zipper pocket on the left chest", the labeling results are: A0 (add) = "on / p left chest / n", A1 (add) = "one / m zipper / n pocket / n". Among them, A0 represents the place where the action occurs, and A1 represents the recipient of the action. The semantic role labeling results are stored in a JSON object, for example: {"action"; "add", "location"; "left chest", "target"; "zipper pocket"}.

[0076] Use the SPARQL query language to query the pre-built clothing craft domain ontology knowledge base. For example, based on the key parameter "pocket" and the "target" in the semantic role annotation data; "zipper pocket", query the concepts related to "pocket" in the ontology knowledge base, such as "patch pocket", "dig pocket", "zipper pocket", etc., as well as the hierarchical relationships and attributes between them. Store the query results in a list to form ontology knowledge data.

[0077] According to the ontology knowledge data obtained in step S413, the semantic role annotation data obtained in step S412 is structured to form a structured demand intent. For example, "zipper pocket" is mapped to the corresponding concept in the ontology knowledge base, and its attributes, such as "pocket type", "zipper length", etc., are extracted. Finally, the structured demand intent is represented as a JSON object, for example: {"action"; "increase", "location"; "left chest", "target"; "pocket", "pocket_type"; "zipper pocket", "zipper_length"; "10cm"}.

[0078] Read the information of each module from the MongoDB process module library constructed in step S35, including the module name, function description, input and output parameters, etc. Use the TF-IDF algorithm to extract the feature vector of the module description text. At the same time, one-hot encode the input and output parameters of the module to form a parameter feature vector. Concatenate the text feature vector and the parameter feature vector to form a comprehensive feature vector of the module. Store the feature vectors of all modules in a list to form a module feature vector set.

[0079] The structured requirement intent obtained in step S414 is converted into a feature vector, for example, Word2Vec is used to convert the requirement keywords into word vectors, and then the word vectors are averaged to obtain the requirement feature vector. The cosine similarity between the requirement feature vector and each module feature vector in the module feature vector set obtained in step S415 is calculated to obtain a module semantic similarity list, where the list elements include the module ID and the corresponding similarity score.

[0080] Use a pre-trained machine learning model (e.g., gradient boosting decision tree model) to score the modules. The input of the model is the feature vector of the structured requirement intent obtained in step S414 and the module feature vector obtained in step S415, and the output is the matching score of the module. The matching scores of all modules are stored in a list to form a module matching score list, and the list elements include the module ID and the corresponding matching score.

[0081] The module semantic similarity list obtained in step S416 and the module matching score list obtained in step S417 are sorted from high to low according to the similarity or score. A threshold is set, such as 0.8, to filter out modules whose semantic similarity and module matching score are both higher than the threshold. If the new design requirement parameter set obtained in step S41 contains a specific parameter constraint, such as "pocket type: zipper pocket", further filter out modules that meet the parameter constraint. Sort the filtered modules according to the comprehensive score (e.g., the weighted average of the semantic similarity and module matching score) to obtain the final candidate module set, and store it in a list, where the list element is the module ID.

[0082] Preferably, step S5 comprises the following steps: Step S51: constructing a simulation environment according to the process element model to obtain a virtual simulation environment; Step S52: importing the reorganized process flow into the virtual simulation environment, performing correlation on the relevant processes, and obtaining the imported process flow; Step S53: simulate and run the imported process flow to obtain original simulation data; Step S54: Analyze and evaluate the original simulation data to obtain a simulation evaluation report; Step S55: Optimize and adjust parameters of the reorganization process flow according to the simulation evaluation report to obtain simulation results.

[0083] In an embodiment of the present invention, Clo3D software is used to build a virtual simulation environment. The process metamodel constructed in step S15 is read from the Neo4j database to obtain process parameters, module information, and connection relationships between modules. According to the clothing style information in the process metamodel, a virtual human model and clothing style are created in Clo3D. According to the fabric parameters in the process metamodel, the corresponding virtual fabric is imported into Clo3D. According to the process parameters defined in the process metamodel, the parameters of virtual equipment such as a virtual sewing machine and a virtual cutting table in Clo3D are configured, such as sewing speed, stitch length, cutting accuracy, etc.

[0084] Convert the reorganized process flow generated in step S44 (directed graph stored in Neo4j database) into a sequence of instructions that can be recognized by Clo3D. For example, convert the "cutting module" into the cutting operation instruction in Clo3D, specify the virtual fabric, cutting path and cutting parameters for cutting. Convert the "sewing module" into the sewing operation instruction in Clo3D, specify the parameters such as the virtual garment pieces, sewing thread type, stitch length, sewing speed, etc. Add all instructions to the simulation process of Clo3D in sequence to form the imported process flow.

[0085] Run the process flow imported in step S52 in the Clo3D virtual simulation environment. During the simulation, record the execution time, material consumption (e.g., fabric usage, sewing thread usage), and product quality parameters (e.g., stitching strength, dimensional deviation) of each process step. Store all recorded data in a CSV file to form the original simulation data.

[0086] Read the original simulation data (CSV file) generated in step S53. Calculate the total production time, total material consumption, and the average and standard deviation of each product quality index. Compare the calculation results with the pre-defined quality standards and cost targets to evaluate the feasibility, efficiency, and economy of the process. Organize the evaluation results into a simulation evaluation report, which includes the values ​​of each index, charts, and analysis and summary of the simulation results.

[0087] According to the simulation evaluation report generated in step S54, if it is found that the simulation results do not meet the preset quality standards or cost targets, it is necessary to optimize and adjust the parameters of the reorganized process flow. For example, if it is found that the production time is too long, you can try to increase the sewing speed of the virtual sewing machine, or adjust the process flow so that certain steps are executed in parallel. Parameter optimization can use optimization algorithms such as genetic algorithms to minimize production time and material consumption and maximize product quality. Apply the optimized parameters to the Clo3D simulation environment, rerun the simulation, and store the optimized simulation results in the database, and finally obtain the simulation results, including the optimized process parameters and the corresponding simulation data.

[0088] Preferably, the present invention also provides a clothing process modular design system based on adjustable parameters, which is used to execute the clothing process modular design method based on adjustable parameters as described above, and the clothing process modular design system based on adjustable parameters includes: A metamodel construction unit is used to obtain a basic process parameter set; define a process association relationship according to the basic process parameter set to obtain a process association diagram; and construct a metamodel according to the process association diagram to obtain a process metamodel; The parametric design unit is used to obtain a target style parameter set; map the parameters in the target style parameter set to the process meta-model, adjust the process parameters, and obtain an adjusted process parameter set; optimize the process flow according to the adjusted process parameter set to obtain a parametric process model; The modular decomposition unit is used to perform process analysis on the parameterized process model to obtain a process flow chart; divide the process flow chart into module boundaries to obtain a module boundary set; extract module information from the parameterized process model according to the module boundary set, and construct a process module library to obtain a process module library; The module reorganization unit is used to obtain a new design requirement parameter set; perform module matching on the process module library according to the new design requirement parameter set to obtain a candidate module set; perform process reorganization optimization according to the process meta-model and the candidate module set to obtain a reorganized process flow; The process verification unit is used to perform model process simulation on the reorganized process flow according to the process meta-model to obtain original simulation data; perform process verification on the original simulation data to obtain simulation results to realize the modular design task of clothing process.

[0089] Therefore, the embodiments should be regarded as illustrative and non-restrictive from all points, and the scope of the present invention is limited by the appended claims rather than the above description, and it is therefore intended that all changes falling within the meaning and range of equivalent elements of the application documents are included in the present invention.

[0090] The above description is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but should conform to the widest scope consistent with the principles and novel features invented herein.

Claims

1. A modular design method for garment technology based on adjustable parameters, characterized in that: The following steps are involved: Step S1: Obtain a basic process parameter set; Define the process association relationship according to the basic process parameter set to obtain a process association diagram; The metamodel is constructed according to the process association diagram to obtain the process metamodel; Step S2: Obtain a target style parameter set; Mapping the parameters in the target style parameter set to the process meta-model, adjusting the process parameters, and obtaining the adjusted process parameter set; Optimize the process flow according to the adjusted process parameter set to obtain a parameterized process model; Step S3: performing process flow analysis on the parameterized process model to obtain a process flow chart; Divide the process flow chart into module boundaries to obtain a module boundary set; Extracting module information from the parameterized process model according to the module boundary set, and constructing a process module library to obtain a process module library; Step S4: obtaining a new design requirement parameter set; performing module matching on the process module library according to the new design requirement parameter set to obtain a candidate module set; performing process reorganization optimization according to the process meta-model and the candidate module set to obtain a reorganized process flow; Step S5: performing model process simulation on the reorganized process flow according to the process element model to obtain original simulation data; The original simulation data is processed and the simulation results are obtained to realize the modular design task of clothing process.

2. The parameter-adjustable clothing process modular design method according to claim 1 is characterized in that: Step S1 includes the following steps: Step S11: collecting basic process parameters according to common clothing processes to obtain a basic process parameter set; Step S12: defining the process association relationship according to the basic process parameter set to obtain a process association diagram; Step S13: abstracting the process module of the process association diagram to obtain an abstract process module library; Step S14: parametrically expressing the abstract process module library to obtain a parameterized process module library; Step S15: constructing a meta-model for the parameterized process module library, the process association diagram, and the basic process parameter set to obtain a process meta-model.

3. The parameter-adjustable clothing process modular design method according to claim 2 is characterized in that: Step S12 includes the following steps: Step S121: construct a knowledge graph according to the basic process parameter set to obtain a clothing process knowledge graph; Step S122: creating a process node for the basic process parameter set to obtain a process node set; Step S123: performing relationship reasoning on the process node set according to the clothing process knowledge graph to obtain a reasoning relationship set; Step S124: constructing association relationships for the process node set according to the inference relationship set to obtain an initial process association graph; Step S125: Manually review and modify the initial process association diagram to obtain a process association diagram.

4. The parameter-adjustable clothing process modular design method according to claim 1 is characterized in that: Step S2 includes the following steps: Step S21: acquiring design parameters for a target style to obtain a target style parameter set; Step S22: Mapping the parameters in the target style parameter set to the process meta-model to obtain a mapping parameter set; Step S23: adjusting the process parameters of the process element model according to the mapping parameter set to obtain an adjusted process parameter set; Step S24: selecting a process module for the process element model according to the adjusted process parameter set, and generating a process flow to obtain an initial process flow; Step S25: Optimize the initial process flow to obtain a parameterized process model.

5. The parameter-adjustable clothing process modular design method according to claim 1 is characterized in that: Step S3 includes the following steps: Step S31: performing process flow analysis on the parameterized process model to obtain a process flow chart; Step S32: identifying key nodes of the process flow chart to obtain a key node set; Step S33: dividing the process flow chart into module boundaries according to the key node set to obtain a module boundary set; Step S34: extracting module information from the parameterized process model according to the module boundary set to obtain a module information set; Step S35: construct a process module library for the module information set to obtain a process module library.

6. The method for modular design of garment technology based on adjustable parameters according to claim 5, characterized in that: Step S33 includes the following steps: Step S331: performing initial module segmentation on the process flow chart according to the key node set to obtain an initial module segment set; Step S332: performing module complexity evaluation on the initial module segment set to obtain a module complexity index set; Step S333: using a preset complexity threshold to identify high-complexity modules in the module complexity index set, and obtaining a high-complexity module segment set; Step S334: identifying split points of the high-complexity module segment set to obtain split point data; Step S335: redrawing the boundary of the high-complexity module segment set according to the split point data to obtain boundary redrawing data; Step S336: adjusting the module boundaries of the high-complexity module segment set according to the boundary redrawing data to obtain an adjusted module segment set; Step S337: Determine the module boundary according to the adjusted module segment set and the initial module segment set to obtain a module boundary set.

7. The parameter-adjustable clothing process modular design method according to claim 1 is characterized in that: Step S4 includes the following steps: Step S41: obtaining a new design requirement parameter set; performing module matching on the process module library according to the new design requirement parameter set to obtain a candidate module set; Step S42: Sort the candidate module set according to the process meta-model to obtain a module sequence; Step S43: Connect the module sequence according to the process element model to obtain a module network; Step S44: verify the module network process, and perform process reorganization and optimization to obtain a reorganized process flow.

8. The parameter-adjustable clothing process modular design method according to claim 7 is characterized in that: Step S41 includes the following steps: Step S411: performing text parsing and keyword extraction on the new design requirement parameter set to obtain the key requirement parameters; Step S412: performing semantic role labeling on the key parameters of the requirements to obtain semantic role labeling data; Step S413: querying the pre-built clothing craft domain ontology knowledge base according to the required key parameters and the semantic role annotation data to obtain ontology knowledge data; Step S414: Structuring the demand intent of the ontology knowledge data to obtain structured demand intent; Step S415: constructing module feature vectors according to the process module library to obtain a module feature vector set; Step S416: Calculate the semantic similarity between the requirements and the modules according to the structured requirements intention and the module feature vector set to obtain a module semantic similarity list; Step S417: Perform module scoring based on machine learning according to the structured demand intention and the module feature vector set to obtain a module matching score list; Step S418: Screening and sorting candidate modules in the module semantic similarity list and the module matching score list to obtain a candidate module set.

9. The parameter-adjustable clothing process modular design method according to claim 1 is characterized in that: Step S5 includes the following steps: Step S51: constructing a simulation environment according to the process element model to obtain a virtual simulation environment; Step S52: importing the reorganized process flow into the virtual simulation environment, performing correlation on the relevant processes, and obtaining the imported process flow; Step S53: simulate and run the imported process flow to obtain original simulation data; Step S54: Analyze and evaluate the original simulation data to obtain a simulation evaluation report; Step S55: Optimize and adjust parameters of the reorganization process flow according to the simulation evaluation report to obtain simulation results.

10. A parameter-adjustable clothing process modular design system, characterized in that: Used to execute the parameter-adjustable clothing process modular design method according to claim 1, the parameter-adjustable clothing process modular design system comprises: A metamodel construction unit is used to obtain a basic process parameter set; define a process association relationship according to the basic process parameter set to obtain a process association diagram; and construct a metamodel according to the process association diagram to obtain a process metamodel; The parametric design unit is used to obtain a target style parameter set; map the parameters in the target style parameter set to the process meta-model, adjust the process parameters, and obtain an adjusted process parameter set; optimize the process flow according to the adjusted process parameter set to obtain a parametric process model; The modular decomposition unit is used to perform process analysis on the parameterized process model to obtain a process flow chart; divide the process flow chart into module boundaries to obtain a module boundary set; extract module information from the parameterized process model according to the module boundary set, and construct a process module library to obtain a process module library; The module reorganization unit is used to obtain a new design requirement parameter set; perform module matching on the process module library according to the new design requirement parameter set to obtain a candidate module set; perform process reorganization optimization according to the process meta-model and the candidate module set to obtain a reorganized process flow; The process verification unit is used to perform model process simulation on the reorganized process flow according to the process meta-model to obtain original simulation data; perform process verification on the original simulation data to obtain simulation results to realize the modular design task of clothing process.

Images