A clothing pattern production monitoring method and system based on design process

Through the combination of AI-CAD model and virtual fitting system, the automated optimization and real-time monitoring of clothing design and plate making are achieved, solving the problems of disconnection between design and production and insufficient real-time monitoring in traditional plate making, and improving production efficiency and product quality.

CN118779938BActive Publication Date: 2024-12-03NANTONG RUDONG FENGBO CLOTHING CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411244768.3
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-09-06
Publication Date
2024-12-03
Estimated Expiration
2044-09-06

AI Technical Summary

Technical Problem

The design and production are disconnected from the production process of traditional clothing, resulting in frequent errors, low production efficiency, and lack of real-time monitoring and feedback mechanisms, which affect product quality and market competitiveness.

Method used

Using a clothing plate making production monitoring method based on the design process, the design patterns are automatically optimized through the AI-CAD model, combined with a virtual fitting system and real-time monitoring sensor, to achieve seamless docking and real-time monitoring of design and plate making, and to utilize deep learning and enhanced learning algorithms Build a multi-level AI-CAD model, automatically recommend design parameters, and optimize production parameters through feedback control system.

Benefits of technology

It significantly improves the accuracy and consistency of design drawings, reduces the disconnection between design and plate making, realizes real-time monitoring and feedback, improves the stability of the production process and product quality, and reduces resource waste and production cycles.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118779938B_ABST
    Figure CN118779938B_ABST
Patent Text Reader

Abstract

The present invention discloses a clothing plate-making production monitoring method and system based on a design process, including constructing clothing design drawings; constructing an AI-CAD model; generating optimized clothing design drawings; directly docking the optimized clothing design drawings with the plate-making module in the plate-making equipment, the plate-making equipment performs clothing plate-making according to the optimized clothing design drawings, and the plate-making equipment is equipped with an adaptive cutting tool and a real-time monitoring and adjustment function; the production monitoring sensor installed on the plate-making equipment monitors the cutting accuracy, material usage and equipment operation status in the plate-making process in real time; the plate-making equipment has a built-in central control system, which analyzes production data in real time, detects abnormal conditions in the production process, predicts potential problems and provides optimization suggestions; optimizes production parameters in real time until the clothing plate-making is completed. The present invention significantly optimizes the clothing production process and provides strong support for the intelligent and automated development of the clothing industry.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of clothing plate making, and in particular to a clothing plate making production monitoring method and system based on a design process. Background Art

[0002] With the continuous development of the clothing industry, clothing pattern making, as a crucial link in the clothing production process, directly affects the quality and production efficiency of clothing. The traditional clothing pattern making process usually involves multiple links, including design, pattern making and production. However, there are often problems with the collaboration and information transmission between these links, resulting in low production efficiency and frequent errors, which ultimately affect product quality and market competitiveness.

[0003] First, in the traditional clothing pattern making process, design and production are often separate links. After the designer completes the design, he needs to manually convert the design drawings into plate making data, which is prone to human errors and inaccurate data transmission. This disconnect between design and production leads to frequent errors in the plate making process, affecting the accuracy of plate making and the quality of the final product. Manual conversion of design drawings is not only time-consuming and labor-intensive, but also increases the risk of errors, especially in the case of complex design drawings and diverse fabric choices, where the errors are more significant.

[0004] Secondly, the lack of real-time monitoring and feedback mechanism is another major problem in the traditional clothing pattern making process. During the production process, manual monitoring and post-inspection are often relied on. This method cannot timely discover and solve problems that arise in the production process, resulting in unstable production and fluctuations in product quality. In addition, production data cannot be collected and analyzed in real time, making it difficult to effectively optimize and manage the production process. Even if adjustments are made after problems arise, delays will cause a lot of resource waste and extend the production cycle.

[0005] At present, although some companies have introduced computer-aided design (CAD) systems to improve design efficiency, there are still obvious deficiencies in the seamless connection and real-time monitoring of design and plate-making data. Although the application of CAD systems has reduced the conversion time between design and plate-making to a certain extent, information loss or errors may still occur during data transmission and conversion. In addition, the CAD system has limited capabilities in real-time feedback and production process monitoring, making it difficult to achieve automated control and optimization of the production process.

[0006] Therefore, how to provide a clothing pattern production monitoring method and system based on the design process is a problem that technical personnel in this field urgently need to solve. Summary of the invention

[0007] One purpose of the present invention is to propose a clothing pattern production monitoring method and system based on the design process. The present invention significantly optimizes the clothing production process and provides strong support for the intelligent and automated development of the clothing industry.

[0008] A garment pattern production monitoring method based on a design process according to an embodiment of the present invention comprises the following steps:

[0009] S1. Collect fabric selection data, size specification data and style design data in clothing design, and construct clothing design patterns;

[0010] S2. Build an AI-CAD model, which automatically recommends design, modifies or optimizes parameters of clothing design drawings based on historical clothing design drawing design data and fashion trends;

[0011] S3. Import the clothing design drawings optimized by the AI-CAD model into the virtual fitting system to simulate the real wearing effect. The virtual fitting system detects the comfort and visual effect of the design, automatically optimizes the design defects, and generates the optimized clothing design drawings;

[0012] S4. The optimized clothing design pattern is directly connected to the plate-making module in the plate-making equipment. The plate-making equipment makes clothing plates according to the optimized clothing design pattern. The plate-making equipment is equipped with an adaptive cutting tool and a real-time monitoring and adjustment function.

[0013] S5. The production monitoring sensor installed on the plate-making equipment monitors the cutting accuracy, material usage and equipment operation status in the plate-making process in real time, and transmits data in real time;

[0014] S6. The plate-making equipment has a built-in central control system that analyzes production data in real time, detects abnormal conditions in the production process, predicts potential problems and provides optimization suggestions;

[0015] S7. Based on the data analysis results, the feedback and control system sends adjustment instructions to the plate-making equipment to optimize the production parameters in real time until the clothing plate-making is completed.

[0016] Optionally, the S1 specifically includes the following sub-steps:

[0017] S11. Obtain fabric information required for clothing design through database retrieval or manual input , including fabric type, color, texture and elasticity parameters:

[0018] ;

[0019] in, represents the weight of the m-th fabric, Indicates the type of fabric. Indicates the color of the fabric. Indicates the texture of the fabric. Indicates the elastic parameters of the fabric. represents other correction factors that affect fabric selection, and M represents the total number of fabric types;

[0020] S12. Obtain the size specifications required for clothing design through measurement or database retrieval , including length, sleeve length, shoulder width, chest circumference and waist circumference:

[0021] ;

[0022] in, represents the weight of the nth size specification, Indicates the length of the garment. Indicates sleeve length. Indicates shoulder width. Indicates chest circumference, They represent the correction coefficients for sleeve length, shoulder width, and chest circumference, respectively. It represents the correction coefficient of other dimension specification parameters, and N represents the total number of dimension specification parameters;

[0023] S13. Obtaining style information of clothing design through database retrieval, manual input or automatic generation , including collar, sleeves, hem and decorative elements:

[0024] ;

[0025] in, represents the weight of the p-th style design parameter, Indicates collar type, Indicates sleeve type, Indicates the hem of a garment. , They represent the correction coefficients for sleeve type and hem respectively. represents the correction coefficient of other style design parameters, K represents the total number of style design parameters, and P represents the number of all style design elements;

[0026] S14. Based on the acquired fabric information , Dimensions and style information , use CAD to generate clothing design drawings P(x):

[0027] ;

[0028] in, Respectively represent the weight coefficients of fabric selection data, size specification data and style design data, It represents the comprehensive correction coefficient in the process of design pattern generation, I and J represent fabric selection data and size specification data respectively, and x represents the specific design element.

[0029] Optionally, S2 specifically includes the following sub-steps:

[0030] S21. Obtain a large number of historical clothing design patterns through database retrieval or historical data storage ;

[0031] S22. Obtain current and future trend data based on market research and trend analysis ;

[0032] S23. Use a combination of deep learning and reinforcement learning to build a multi-level and multi-module AI-CAD model:

[0033] Image generation module, which generates representative clothing design patterns based on generative adversarial networks;

[0034] The generated clothing design patterns are optimized using the reinforcement learning algorithm, and the parameters are adjusted in multiple dimensions based on the comfort, visual effects and market feedback of the clothing to generate optimized design parameters;

[0035] S24, AI-CAD model automatically recommends design parameters based on training and optimization results ;

[0036] S25. Automatically modify or optimize the clothing design drawing according to the design parameters recommended by the AI-CAD model to generate a final optimized design drawing P'(x).

[0037] Optionally, the automatically recommended design parameters Including fabric selection, size specifications and style design;

[0038] ;

[0039] in, Respectively represent the weights of fabric selection data, size specification data and style design data, Do not indicate fabric selection data, size specification data and style design data, They represent the correction factors for fabric selection, size specifications and style design data respectively.

[0040] Optionally, the S23 specifically includes the following sub-steps:

[0041] S231, construct an image generation module, and use a generative adversarial network to generate basic clothing design patterns. The generative adversarial network includes a generator G and a discriminator D. The generator G generates basic clothing design patterns through a noise vector z, and the discriminator D identifies the generated patterns:

[0042] ;

[0043] Among them, G(z) represents the basic clothing design pattern generated by the generator, and is the weight matrix of the generator, represents the activation function, and is the bias vector of the generator, z is the input noise vector;

[0044] S232, generating an adversarial network through adversarial training, so that the generator G and the discriminator D compete with each other for optimization, and generate realistic clothing design patterns;

[0045] S233, on the basis of generating realistic clothing design drawings, further generating clothing design drawings including detail design, the input is realistic clothing design drawings and detail information c:

[0046] ;

[0047] Among them, G(z|c) represents the detailed clothing design pattern generated by the conditional generator, and is the weight matrix of the conditional generator, and is the bias vector of the conditional generator, [z, c] represents the concatenated vector of the noise vector and the detail information;

[0048] S234, through conditional adversarial training, the generator G and the discriminator D are optimized under the detail information c to generate a clothing design pattern containing detail design;

[0049] S235. Based on the clothing design pattern containing detailed design, the reinforcement learning algorithm is applied to optimize the clothing design pattern, and the reward function R is constructed. The parameters are adjusted based on the comfort, visual effect and market feedback of the clothing:

[0050] ;

[0051] Among them, R is the reward function, , , are the weights of comfort, visual effect and market feedback, represents the comfort score of the i-th sample, represents the visual effect score of the jth sample, represents the market feedback score of the kth sample, N, , L represent the total number of samples of comfort, visual effect and market feedback respectively;

[0052] S236. Use the policy gradient algorithm to optimize the design parameters and update the strategy :

[0053] ;

[0054] in, represents the policy gradient, represents the policy parameters, Represents a complete design generation process. and denote the action and state at time step t, respectively. Represents the cumulative reward of the entire generation process.

[0055] Optionally, the training objective function of the generative adversarial network is as follows:

[0056] ;

[0057] Among them, V(D, G) is the training objective function, represents the expected value, represents the real data distribution, represents the distribution of the noise vector, D(x) is the discriminator's recognition result for the real clothing design pattern, and D(G(z)) is the discriminator's recognition result for the generated pattern;

[0058] The training objective function of the conditional generator to generate the adversarial network is as follows:

[0059] ;

[0060] Among them, V(D, G|c) is the conditional training objective function, represents the conditional data distribution, D(x|c) is the recognition result of the conditional discriminator on the real clothing design pattern, and D(G(z|c)) is the recognition result of the conditional discriminator on the generated pattern.

[0061] Optionally, S3 specifically includes the following sub-steps:

[0062] S31, importing the optimized clothing design pattern P'(x) into the virtual fitting system;

[0063] S32. Establishing a three-dimensional virtual human body model based on the user's body measurement data ;

[0064] S33, applying the optimized clothing design pattern P'(x) to the virtual human body model , using physical simulation algorithms to simulate the wearing effect of clothing and generate preliminary wearing simulation results ;

[0065] S34, virtual fitting system based on wearing simulation results , evaluate the comfort C and visual effect V of clothing:

[0066] ;

[0067] ;

[0068] Among them, C represents the comfort score, V represents the visual effect score, represent the comfort level and error correction coefficient respectively, Represent visual effects and market feedback correction coefficients respectively, Respectively represent the simulation results of comfort and visual effect, denote the error and feedback parameters respectively;

[0069] S35. Based on the comfort score C and the visual effect score V, the virtual fitting system identifies design defects, applies the optimization algorithm to make adjustments, and generates optimization suggestions O:

[0070] ;

[0071] Among them, O represents optimization suggestions, , Respectively represent the optimization weight coefficients of comfort and visual effect, and Respectively represent the current comfort and visual effect parameters, represents other correction factors, The total number of optimization parameters for the optimization suggestion O;

[0072] S36. Update the clothing design pattern according to the optimization suggestion O, and generate an updated design pattern :

[0073] ;

[0074] in, Indicates the updated design drawing. represents the kth optimization parameter, represents the corresponding weight, represents the adjustment factor, Indicates the specific parameters of the optimization suggestion. is the total number of optimization parameters of the design pattern.

[0075] Optionally, the S4 specifically includes the following sub-steps:

[0076] S41, the updated design drawings Import the plate-making module in the plate-making equipment through the data interface;

[0077] S42, the plate-making device receives the clothing design drawing: the control system in the plate-making device parses and receives the clothing design drawing data, and generates the control instructions required for plate-making ,in Specific instruction sets for each plate-making step;

[0078] S43, the plate making equipment receives the control instruction Performing clothing pattern making, using an adaptive cutting tool to perform a cutting operation, the operating parameters of the cutting tool include a cutting speed v, a cutting depth d and a cutting path p;

[0079] S44, sensors on the plate-making equipment monitor key parameters of the cutting process in real time, including cutting accuracy, material usage m and equipment operating status s, and record and transmit monitoring data ;

[0080] S45, based on real-time monitoring data , using the feedback control system to adjust the cutting tool in real time, the cutting tool adjustment parameters include cutting speed correction, cutting depth correction and cutting path correction;

[0081] S46. Under real-time monitoring and feedback control, the plate-making equipment completes all cutting operations, generates clothing patterns that meet design requirements, and stores the final plate-making data in the storage system within the plate-making equipment, records all operating parameters and monitoring data during the plate-making process, and generates a final plate-making report. .

[0082] A clothing pattern production monitoring system based on design process, including the following modules:

[0083] Design module, which is used to create and optimize clothing design drawings, including fabric selection, size specifications and style design, and directly connects with the pattern making module;

[0084] Plate making module, automated plate making equipment, makes clothing plates according to optimized clothing design drawings, equipped with adaptive cutting tools and real-time monitoring and adjustment functions;

[0085] Virtual fitting module: import clothing design drawings into the virtual fitting system to simulate real wearing effects, evaluate and optimize design defects;

[0086] Production monitoring module, a production monitoring sensor installed on the plate-making equipment, monitors key parameters of the plate-making process in real time, including cutting accuracy, material usage and equipment operating status;

[0087] Feedback and control module, real-time data processing and analysis system, sends adjustment instructions to plate-making equipment according to the real-time analysis results of production data, and optimizes production parameters in real time;

[0088] Information management module, data storage system and visual management platform, used to safely store all production data, support historical data query and traceability, and display the real-time status of the production process and historical data analysis results;

[0089] The integrated interface module, data interface and control interface are used for data transmission between the design module and the plate-making module, and for instruction transmission between the feedback and control module and the plate-making equipment.

[0090] The beneficial effects of the present invention are:

[0091] (1) The present invention adopts an AI-CAD model to directly connect clothing design drawings with plate-making equipment, thereby avoiding errors in the manual conversion process. The AI-CAD model automatically recommends design, modification or optimization of clothing design drawing parameters based on historical clothing design drawing data and fashion trends, thereby improving the accuracy and consistency of the design drawings. A multi-level and multi-module AI-CAD model is constructed by combining deep learning and reinforcement learning methods, which realizes automatic optimization and recommendation of design parameters and significantly reduces the disconnection between design and plate-making.

[0092] (2) The present invention realizes real-time monitoring of the plate-making production process through sensor technology and data acquisition and processing technology. The sensors on the plate-making equipment monitor the cutting accuracy, material usage and equipment operation status in real time, record and transmit the monitoring data, and use the feedback control system to adjust the cutting tool in real time to ensure the accuracy and stability of the cutting operation, effectively improving the stability of the production process and product quality, and avoiding the waste of resources and extended production cycle caused by delayed adjustment in traditional technologies.

[0093] (3) The present invention imports clothing design drawings into the virtual fitting system, simulates the wearing effect of clothing by establishing a three-dimensional virtual human body model and a physical simulation algorithm, and generates preliminary wearing simulation results. The virtual fitting system evaluates the comfort and visual effect of clothing based on the wearing simulation results, identifies design defects and generates optimization suggestions, and generates optimized design drawings to ensure the comfort and aesthetics of the design. This not only improves the accuracy of the design, but also significantly reduces the trial and error cost and time in actual production. BRIEF DESCRIPTION OF THE DRAWINGS

[0094] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0095] Figure 1 The present invention provides a flow chart of a method and system for monitoring the production of clothing pattern making based on a design process. DETAILED DESCRIPTION

[0096] The present invention will now be described in further detail with reference to the accompanying drawings. These drawings are simplified schematic diagrams, which only illustrate the basic structure of the present invention in a schematic manner, and therefore only show the components related to the present invention.

[0097] refer to Figure 1 , a clothing pattern production monitoring method based on a design process, comprising the following steps:

[0098] S1. Collect fabric selection data, size specification data and style design data in clothing design, and construct clothing design patterns;

[0099] S2. Build an AI-CAD model, which automatically recommends design, modifies or optimizes parameters of clothing design drawings based on historical clothing design drawing design data and fashion trends;

[0100] S3. Import the clothing design drawings optimized by the AI-CAD model into the virtual fitting system to simulate the real wearing effect. The virtual fitting system detects the comfort and visual effect of the design, automatically optimizes the design defects, and generates the optimized clothing design drawings;

[0101] S4. The optimized clothing design pattern is directly connected to the plate-making module in the plate-making equipment. The plate-making equipment makes clothing plates according to the optimized clothing design pattern. The plate-making equipment is equipped with an adaptive cutting tool and a real-time monitoring and adjustment function.

[0102] S5. The production monitoring sensor installed on the plate-making equipment monitors the cutting accuracy, material usage and equipment operation status in the plate-making process in real time, and transmits data in real time;

[0103] S6. The plate-making equipment has a built-in central control system that analyzes production data in real time, detects abnormal conditions in the production process, predicts potential problems and provides optimization suggestions;

[0104] S7. Based on the data analysis results, the feedback and control system sends adjustment instructions to the plate-making equipment to optimize the production parameters in real time until the clothing plate-making is completed.

[0105] In this implementation, S1 specifically includes the following sub-steps:

[0106] S11. Obtain fabric information required for clothing design through database retrieval or manual input , including fabric type, color, texture and elasticity parameters:

[0107] ;

[0108] in, represents the weight of the m-th fabric, Indicates the type of fabric. Indicates the color of the fabric. Indicates the texture of the fabric. Indicates the elastic parameters of the fabric. represents other correction factors that affect fabric selection, and M represents the total number of fabric types;

[0109] S12. Obtain the size specifications required for clothing design through measurement or database retrieval , including length, sleeve length, shoulder width, chest circumference and waist circumference:

[0110] ;

[0111] in, represents the weight of the nth size specification, Indicates the length of the garment. Indicates sleeve length. Indicates shoulder width. Indicates chest circumference, They represent the correction coefficients for sleeve length, shoulder width, and chest circumference, respectively. It represents the correction coefficient of other dimension specification parameters, and N represents the total number of dimension specification parameters;

[0112] S13. Obtaining style information of clothing design through database retrieval, manual input or automatic generation , including collar, sleeves, hem and decorative elements:

[0113] ;

[0114] in, represents the weight of the p-th style design parameter, Indicates collar type, Indicates sleeve type, Indicates the hem of a garment. , They represent the correction coefficients for sleeve type and hem respectively. represents the correction coefficient of other style design parameters, K represents the total number of style design parameters, and P represents the number of all style design elements;

[0115] S14. Based on the acquired fabric information , Dimensions and style information , use CAD to generate clothing design drawings P(x):

[0116] ;

[0117] in, Respectively represent the weight coefficients of fabric selection data, size specification data and style design data, It represents the comprehensive correction coefficient in the process of design pattern generation, I and J represent fabric selection data and size specification data respectively, and x represents the specific design element.

[0118] In this implementation, S2 specifically includes the following sub-steps:

[0119] S21. Obtain a large number of historical clothing design patterns through database retrieval or historical data storage ;

[0120] S22. Obtain current and future trend data based on market research and trend analysis ;

[0121] S23. Use a combination of deep learning and reinforcement learning to build a multi-level and multi-module AI-CAD model:

[0122] Image generation module, which generates representative clothing design patterns based on generative adversarial networks;

[0123] The generated clothing design patterns are optimized using the reinforcement learning algorithm, and the parameters are adjusted in multiple dimensions based on the comfort, visual effects and market feedback of the clothing to generate optimized design parameters;

[0124] This implementation method constructs a multi-level and multi-module AI-CAD model by combining deep learning and reinforcement learning. It uses generative adversarial networks and reinforcement learning algorithms to automatically generate and optimize design drawings. The generative adversarial network can generate realistic preliminary design drawings through adversarial training of the generator and the discriminator; the reinforcement learning algorithm constructs a reward function and adjusts parameters according to the comfort, visual effects and market feedback of the clothing, thereby achieving automatic optimization of the design drawings. This not only improves design efficiency, but also reduces human errors and improves the accuracy and innovation of the design.

[0125] S24, AI-CAD model automatically recommends design parameters based on training and optimization results ;

[0126] S25. Automatically modify or optimize the clothing design drawing according to the design parameters recommended by the AI-CAD model to generate a final optimized design drawing P'(x).

[0127] In this embodiment, the design parameters are automatically recommended. Including fabric selection, size specifications and style design;

[0128] ;

[0129] in, Respectively represent the weights of fabric selection data, size specification data and style design data, Do not indicate fabric selection data, size specification data and style design data, They represent the correction factors for fabric selection, size specifications and style design data respectively.

[0130] In this implementation, S23 specifically includes the following sub-steps:

[0131] S231, construct an image generation module, and use a generative adversarial network to generate basic clothing design patterns. The generative adversarial network includes a generator G and a discriminator D. The generator G generates basic clothing design patterns through a noise vector z, and the discriminator D identifies the generated patterns:

[0132] ;

[0133] Among them, G(z) represents the basic clothing design pattern generated by the generator, and is the weight matrix of the generator, represents the activation function, and is the bias vector of the generator, z is the input noise vector;

[0134] S232, generating an adversarial network through adversarial training, so that the generator G and the discriminator D compete with each other for optimization, and generate realistic clothing design patterns;

[0135] S233, on the basis of generating realistic clothing design drawings, further generating clothing design drawings including detail design, the input is realistic clothing design drawings and detail information c:

[0136] ;

[0137] Among them, G(z|c) represents the detailed clothing design pattern generated by the conditional generator, and is the weight matrix of the conditional generator, and is the bias vector of the conditional generator, [z, c] represents the concatenated vector of the noise vector and the detail information;

[0138] S234, through conditional adversarial training, the generator G and the discriminator D are optimized under the detail information c to generate a clothing design pattern containing detail design;

[0139] S235. Based on the clothing design pattern containing detailed design, the reinforcement learning algorithm is applied to optimize the clothing design pattern, and the reward function R is constructed. The parameters are adjusted based on the comfort, visual effect and market feedback of the clothing:

[0140] ;

[0141] Among them, R is the reward function, , , are the weights of comfort, visual effect and market feedback, represents the comfort score of the i-th sample, represents the visual effect score of the jth sample, represents the market feedback score of the kth sample, N, , L represent the total number of samples of comfort, visual effect and market feedback respectively;

[0142] S236. Use the policy gradient algorithm to optimize the design parameters and update the strategy :

[0143] ;

[0144] in, represents the policy gradient, represents the policy parameters, Represents a complete design generation process. and denote the action and state at time step t, respectively. Represents the cumulative reward of the entire generation process.

[0145] In this implementation, the training objective function of the generative adversarial network is as follows:

[0146] ;

[0147] Among them, V(D, G) is the training objective function, represents the expected value, represents the real data distribution, represents the distribution of the noise vector, D(x) is the discriminator's recognition result for the real clothing design pattern, and D(G(z)) is the discriminator's recognition result for the generated pattern;

[0148] The training objective function of the conditional generator generative adversarial network is as follows:

[0149] ;

[0150] Among them, V(D, G|c) is the conditional training objective function, represents the conditional data distribution, D(x|c) is the recognition result of the conditional discriminator on the real clothing design pattern, and D(G(z|c)) is the recognition result of the conditional discriminator on the generated pattern.

[0151] In this embodiment, the AI-CAD model can automatically recommend design parameters through a deep learning algorithm based on a large amount of historical design drawing data and current trends. It not only takes into account the aesthetics of the design, but also combines market feedback and trends to ensure that the design meets market demand. Through the optimization of the enhanced learning algorithm, the comfort and visual effects of the design are further improved to ensure that the design quality reaches the best state.

[0152] The multi-level and multi-module AI-CAD model has the ability of real-time feedback and self-optimization. During the design process, the system can analyze the generated design drawings in real time and adjust and optimize the design parameters in a timely manner based on the feedback. Through continuous training and optimization, the AI-CAD model can improve itself, improve the quality and efficiency of the design, ensure the efficiency and dynamic optimization capabilities of the design process, and quickly respond to market changes and user needs.

[0153] In the traditional design process, there is often information asymmetry and disconnection between the design and production links, resulting in production errors and inefficiency. Through the AI-CAD model of the present invention, the design drawings can be directly connected to the plate-making equipment, avoiding the errors that may occur during manual conversion. The design drawings generated by the AI-CAD model are accurate and detailed and can be directly used for plate-making production, ensuring that the design intent is accurately realized in the production link, reducing the disconnection between design and production.

[0154] To sum up, the AI-CAD model in this implementation significantly improves the intelligence level and efficiency of the design process, optimizes the design quality and market adaptability, has real-time feedback and self-optimization capabilities, reduces the disconnection between design and production, saves time and costs, and reflects extremely high application value and market potential.

[0155] In this implementation, S3 specifically includes the following sub-steps:

[0156] S31, importing the optimized clothing design pattern P'(x) into the virtual fitting system;

[0157] S32. Establishing a three-dimensional virtual human body model based on the user's body measurement data ;

[0158] S33, applying the optimized clothing design pattern P'(x) to the virtual human body model , using physical simulation algorithms to simulate the wearing effect of clothing and generate preliminary wearing simulation results ;

[0159] S34, virtual fitting system based on wearing simulation results , evaluate the comfort C and visual effect V of clothing:

[0160] ;

[0161] ;

[0162] Among them, C represents the comfort score, V represents the visual effect score, represent the comfort level and error correction coefficient respectively, Represent visual effects and market feedback correction coefficients respectively, Respectively represent the simulation results of comfort and visual effect, denote the error and feedback parameters respectively;

[0163] S35. Based on the comfort score C and the visual effect score V, the virtual fitting system identifies design defects, applies the optimization algorithm to make adjustments, and generates optimization suggestions O:

[0164] ;

[0165] Among them, O represents optimization suggestions, , Respectively represent the optimization weight coefficients of comfort and visual effect, and Respectively represent the current comfort and visual effect parameters, represents other correction factors, The total number of optimization parameters for the optimization suggestion O;

[0166] S36. Update the clothing design pattern according to the optimization suggestion O, and generate an updated design pattern :

[0167] ;

[0168] in, Indicates the updated design drawing. represents the kth optimization parameter, represents the corresponding weight, represents the adjustment factor, Indicates the specific parameters of the optimization suggestion. is the total number of optimization parameters of the design pattern.

[0169] In this implementation, S4 specifically includes the following sub-steps:

[0170] S41, the updated design drawings Import the plate-making module in the plate-making equipment through the data interface;

[0171] S42, the plate-making device receives the clothing design drawing: the control system in the plate-making device parses and receives the clothing design drawing data, and generates the control instructions required for plate-making ,in Specific instruction sets for each plate-making step;

[0172] S43, the plate making equipment receives the control instruction Performing clothing pattern making, using an adaptive cutting tool to perform a cutting operation, the operating parameters of the cutting tool include a cutting speed v, a cutting depth d and a cutting path p;

[0173] S44, sensors on the plate-making equipment monitor key parameters of the cutting process in real time, including cutting accuracy, material usage m and equipment operating status s, and record and transmit monitoring data ;

[0174] S45, based on real-time monitoring data , using the feedback control system to adjust the cutting tool in real time, the cutting tool adjustment parameters include cutting speed correction, cutting depth correction and cutting path correction;

[0175] S46. Under real-time monitoring and feedback control, the plate-making equipment completes all cutting operations, generates clothing patterns that meet design requirements, and stores the final plate-making data in the storage system within the plate-making equipment, records all operating parameters and monitoring data during the plate-making process, and generates a final plate-making report. .

[0176] A clothing pattern production monitoring system based on design process, including the following modules:

[0177] Design module, which is used to create and optimize clothing design drawings, including fabric selection, size specifications and style design, and directly connects with the pattern making module;

[0178] Plate making module, automated plate making equipment, makes clothing plates according to optimized clothing design drawings, equipped with adaptive cutting tools and real-time monitoring and adjustment functions;

[0179] Virtual fitting module: import clothing design drawings into the virtual fitting system to simulate real wearing effects, evaluate and optimize design defects;

[0180] Production monitoring module, a production monitoring sensor installed on the plate-making equipment, monitors key parameters of the plate-making process in real time, including cutting accuracy, material usage and equipment operating status;

[0181] Feedback and control module, real-time data processing and analysis system, sends adjustment instructions to plate-making equipment according to the real-time analysis results of production data, and optimizes production parameters in real time;

[0182] Information management module, data storage system and visual management platform, used to safely store all production data, support historical data query and traceability, and display the real-time status of the production process and historical data analysis results;

[0183] The integrated interface module, data interface and control interface are used for data transmission between the design module and the plate-making module, and for instruction transmission between the feedback and control module and the plate-making equipment. Example

[0184] A well-known clothing brand hopes to improve the design and production efficiency of its new clothing, reduce production errors, and improve product quality by optimizing the design process and plate-making production monitoring system. The brand's design team encountered some challenges at the spring new model release conference: the traditional clothing plate-making method led to a disconnect between design and production, and the lack of real-time monitoring and feedback mechanism during the production process led to unstable finished product quality and low production efficiency. In order to solve these problems, they decided to use the present invention for experimentation.

[0185] At the spring new model release conference, the brand decided to design a fashionable dress and applied the method of the present invention to monitor the entire design process and plate production. The designer first used the AI-CAD model to design according to historical data and current fashion trends, then optimized the design through the virtual fitting system, and finally imported the optimized design pattern into the plate making equipment for production.

[0186] Designers input spring fashion trend data and historical dress design pattern data into the AI-CAD model, and the system automatically generates preliminary design patterns through a combination of deep learning and reinforcement learning. This process uses a generative adversarial network to generate preliminary patterns and further optimizes the details through conditional adversarial training.

[0187] The generated preliminary design showed some potential design problems, such as the length of the skirt and the design of the neckline were not fashionable enough. Through the virtual fitting system, the designer imported the design drawings, built a three-dimensional virtual human model, and simulated the wearing effect. The virtual fitting system evaluated the comfort and visual effect of the clothing and found several specific problems: the length of the skirt needed to be shortened by 3 cm, and the curvature of the neckline needed to be adjusted to enhance the visual effect.

[0188] Based on the feedback from the virtual fitting system, the AI-CAD model generated optimization suggestions, automatically modified the design drawings, and generated the final optimized design drawings. This process ensures the comfort and aesthetics of the design through physical simulation algorithms and optimization algorithms.

[0189] The optimized design pattern is directly imported into the plate-making module of the plate-making equipment through the data interface. After receiving the design pattern, the plate-making equipment generates the control instructions required for plate-making and uses the adaptive cutting tool to perform the cutting operation. During the cutting process, the sensor monitors the cutting accuracy, material usage and equipment operation status in real time, and transmits the data to the feedback control system.

[0190] During the production process, the feedback control system adjusts the operating parameters of the cutting tool in real time based on real-time monitoring data, ensuring cutting accuracy and optimal material use. Finally, the plate-making equipment completes all cutting operations and generates a garment pattern that meets the design requirements. All operating parameters and monitoring data are recorded to generate a final plate-making report.

[0191] In order to prove the effectiveness of the method of the present invention, the design team selected two groups of training samples for comparative experiments: one group adopted the traditional design and plate making method, and the other group adopted the method of the present invention.

[0192] In the traditional method, designers manually converted the design drawings into plate-making data, and the production process relied on manual monitoring and post-inspection. The whole process took 14 days, 4 errors occurred during the production process, and the qualified rate of the finished product was 85.87%.

[0193] In the method of the present invention, the designer uses the AI-CAD model to automatically generate the design drawings, optimizes the design through the virtual fitting system, and the plate-making equipment completes the production under real-time monitoring and feedback control. The whole process took 8 days, and there was 1 error during the production process. The qualified rate of the finished product was 98.21%.

[0194] The specific data are shown in Table 1 below:

[0195] Table 1 Comparison table between traditional method and the method of the present invention

[0196] project Traditional methods Method of the present invention Design and plate making time 14 days 8 days Errors in the production process 4 times 1 time Finished product qualification rate 85.87% 98.21% Data transmission accuracy Low high Real-time monitoring and feedback mechanism none have Production efficiency Low high

[0197] The data in Table 1 fully demonstrate the significant advantages of the method of the present invention in improving production efficiency, reducing production errors and improving the quality of finished products.

[0198] Taking the design of a dress as an example, the designer first input historical dress design patterns and 2023 spring fashion trend data into the AI-CAD model. The system generated preliminary design patterns through deep learning algorithms, including fabric selection, size specifications and style design.

[0199] The designer imported the preliminary design into the virtual fitting system, which simulated the wearing effect of the dress based on the model's 3D human body model. Through physical simulation, the system found several design problems, such as the hem being too long and the curvature of the neckline not conforming to the current fashion trend. The system gave optimization suggestions: shorten the hem by 3 cm and adjust the curvature of the neckline to make it more fashionable.

[0200] According to the optimization suggestions of the virtual fitting system, the AI-CAD model automatically modified the design drawing and generated an optimized design drawing. The optimized design drawing is directly imported into the plate-making equipment, which generates control instructions based on the design drawing and uses adaptive cutting tools for precise cutting. The entire cutting process is carried out under the real-time monitoring of the sensor, and the system makes feedback adjustments based on real-time data to ensure the accuracy and stability of the cutting.

[0201] During the production process, the plate-making equipment recorded all operating parameters and monitoring data and generated the final plate-making report. The final finished product passed the quality inspection with a pass rate of 98.21%, which is much higher than the 85.87% of the traditional method.

[0202] Through the design and production process of the new spring dress, the design team fully verified the feasibility and effectiveness of the method of the present invention. The system not only greatly improved the efficiency of design and production, but also significantly improved product quality and reduced production errors, demonstrating its great potential in clothing pattern making and production.

[0203] The present invention adopts an AI-CAD model to directly connect clothing design drawings with plate-making equipment, avoiding errors in the manual conversion process. The AI-CAD model automatically recommends design, modification or optimization of parameters of clothing design drawings based on historical clothing design drawing data and fashion trends, thereby improving the accuracy and consistency of the design drawings. A multi-level and multi-module AI-CAD model is constructed by combining deep learning and reinforcement learning, which realizes automatic optimization and recommendation of design parameters and significantly reduces the disconnection between design and plate-making.

[0204] The present invention realizes real-time monitoring of the plate-making production process through sensor technology and data acquisition and processing technology. The sensors on the plate-making equipment monitor the cutting accuracy, material usage and equipment operation status in real time, record and transmit the monitoring data, and use the feedback control system to adjust the cutting tool in real time to ensure the accuracy and stability of the cutting operation, effectively improving the stability of the production process and product quality, and avoiding the waste of resources and extended production cycle caused by delayed adjustment in traditional technologies.

[0205] The present invention imports clothing design drawings into the virtual fitting system, simulates the wearing effect of the clothing by establishing a three-dimensional virtual human body model and a physical simulation algorithm, and generates preliminary wearing simulation results. The virtual fitting system evaluates the comfort and visual effect of the clothing based on the wearing simulation results, identifies design defects and generates optimization suggestions, and generates optimized design drawings to ensure the comfort and aesthetics of the design, which not only improves the accuracy of the design, but also significantly reduces the trial and error cost and time in actual production.

[0206] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A method for monitoring the production of clothing pattern making based on a design process, characterized in that: The following steps are involved: S1. Collect fabric selection data, size specification data and style design data in clothing design, and construct clothing design patterns; S2. Build an AI-CAD model, which automatically recommends design, modifies or optimizes parameters of clothing design drawings based on historical clothing design drawing design data and fashion trends; S3. Import the clothing design drawings optimized by the AI-CAD model into the virtual fitting system to simulate the real wearing effect. The virtual fitting system detects the comfort and visual effect of the design, automatically optimizes the design defects, and generates the optimized clothing design drawings; S4. The optimized clothing design pattern is directly connected to the plate-making module in the plate-making equipment. The plate-making equipment makes clothing plates according to the optimized clothing design pattern. The plate-making equipment is equipped with an adaptive cutting tool and a real-time monitoring and adjustment function. S5. The production monitoring sensor installed on the plate-making equipment monitors the cutting accuracy, material usage and equipment operation status in the plate-making process in real time, and transmits data in real time; S6. The plate-making equipment has a built-in central control system that analyzes production data in real time, detects abnormal conditions in the production process, predicts potential problems and provides optimization suggestions; S7. Based on the data analysis results, the feedback and control system sends adjustment instructions to the plate-making equipment to optimize the production parameters in real time until the garment plate-making is completed; The S3 specifically includes the following sub-steps: S31, importing the optimized clothing design pattern P'(x) into the virtual fitting system; S32. Establishing a three-dimensional virtual human body model based on the user's body measurement data ; S33, applying the optimized clothing design pattern P'(x) to the virtual human body model , using physical simulation algorithms to simulate the wearing effect of clothing and generate preliminary wearing simulation results ; S34, virtual fitting system based on wearing simulation results , evaluate the comfort C and visual effect V of clothing: ; ; Among them, C represents the comfort score, V represents the visual effect score, represent the comfort level and error correction coefficient respectively, Represent visual effects and market feedback correction coefficients respectively, Respectively represent the simulation results of comfort and visual effect, denote the error and feedback parameters respectively; S35. Based on the comfort score C and the visual effect score V, the virtual fitting system identifies design defects, applies the optimization algorithm to make adjustments, and generates optimization suggestions O: ; Among them, O represents optimization suggestions, , Respectively represent the optimization weight coefficients of comfort and visual effect, and Respectively represent the current comfort and visual effect parameters, represents other correction factors, The total number of optimization parameters for the optimization suggestion O; S36. Update the clothing design pattern according to the optimization suggestion O, and generate an updated design pattern : ; in, Indicates the updated design drawing. represents the kth optimization parameter, represents the corresponding weight, represents the adjustment factor, Indicates the specific parameters of the optimization suggestion. is the total number of optimization parameters of the design pattern; The S4 specifically includes the following sub-steps: S41, the updated design drawings Import the plate-making module in the plate-making equipment through the data interface; S42, the plate-making device receives the clothing design drawing: the control system in the plate-making device parses and receives the clothing design drawing data, and generates the control instructions required for plate-making ,in Specific instruction sets for each plate-making step; S43, the plate making equipment receives the control instruction Performing clothing pattern making, using an adaptive cutting tool to perform a cutting operation, the operating parameters of the cutting tool include a cutting speed v, a cutting depth d and a cutting path p; S44, sensors on the plate-making equipment monitor key parameters of the cutting process in real time, including cutting accuracy, material usage m and equipment operating status s, and record and transmit monitoring data ; S45, based on real-time monitoring data , using the feedback control system to adjust the cutting tool in real time, the cutting tool adjustment parameters include cutting speed correction, cutting depth correction and cutting path correction; S46. Under real-time monitoring and feedback control, the plate-making equipment completes all cutting operations, generates clothing patterns that meet design requirements, and stores the final plate-making data in the storage system within the plate-making equipment, records all operating parameters and monitoring data during the plate-making process, and generates a final plate-making report. .

2. A method for monitoring clothing plate production based on design process according to claim 1, characterized in that: The S1 specifically includes the following sub-steps: S11. Obtain fabric information required for clothing design through database retrieval or manual input , including fabric type, color, texture and elasticity parameters: ; in, represents the weight of the m-th fabric, Indicates the type of fabric. Indicates the color of the fabric. Indicates the texture of the fabric. Indicates the elastic parameters of the fabric. represents other correction factors that affect fabric selection, and M represents the total number of fabric types; S12. Obtain the size specifications required for clothing design through measurement or database retrieval , including length, sleeve length, shoulder width, chest circumference and waist circumference: ; in, represents the weight of the nth size specification, Indicates the length of the garment. Indicates sleeve length. Indicates shoulder width. Indicates chest circumference, They represent the correction coefficients for sleeve length, shoulder width, and chest circumference, respectively. It represents the correction coefficient of other dimension specification parameters, and N represents the total number of dimension specification parameters; S13. Obtaining style information of clothing design through database retrieval, manual input or automatic generation , including collar, sleeves, hem and decorative elements: ; in, represents the weight of the p-th style design parameter, Indicates collar type, Indicates sleeve type, Indicates the hem of a garment. , They represent the correction coefficients for sleeve type and hem respectively. represents the correction coefficient of other style design parameters, K represents the total number of style design parameters, and P represents the number of all style design elements; S14. Based on the acquired fabric information , Dimensions and style information , use CAD to generate clothing design drawings P(x): ; in, Respectively represent the weight coefficients of fabric selection data, size specification data and style design data, It represents the comprehensive correction coefficient in the process of design pattern generation, I and J represent fabric selection data and size specification data respectively, and x represents the specific design element.

3. The method for monitoring the production of clothing pattern making based on the design process according to claim 2, characterized in that: The S2 specifically includes the following sub-steps: S21. Obtain a large number of historical clothing design patterns through database retrieval or historical data storage ; S22. Obtain current and future trend data based on market research and trend analysis ; S23. Use a combination of deep learning and reinforcement learning to build a multi-level and multi-module AI-CAD model: Image generation module, which generates representative clothing design patterns based on generative adversarial networks; The generated clothing design patterns are optimized using the reinforcement learning algorithm, and the parameters are adjusted in multiple dimensions based on the comfort, visual effects and market feedback of the clothing to generate optimized design parameters; S24, AI-CAD model automatically recommends design parameters based on training and optimization results ; S25. Automatically modify or optimize the clothing design drawing according to the design parameters recommended by the AI-CAD model to generate a final optimized design drawing P'(x).

4. The method for monitoring clothing plate production based on design process according to claim 3, characterized in that: The automatically recommended design parameters Including fabric selection, size specifications and style design; ; in, Respectively represent the weights of fabric selection data, size specification data and style design data, Do not indicate fabric selection data, size specification data and style design data, They represent the correction factors for fabric selection, size specifications and style design data respectively.

5. The method for monitoring clothing plate production based on design process according to claim 3, characterized in that: The S23 specifically includes the following sub-steps: S231, construct an image generation module, and use a generative adversarial network to generate basic clothing design patterns. The generative adversarial network includes a generator G and a discriminator D. The generator G generates basic clothing design patterns through a noise vector z, and the discriminator D identifies the generated patterns: ; Among them, G(z) represents the basic clothing design pattern generated by the generator, and is the weight matrix of the generator, represents the activation function, and is the bias vector of the generator, z is the input noise vector; S232, generating an adversarial network through adversarial training, so that the generator G and the discriminator D compete with each other for optimization, and generate realistic clothing design patterns; S233, on the basis of generating realistic clothing design drawings, generating clothing design drawings including detail design, the input is realistic clothing design drawings and detail information c: ; Among them, G(z|c) represents the detailed clothing design pattern generated by the conditional generator, and is the weight matrix of the conditional generator, and is the bias vector of the conditional generator, [z, c] represents the concatenated vector of the noise vector and the detail information; S234, through conditional adversarial training, the generator G and the discriminator D are optimized under the detail information c to generate a clothing design pattern containing detail design; S235. Based on the clothing design pattern containing detailed design, the reinforcement learning algorithm is applied to optimize the clothing design pattern, and the reward function R is constructed. The parameters are adjusted based on the comfort, visual effect and market feedback of the clothing: ; Among them, R is the reward function, , , are the weights of comfort, visual effect and market feedback, represents the comfort score of the i-th sample, represents the visual effect score of the jth sample, represents the market feedback score of the kth sample, N, , L represent the total number of samples of comfort, visual effect and market feedback respectively; S236. Use the policy gradient algorithm to optimize the design parameters and update the strategy : ; in, represents the policy gradient, represents the policy parameters, Represents a complete design generation process. and denote the action and state at time step t, respectively. Represents the cumulative reward of the entire generation process.

6. The method for monitoring the production of clothing pattern making based on the design process according to claim 5, characterized in that: The training objective function of the generative adversarial network is as follows: ; Among them, V(D, G) is the training objective function, represents the expected value, represents the real data distribution, represents the distribution of the noise vector, D(x) is the discriminator's recognition result for the real clothing design pattern, and D(G(z)) is the discriminator's recognition result for the generated pattern; The training objective function of the conditional generator to generate the adversarial network is as follows: ; Among them, V(D, G|c) is the conditional training objective function, represents the conditional data distribution, D(x|c) is the recognition result of the conditional discriminator on the real clothing design pattern, and D(G(z|c)) is the recognition result of the conditional discriminator on the generated pattern.

7. A clothing pattern production monitoring system based on design process, characterized in that: Includes the following modules: Design module, which is used to create and optimize clothing design drawings, including fabric selection, size specifications and style design, and directly connects with the pattern making module; Plate making module, automated plate making equipment, makes clothing plates according to optimized clothing design drawings, equipped with adaptive cutting tools and real-time monitoring and adjustment functions; Virtual fitting module: import clothing design drawings into the virtual fitting system to simulate real wearing effects, evaluate and optimize design defects; Production monitoring module, a production monitoring sensor installed on the plate-making equipment, monitors key parameters of the plate-making process in real time, including cutting accuracy, material usage and equipment operating status; Feedback and control module, real-time data processing and analysis system, sends adjustment instructions to plate-making equipment according to the real-time analysis results of production data, and optimizes production parameters in real time; Information management module, data storage system and visual management platform, used to safely store all production data, support historical data query and traceability, and display the real-time status of the production process and historical data analysis results; The integrated interface module, data interface and control interface are used for data transmission between the design module and the plate-making module, and for instruction transmission between the feedback and control module and the plate-making equipment.