Intelligent control system and method in garment production process

Abstract

The invention relates to the technical field of intelligent garment production, and discloses an intelligent control system and method in a garment production process, and the system comprises a design and development module, a material management module, a production management module, a quality control module, a logistics and distribution module, and an information management module. The design and development module is used for costume design, template manufacturing, style design and pattern design, converts a design file into a production instruction through a seamless conversion algorithm, carries out virtual fitting of the costume based on high-precision physical simulation, and helps a designer and a customer to carry out accurate verification and adjustment before production; the material management module utilizes an RFID tag and bar code technology. Through the intelligent production scheduling and process control module, intelligent scheduling and real-time process parameter adjustment in the production process are achieved, the production efficiency is remarkably improved, resource utilization is optimized, shutdown and resource waste in the production process are reduced, and the output capacity and economic benefits of the whole production line are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent clothing production, and in particular to an intelligent control system and method in a clothing production process. Background Art

[0002] In the clothing production process, ensuring product quality is crucial. Product quality directly affects the quality and market competitiveness of clothing. Usually, clothing products need to be inspected for defects. However, with the development of the clothing industry and the diversification of market demand, traditional clothing production methods face many challenges, resulting in the following problems:

[0003] In the traditional clothing production process, production scheduling and process control mainly rely on manual experience and fixed production plans, resulting in low production efficiency, easy production line bottlenecks, and inability to respond to market demand fluctuations in a timely manner. In addition, frequent production stoppages and serious waste of resources limit the output capacity and economic benefits of the overall production line.

[0004] Traditional clothing production quality control mainly relies on manual inspection, lacks systematic quality monitoring methods, and is inefficient, prone to missed inspections and misjudgments, resulting in uneven product quality and a high defective rate.

[0005] In the traditional material management and cutting process, the management of fabrics and accessories usually relies on manual records and empirical judgments. The low accuracy of inventory management and material tracking will cause material waste, increase production costs, and low resource utilization.

[0006] Therefore, those skilled in the art provide an intelligent control system and method for a clothing production process to solve the problems raised in the above background technology. Summary of the invention

[0007] In view of the deficiencies of the prior art, the present invention provides an intelligent control system and method in a clothing production process to solve the problems raised in the above background technology.

[0008] To achieve the above objectives, the present invention is implemented through the following technical solutions: an intelligent control system in the clothing production process, including a design and development module, a material management module, a production management module, a quality control module, a logistics and distribution module, and an information management module;

[0009] The design and development module is used for clothing design, sample making, style design and pattern design, and converts design files into production instructions through seamless conversion algorithms. At the same time, based on high-precision physical simulation, it conducts virtual try-on of clothing to help designers and customers conduct precise verification and adjustments before production;

[0010] The material management module uses RFID tags and barcode technology to achieve environmental adaptability management, inventory optimization and material tracking of fabrics and accessories, supports accurate data collection and backup, and uses multifunctional cutting equipment to support automatic cutting of complex fabrics;

[0011] The production management module generates optimized production plans based on real-time data and predictive analysis, and realizes automatic adjustment of process parameters during the production process through adaptive control algorithms;

[0012] The quality control module combines multiple detection technologies and artificial intelligence-assisted identification to monitor production quality in real time, detect and feedback quality problems in a timely manner, and collect and process production data in real time through data verification, cleaning and edge computing, so as to conduct complex analysis in combination with cloud computing;

[0013] The logistics and distribution module uses advanced path planning and navigation technology to automatically manage the storage and distribution of finished products, integrate dynamic order adjustment functions, and work in collaboration with production and logistics systems;

[0014] The information management module monitors and manages workshop production activities in real time, supporting efficient data processing and simplified operations.

[0015] Preferably, the design and development module includes a CAD unit, a CAM unit and a virtual fitting unit; the CAD unit is used for clothing design, sample making and style design, and helps designers to make accurate clothing designs through computer-aided design tools; the CAM unit is responsible for converting CAD designs into production instructions, and supports automatic cutting and production; the virtual fitting unit realizes virtual fitting through high-precision physical simulation, and helps designers and customers to preview and adjust designs before production;

[0016] The material management module includes an intelligent storage unit and an automatic cutting unit; the intelligent storage unit uses RFID tags and barcode technology to achieve environmental adaptability management of fabrics and accessories, inventory optimization and real-time material tracking, and supports data collection and backup; the automatic cutting unit uses multifunctional cutting equipment to support automatic cutting of complex fabrics;

[0017] The production management module includes an intelligent production scheduling unit and a process control unit; the intelligent production scheduling unit generates an optimized production plan based on real-time data and predictive analysis; the process control unit automatically adjusts various process parameters in the production process through an adaptive control algorithm.

[0018] Preferably, the quality control module includes an online detection unit and a data processing unit; the online detection unit combines multiple detection technologies and artificial intelligence-assisted identification to monitor production quality in real time, and promptly detect and feedback quality problems; the data processing unit collects and processes production data in real time through data verification, cleaning and edge computing, and combines cloud computing to perform complex data analysis to support quality improvement and decision-making;

[0019] The logistics and distribution module includes a path planning unit and an order management unit; the path planning unit uses advanced navigation technology to perform automated warehousing and distribution management of finished products; the order management unit integrates a dynamic order adjustment function, coordinates production and logistics systems, and optimizes order fulfillment and distribution processes in real time.

[0020] Preferably, the CAD unit generates and represents complex curves and surfaces in clothing design using NURBS of curve and surface design, and its NURBS formula is:

[0021]

[0022] The base function is recursively defined as:

[0023]

[0024] Where C(u) is the curve point under parameter u, P i is the control point, w i is the control point P i The corresponding weight, N i,p (u) is the i-th basis function, P is the order of the B-spline;

[0025] The linear programming algorithm for tool path optimization in the CAM unit is used to optimize the tool path in the cutting process, reduce cutting time and material waste, and its linear programming model is:

[0026] Objective function:

[0027]

[0028] Where Z is the total cost of the tool path, c i is the cost coefficient of path i, x i is a binary variable;

[0029] Constraints:

[0030]

[0031] Among them, a ij is the contribution coefficient of path i to constraint j, b j is the demand of constraint j;

[0032] The virtual fitting unit uses a physics-based cloth simulation algorithm to simulate the behavior of cloth in a virtual environment. The cloth stress-strain relationship of the finite element method is:

[0033] Stress-strain equation:

[0034] σ=D·ε,

[0035] Where σ is the stress tensor, D is the material stiffness matrix, and ε is the strain tensor;

[0036] Dynamic equilibrium equation:

[0037]

[0038] Where M is the mass matrix, C is the damping matrix, K is the stiffness matrix, u is the displacement vector, and ü are velocity and acceleration vectors, and F is the external force vector.

[0039] Preferably, the intelligent storage unit is based on the RFID inventory optimization and material tracking algorithm, and the RFID tag positioning algorithm can estimate the speed and position of the tag by utilizing the Doppler effect, and its algorithm formula is:

[0040]

[0041] Among them, f d is the Doppler shift, v r is the radial velocity of the tag relative to the reader, v is the velocity of the tag, λ is the wavelength of the signal, and θ is the angle between the tag moving direction and the signal propagation direction;

[0042] The inventory optimization algorithm is used to optimize storage space and material storage location. Its objective function is:

[0043]

[0044] Where Z is the total pickup time or cost, c ij is the time or cost factor for storing material j at location i, x ij is a binary variable indicating whether material j is stored in location i;

[0045] Constraints:

[0046]

[0047] Each location can store at most a single material, and all materials must be stored in a single location.

[0048] Preferably, the automatic cutting unit uses cutting path optimization and material utilization rate algorithms to find the optimal path during the cutting process, maximize material utilization rate and minimize tool movement distance;

[0049] The cutting path optimization algorithm is performed by a genetic algorithm;

[0050] The steps of the genetic algorithm include selection, crossover and mutation:

[0051] a. Initial population generation: randomly generate a single group of clipping paths;

[0052] b. Fitness function calculation: Evaluate the quality of each path. The fitness function can be expressed as the total length of the cutting path or the material utilization rate:

[0053]

[0054] Among them, P i is path i, L i is the total length of path i, W i is the material waste of path i, α is the penalty coefficient used to balance path length and material waste;

[0055] c. Selection: Select excellent paths to enter the next generation according to fitness;

[0056] d. Crossover: Generate new paths through crossover operations:

[0057] Crossover(P1,P2)=Combine(P1,P2),

[0058] eMutation: Randomly change part of the path to introduce diversity:

[0059] Mutation(P i )=Random_Alteration(P i ),

[0060] f. Iteration: Repeat the above steps until the optimal path is reached;

[0061] The material utilization optimization algorithm maximizes material utilization and reduces waste when cutting fabrics through a two-dimensional rectangular packing problem;

[0062] Problem Model:

[0063] Given a rectangular material W×H and a rectangle w to be cut i ×h i , find the arrangement method to maximize the utilization rate of materials;

[0064] Objective function:

[0065]

[0066] Where U is the material utilization rate, x iis a binary variable indicating whether rectangle i is placed;

[0067] Constraints:

[0068]

[0069] Preferably, the process control unit uses an adaptive control algorithm to adjust the process parameters in the production process in real time, and the algorithm formula is as follows:

[0070] Objective function:

[0071]

[0072] Among them, y(k) is the output value at the kth step, y ref (k) is the desired reference value, u(k) is the control input, Δu(k) is the change in the control input, Q and R are weight matrices;

[0073] Constraints:

[0074] u min ≤u(k)≤u max ,

[0075] The control input is within the allowable range;

[0076] y min ≤y(k)≤y max ,

[0077] The process parameter output meets the upper and lower limits of process requirements.

[0078] Preferably, the Canny edge detection algorithm in the image processing in the online detection unit is used to detect the edges in the image to help identify small defects in the fabric or sewing, and its algorithm steps are:

[0079] Gaussian filtering:

[0080] I'(x,y)=I(x,y)*G(x,y,σ),

[0081] Where I(x,y) is the input image, G(x,y,σ(is the Gaussian kernel function, and σ controls the width of the filter;

[0082] Gradient calculation:

[0083]

[0084] Among them, G x and G y are the horizontal and vertical gradients of the image;

[0085] Gradient magnitude and direction:

[0086]

[0087] Where G is the gradient magnitude and θ is the gradient direction.

[0088] Preferably, the order management unit dynamically adjusts the order allocation and path planning based on a dynamic scheduling algorithm based on reinforcement learning, and its algorithm formula is:

[0089] Update the Q value of the strategy:

[0090]

[0091] Among them, Q(s,a) is the Q value of executing action a in state s, α is the learning rate, r is the immediate reward after executing the action, and γ is the discount factor. is the maximum expected return of the next state s′;

[0092] Use the ∈-greedy strategy to balance exploration and exploitation:

[0093]

[0094] where ∈ is the exploration probability and |A(s)| is the number of actions that can be performed in state s.

[0095] The intelligent control method in the clothing production process includes the following steps:

[0096] S1. Design clothing styles, samples and patterns, and then the designer generates curves and surfaces through the NURBS algorithm;

[0097] S2. Seamlessly convert CAD design files into production instructions and optimize cutting paths;

[0098] S3. Using the virtual fitting unit, designers and customers can preview and adjust designs before production by simulating the behavior of fabrics in a virtual environment through high-precision physical simulation and finite element method.

[0099] S4, using intelligent storage units to manage fabrics and accessories through RFID tags and barcode technology, and RFID tag positioning algorithms to estimate the speed and position of tags through the Doppler effect;

[0100] S5. Use the multifunctional cutting device to perform cutting operations according to the design instructions;

[0101] S6. Generate optimized production plans based on real-time data and predictive analysis;

[0102] S7, using adaptive control algorithms to adjust process parameters in real time during production;

[0103] S8, combining multiple detection technologies and AI-assisted recognition to monitor production quality in real time, detect edges in images, and identify defects in fabrics or sewing;

[0104] S9, through data verification, cleaning and edge computing, real-time collection and processing of production data, combined with cloud computing for complex data analysis;

[0105] S10. Use advanced navigation technology to automatically manage the storage and distribution of finished products and plan the optimal distribution route;

[0106] S11. Dynamic scheduling algorithm based on reinforcement learning adjusts order allocation and route planning in real time to optimize order fulfillment and delivery processes.

[0107] The present invention provides an intelligent control system and method in the clothing production process, which has the following beneficial effects:

[0108] 1. The present invention realizes intelligent scheduling of the production process and real-time adjustment of process parameters through intelligent production scheduling and process control modules, significantly improves production efficiency, optimizes resource utilization, reduces shutdowns and resource waste in the production process, and improves the output capacity and economic benefits of the overall production line.

[0109] 2. The present invention combines multiple detection technologies and artificial intelligence-assisted identification through a quality control module to monitor quality problems in the production process in real time, ensuring that each product meets high quality standards. Through the collaborative work of the online detection unit and the data processing unit, the system can provide timely feedback on quality problems, significantly reduce the defective rate, and improve product consistency and reliability.

[0110] 3. The present invention realizes precise management and optimal utilization of materials through the combination of intelligent warehouse management and automatic cutting units, and the application of cutting path optimization algorithm and material utilization optimization algorithm improves material utilization, reduces material waste, reduces production costs, and improves the resource utilization efficiency of the enterprise. BRIEF DESCRIPTION OF THE DRAWINGS

[0111] Figure 1 is a system diagram of the present invention;

[0112] Figure 2 It is a schematic diagram of the design and development module of the present invention;

[0113] Figure 3 It is a schematic diagram of a material management module of the present invention;

[0114] Figure 4 It is a schematic diagram of the production management module of the present invention;

[0115] Figure 5 It is a schematic diagram of the quality control module of the present invention;

[0116] Figure 6 It is a schematic diagram of the logistics and distribution module of the present invention;

[0117] Figure 7 It is a flow chart of the present invention. DETAILED DESCRIPTION

[0118] In order to make the technical personnel in the technical field understand the scheme of the present invention, the technical scheme in the embodiment of the present invention will be clearly and completely described below in combination with the drawings in the embodiment of the present invention. Obviously, the described embodiment is a partial embodiment of the present invention, not a complete embodiment. Based on the embodiment of the present invention, other embodiments obtained by ordinary technicians in the field without creative work should fall within the scope of protection of the present invention.

[0119] The present invention is described in detail below in conjunction with the accompanying drawings:

[0120] Example:

[0121] Please see attached Figure 1 - Attachment Figure 7 The embodiment of the present invention provides an intelligent control system in the clothing production process, including a design and development module, a material management module, a production management module, a quality control module, a logistics and distribution module, and an information management module; the design and development module is used for clothing design, sample making, style design, and pattern design, and converts design files into production instructions through a seamless conversion algorithm, and performs virtual try-ons of clothing based on high-precision physical simulation, helping designers and customers to accurately verify and adjust before production; the material management module uses RFID tags and barcode technology to achieve environmental adaptability management, inventory optimization, and material tracking of fabrics and accessories, supports accurate data collection and backup, and uses multifunctional cutting equipment to support complex fabrics. The production management module generates optimized production plans based on real-time data and predictive analysis, and realizes automatic adjustment of process parameters in the production process through adaptive control algorithms; the quality control module combines multiple detection technologies and artificial intelligence-assisted identification to monitor production quality in real time, detect and feedback quality problems in a timely manner, and collect and process production data in real time through data verification, cleaning and edge computing, so as to combine cloud computing for complex analysis; the logistics and distribution module uses advanced path planning and navigation technology to automatically manage the warehousing and distribution of finished products, integrate dynamic order adjustment functions, and work in collaboration with production and logistics systems; the information management module monitors and manages workshop production activities in real time, and supports efficient data processing and operation simplification.

[0122] Designers in the design and development module can accurately transform ideas into manufacturable products, reducing the deviation between the design and the actual finished product. The virtual try-on function improves the efficiency of design verification, allowing designers and customers to quickly adjust the design before production, shortening the development cycle, reducing the risk of rework, and improving customer satisfaction. The material management module reduces material waste and improves material utilization through precise inventory management and automatic cutting technology, making the material in and out of the warehouse more efficient and accurate, reducing inventory costs and reducing material losses. The production management module significantly improves production efficiency through intelligent scheduling and real-time process control, reduces bottlenecks and shutdowns in the production process, improves product quality, and reduces the impact of human intervention on the production process. The quality control module makes the monitoring of production quality more accurate and efficient, and can promptly discover and correct quality problems in production and reduce the defective rate. The logistics and distribution module improves logistics and distribution efficiency, reduces distribution time and cost, improves the accuracy and timeliness of order fulfillment, and enhances the overall responsiveness of the supply chain. The information management module simplifies the operating process, reduces the possibility of human errors, ensures the accuracy and consistency of data, and improves the transparency and controllability of the production system.

[0123] The design and development module includes CAD unit, CAM unit and virtual fitting unit; CAD unit is used for clothing design, sample making and style design, helping designers to make precise clothing design through computer-aided design tools; CAM unit is responsible for converting CAD design into production instructions, supporting automatic cutting and production; virtual fitting unit realizes virtual try-on through high-precision physical simulation, helping designers and customers preview and adjust designs before production; material management module includes intelligent storage unit and automatic cutting unit; intelligent storage unit uses RFID tags and barcode technology to realize environmental adaptability management of fabrics and accessories, inventory optimization and real-time material tracking, and supports data collection and backup; automatic cutting unit uses multi-functional cutting equipment to support automatic cutting of complex fabrics; production management module includes intelligent scheduling unit and process control unit; the intelligent scheduling unit generates optimized production plans based on real-time data and predictive analysis; the process control unit automatically adjusts the process parameters in the production process through an adaptive control algorithm; the quality control module includes an online detection unit and a data processing unit; the online detection unit combines multiple detection technologies and artificial intelligence-assisted identification to monitor production quality in real time, and promptly detect and feedback quality problems; the data processing unit collects and processes production data in real time through data verification, cleaning and edge computing, and combines cloud computing for complex data analysis to support quality improvement and decision-making; the logistics and distribution module includes a path planning unit and an order management unit; the path planning unit uses advanced navigation technology to perform automated warehousing and distribution management of finished products; the order management unit integrates dynamic order adjustment functions, coordinates production and logistics systems, and optimizes order fulfillment and distribution processes in real time.

[0124] The CAD unit shortens the design cycle, improves the accuracy and efficiency of design, and reduces the errors and time costs of manual design; the CAM unit reduces errors in manual operations, improves production efficiency and the consistency of finished product quality; the virtual fitting unit speeds up the design verification process, reduces the risk of rework in the later stages of production, improves customer satisfaction, and shortens product time to market; the intelligent warehousing unit reduces the complexity of inventory management, improves inventory turnover, reduces inventory backlogs and material losses, and achieves efficient supply chain management; the automatic cutting unit reduces the errors and labor intensity of manual cutting, increases production speed, and improves production efficiency and finished product quality; the intelligent scheduling unit reduces bottlenecks and shutdowns in the production process, thereby improving the overall efficiency of the production line; the process control unit reduces the uncertainty of human intervention, reduces scrap rates and rework costs; the online inspection unit reduces the delay of traditional quality inspections and significantly reduces the defective rate; the data processing unit improves the speed and accuracy of data processing; the path planning unit reduces delivery time and costs, improves the efficiency of logistics management, and improves customer satisfaction and the overall efficiency of the supply chain; the order management unit improves the accuracy and timeliness of order processing and reduces the risk of delayed delivery.

[0125] The CAD unit uses NURBS of curve and surface design to generate and represent complex curves and surfaces in clothing design. Its NURBS formula is:

[0126]

[0127] The base function is recursively defined as:

[0128]

[0129] Where C(u) is the curve point under parameter u, P i is the control point, w i is the control point P i The corresponding weight, N i,p (u) is the i-th basis function, and P is the order of the B-spline.

[0130] The NURBS curve formula is used to calculate the curve point C(u) under the parameter u. The NURBS curve passes through the control point P i , weight w i And basis function N i,p It can accurately describe curves of various shapes, including straight lines, arcs and complex free curves. The advantage of NURBS lies in its flexibility and accuracy, and it can represent geometric shapes and complex organic curves.

[0131] The linear programming algorithm for tool path optimization in the CAM unit is used to optimize the tool path in the cutting process, reduce cutting time and material waste, and its linear programming model is:

[0132] Objective function:

[0133]

[0134] Where Z is the total cost of the tool path, c i is the cost coefficient of path i, x i is a binary variable;

[0135] Constraints:

[0136]

[0137] Among them, a ij is the contribution coefficient of path i to constraint j, b j is the demand of constraint j.

[0138] The recursive definition of basis functions is used to calculate the basis functions in NURBS curves. i,p (u) The basis function determines each control point P i The influence of the curve at parameter u produces different weights at different parameter values ​​u, so that the NURBS curve can pass through or approach the control point smoothly and accurately.

[0139] The virtual fitting unit uses a physics-based cloth simulation algorithm to simulate the behavior of cloth in a virtual environment. The cloth stress-strain relationship of the finite element method is as follows:

[0140] Stress-strain equation:

[0141] σ=D·ε,

[0142] Where σ is the stress tensor, D is the material stiffness matrix, and ε is the strain tensor;

[0143] Dynamic equilibrium equation:

[0144]

[0145] Where M is the mass matrix, C is the damping matrix, K is the stiffness matrix, u is the displacement vector, and ü are velocity and acceleration vectors, and F is the external force vector.

[0146] In virtual fitting, the stress-strain equation is used to calculate how the force on the fabric during wearing affects its deformation, helping to generate the real deformation effect of the fabric, making the virtual fitting effect closer to the real situation;

[0147] The dynamic balance equation is used to simulate the dynamic behavior of fabrics in virtual fitting, such as the swaying of fabrics when walking, fluttering in the wind, or responding to external forces. By solving the equation, the real physical behavior of fabrics in motion can be accurately simulated, and the dynamic effect of fabrics can be displayed, making the virtual fitting experience more realistic. It can preview the performance of clothing in different actions and environments, helping designers and customers better understand the performance characteristics of fabrics.

[0148] The intelligent storage unit is based on RFID inventory optimization and material tracking algorithms. The RFID tag positioning algorithm can estimate the speed and position of the tag by using the Doppler effect. The algorithm formula is:

[0149]

[0150] Among them, f d is the Doppler shift, v r is the radial velocity of the tag relative to the reader, v is the velocity of the tag, λ is the wavelength of the signal, and θ is the angle between the tag movement direction and the signal propagation direction.

[0151] Through the Doppler effect, the system can detect the movement of the tag in the warehouse. Specifically, when the RFID tag moves in the warehouse, the Doppler effect causes the RFID reader to detect the frequency change of the tag's transmitted signal. Based on the frequency change, the system can calculate the radial velocity v of the tag. r and the actual direction of motion θ;

[0152] Using the frequency change data received by multiple readers, the precise location of the RFID tag can be further calculated through triangulation.

[0153] By calculating the radial velocity v of the tag r , the movement speed and direction of the tag can be understood in real time, and the movement of materials in the warehouse can be tracked.

[0154] The inventory optimization algorithm is used to optimize storage space and material storage location. Its objective function is:

[0155]

[0156] Where Z is the total pickup time or cost, c ij is the time or cost factor for storing material j at location i, x ij is a binary variable indicating whether material j is stored in location i;

[0157] Constraints:

[0158]

[0159] Each location can store at most a single material, and all materials must be stored in a single location.

[0160] The objective function is used to minimize the total picking time or total cost in the warehousing system, by optimizing the storage location of materials, reducing the time workers spend picking up goods in the warehouse, or reducing the cost of material storage and retrieval;

[0161] The objective function is to complete the material allocation plan through the optimization algorithm so that the materials are stored in the warehouse in a way that minimizes the total pickup time or minimizes the total storage and retrieval cost, which can improve the warehouse's operating efficiency, reduce the time personnel spend moving around the warehouse, and reduce logistics costs;

[0162] The role of constraints is to ensure that at each location i in the warehouse, at most one material j is stored, to ensure that each material has a clear storage location, to prevent materials from being scattered in multiple locations, causing confusion and increased time when picking up goods, and to ensure the orderliness and efficiency of inventory management.

[0163] The automatic cutting unit uses cutting path optimization and material utilization algorithms to find the optimal path during the cutting process, maximize material utilization and minimize tool movement distance;

[0164] The cutting path optimization algorithm is performed by genetic algorithm;

[0165] The steps of genetic algorithm include selection, crossover and mutation:

[0166] a. Initial population generation: randomly generate a single group of clipping paths;

[0167] b. Fitness function calculation: Evaluate the quality of each path. The fitness function can be expressed as the total length of the cutting path or the material utilization rate:

[0168]

[0169] Among them, P i is path i, L i is the total length of path i, W i is the material waste of path i, α is the penalty coefficient used to balance path length and material waste;

[0170] c. Selection: Select excellent paths to enter the next generation according to fitness;

[0171] d. Crossover: Generate new paths through crossover operations:

[0172] Crossover(P1,P2)=Combine(P1,P2),

[0173] eMutation: Randomly change part of the path to introduce diversity:

[0174] Mutation(P i )=Random_Alteration(P i ),

[0175] f. Iteration: Repeat the above steps until convergence to the optimal path.

[0176] The fitness function is used to evaluate the quality of each cutting path, and the genetic algorithm uses the fitness function to measure the quality of the path, which comprehensively considers the length of the cutting path and the waste of material. Therefore, the genetic algorithm tends to minimize material waste and reduce the path of tool movement distance, thereby improving material utilization and cutting efficiency.

[0177] The role of the crossover operation formula Through the crossover operation, the new path will inherit the good characteristics of the parent path and produce a better next-generation path, which helps to gradually form better solutions in the genetic algorithm, increase the diversity of the algorithm, and avoid falling into the local optimum;

[0178] The mutation operation formula works by applying the mutation operation to the path P. i , increase the diversity of the population, avoid the algorithm from converging to a suboptimal solution too early, prevent the algorithm from falling into a local optimal solution, and increase the chance of finding a global optimal solution, ensuring continuous optimization of the clipping path.

[0179] The material utilization optimization algorithm maximizes material utilization and reduces waste when cutting fabrics through a two-dimensional rectangular packing problem;

[0180] Problem Model:

[0181] Given a rectangular material W×H and a rectangle w to be cut i ×h i , find the arrangement method to maximize the utilization rate of materials;

[0182] Objective function:

[0183]

[0184] Where U is the material utilization rate, x i is a binary variable indicating whether rectangle i is placed;

[0185] Constraints:

[0186]

[0187] The role of the objective function is to maximize the material utilization rate U. The algorithm strives to maximize the use of the material surface during cutting, so that more rectangles can be cut and used, which can reduce the unused material area, reduce waste, and improve the economy and material efficiency of the cutting process;

[0188] The role of constraints is to ensure that during the clipping process, no attempt is made to place a rectangle onto the material that exceeds the total area of ​​the material, and all rectangles are forced to be arranged within the bounds of the material.

[0189] The process control unit uses an adaptive control algorithm to adjust the process parameters in the production process in real time. The algorithm formula is as follows:

[0190] Objective function:

[0191]

[0192] Among them, y(k) is the output value at the kth step, y ref (k) is the desired reference value, u(k) is the control input, Δu(k) is the change in the control input, Q and R are weight matrices;

[0193] Constraints:

[0194] u min ≤u(k)≤u max ,

[0195] The control input is within the allowable range;

[0196] y min ≤y(k)≤y max ,

[0197] The process parameter output meets the upper and lower limits of process requirements.

[0198] The objective function is to make the actual value of the process parameter close to the target value through the adaptive control algorithm, reduce the deviation in the production process, and at the same time, control the stability when adjusting the process parameters to avoid instability caused by excessive adjustment. The objective function helps maintain the stability of the production process, ensure the consistency of product quality, and reduce possible quality fluctuations;

[0199] The role of constraints is to ensure that the actual output of process parameters remains within an acceptable range and prevent situations that exceed process requirements. This is crucial to maintaining product quality and production safety. Through constraints, the system can ensure product consistency and the reliability of the production process.

[0200] The Canny edge detection algorithm in the image processing of the online detection unit is used to detect the edges in the image and help identify small defects in the fabric or sewing. The algorithm steps are as follows:

[0201] Gaussian filtering:

[0202] I'(x,y)=I(x,y)*G(x,y,σ),

[0203] Where I(x,y) is the input image, G(x,y,σ) is the Gaussian kernel function, and σ controls the width of the filter;

[0204] Gradient calculation:

[0205]

[0206] Among them, G x and G y are the horizontal and vertical gradients of the image;

[0207] Gradient magnitude and direction:

[0208]

[0209] Where G is the gradient magnitude and θ is the gradient direction.

[0210] The role of Gaussian filtering: In the process of image processing, noise can cause false edges, affecting the accuracy of edge detection. Through Gaussian filtering, high-frequency noise in the image is smoothed out, and the accuracy of edge detection is improved, which is crucial for identifying real defects in fabrics or sewing in online detection, avoiding misjudgment caused by noise interference;

[0211] The role of gradient calculation: The edge of an image usually corresponds to a region where the grayscale value changes rapidly. By calculating the gradient of the image in the horizontal and vertical directions, the algorithm can identify the changing area and find the potential edge, laying the foundation for subsequent edge positioning and helping to identify subtle defects in fabrics or sewing.

[0212] The role of gradient magnitude and direction By calculating the gradient magnitude and direction, the algorithm can identify that the pixel is an edge point and determine the direction of the edge, which is crucial for subsequent edge connection and further defect detection. It ensures that the detected edges are accurate and coherent, which helps to accurately identify small defects in fabrics or sewing.

[0213] The order management unit dynamically adjusts the order allocation and path planning based on the dynamic scheduling algorithm of reinforcement learning. The algorithm formula is:

[0214] Update the Q value of the strategy:

[0215]

[0216] Among them, Q(s,a) is the Q value of executing action a in state s, α is the learning rate, r is the immediate reward after executing the action, and γ is the discount factor. is the maximum expected return of the next state s′;

[0217] Use the ∈-greedy strategy to balance exploration and exploitation:

[0218]

[0219] where ∈ is the exploration probability and |A(s)| is the number of actions that can be performed in state s.

[0220] The Q-value update formula is used to update the Q-value, which is the Q-learning step in the reinforcement learning algorithm. The Q-value represents the expected return of selecting action a under a given state s. The function of the formula is to enable the algorithm to gradually learn the best action strategy by continuously updating the Q-value, and to make the best decision when facing different orders and logistics scenarios;

[0221] The ε-greedy strategy selects the currently known optimal strategy in most cases, while occasionally trying new random actions, ensuring that the system can fully explore all possible strategy spaces in the early stages of learning and focus on using the best strategies learned in the later stages. In order management, the strategy enables the system to make full use of existing order allocation and path planning experience, and can respond to new problems and changes that may arise by constantly trying new solutions, thereby improving the optimization effect of the overall scheduling strategy.

[0222] The intelligent control method in the clothing production process includes the following steps:

[0223] S1. Design clothing styles, samples and patterns, and then the designer generates curves and surfaces through the NURBS algorithm;

[0224] S2. Seamlessly convert CAD design files into production instructions and optimize cutting paths;

[0225] S3. Using the virtual fitting unit, designers and customers can preview and adjust designs before production by simulating the behavior of fabrics in a virtual environment through high-precision physical simulation and finite element method.

[0226] S4, using intelligent storage units to manage fabrics and accessories through RFID tags and barcode technology, and RFID tag positioning algorithms to estimate the speed and position of tags through the Doppler effect;

[0227] S5. Use the multifunctional cutting device to perform cutting operations according to the design instructions;

[0228] S6. Generate optimized production plans based on real-time data and predictive analysis;

[0229] S7, using adaptive control algorithms to adjust process parameters in real time during production;

[0230] S8, combining multiple detection technologies and AI-assisted recognition to monitor production quality in real time, detect edges in images, and identify defects in fabrics or sewing;

[0231] S9, through data verification, cleaning and edge computing, real-time collection and processing of production data, combined with cloud computing for complex data analysis;

[0232] S10. Use advanced navigation technology to automatically manage the storage and distribution of finished products and plan the optimal distribution route;

[0233] S11. Dynamic scheduling algorithm based on reinforcement learning adjusts order allocation and route planning in real time to optimize order fulfillment and delivery processes.

[0234] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. Intelligent control system in the clothing production process, characterized by: Including design and development module, material management module, production management module, quality control module, logistics and distribution module, information management module; The design and development module is used for clothing design, sample making, style design and pattern design, and converts design files into production instructions through seamless conversion algorithms. At the same time, based on high-precision physical simulation, it conducts virtual try-on of clothing to help designers and customers conduct precise verification and adjustments before production; The material management module uses RFID tags and barcode technology to achieve environmental adaptability management, inventory optimization and material tracking of fabrics and accessories, supports accurate data collection and backup, and uses multifunctional cutting equipment to support automatic cutting of complex fabrics; The production management module generates optimized production plans based on real-time data and predictive analysis, and realizes automatic adjustment of process parameters during the production process through adaptive control algorithms; The quality control module combines multiple detection technologies and artificial intelligence-assisted identification to monitor production quality in real time, detect and feedback quality problems in a timely manner, and collect and process production data in real time through data verification, cleaning and edge computing, so as to conduct complex analysis in combination with cloud computing; The logistics and distribution module uses advanced path planning and navigation technology to automatically manage the storage and distribution of finished products, integrate dynamic order adjustment functions, and work in coordination with production and logistics systems; The information management module monitors and manages workshop production activities in real time, supporting efficient data processing and simplified operations.

2. The intelligent control system in the clothing production process according to claim 1 is characterized in that: The design and development module includes a CAD unit, a CAM unit and a virtual fitting unit; the CAD unit is used for clothing design, sample making and style design, helping designers to make accurate clothing designs through computer-aided design tools; the CAM unit is responsible for converting CAD designs into production instructions, supporting automatic cutting and production; the virtual fitting unit realizes virtual fitting through high-precision physical simulation, helping designers and customers to preview and adjust designs before production; The material management module includes an intelligent storage unit and an automatic cutting unit; the intelligent storage unit uses RFID tags and barcode technology to achieve environmental adaptability management of fabrics and accessories, inventory optimization and real-time material tracking, and supports data collection and backup; the automatic cutting unit uses multifunctional cutting equipment to support automatic cutting of complex fabrics; The production management module includes an intelligent production scheduling unit and a process control unit; the intelligent production scheduling unit generates an optimized production plan based on real-time data and predictive analysis; The process control unit automatically adjusts various process parameters during the production process through an adaptive control algorithm.

3. The intelligent control system in the clothing production process according to claim 1 is characterized in that: The quality control module includes an online detection unit and a data processing unit; the online detection unit combines multiple detection technologies and artificial intelligence-assisted identification to monitor production quality in real time, and promptly detect and feedback quality problems; the data processing unit collects and processes production data in real time through data verification, cleaning and edge computing, and combines cloud computing to perform complex data analysis to support quality improvement and decision-making; The logistics and distribution module includes a path planning unit and an order management unit; the path planning unit uses advanced navigation technology to perform automated warehousing and distribution management of finished products; the order management unit integrates a dynamic order adjustment function, coordinates production and logistics systems, and optimizes order fulfillment and distribution processes in real time.

4. The intelligent control system in the clothing production process according to claim 2 is characterized in that: The CAD unit uses NURBS of curve and surface design to generate and represent complex curves and surfaces in clothing design, and its NURBS formula is: The base function is recursively defined as: Where C(u) is the curve point under parameter u, P i is the control point, w i is the control point P i The corresponding weight, N i,p (u) is the i-th basis function, P is the order of the B-spline; The linear programming algorithm for tool path optimization in the CAM unit is used to optimize the tool path in the cutting process, reduce cutting time and material waste, and its linear programming model is: Objective function: Where Z is the total cost of the tool path, c i is the cost coefficient of path i, x i is a binary variable; Constraints: Among them, a ij is the contribution coefficient of path i to constraint j, b j is the demand of constraint j; The virtual fitting unit uses a physics-based cloth simulation algorithm to simulate the behavior of cloth in a virtual environment. The cloth stress-strain relationship of the finite element method is: Stress-strain equation: σ=D·ε, Where σ is the stress tensor, D is the material stiffness matrix, and ε is the strain tensor; Dynamic equilibrium equation: Where M is the mass matrix, C is the damping matrix, K is the stiffness matrix, u is the displacement vector, and are velocity and acceleration vectors, and F is the external force vector.

5. The intelligent control system in the clothing production process according to claim 2 is characterized in that: The intelligent storage unit is based on the RFID inventory optimization and material tracking algorithm. The RFID tag positioning algorithm can estimate the speed and position of the tag by utilizing the Doppler effect. The algorithm formula is: Among them, f d is the Doppler shift, v r is the radial velocity of the tag relative to the reader, v is the velocity of the tag, λ is the wavelength of the signal, and θ is the angle between the tag moving direction and the signal propagation direction; The inventory optimization algorithm is used to optimize storage space and material storage location. Its objective function is: Where Z is the total pickup time or cost, c ij is the time or cost factor for storing material j at location i, x ij is a binary variable indicating whether material j is stored in location i; Constraints: Each location can store at most a single material, and all materials must be stored in a single location.

6. The intelligent control system in the clothing production process according to claim 2 is characterized in that: The automatic cutting unit uses cutting path optimization and material utilization rate algorithms to find the optimal path during the cutting process, maximize material utilization rate and minimize tool movement distance; The cutting path optimization algorithm is performed by a genetic algorithm; The steps of the genetic algorithm include selection, crossover and mutation: a. Initial population generation: randomly generate a single group of clipping paths; b. Fitness function calculation: Evaluate the quality of each path. The fitness function can be expressed as the total length of the cutting path or the material utilization rate: Among them, P i is path i, L i is the total length of path i, W i is the material waste of path i, α is the penalty coefficient used to balance path length and material waste; c. Selection: Select excellent paths to enter the next generation according to fitness; d. Crossover: Generate new paths through crossover operations: Crossover(P1,P2)=Combine(P1,P2), eMutation: Randomly change part of the path to introduce diversity: Mutation(P i )=Random_Alteration(P i ), f. Iteration: Repeat the above steps until the optimal path is reached; The material utilization optimization algorithm maximizes material utilization and reduces waste when cutting fabrics through a two-dimensional rectangular packing problem; Problem Model: Given a rectangular material W×H and a rectangle w to be cut i ×h i , find the arrangement method to maximize the utilization rate of materials; Objective function: Where U is the material utilization rate, x i is a binary variable indicating whether rectangle i is placed; Constraints:

7. The intelligent control system in the clothing production process according to claim 2 is characterized in that: The process control unit uses an adaptive control algorithm to adjust the process parameters in the production process in real time. The algorithm formula steps are: Objective function: Among them, y ( k) is the output value at step k, y ref (k ) is the expected reference value, u(k ) is the control input, Δu(k) is the change in the control input, Q and R are weight matrices; Constraints: in min ≤u(k)≤u max , The control input is within the allowable range; and min ≤y(k)≤y max , The process parameter output meets the upper and lower limits of process requirements.

8. The intelligent control system in the clothing production process according to claim 3 is characterized in that: The Canny edge detection algorithm in the image processing in the online detection unit is used to detect the edges in the image and help identify small defects in fabrics or sewing. The algorithm steps are as follows: Gaussian filtering: I'(x,y)=I(x,y)*G(x,y,σ), Among them, I(x,y) is the input image, G ( x,y,σ) is the Gaussian kernel function, σ controls the width of the filter; Gradient calculation: Among them, G x and G y are the horizontal and vertical gradients of the image; Gradient magnitude and direction: Where G is the gradient magnitude and θ is the gradient direction.

9. The intelligent control system in the clothing production process according to claim 3 is characterized in that: The order management unit dynamically adjusts the order allocation and path planning based on the dynamic scheduling algorithm of reinforcement learning, and its algorithm formula is: Update the Q value of the strategy: Among them, Q(s,a) is the Q value of executing action a in state s, α is the learning rate, r is the immediate reward after executing the action, and γ is the discount factor. is the maximum expected return of the next state s′; Use the ∈-greedy strategy to balance exploration and exploitation: where ∈ is the exploration probability and |A(s)| is the number of actions that can be performed in state s.

10. An intelligent control method in a clothing production process, according to the intelligent control system in a clothing production process according to any one of claims 1 to 9, characterized in that: The steps include: S1. Design clothing styles, samples and patterns, and then the designer generates curves and surfaces through the NURBS algorithm; S2. Seamlessly convert CAD design files into production instructions and optimize cutting paths; S3. Using the virtual fitting unit, designers and customers can preview and adjust designs before production by simulating the behavior of fabrics in a virtual environment through high-precision physical simulation and finite element method. S4, using intelligent storage units to manage fabrics and accessories through RFID tags and barcode technology, and RFID tag positioning algorithms to estimate the speed and position of tags through the Doppler effect; S5. Use the multifunctional cutting equipment to perform cutting operations according to the design instructions; S6. Generate optimized production plans based on real-time data and predictive analysis; S7, using adaptive control algorithms to adjust process parameters in real time during production; S8, combining multiple detection technologies and AI-assisted recognition to monitor production quality in real time, detect edges in images, and identify defects in fabrics or sewing; S9, through data verification, cleaning and edge computing, real-time collection and processing of production data, combined with cloud computing for complex data analysis; S10. Use advanced navigation technology to automatically manage the storage and distribution of finished products and plan the optimal distribution route; S11. Dynamic scheduling algorithm based on reinforcement learning adjusts order allocation and route planning in real time to optimize order fulfillment and delivery processes.

Images