Medical auxiliary sock production plan generation method and system based on artificial intelligence

Abstract

The invention relates to the technical field of production design optimization and verification, in particular to a medical auxiliary sock production plan generation method and system based on artificial intelligence, and the method comprises the steps: obtaining the expected order data of medical auxiliary socks, including order demand information, product specification information and delivery requirement information, dividing the expected order data into a plurality of subclasses, obtaining M groups of order subclass data; obtaining a production plan constraint condition, and obtaining M groups of to-be-produced order data according to the inventory data and the M groups of order subclass data; generating a plurality of candidate production strategies according to the M groups of to-be-produced order data and the production plan constraint conditions; establishing a medical auxiliary sock production digital twinborn model; and simulating and verifying different candidate production strategies through the medical auxiliary sock production digital twin model to obtain virtual delivery time of each order to be produced, and selecting an optimal production plan according to the virtual delivery time. According to the method, through simulation verification of the digital twinborn model, the decision transparency and predictability of the production plan can be enhanced.

Description

Technical Field

[0001] The present invention relates to the technical field of production design optimization and verification, and in particular to a method and system for generating a production plan for medical auxiliary socks based on artificial intelligence. Background Art

[0002] Medical assistive socks are elastic socks that can reduce venous pressure in the lower limbs by applying appropriate pressure to the legs. They have achieved remarkable results in relieving leg discomfort. Therefore, they have attracted widespread attention from consumers, and the production demand for medical assistive socks has gradually increased. The production of medical assistive socks is characterized by small batches and customization, and requires strict quality testing to ensure that the products meet the standards. This production feature poses a challenge to the traditional large-scale, standardized production model. Traditional production planning methods have been unable to efficiently respond to the growing market demand. There is an urgent need for an efficient method for generating production plans for medical assistive socks to better coordinate production resources, optimize production processes, and improve production efficiency.

[0003] The continuous development of artificial intelligence technology provides technical support and guarantee for the generation of intelligent production plans. At present, the production orders of medical auxiliary socks are characterized by diverse styles and large differences in delivery time. Traditional production plan generation methods usually focus on optimizing the production efficiency of the production line, but often ignore the differences between orders, resulting in overdue orders or unreasonable production scheduling while pursuing maximum production capacity. In addition, most traditional methods rely on improving the efficiency of a single production line and lack the ability to flexibly adapt to diversified order demands and dynamic delivery times. For complex production tasks with diverse order styles and different delivery times, traditional methods often cannot provide comprehensive data support, making it difficult for decision makers to rationally evaluate the pros and cons of different strategies, increasing the risk and uncertainty of decision-making.

[0004] Therefore, a method and system for generating production plan of medical auxiliary socks based on artificial intelligence was proposed. Summary of the invention

[0005] The purpose of the present invention is to provide a method and system for generating a production plan for medical auxiliary socks based on artificial intelligence, including: obtaining expected orders for medical auxiliary socks, including unscheduled orders and forecast orders; obtaining expected order data according to expected orders, including order demand information, product specification information and delivery demand information; dividing the order data into multiple groups of sub-category data, obtaining multiple groups of to-be-produced order data according to inventory conditions and multiple groups of order sub-category data, and generating multiple candidate production strategies in combination with constraints such as delivery time, production batch, equipment capacity, process dependency and inventory. By establishing a digital twin model for the production of medical auxiliary socks, simulating the entire production process, including production, quality inspection and inventory management processes, the virtual delivery time of each to-be-produced order is obtained. After comparing the virtual delivery time with the expected delivery time, overdue and inventory orders are identified, the loss cost is calculated, and finally the candidate production strategy with the lowest loss cost is selected as the optimal production plan.

[0006] To achieve the above object, the present invention provides the following technical solutions: A method for generating a production plan for medical auxiliary socks based on artificial intelligence, comprising: Obtaining expected orders for designated medical auxiliary socks, wherein the expected orders include unscheduled orders and predicted orders; obtaining expected order data according to the expected orders, wherein the expected order data includes order demand information, product specification information, and delivery demand information; determining order priority according to the order demand information and the delivery demand information; the predicted orders are predicted based on historical order data of the medical auxiliary socks; Dividing the expected order data into M groups of order sub-category data; Obtaining production plan constraints for the medical auxiliary socks, including delivery time constraints, production line minimum start-up batch constraints, equipment production capacity constraints, process sequence dependency constraints, and inventory constraints; According to the inventory data and the order sub-category data of group M, M groups of to-be-produced order data are obtained; Generate multiple candidate production strategies according to the M groups of to-be-produced order data and the production plan constraint conditions; Establishing a digital twin model for the production of medical auxiliary socks; the digital twin model for the production of medical auxiliary socks is used to simulate the complete production process of the medical auxiliary socks, including the production process, quality inspection process and inventory management process; The complete production process of the candidate production strategy is simulated by the medical assistive socks production digital twin model to obtain the virtual delivery time of each order to be produced; and the optimal production plan is obtained based on the virtual delivery time.

[0007] Further, the expected order data is divided into M groups of order subcategory data including: the product specification information includes product size, product style, product color and product pressure level; the orders with the same product size, product style, product color and product pressure level are divided into a group to obtain M groups of order subcategory data.

[0008] Furthermore, the acquisition of the production plan constraint conditions includes: the delivery time constraint is determined according to the expected delivery time of the delivery demand information; the minimum start-up batch constraint of the production line is acquired according to historical production line data; the equipment production capacity constraint is acquired according to historical equipment production capacity data; the process sequence dependency constraint is acquired according to the production line flow chart; the inventory constraint is acquired according to historical inventory data.

[0009] Further, the M groups of to-be-produced order data are obtained according to the inventory data and the M groups of order subcategory data, including: the inventory data includes inventory product specification information; the inventory product specification information includes inventory product size, inventory product style, inventory product color and inventory product pressure level; orders with the same inventory product size, inventory product style, inventory product color and inventory product pressure level are divided into one group to obtain M groups of inventory subcategory data; According to the M groups of inventory sub-category data and the M groups of order sub-category data, the inventory quantity in each group of the inventory sub-category data is compared with the demand quantity in the corresponding order sub-category data; if the demand quantity is greater than the inventory quantity, the difference between the demand quantity and the inventory quantity is used as the order to be produced, and a column of order matching data is added to the inventory data to record the order number corresponding to the inventory, so as to form inventory matching order data; the order matching data is generated according to the matching result of the product specification information and the inventory product specification information, and combined with the order priority; if the demand quantity is not greater than the inventory quantity, the order data to be produced is empty, and the order quantity to be produced is 0; Eliminate the order data corresponding to the order matching data from the corresponding order sub-category data to obtain the expected production order data; obtain the configuration data of Q production lines, including applicable product specification data and process capability data; Match the configuration data with the expected to-be-produced order data to obtain Q groups of order combination data; the order combination data includes the order subcategory data required to be produced by each of the production lines; sort the Q groups of order combination data from high to low according to the order priority to obtain Q groups of to-be-produced order data.

[0010] Furthermore, generating a plurality of candidate production strategies based on the M groups of production order data and the production plan constraints includes: dividing each process in the production process and the quality inspection process into independent processes and shared processes according to the production line flow chart to obtain process classification data; generating a plurality of candidate production strategies based on the process classification data and in combination with the production plan constraints using an intelligent optimization algorithm, each candidate production strategy including the production time and production sequence of each independent process and the shared process.

[0011] Furthermore, using the intelligent optimization algorithm to generate a plurality of candidate production strategies includes: Initializing a production strategy population, wherein each production strategy in the production strategy population includes the production time and production sequence of each process on each production line; defining the fitness function of the intelligent optimization algorithm according to the total queue time and the capacity utilization rate; The iterative process of the intelligent optimization algorithm includes the following steps: According to the production time, production sequence and queue time calculated by the queue time prediction model in each of the production strategies, the fitness value of each individual is calculated; according to the fitness value, the individual with the fitness value greater than the preset threshold is selected as the parent; based on the parent, a new child production strategy is generated through a crossover operation; a mutation operation is introduced into the child production strategy to randomly adjust the production sequence and / or production time of the process to obtain the mutated child production strategy; in each iteration, the fitness value of each generated mutated child production strategy is calculated according to the fitness function, and the production strategy population is updated according to the fitness value; it is determined whether the stop condition is met; when the stop condition is not met, the iterative process is repeated; when the stop condition is met, a plurality of candidate production strategies are generated; each candidate production strategy is used to determine the production time and production sequence of each process of each production line for the corresponding order data to be produced.

[0012] Furthermore, the establishment process of the digital twin model of medical auxiliary socks production is specifically as follows: creating a virtual physical model for real-time simulation of the equipment status and process parameters in the production process; building a logical model for analyzing and judging key indicators in the production process and detecting abnormal conditions; building a simulation model, which obtains evaluation indicators of different production strategies by mapping the equipment status and process parameters in the virtual physical model; Constructing a data model; the data model is used to provide data input for the virtual physical model, the logical model and the simulation model.

[0013] Furthermore, the production process data includes: process parameters, production speed of each process, queue status of each process and production result data.

[0014] Furthermore, obtaining the optimal production plan based on the virtual delivery time includes: comparing the virtual delivery time with the expected delivery time in the delivery demand information to identify overdue orders and inventory orders; calculating the overdue duration of the overdue orders and the inventory duration of the inventory orders based on the virtual delivery time and the expected delivery time, and generating loss costs based on the overdue orders, the overdue duration, the inventory orders and the inventory duration; and selecting the candidate production strategy with the smallest loss cost as the optimal production plan.

[0015] A medical auxiliary socks production plan generation system based on artificial intelligence, comprising: An order data acquisition module is configured to acquire expected orders for a specified medical auxiliary socks, wherein the expected orders include unscheduled orders and predicted orders; acquire expected order data according to the expected orders, wherein the expected order data includes order demand information, product specification information, and delivery demand information; determine order priority according to the order demand information and the delivery demand information; and the predicted order is predicted based on the historical order data of the medical auxiliary socks; An order classification module, dividing the expected order data into M groups of order sub-category data; A production plan constraint management module is used to obtain production plan constraint conditions of the medical auxiliary socks, including delivery time constraint, production line minimum start-up batch constraint, equipment production capacity constraint, process sequence dependency constraint and inventory constraint; A module for generating orders to be produced obtains M groups of data of orders to be produced according to the inventory data and the order sub-category data of the M groups; A production strategy generation module generates a plurality of candidate production strategies according to the M groups of to-be-produced order data and the production plan constraint conditions; A digital twin model establishment module is used to establish a digital twin model for the production of medical auxiliary socks; the digital twin model for the production of medical auxiliary socks is used to simulate the complete production process of the medical auxiliary socks, including the production process, quality inspection process and inventory management process; A production strategy simulation module, which simulates the complete production process of the candidate production strategy through the medical auxiliary socks production digital twin model to obtain the virtual delivery time of each order to be produced; The production plan acquisition module obtains the optimal production plan according to the virtual delivery time.

[0016] Compared with the prior art, the present invention has the following beneficial effects: 1. The present invention proposes a digital twin model for the production of medical auxiliary socks, which can monitor the equipment status and process parameters in real time during the production process by accurately simulating the physical characteristics, business logic and production rules of production equipment and processes. The combination of these models can not only effectively analyze and evaluate the key indicators in the production process, but also detect potential abnormalities, and adjust the process parameters in time according to the preset production rules, thereby avoiding production interruptions caused by equipment failure or process fluctuations. In addition, the digital twin model uses IoT devices and sensor groups to collect real-time data, including various key data such as inventory, production process and quality inspection process, and standardizes them through data models to provide high-quality data support for virtual physical models, logical models and simulation models. This closed-loop data flow enables real-time feedback and precise control of every link in the production process, greatly improving the transparency of production. Through the digital twin model, the execution effects and potential risks of different candidate production strategies can be more clearly understood, and more data basis can be provided when making decisions. This not only makes the adjustment of production plans more flexible and efficient, but also greatly enhances the predictability of the entire production process, thereby reducing decision-making risks and ensuring the stability and reliability of production plans.

[0017] 2. The present invention provides an efficient and dynamic solution for the production plan of medical auxiliary socks by combining intelligent optimization algorithms with machine learning technology. First, the queue time prediction model is used for training based on historical production line data, which can accurately predict the queue time of each production line and each process, thereby optimizing the scheduling and resource allocation of the production plan and reducing production bottlenecks and delays. In the process of generating candidate production strategies, the iterative optimization method of the genetic algorithm is adopted. Through crossover and mutation operations, different production strategy combinations can be explored, and the effect of each solution can be evaluated through the fitness function. It provides a clear and feasible strategy basis for the subsequent application of the digital twin model, avoids the calculation and decision-making difficulties caused by the infinite strategy space, and makes the generation process of candidate production strategies more efficient and controllable. Genetic algorithms can effectively deal with complex production constraints, multi-objective optimization problems and huge solution spaces, ensuring that the generated candidate production strategies not only meet production requirements, but also adapt to dynamic changes in the production environment, thereby providing accurate, reliable and efficient solutions, providing strong guarantees for the adjustment and execution of production plans, and helping to reduce the decision-making risks of production plans.

[0018] 3. The present invention combines the digital twin model to simulate the production process of medical auxiliary socks according to the candidate production strategy in a virtual environment, and compares the simulated virtual delivery time with the expected delivery time. The digital twin model can truly reflect the dynamic changes of each link in the production process, help identify overdue orders and inventory orders under each candidate production strategy, and effectively calculate the overdue time and inventory time. At the same time, the digital twin model can identify and avoid potential production risks in advance by providing a global view of the entire production process. Through the comparative analysis of virtual simulation and actual data, the generated loss cost can more accurately reflect the actual benefits of the candidate production strategy. Finally, by selecting the candidate production strategy with the smallest loss cost as the optimal production plan, not only the configuration of production resources is optimized, but also the execution effect and risk of each candidate production strategy can be clearly seen, so that more reasonable decisions can be made to ensure the stability and reliability of the production plan. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 A flow chart of a method for generating a production plan for medical auxiliary socks based on artificial intelligence provided by an embodiment of the present invention; Figure 2 A production line flow chart provided for an embodiment of the present invention; Figure 3 A flow chart of a method for generating multiple candidate production strategies using an intelligent optimization algorithm provided by an embodiment of the present invention; Figure 4 A structural diagram of a medical auxiliary socks production plan generation system based on artificial intelligence provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0020] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0021] Elastic stockings are a major type of medical auxiliary stockings, mainly used to promote blood circulation and relieve leg discomfort. Medical auxiliary stockings have different pressure levels, sizes and styles to meet the personalized needs of users. In the production of medical auxiliary stockings, different orders may require different specifications such as styles, sizes, colors and pressure levels. In order to ensure the effect of the product and user experience, the quality consistency of each batch of products must be strictly controlled during production. However, in the face of diversified orders and complex production requirements, the production management of medical auxiliary stockings faces huge challenges. The production of medical auxiliary stockings often involves multiple processes, such as weaving, suturing, pressure testing, quality inspection and packaging, and the process and production sequence of each process need to be precisely controlled. In the face of these needs, if there is a lack of an intelligent production planning system, there will not only be a risk of delayed delivery, but also inventory backlogs or waste of resources.

[0022] See also Figures 1 to 4 The present invention provides a method and system for generating a production plan for medical auxiliary socks based on artificial intelligence. The technical solution is specifically described below through two embodiments.

[0023] Embodiment 1: Company A is a manufacturer of medical auxiliary socks. Its medical auxiliary socks are sold in a pre-sale mode, that is, before formal production, production plans and stocking are planned based on consumers' pre-order demand. Company A has applied a method for generating production plans for medical auxiliary socks based on artificial intelligence, such as Figure 1 As shown; See also Figure 1 S10 in which expected orders for the specified medical auxiliary socks are obtained, the expected orders including unscheduled orders and predicted orders; expected order data are obtained according to the expected orders, referring to Table 1, the expected order data including order demand information, product specification information and delivery demand information; order priority is determined according to the order demand information and the delivery demand information; the predicted order is predicted based on the historical order data of the medical auxiliary socks; Specifically, the order priority is determined comprehensively based on the order quantity in the order demand information, the customer category and the delivery time in the delivery demand information.

[0024] Table 1 Example table of expected order data

[0025] In this embodiment, the designated medical auxiliary socks refer to specific medical auxiliary socks products (specifically a certain type of elastic socks) that are the production scheduling targets. Unscheduled orders are specifically orders that customers have placed orders but have not yet been scheduled for production. Forecast orders refer to orders generated based on historical sales data, market demand forecasts, and customer pre-order trends, which represent estimates of demand that is expected to be generated within a specified time in the future. These orders have not yet been actually placed by customers, but potential order demand is predicted based on factors such as past order data, seasonality, and promotional activities. As a feasible implementation method in the embodiment, first collect the data for the same period in the past five years and record the data at a daily time resolution. Through data cleaning and preprocessing, remove outliers and missing values ​​to obtain historical standard order data. Then, a demand forecasting model is constructed based on standard order data using a machine learning model (such as an XGBoost model) or a time series analysis method (such as ARIMA or Prophet). The demand forecasting model is updated regularly and combined with the latest real-time order data for rolling forecasts to generate order demand data for a period of time in the future.

[0026] See also Figure 1 S20 in which the expected order data is divided into M groups of order sub-category data; Further, the expected order data is divided into M groups. The order sub-category data includes: The product specification information includes product size, product style, product color and product pressure level; Orders with the same product size, product style, product color and product pressure level are divided into one group to obtain M groups of order subcategory data; the calculation formula of M is: ; in, is the number of categories of the product size, is the number of categories of the product style, is the number of product color categories, It is the category number of the pressure level of the product.

[0027] Table 2 Example table of expected order data grouping

[0028] In this embodiment, the product sizes are divided into three types: S, M and L, the product styles are divided into two types: open toe and closed toe, the product colors are divided into two types: skin color and black, and the product pressure levels are divided into three types: level one, level two and level three.

[0029] but ; The expected order data is divided into 36 groups of order subcategory data, and the order quantity of each group is summarized and integrated to obtain the total demand of each subcategory, as shown in Table 2.

[0030] By classifying the expected order data according to product size, product style, product color and product pressure level, the orders can be managed in a refined manner, fully considering the impact of different product specifications on the production plan. This classification method can achieve batch production of products of the same specification, reduce the number of production line switching, reduce the time and cost consumption caused by frequent adjustments to equipment and process parameters during the production process, and improve production efficiency. At the same time, clear order classification helps to accurately predict the demand for various products and avoid waste of resources or production delays caused by excessive or insufficient inventory.

[0031] See also Figure 1 S30 in which the production plan constraints of the medical auxiliary socks are obtained, including delivery time constraints, production line minimum start-up batch constraints, equipment production capacity constraints, process sequence dependency constraints and inventory constraints; Furthermore, obtaining the production plan constraint conditions includes: The delivery time constraint is determined according to the expected delivery time of the delivery demand information; The minimum startup batch constraint of the production line is obtained based on historical production line data; The equipment production capacity constraint is obtained based on historical equipment production capacity data; The process sequence dependency constraints are based on the production line flow chart (such as Figure 2 As shown) obtain; The inventory constraint is obtained based on historical inventory data.

[0032] By obtaining the constraints of the production plan from historical data and related information, including delivery time constraints, production line minimum start-up batch constraints, equipment production capacity constraints, process sequence dependency constraints and inventory constraints, the production plan can be made more accurate and reliable. The reference value of historical data is utilized to truly reflect the actual capacity and limitations in the production process, avoid the idealization of production strategies, and make the production plan more feasible and practical. By fully considering various constraints, production delays, resource waste and unreasonable production scheduling caused by incomplete or inaccurate information are avoided.

[0033] See also Figure 1 S40, obtaining M groups of to-be-produced order data according to the inventory data and the M groups of order subcategory data; Further, obtaining the M groups of to-be-produced order data according to the inventory data and the M groups of order subcategory data includes: The inventory data includes inventory product specification information; the inventory product specification information includes inventory product size, inventory product style, inventory product color and inventory product pressure level; orders with the same inventory product size, inventory product style, inventory product color and inventory product pressure level are divided into one group to obtain M groups of inventory sub-category data; According to the M groups of inventory subcategory data and the M groups of order subcategory data, compare the inventory quantity in each group of the inventory subcategory data with the demand quantity in the corresponding order subcategory data; if the demand quantity is greater than the inventory quantity, take the difference between the demand quantity and the inventory quantity as the order to be produced, add a column of order matching data to the inventory data, which is used to record the order number corresponding to the inventory, and generate inventory matching order data, as shown in Table 3. This data table comprehensively records the order demand, inventory status and the order number matching the inventory, which is convenient for inventory management and production scheduling. Among them, the newly added "order matching data" column is used to identify the correspondence between each order and inventory. The order matching data is generated based on the matching results of the product specification information and the inventory product specification information, and combined with the order priority; if the demand quantity is less than or equal to the inventory quantity, the order data to be produced is empty, and the order quantity to be produced is 0; Table 3 Inventory matching order data example table

[0034] Eliminate the order data corresponding to the order matching data from the corresponding order subcategory data to obtain expected production order data; Obtain the configuration data of Q production lines, including applicable product specification data and process capability data; Q is the number of production lines in the first process.

[0035] By accurately matching inventory data with order subclass data, this method can accurately determine the order quantity that needs to be produced, avoid unnecessary production and inventory backlog, and improve the efficiency of inventory management and production scheduling. By classifying inventory and orders according to product specifications (size, style, color and pressure level), refined management of inventory and order data helps optimize production plans. Comparing inventory with order demand, giving priority to orders that can be met by inventory, reducing the production burden; for orders with insufficient inventory, generating pending orders based on demand difference, and sorting them according to order priority to ensure that important or urgent orders are given priority. Matching the configuration data of the production line with the expected pending order data enables each production line to produce the most suitable order according to its own process capabilities and applicable product specifications, thereby improving the utilization rate and production efficiency of the production line. Finally, by sorting by order priority, ensuring that high-priority orders are produced first, reducing the production cycle, improving delivery timeliness and customer satisfaction.

[0036] Match the configuration data with the expected to-be-produced order data to obtain Q groups of order combination data; the order combination data includes the order subcategory data required to be produced by each of the production lines; sort the Q groups of order combination data from high to low according to the order priority to obtain Q groups of to-be-produced order data, as shown in Table 4, which is one group of to-be-produced order data (corresponding to a first-level pressure level production line).

[0037] Table 4 Example table of pending production order data

[0038] In this embodiment, Q is 3, and there are 3 production lines that can be used for the production of a certain type of elastic stockings. Figure 2 The entire production system is divided into three independent main production lines according to the pressure level (level 1, level 2 and level 3). Each production line is responsible for the full process production of one pressure level, as shown in Table 5, including the following complete processes: Knitting process: Knitting socks for specific pressure levels to ensure that the pressure parameters meet the specification requirements.

[0039] Forming and shaping process: Forming and shaping according to size (S, M and L) to ensure that the socks meet the requirements.

[0040] Sewing process: Sewing is performed according to the style (open toe and closed toe) to meet the needs of different styles.

[0041] Pressure testing process: The socks are pressure tested according to the pressure levels (level one, level two and level three) to ensure that the pressure values ​​meet the standards.

[0042] Cleaning and disinfection procedures: Clean and disinfect by color (black and skin color) to avoid staining problems.

[0043] Quality inspection and packaging process: uniform quality inspection and packaging to ensure that the finished products meet factory requirements.

[0044] Table 5 Configuration data table of each process of production line

[0045] See also Figure 1 In S50, a plurality of candidate production strategies are generated according to the M groups of to-be-produced order data and the production plan constraint conditions; Furthermore, generating a plurality of candidate production strategies according to the M groups of to-be-produced order data and the production plan constraint conditions includes: According to the production line flow chart, Figure 2 As shown, each process in the production process and the quality inspection process is divided into independent processes and shared processes to obtain process classification data; In the production process of this embodiment, except for the quality inspection and packaging processes, all other processes are independent processes. These independent processes are set according to the size, style, pressure level or color of the product to ensure the consistency of each specification of products during the production process and meet the quality standards. In this embodiment, independent processes and shared processes are clearly distinguished. The constraints of independent processes and shared processes in production are different. Among them, the scheduling and production sequence of independent processes are only limited to the single production line to which they belong, and their production time and queuing time are mainly affected by the production capacity of the production line. The production time and sequence of shared processes are not only limited by the production capacity of a specific production line, but also need to consider the resource usage of other production lines. Therefore, in the optimization process, the resource allocation and coordination between different production lines must be comprehensively considered.

[0046] According to the process classification data and in combination with the production plan constraints, an intelligent optimization algorithm is used to generate a plurality of candidate production strategies, each of which includes the production time and production sequence of each independent process and the shared process.

[0047] By dividing each process in the production process and quality inspection process into independent processes and shared processes, combined with production plan constraints, and using intelligent optimization algorithms to generate multiple candidate production strategies, efficient scheduling of production processes and reasonable allocation of resources are achieved. The classification of independent processes and shared processes enables candidate production strategies to be personalized according to the characteristics of different processes, avoiding waste of resources and excessive switching of production lines, and improving production flexibility and efficiency. At the same time, the intelligent optimization algorithm comprehensively considers production constraints and generates multiple solutions to meet different production needs, providing flexible options for production management.

[0048] Furthermore, using the intelligent optimization algorithm to generate a plurality of candidate production strategies includes: Using a machine learning algorithm to establish a queue time prediction model based on historical production line data, the queue time prediction model is used to predict the expected queue time of each process on each production line; See also Figure 3 S501 in which a production strategy population is initialized, wherein each production strategy in the production strategy population includes a production time and a production sequence for each process on each production line; See also Figure 3 In S502, a fitness function of the intelligent optimization algorithm is defined according to the total queue time and the capacity utilization rate; The iterative process of the intelligent optimization algorithm includes the following steps: See also Figure 3S503, calculating the fitness value of each individual according to the production time, production sequence in each of the production strategies and the queue time calculated by the queue time prediction model; The fitness value calculation formula is as follows: ; in, is the fitness value, is the first weight, is the second weight, is the capacity utilization rate, is the total queue time, is the maximum possible queue time; The calculation formula for capacity utilization is: ; in, It represents the sum of actual production time of all production lines; It represents the total available production time of all production lines during the planning period; See also Figure 3 In S504, according to the fitness value, select an individual whose fitness value is greater than a preset threshold as a parent; See also Figure 3 In S505, based on the parent generation, a new child generation production strategy is generated through a crossover operation; See also Figure 3 In S506, a mutation operation is introduced into the child production strategy to randomly adjust the production sequence (order production sequence) or production time of the process, or the production sequence and production time can be adjusted at the same time to obtain the mutated child production strategy; in the optimization process, by randomly adjusting the production sequence of the process, the order of different orders can be adjusted. For example, suppose the original production sequence is that order A performs the weaving process first, and order B performs the weaving process later. After random adjustment, the order may be modified to order B performs weaving first, and order A performs weaving later. Through this order adjustment, the intelligent optimization algorithm can generate a variety of solutions.

[0049] See also Figure 3 S507 in which, in each iteration, the fitness value of each generated mutated offspring production strategy is calculated according to the fitness function; see Figure 3 In S508, the production strategy population is updated according to the fitness value; See also Figure 3 In S509, it is determined whether a stop condition is met; when the stop condition is not met, the iterative process is repeated; in this embodiment, the stop condition is that the maximum number of iterations reaches 1000 times.

[0050] See also Figure 3 In S510, a plurality of candidate production strategies are generated when the stop condition is met; each candidate production strategy is used to determine the production time and production sequence of each process of each production line for the corresponding order data to be produced.

[0051] See also Figure 1 S60 in which a digital twin model for the production of medical auxiliary socks is established; the digital twin model for the production of medical auxiliary socks is used to simulate the complete production process of the medical auxiliary socks, including the production process, the quality inspection process and the inventory management process; In this embodiment, the number of candidate production strategies generated is a preset value, such as 20 strategies with high fitness. The specific number can be adjusted according to the computing power of the digital twin model, simulation time requirements and user needs.

[0052] Through multiple rounds of iteration and screening, combined with optimization operations such as crossover and mutation, the algorithm finally obtains multiple high-quality candidate production strategies to be simulated. These candidate production strategies provide different optimization combinations under the premise of meeting the capacity and queuing constraints, providing rich and high-quality strategy selection for the subsequent simulation and verification of the digital twin model. This candidate production strategy generation method provides the digital twin model with limited and high-quality candidate production strategies for simulation, thereby avoiding excessive computational complexity and decision-making difficulties in the infinite solution space. By using intelligent optimization algorithms to generate a variety of optimized production strategies, the best solutions are selected from the huge possible solution space. These strategies have been evaluated by the fitness function, taking into account key indicators such as capacity utilization and total queuing time, ensuring that each solution is the best choice under the existing constraints. The digital twin model only needs to simulate these limited and high-quality solutions, which not only improves the computational efficiency, but also enhances the controllability of decision-making, effectively reduces decision-making risks, and ensures the stability and reliability of the production plan. This not only optimizes resource allocation and production processes, but also improves the efficiency and competitiveness of the entire production system.

[0053] Furthermore, the process of establishing the digital twin model of the production of medical assistive socks is specifically as follows: creating a virtual physical model based on the physical characteristics of the production equipment and production process of the medical assistive socks, and the virtual physical model is used to simulate the equipment status and process parameters in the production process in real time; constructing a logical model based on the business logic, production rules and quality control standards of the medical assistive socks in production, and the logical model is used to analyze and judge the key indicators in the production process, detect abnormal situations, and adjust the process parameters according to the production rules; constructing a simulation model based on the operation status of the virtual physical model, and the simulation model obtains evaluation indicators of different production strategies by mapping the equipment status and the process parameters in the virtual physical model; collecting real-time data through Internet of Things devices and sensor groups, including inventory data, production process data, quality inspection process data and external environment data; preprocessing the real-time data to obtain standard real-time data; constructing a data model based on the standard real-time data; the data model is used to provide data input for the virtual physical model, the logical model and the simulation model.

[0054] By creating a virtual physical model based on the physical characteristics of production equipment and processes, the operating status and process parameters of production equipment can be truly reproduced, and equipment performance and production progress can be monitored in real time. Secondly, by building a logical model based on business logic, production rules and quality control standards, key indicators in the production process can be analyzed and judged, abnormal situations can be discovered and warned in a timely manner, and process parameters can be automatically adjusted according to preset production rules to ensure the stability of the production process and product quality. In addition, by mapping the equipment status and process parameters in the virtual physical model and building a simulation model, the effects of different candidate production strategies can be evaluated, objective evaluation indicators can be provided, and data support can be provided for optimizing production strategies. Finally, the real-time data collected by IoT devices and sensors is preprocessed and standardized to build a data model to provide accurate and real-time data input for the virtual physical model, logical model and simulation model. This multi-level model construction method combines physical characteristics, logical rules and real-time data, enabling the digital twin model to comprehensively and dynamically simulate and optimize the production process, enhance the predictability of the production process and the transparency of decision-making, reduce production risks, and optimize resource allocation.

[0055] Furthermore, the production process data includes: production equipment status, process parameters, production speed of each process, queue status of each process and production result data. This helps to ensure that the digital twin model can accurately and comprehensively reflect the actual production situation, thereby achieving efficient monitoring and optimized management, and can obtain important information about the production process in a more targeted manner.

[0056] See also Figure 1In S70, the complete production process of the candidate production strategy is simulated by the medical auxiliary socks production digital twin model to obtain the virtual delivery time of each order to be produced; and the optimal production plan is obtained according to the virtual delivery time.

[0057] Furthermore, obtaining the optimal production plan based on the virtual delivery time includes: comparing the virtual delivery time with the expected delivery time in the delivery demand information, identifying overdue orders and inventory orders (orders produced in advance to be stored); calculating the overdue duration of the overdue orders and the inventory duration of the inventory orders (the duration required to be stored in inventory) based on the virtual delivery time and the expected delivery time, and generating loss costs based on the overdue orders, the overdue duration, the inventory orders and the inventory duration; each candidate production strategy corresponds to a loss cost; and selecting the candidate production strategy with the smallest loss cost as the optimal production plan.

[0058] By comparing the virtual delivery time with the expected delivery time, identifying overdue orders and inventory orders, and calculating their overdue duration and inventory duration, the potential loss costs of each candidate production strategy can be comprehensively evaluated. The loss cost is generated according to the number of overdue orders, the overdue duration, the number of inventory orders, and the inventory duration, so that the selection of candidate production strategies is not only based on production capacity and efficiency, but also takes into account the impact on timely order delivery and inventory management. Finally, by selecting the candidate production strategy with the smallest loss cost as the optimal production plan, the financial losses caused by overdue delivery and inventory backlog are effectively reduced, the execution efficiency of the production plan and the reliability of delivery are improved, the rational use of resources is ensured, and customer satisfaction is improved.

[0059] In order to verify the actual effect of the scheme proposed by the present invention, a set of comparative experiments was conducted. The experimental group adopted a method for generating a production plan for medical auxiliary socks based on artificial intelligence proposed by the present invention. In the control group 1, the traditional fixed rule scheduling method was used to formulate the production plan. The production was mainly scheduled according to the order of arrival of the order, without predicting the order demand, and without considering the operation status of the machine and the priority of the order. According to the principle of first come first served, all orders were arranged for production in the order of receipt, and the production tasks remained unchanged during the production process, and no dynamic adjustment was made according to the actual production status and resource utilization. In the control group 2, a production plan generation method based on traditional machine learning prediction was adopted, and the future order demand and machine operation status were predicted by the machine learning model to assist in the formulation of production scheduling. First, the historical order data was used to predict the future order demand through the time series model, so as to arrange the production resources in advance; then, the historical operation data of the machine was used to use a variety of machine learning algorithms to predict the failure probability of each machine, and the algorithm with the best prediction effect was selected to ensure that the equipment with high failure risk undertook non-critical tasks in the scheduling; and the classification or regression model was used to prioritize the orders. Finally, based on the prediction results of machine learning (demand, failure probability, and priority), a greedy algorithm is used to generate a schedule.

[0060] Table 6 Comparative experimental data table

[0061] Referring to Table 6, from the experimental results, the scheme of the present invention (experimental group) shows significant advantages in all indicators, proving the actual effect of the production plan generation method based on artificial intelligence in optimizing production efficiency, reducing costs, and improving delivery on-time rate. In particular, it is significantly better than the two control groups in terms of on-time delivery rate, average delay time of delayed orders, inventory cost, and timely rate of high-priority orders.

[0062] The present invention provides a method for generating a production plan for medical assistive socks based on artificial intelligence. By combining order forecasting, production constraints, a digital twin model and an artificial intelligence algorithm, an intelligent, efficient and flexible production plan generation system is constructed.

[0063] First, by obtaining unscheduled orders and orders predicted based on historical data, we can fully grasp the future production demand. Dividing the expected order data into M groups of order sub-category data is helpful to classify and manage orders with different product specifications and delivery requirements, and improve the accuracy of the production plan. Obtaining production plan constraints in various aspects such as equipment, process and inventory, combining inventory data and order sub-category data, accurately obtain the number of orders to be produced, avoid overproduction or inventory backlog, and improve inventory management efficiency. According to the data of pending orders and production plan constraints, artificial intelligence algorithms are used to generate a variety of feasible candidate production strategies to provide more options for decision-making; at the same time, a digital twin model for the production of medical auxiliary socks is established to fully simulate the production process, including production, quality inspection and inventory management, and obtain the virtual delivery time of each pending order by simulating the execution process of each candidate production strategy. Through the simulation of the digital twin model, managers can clearly see the execution effects and potential risks of different candidate production strategies, and select the production strategy that best meets the delivery requirements and has reasonable resource allocation according to the virtual delivery time to ensure the stability and reliability of the production plan.

[0064] This method can detect and solve potential problems before actual production, reduce production delays and quality problems, reasonably allocate production resources, and avoid resource waste and cost increase caused by unreasonable production scheduling. By coping with the diversity and complexity of orders, meeting the needs of precise production, comprehensively considering production constraints, using digital twin models to improve decision-making quality, and combining artificial intelligence to optimize production strategies, it can ultimately reduce production risks and costs and enhance the predictability and transparency of decisions.

[0065] Embodiment 2: Factory B is a manufacturer specializing in the production of medical assistive socks. With the increase in market demand, especially during promotional activities, Factory B needs to process a large number of orders in a short period of time. Orders from different customers vary significantly in specifications, quantity, and delivery time. In order to improve production efficiency, arrange inventory reasonably, and ensure on-time delivery, Factory B decided to introduce an artificial intelligence-based medical assistive socks production plan generation system, such as Figure 4 As shown, including: An order data acquisition module is configured to acquire expected orders for a specified medical auxiliary socks, wherein the expected orders include unscheduled orders and predicted orders; acquire expected order data according to the expected orders, wherein the expected order data includes order demand information, product specification information, and delivery demand information; determine order priority according to the order demand information and the delivery demand information; and the predicted order is predicted based on the historical order data of the medical auxiliary socks; An order classification module, dividing the expected order data into M groups of order sub-category data; A production plan constraint management module is used to obtain production plan constraint conditions of the medical auxiliary socks, including delivery time constraint, production line minimum start-up batch constraint, equipment production capacity constraint, process sequence dependency constraint and inventory constraint; A module for generating orders to be produced obtains M groups of data of orders to be produced according to the inventory data and the order sub-category data of the M groups; A production strategy generation module generates a plurality of candidate production strategies according to the M groups of to-be-produced order data and the production plan constraint conditions; A digital twin model establishment module is used to establish a digital twin model for the production of medical auxiliary socks; the digital twin model for the production of medical auxiliary socks is used to simulate the complete production process of the medical auxiliary socks, including the production process, quality inspection process and inventory management process; A production strategy simulation module, which simulates the complete production process of the candidate production strategy through the medical auxiliary socks production digital twin model to obtain the virtual delivery time of each order to be produced; The production plan acquisition module obtains the optimal production plan according to the virtual delivery time.

[0066] Further, the expected order data is divided into M groups. The order sub-category data includes: The product specification information includes product size, product style, product color and product pressure level; Grouping orders with the same product size, product style, product color, and product pressure level into one group, to obtain M groups of order subcategory data; The calculation formula of M is: ; in, is the number of categories of the product size, is the number of categories of the product style, is the number of product color categories, It is the category number of the pressure level of the product.

[0067] In this embodiment, the product sizes are divided into four types: S, M, L and XL, the product style is only one, the product color is black, and the product pressure levels are divided into three types: level one, level two and level three.

[0068] but, ; The expected order data is divided into 12 groups of order subcategory data, and the order quantity of each group is aggregated and integrated to obtain the total demand of each subcategory.

[0069] Furthermore, obtaining the production plan constraint conditions includes: The delivery time constraint is determined according to the expected delivery time of the delivery demand information; The minimum startup batch constraint of the production line is obtained based on historical production line data; The equipment production capacity constraint is obtained based on historical equipment production capacity data; The process sequence dependency constraint is obtained according to the production line flow chart; The inventory constraint is obtained based on historical inventory data.

[0070] Further, obtaining the M groups of to-be-produced order data according to the inventory data and the M groups of order subcategory data includes: The inventory data includes inventory product specification information; the inventory product specification information includes inventory product size, inventory product style, inventory product color and inventory product pressure level; orders with the same inventory product size, inventory product style, inventory product color and inventory product pressure level are divided into one group to obtain M groups of inventory sub-category data; According to the M groups of inventory sub-category data and the M groups of order sub-category data, the inventory quantity in each group of the inventory sub-category data is compared with the demand quantity in the corresponding order sub-category data; if the demand quantity is greater than the inventory quantity, the difference between the demand quantity and the inventory quantity is used as the order to be produced, and a column of order matching data is added to the inventory data to record the order number corresponding to the inventory, so as to form inventory matching order data; the order matching data is generated according to the matching result of the product specification information and the inventory product specification information, and combined with the order priority; if the demand quantity is not greater than the inventory quantity, the order data to be produced is empty, and the order quantity to be produced is 0; Eliminate the order data corresponding to the order matching data from the corresponding order subcategory data to obtain expected production order data; Obtain configuration data for Q production lines, including applicable product specification data and process capability data; Match the configuration data with the expected to-be-produced order data to obtain Q groups of order combination data; the order combination data includes the order subcategory data required to be produced by each of the production lines; sort the Q groups of order combination data from high to low according to the order priority to obtain Q groups of to-be-produced order data.

[0071] In this embodiment, Q is 3, and there are 3 production lines that can be used for the production of a certain type of elastic stockings. Figure 2 The entire production system is divided into 4 independent main production lines according to product sizes (S, M, L and XL). Each production line is responsible for the full production of one size, including the following complete processes: Knitting process: Knitting of socks of specific sizes (S, M, L and XL).

[0072] Forming and shaping process: Forming and shaping according to size (S, M, L and XL) to ensure that the socks meet the requirements.

[0073] Suturing process: Suturing is performed.

[0074] Pressure testing process: The socks are pressure tested according to the pressure levels (level one, level two and level three) to ensure that the pressure values ​​meet the standards.

[0075] Cleaning and disinfection process: Clean and disinfect.

[0076] Quality inspection and packaging process: uniform quality inspection and packaging to ensure that the finished products meet factory requirements.

[0077] Furthermore, generating a plurality of candidate production strategies according to the M groups of to-be-produced order data and the production plan constraint conditions includes: According to the production line flow chart, each process in the production process and quality inspection process is divided into independent processes and shared processes to obtain process classification data; According to the process classification data and in combination with the production plan constraints, an intelligent optimization algorithm is used to generate a plurality of candidate production strategies, each of which includes the production time and production sequence of each independent process and the shared process.

[0078] Furthermore, using the intelligent optimization algorithm to generate a plurality of candidate production strategies includes: Using a machine learning algorithm to establish a queue time prediction model based on historical production line data, the queue time prediction model is used to predict the expected queue time of each process on each production line; Initializing a production strategy population, wherein each production strategy in the production strategy population includes a production time and a production sequence for each process on each production line; Defining the fitness function of the intelligent optimization algorithm according to the total queue time and the capacity utilization rate; The iterative process of the intelligent optimization algorithm includes the following steps: Calculate the fitness value of each individual according to the production time, production sequence in each of the production strategies and the queue time calculated by the queue time prediction model; According to the fitness value, selecting an individual whose fitness value is greater than a preset threshold as a parent; Based on the parent generation, a new child generation production strategy is generated through a crossover operation; Introducing a mutation operation into the child production strategy, randomly adjusting the production sequence or production time of the process, or adjusting the production sequence and production time at the same time, to obtain a mutated child production strategy; In each iteration, the fitness value of each generated mutated offspring production strategy is calculated according to the fitness function, and the production strategy population is updated according to the fitness value; whether a stop condition is met is determined; when the stop condition is not met, the iterative process is repeated; When the stop condition is met, a plurality of candidate production strategies are generated; each candidate production strategy is used to determine the production time and production sequence of each process of each production line for the corresponding order data to be produced.

[0079] Furthermore, the process of establishing the digital twin model of medical auxiliary socks production is specifically as follows: Creating a virtual physical model based on the physical characteristics of the production equipment and production process of the medical auxiliary socks, wherein the virtual physical model is used to simulate the equipment status and process parameters in the production process in real time; A logic model is constructed according to the business logic, production rules and quality control standards of the medical auxiliary socks in production, and the logic model is used to analyze and judge key indicators in the production process, detect abnormal conditions, and adjust the process parameters according to the production rules; Building a simulation model according to the operation status of the virtual physical model, wherein the simulation model obtains evaluation indicators of different production strategies by mapping the equipment status and the process parameters in the virtual physical model; Real-time data is collected through Internet of Things devices and sensor groups, including inventory data, production process data, quality inspection process data and external environment data; the real-time data is preprocessed to obtain standard real-time data; a data model is constructed based on the standard real-time data; the data model is used to provide data input for the virtual physical model, the logical model and the simulation model.

[0080] Furthermore, the production process data includes: production equipment status, process parameters, production speed of each process, queue status of each process and production result data.

[0081] Further, obtaining the optimal production plan according to the virtual delivery time includes: Compare the virtual delivery time with the expected delivery time in the delivery demand information to identify overdue orders and inventory orders; calculate the overdue duration of the overdue orders and the inventory duration of the inventory orders based on the virtual delivery time and the expected delivery time, and generate loss costs based on the overdue orders, the overdue duration, the inventory orders and the inventory duration; select the candidate production strategy with the smallest loss cost as the optimal production plan.

[0082] 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. A method for generating a production plan for medical auxiliary socks based on artificial intelligence, characterized in that: include: Obtaining expected orders for designated medical auxiliary socks, wherein the expected orders include unscheduled orders and predicted orders; obtaining expected order data according to the expected orders, wherein the expected order data includes order demand information, product specification information, and delivery demand information; determining order priority according to the order demand information and the delivery demand information; the predicted orders are predicted based on historical order data of the medical auxiliary socks; Dividing the expected order data into M groups of order sub-category data; Obtaining production plan constraints for the medical auxiliary socks, including delivery time constraints, production line minimum start-up batch constraints, equipment production capacity constraints, process sequence dependency constraints, and inventory constraints; According to the inventory data and the order sub-category data of group M, M groups of to-be-produced order data are obtained; Generate multiple candidate production strategies according to the M groups of to-be-produced order data and the production plan constraints; Establish a digital twin model for the production of medical auxiliary socks; The medical auxiliary socks production digital twin model is used to simulate the complete production process of the medical auxiliary socks, including the production process, quality inspection process and inventory management process; Simulating the complete production process of the candidate production strategy through the medical auxiliary socks production digital twin model to obtain the virtual delivery time of each order to be produced; An optimal production plan is obtained according to the virtual delivery time.

2. The method for generating a production plan for medical auxiliary socks based on artificial intelligence according to claim 1, characterized in that: The expected order data is divided into M groups of order subcategory data including: the product specification information includes product size, product style, product color and product pressure level; the orders with the same product size, product style, product color and product pressure level are divided into a group to obtain M groups of order subcategory data.

3. The method for generating a production plan for medical auxiliary socks based on artificial intelligence according to claim 1, characterized in that: The acquisition of the production plan constraint conditions includes: the delivery time constraint is determined according to the expected delivery time of the delivery demand information; the minimum startup batch constraint of the production line is acquired according to the historical production line data; the equipment production capacity constraint is acquired according to the historical equipment production capacity data; the process sequence dependency constraint is acquired according to the production line flow chart; the inventory constraint is acquired according to the historical inventory data.

4. The method for generating a production plan for medical auxiliary socks based on artificial intelligence according to claim 1, characterized in that: According to the inventory data and the M groups of order sub-category data, the M groups of to-be-produced order data are obtained, including: the inventory data includes inventory product specification information; the inventory product specification information includes inventory product size, inventory product style, inventory product color and inventory product pressure level; orders with the same inventory product size, inventory product style, inventory product color and inventory product pressure level are divided into one group to obtain M groups of inventory sub-category data; According to the M groups of inventory sub-category data and the M groups of order sub-category data, the inventory quantity in each group of the inventory sub-category data is compared with the demand quantity in the corresponding order sub-category data; if the demand quantity is greater than the inventory quantity, the difference between the demand quantity and the inventory quantity is used as the order to be produced, and a column of order matching data is added to the inventory data to record the order number corresponding to the inventory, so as to form inventory matching order data; the order matching data is generated according to the matching result of the product specification information and the inventory product specification information, and combined with the order priority; if the demand quantity is not greater than the inventory quantity, the order data to be produced is empty, and the order quantity to be produced is 0; Eliminate the order data corresponding to the order matching data from the corresponding order sub-category data to obtain the expected production order data; obtain the configuration data of Q production lines, including applicable product specification data and process capability data; Match the configuration data with the expected to-be-produced order data to obtain Q groups of order combination data; the order combination data includes the order subcategory data required to be produced by each of the production lines; sort the Q groups of order combination data from high to low according to the order priority to obtain Q groups of to-be-produced order data.

5. The method for generating a production plan for medical auxiliary socks based on artificial intelligence according to claim 1, characterized in that: Generating a plurality of candidate production strategies based on the M groups of to-be-produced order data and the production plan constraints includes: dividing each process in the production process and the quality inspection process into independent processes and shared processes according to the production line flow chart to obtain process classification data; generating a plurality of candidate production strategies based on the process classification data and in combination with the production plan constraints using an intelligent optimization algorithm, each of the candidate production strategies including the production time and production sequence of each independent process and the shared process.

6. The method for generating a production plan for medical auxiliary socks based on artificial intelligence according to claim 5, characterized in that: Using the intelligent optimization algorithm to generate a plurality of candidate production strategies includes: Initializing a production strategy population, wherein each production strategy in the production strategy population includes the production time and production sequence of each process on each production line; defining the fitness function of the intelligent optimization algorithm according to the total queue time and the capacity utilization rate; The iterative process of the intelligent optimization algorithm includes the following steps: According to the production time, production sequence and queue time calculated by the queue time prediction model in each of the production strategies, the fitness value of each individual is calculated; according to the fitness value, the individual with the fitness value greater than the preset threshold is selected as the parent; based on the parent, a new child production strategy is generated through a crossover operation; a mutation operation is introduced into the child production strategy to randomly adjust the production sequence and / or production time of the process to obtain the mutated child production strategy; in each iteration, the fitness value of each generated mutated child production strategy is calculated according to the fitness function, and the production strategy population is updated according to the fitness value; it is determined whether the stop condition is met; when the stop condition is not met, the iterative process is repeated; when the stop condition is met, a plurality of candidate production strategies are generated; each candidate production strategy is used to determine the production time and production sequence of each process of each production line for the corresponding order data to be produced.

7. The method for generating a production plan for medical auxiliary socks based on artificial intelligence according to claim 1, characterized in that: The process of establishing the digital twin model of medical assistive socks production is specifically as follows: creating a virtual physical model for real-time simulation of equipment status and process parameters in the production process; building a logical model for analyzing and judging key indicators in the production process and detecting abnormal conditions; building a simulation model, which obtains evaluation indicators of different production strategies by mapping the equipment status and process parameters in the virtual physical model; and building a data model; the data model is used to provide data input for the virtual physical model, the logical model, and the simulation model.

8. The method for generating a production plan for medical auxiliary socks based on artificial intelligence according to claim 7, characterized in that: The production process data include: process parameters, production speed of each process, queue status of each process and production result data.

9. The method for generating a production plan for medical auxiliary socks based on artificial intelligence according to claim 1, characterized in that: Obtaining the optimal production plan based on the virtual delivery time includes: comparing the virtual delivery time with the expected delivery time in the delivery demand information to identify overdue orders and inventory orders; calculating the overdue duration of the overdue orders and the inventory duration of the inventory orders based on the virtual delivery time and the expected delivery time, and generating loss costs based on the overdue orders, the overdue duration, the inventory orders and the inventory duration; and selecting the candidate production strategy with the smallest loss cost as the optimal production plan.

10. A medical auxiliary socks production plan generation system based on artificial intelligence, characterized in that: include: An order data acquisition module is configured to acquire expected orders for a specified medical auxiliary socks, wherein the expected orders include unscheduled orders and predicted orders; acquire expected order data according to the expected orders, wherein the expected order data includes order demand information, product specification information, and delivery demand information; determine order priority according to the order demand information and the delivery demand information; and the predicted order is predicted based on the historical order data of the medical auxiliary socks; An order classification module, dividing the expected order data into M groups of order sub-category data; A production plan constraint management module is used to obtain production plan constraint conditions of the medical auxiliary socks, including delivery time constraint, production line minimum start-up batch constraint, equipment production capacity constraint, process sequence dependency constraint and inventory constraint; A module for generating orders to be produced obtains M groups of data of orders to be produced according to the inventory data and the order sub-category data of the M groups; A production strategy generation module generates a plurality of candidate production strategies according to the M groups of to-be-produced order data and the production plan constraint conditions; Digital twin model building module, to build a digital twin model for the production of medical auxiliary socks; The medical auxiliary socks production digital twin model is used to simulate the complete production process of the medical auxiliary socks, including the production process, quality inspection process and inventory management process; A production strategy simulation module, which simulates the complete production process of the candidate production strategy through the medical auxiliary socks production digital twin model to obtain the virtual delivery time of each order to be produced; The production plan acquisition module obtains the optimal production plan according to the virtual delivery time.

Images