Foaming shoe production management method and system based on big data

Abstract

This invention discloses a big data-based method and system for managing the production of foam shoes, belonging to the field of production management technology. Specifically, it includes: establishing an order contamination feature database by collecting historical production data; obtaining the contamination propagation patterns of different colored raw materials during the production process and calculating the unit contamination coefficient of each colored raw material; constructing an equipment contamination state evolution model, establishing a mapping relationship table between equipment contamination state and pass rate, and predicting the pass rate distribution of each order by simulating changes in equipment contamination state under different production schedules; and establishing a multi-objective optimization function to comprehensively score candidate production schedules based on order delivery time requirements, selecting the optimal production schedule to drive the production system execution. This invention quantifies the relationship between color contamination and product quality, effectively ensuring stable product quality while significantly improving equipment utilization and production efficiency by optimizing production scheduling and reducing unnecessary cleaning.

Description

Technical Field

[0001] This invention relates to the field of production management technology, specifically to a method and system for managing the production of foam shoes based on big data. Background Technology

[0002] In the manufacturing process of foam shoes, different orders often require the use of EVA raw materials of different colors and hardness. Due to the varying sensitivities to contamination caused by color differences, light-colored orders demand extremely high levels of equipment cleanliness, while dark-colored orders are prone to leaving pigment residue in the equipment after production. This production characteristic necessitates thorough cleaning of the production equipment when switching orders, resulting in significant changeover time and production costs.

[0003] Currently, the industry commonly uses production scheduling methods based on delivery deadlines. This method only considers the time attribute of orders and ignores the substantial impact of color switching on the production system. For example, when producing in the sequence of "white → light gray → dark gray → black," although delivery requirements are met, equipment contamination will continue to accumulate as production progresses. This cumulative contamination effect will result in the following: after completing the light gray order, the equipment contamination index has exceeded the contamination tolerance threshold for the dark gray order. If production continues at this point, the product color will be impure, resulting in quality defects; if production is interrupted for cleaning, production efficiency will be lost.

[0004] Existing production management systems lack the ability to dynamically track equipment contamination status and have not established a quantitative relationship between order color attributes and contamination propagation. This makes it impossible to predict the impact of color switching on product quality when making production scheduling decisions, relying solely on manual experience to determine cleaning times. This approach fails to guarantee product quality stability and makes it difficult to optimize production efficiency. Therefore, there is an urgent need for a production management solution that can quantify contamination propagation patterns and dynamically optimize production scheduling sequences to address these technical challenges. Summary of the Invention

[0005] The purpose of this invention is to provide a method and system for managing the production of foam shoes based on big data, and to solve the following technical problems:

[0006] The existing production management system lacks the ability to dynamically track equipment contamination status and has not established a quantitative relationship between order color attributes and contamination propagation. This makes it impossible to predict the impact of color switching on product quality when making production scheduling decisions, and can only rely on manual experience to arrange cleaning times, which cannot guarantee product quality stability and is difficult to optimize production efficiency.

[0007] The objective of this invention can be achieved through the following technical solutions:

[0008] A big data-based method for managing the production of foam shoes includes the following steps:

[0009] S1. Based on the raw material composition records, production volume records and quality inspection results extracted from historical production data of foam shoe production orders, establish an order contamination feature database;

[0010] S2, Perform pollution load quantitative analysis on the order pollution feature library, statistically analyze the correlation between the production quantity of color raw materials of previous orders and the product qualification rate of subsequent orders in the historical production sequence, and calculate the unit pollution coefficient of each color raw material when it is used as the previous production.

[0011] S3. Based on the unit pollution coefficient and the production quantity of each preceding order in the historical production sequence, calculate the equipment pollution status value after the completion of each preceding order, and establish a mapping relationship table between the equipment pollution status value and the pass rate of each color raw material when it is used for subsequent production.

[0012] S4, obtain the raw material color attribute and production quantity of each order in the order set to be scheduled, and obtain the predicted pass rate distribution of different candidate order scheduling sequences based on the unit pollution coefficient and the pass rate mapping table.

[0013] S5. Based on the predicted pass rate distribution and order delivery time, establish a multi-objective optimization function and give a comprehensive score to each candidate order scheduling sequence; select the candidate order scheduling sequence with the smallest comprehensive score as the final scheduling scheme, output the production execution instruction of the scheme and drive the production system to execute the order production operation.

[0014] As a further aspect of the present invention: the specific process of performing pollution load quantification analysis on the order pollution feature database in step S2 is as follows:

[0015] Order sequence pairs with a direct production sequence relationship are selected from historical production data. Each order sequence pair includes the color attribute of the preceding order and its corresponding production quantity data, as well as the color attribute of the subsequent order and its corresponding quality inspection pass rate data. A linear regression model is established with the production quantity of the preceding order as the independent variable and the pass rate of the subsequent order as the dependent variable. The regression coefficients of the linear regression model are solved by the least squares method, and the regression coefficients are defined as the unit pollution coefficients corresponding to each color raw material.

[0016] As a further aspect of the present invention: in step S3, the specific calculation process for the equipment contamination value is as follows:

[0017] Determine the initial contamination status value based on the initial cleaning records of the production equipment, and process each production order sequentially according to the time order of the historical production sequence.

[0018] For each order in the sequence, the production quantity of the order is multiplied by the unit pollution coefficient of the corresponding color raw material to obtain the pollution increment of the order; the calculated pollution increment is added to the equipment pollution status value at the current moment to obtain the latest equipment pollution status value after the order is completed.

[0019] When a device cleaning event occurs in the production sequence record, the device contamination status value is reset to the initial contamination status value after the cleaning event is completed. The above steps of contamination increment calculation and status value update are repeated until all orders in the historical production sequence are processed, and finally a complete historical device contamination status value sequence is output.

[0020] As a further aspect of the present invention: the specific construction process of the pass rate mapping table in S3 is as follows:

[0021] Collect the pollution status values ​​of each piece of equipment recorded during historical production processes and the corresponding production quality data of subsequent orders; group the equipment pollution status values ​​into numerical intervals, and calculate the actual pass rate of each color raw material in each interval as the actual pass rate in subsequent production; calculate the arithmetic mean of the pass rate data in each equipment pollution status value interval to obtain the standard pass rate corresponding to that interval; establish a mapping relationship between the equipment pollution status value intervals and the corresponding standard pass rates to form a pass rate mapping relationship table with the equipment pollution status value intervals as the index and the standard pass rates of each color raw material as the values.

[0022] As a further aspect of the present invention: the specific process of solving the predicted pass rate distribution in S4 is as follows:

[0023] Obtain the raw material color attribute and production quantity of each order in the set of orders to be scheduled for production. Use an exhaustive method to perform all permutations and combinations on the set of orders to be scheduled for production to generate all possible candidate order scheduling sequences.

[0024] For each candidate order production sequence, initialize the simulated contamination status value of the production equipment, and process each order in the order of the sequence; multiply the production quantity of the current order by the corresponding unit contamination coefficient to obtain the contamination increment, and add the contamination increment to the current simulated contamination status value of the equipment to obtain the updated simulated contamination status value of the equipment; query the pass rate mapping table based on the updated simulated contamination status value of the equipment to obtain the predicted pass rate of the current order;

[0025] Repeat the above steps of updating the simulated pollution status value of the equipment and querying the predicted pass rate until all orders in the candidate order scheduling sequence have been processed, and obtain the predicted pass rate distribution of the candidate order scheduling sequence; perform the same processing procedure on all candidate order scheduling sequences to obtain the predicted pass rate distribution of all candidate order scheduling sequences.

[0026] As a further aspect of the present invention: the specific process of performing comprehensive scoring in S5 is as follows:

[0027] Iterate through each candidate order scheduling sequence. When the predicted pass rate of any order to be scheduled in the sequence is lower than the preset pass rate threshold, insert a machine cleaning mark before the order to be scheduled.

[0028] The total number of equipment cleaning marks in each candidate sequence is counted, and a cleaning cost scoring item is established with the objective function of minimizing the total number of cleaning marks; the difference between the planned completion time and delivery time of each pending production order is calculated, and a delivery timeliness scoring item is established with the objective function of minimizing the total delivery delay time of all pending production orders.

[0029] The cleaning cost score item and the delivery timeliness score item are multiplied by their respective weighting coefficients and then summed to obtain the comprehensive score value of each candidate sequence.

[0030] As a further aspect of the present invention: the specific process for obtaining the planned completion time of each order to be scheduled for production is as follows:

[0031] Set the initial time of the production plan as the current system time, process each order sequentially according to the order order order sequence, and calculate the theoretical production time of the order based on the ratio of the production quantity of the order to the equipment's unit time capacity;

[0032] Add the current system time to the theoretical production time to obtain the preliminary completion time of the order; if the order is marked as requiring equipment cleaning in the production sequence, add the standard time required for equipment cleaning to the preliminary completion time to obtain the final planned completion time of the order; set the final planned completion time as the starting system time of the next order; repeat the above steps until the planned completion time calculation of all orders in the production sequence is completed.

[0033] As a further aspect of the present invention: in S2, the order product qualification rate is the ratio of the number of qualified products determined based on the color inspection results to the total production quantity.

[0034] A big data-based foam shoe production management system, used to implement the aforementioned big data-based foam shoe production management method, includes:

[0035] The data acquisition module is used to extract raw material composition records, production volume records, and quality inspection results from historical production data based on foam shoe production orders, and to establish an order contamination feature database.

[0036] The data analysis module is used to perform pollution load quantification analysis on the order pollution feature library, statistically analyze the correlation between the production quantity of color raw materials of previous orders and the product qualification rate of subsequent orders in the historical production sequence, and calculate the unit pollution coefficient of each color raw material when it is used as the previous production.

[0037] The data optimization module is used to calculate the equipment pollution status value after the completion of each preceding order based on the unit pollution coefficient and the production quantity of each preceding order in the historical production sequence, and to establish a mapping relationship table between the equipment pollution status value and the pass rate of each color raw material when it is used for subsequent production.

[0038] The qualified prediction module is used to obtain the raw material color attributes and production quantity of each order in the set of orders to be scheduled for production, and to obtain the predicted qualified rate distribution of different candidate order scheduling sequences based on the unit pollution coefficient and the qualified rate mapping relationship table.

[0039] The result generation module is used to establish a multi-objective optimization function based on the predicted pass rate distribution and order delivery time, and to give a comprehensive score to each candidate order scheduling sequence; select the candidate order scheduling sequence with the smallest comprehensive score as the final scheduling scheme, output the production execution instruction of the scheme and drive the production system to execute the order production operation.

[0040] The beneficial effects of this invention are:

[0041] 1) This invention, through the establishment of an order contamination feature database and quantitative analysis of contamination load, establishes for the first time a quantitative relationship between color raw materials and contamination propagation in the field of foam shoe production. By calculating the unit contamination coefficient and equipment contamination status value, the dynamic changes in equipment contamination status can be accurately tracked. Based on the establishment of a pass rate mapping table, the production quality risks of different color sequences can be accurately predicted before production scheduling. This data-driven contamination propagation modeling method fundamentally changes the traditional production mode that relies on manual experience to judge quality risks, and realizes the proactive prevention and control of quality problems caused by color contamination. By intelligently inserting cleaning marks into the production sequence, it ensures that each order is produced under equipment conditions that meet its contamination tolerance, effectively eliminating batch quality defects caused by color contamination and significantly improving product quality stability.

[0042] 2) This invention integrates cleaning costs and delivery timeliness into a unified decision-making framework by establishing a multi-objective optimization function. When generating candidate production sequences, it simultaneously considers the predicted pass rate distribution and order delivery time requirements. Through traversal analysis, it automatically identifies necessary cleaning nodes and accurately calculates the cleaning cost score and delivery timeliness score for each sequence. By balancing cleaning frequency with delivery urgency, it minimizes unnecessary cleaning operations while ensuring on-time order delivery, all while maintaining quality standards. This effectively solves the dilemma of conflicting quality requirements and production efficiency in traditional production scheduling, achieving a dual improvement in quality assurance and production efficiency.

[0043] 3) This invention constructs a closed-loop optimization system covering the entire process from data acquisition and model building to production execution. By continuously collecting production quantity and quality inspection results from the actual production process, the system dynamically updates the mapping table between unit contamination coefficient and pass rate, enabling the contamination state evolution model to possess self-learning and self-optimization capabilities. This closed-loop optimization mechanism allows the system to continuously adapt to changes in the production environment, constantly optimizing the accuracy of production scheduling decisions as data accumulates. Simultaneously, by integrating equipment contamination state value calculation and pass rate prediction into the production scheduling decision-making process, intelligent matching of production resources and quality requirements is achieved, providing continuous optimization decision support for production management and effectively improving the overall intelligence level of the production system. Attached Figure Description

[0044] The invention will now be further described with reference to the accompanying drawings.

[0045] Figure 1 This is a schematic diagram of a foam shoe production management method based on big data according to the present invention.

[0046] Figure 2 This is a schematic diagram of the structure of a foam shoe production management system based on big data according to the present invention. Detailed Implementation

[0047] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0048] Please see Figure 1 As shown, this invention is a big data-based method for managing the production of foam shoes, comprising the following steps:

[0049] S1. Based on the raw material composition records, production volume records and quality inspection results extracted from historical production data of foam shoe production orders, establish an order contamination feature database;

[0050] In establishing the order contamination feature database, the first step is to obtain raw historical data from the production management system. This data includes the specific color raw material names used in each order, the actual production quantity, and the inspection results recorded by the quality inspection department. Since different production batches may use different names for the same color (e.g., some records say "snow white" while others say "milky white"), these different descriptions need to be standardized and categorized into a standard color system to ensure that raw materials of the same nature are correctly classified. Next, production quantity data needs to be organized, summarizing the scattered production records of the same order at different time periods, while excluding production interruptions caused by machine malfunctions or other abnormal reasons to ensure data continuity and completeness. Regarding quality data, the inspection reports of each batch of products need to be carefully reviewed, paying particular attention to quality issues caused by cross-contamination of colors. For example, if dark pigment residue from a previous order causes color spots or uneven color in subsequent light-colored products, these specific issues are linked to the corresponding production orders. Through this gradual data collection and organization, a feature database is ultimately formed that includes complete order information, standardized color classification, accurate production quantities, and descriptions of specific quality issues.

[0051] S2, Perform pollution load quantitative analysis on the order pollution feature library, statistically analyze the correlation between the production quantity of color raw materials of previous orders and the product qualification rate of subsequent orders in the historical production sequence, and calculate the unit pollution coefficient of each color raw material when it is used as the previous production.

[0052] First, order combinations with continuous production relationships are screened from the order contamination feature database. Specifically, records where subsequent orders begin production immediately after the completion of a preceding order without equipment cleaning in between are identified. For example, consider this data set: a preceding order produced 5,000 pairs of shoe soles using black raw materials, followed by a subsequent order producing 3,000 pairs using white raw materials. Inspection records show that the white products had a grayish color quality issue. Analyzing a large number of such order combinations reveals that the larger the quantity of the preceding order, the lower the pass rate of the subsequent order tends to be. For instance, when producing 3,000 pairs using black raw materials, the pass rate of the subsequent white orders remains relatively high, but when the production quantity increases to 8,000 pairs, the pass rate of the subsequent white orders drops significantly. Based on this pattern, a mathematical relationship is established between the production quantity of the preceding order and the pass rate of the subsequent order. This relationship quantifies the contamination impact of producing one additional pair of shoe soles on subsequent orders. Calculating the unit contamination coefficient determines how much contamination each unit of production using each color raw material as a preceding material causes to the equipment and ultimately affects the pass rate of subsequent orders.

[0053] S3. Based on the unit pollution coefficient and the production quantity of each preceding order in the historical production sequence, calculate the equipment pollution status value after the completion of each preceding order, and establish a mapping relationship table between the equipment pollution status value and the pass rate of each color raw material when it is used for subsequent production.

[0054] First, the initial state of the equipment needs to be determined, typically by setting the contamination level of equipment that has just undergone thorough cleaning to zero. Then, following the chronological order of historical production orders, the change in equipment contamination level after each order is calculated. For example, when the initial state is zero, an order using black raw materials is produced first, with a quantity of 4,000 pairs. The unit contamination coefficient of black raw materials is known; multiplying the production quantity by the unit contamination coefficient yields the contamination increment caused by this order. This increment is added to the current contamination level of the equipment to obtain the new contamination level after completing the order. The process continues with the next order, such as an order using white raw materials, calculating its contamination increment and adding it to the equipment contamination level, and so on, processing all orders in the historical sequence. If a cleaning record is encountered during the calculation, the equipment contamination level is reset to zero after cleaning. After obtaining a large amount of actual data on equipment contamination level values ​​and corresponding order pass rates, the equipment contamination level values ​​are divided into several continuous intervals. The pass rates for each color raw material within each interval are statistically analyzed as the actual pass rates for subsequent production. By analyzing this data, a correspondence between equipment contamination level intervals and pass rates is established, ultimately forming a complete mapping table.

[0055] By establishing a mapping relationship between equipment contamination status values ​​and pass rates, the accumulation of contamination within equipment can be presented intuitively in numerical form. This contamination status is directly linked to the final product quality, transforming the invisible equipment contamination status into quantifiable management indicators. This allows the production system to accurately predict the quality performance of subsequent production orders based on the current equipment contamination status. A bridge is built between equipment status and product quality, providing a crucial basis for developing scientific production scheduling plans. The system can rationally arrange the production sequence based on the actual equipment status, ensuring both product quality and improved equipment utilization efficiency.

[0056] S4, obtain the raw material color attribute and production quantity of each order in the order set to be scheduled, and obtain the predicted pass rate distribution of different candidate order scheduling sequences based on the unit pollution coefficient and the pass rate mapping table.

[0057] First, detailed information on all pending production orders needs to be obtained, including the planned raw material color and production quantity for each order. For example, there are three pending production orders: Order A to produce 4,000 pairs using black raw material, Order B to produce 3,000 pairs using white raw material, and Order C to produce 2,000 pairs using dark gray raw material. The system will generate all possible production sequence arrangements, such as producing A first, then B, and finally C, or B first, then C, and finally A, etc. For each candidate production sequence, the system simulates the changes in equipment contamination status during actual production: starting from the initial contamination status of the equipment, each order is processed sequentially. The production quantity of each order is multiplied by the unit contamination coefficient corresponding to its raw material to obtain the contamination increment caused by that order, which is then added to the current equipment contamination status value to obtain a new contamination status value. After obtaining the equipment contamination status value for each order during production, the predicted pass rate for that order is determined by querying the pass rate mapping table, based on the current contamination status value and the raw material color of the order. For example, in a certain sequence, when producing a white order, the equipment contamination status value reaches a high level; querying the mapping table reveals that the predicted pass rate for white raw material will be lower under this status value. In this way, the system can generate a complete predicted pass rate distribution for each candidate sequence, clearly showing the quality level that each order may achieve under that sequence.

[0058] Understandably, systematic simulation and prediction can effectively avoid quality problems that frequently occur in traditional production scheduling, such as scheduling light-colored orders when equipment is heavily contaminated, leading to product defects. This predictive mechanism provides a crucial basis for subsequent optimization decisions, enabling production managers to select the production sequence with the best quality performance from multiple feasible options. This ensures stable and reliable product quality while minimizing rework and waste caused by quality issues, thereby comprehensively improving production efficiency and management level.

[0059] S5. Based on the predicted pass rate distribution and order delivery time, establish a multi-objective optimization function and give a comprehensive score to each candidate order scheduling sequence; select the candidate order scheduling sequence with the smallest comprehensive score as the final scheduling scheme, output the production execution instruction of the scheme and drive the production system to execute the order production operation.

[0060] First, each candidate production sequence is evaluated based on the predicted pass rate distribution. For example, a candidate sequence might contain three orders: first, dark gray; then white; and finally black. The system simulation reveals that after completing the dark gray order, the equipment contamination level rises significantly. If production continues on the contamination-sensitive white order, the predicted pass rate will be significantly lower than the preset quality threshold. In this case, the system automatically marks the white order with a cleaning requirement, because the equipment contamination level exceeds the limits that the white raw materials can tolerate. Cleaning is necessary to reset the contamination level to an appropriate level to ensure the white order's product quality meets standards. Next, each candidate sequence is comprehensively evaluated, and the total number of cleaning cycles required is calculated. The number of cleaning cycles directly affects production costs and production speed. Simultaneously, the planned completion time for each order is estimated. Based on the specific production quantity and actual equipment capacity, the precise time schedule for each process is calculated and compared with the agreed delivery time to identify potential delay risks. The number of cleaning cycles and delivery delays are converted into a unified scoring standard, and the two scores are combined into a comprehensive score using a preset importance ratio. For example, one sequence might require an additional cleaning step to ensure on-time delivery, while another sequence might not require cleaning but carries a risk of delay. A comprehensive evaluation is used to balance these two key factors. The production scheduling plan with the optimal comprehensive score is then selected and translated into detailed production instructions, including the order of each order, necessary cleaning procedures, and precise time planning. These instructions are then directly transmitted to the execution system on the production floor.

[0061] By intelligently determining equipment cleaning needs, both product quality stability and delivery timeliness are ensured. It's understandable that when equipment contamination accumulates beyond the tolerance limits of specific raw materials, it directly impacts core product quality indicators; timely cleaning eliminates this potential quality hazard. This mechanism effectively overcomes two common biases in traditional production scheduling: first, excessive cleaning to avoid quality risks leads to resource waste; second, neglecting cleaning in pursuit of efficiency results in quality defects. The systematic evaluation method allows production planning to find the optimal balance between quality control and delivery management, preventing quality risks and optimizing resource allocation. This provides technical support for the scientific and refined management of production, enabling companies to make comprehensive decisions in the actual production environment, thereby achieving continuous improvement in production efficiency.

[0062] In a preferred embodiment of the present invention, the specific process of performing pollution load quantification analysis on the order pollution feature database in step S2 is as follows:

[0063] Order sequence pairs with a direct production sequence relationship are selected from historical production data. Each order sequence pair includes the color attribute of the preceding order and its corresponding production quantity data, as well as the color attribute of the subsequent order and its corresponding quality inspection pass rate data. A linear regression model is established with the production quantity of the preceding order as the independent variable and the pass rate of the subsequent order as the dependent variable. The regression coefficients of the linear regression model are solved by the least squares method, and the regression coefficients are defined as the unit pollution coefficients corresponding to each color raw material.

[0064] In another preferred embodiment of the present invention, the specific calculation process of the equipment pollution value in step S3 is as follows:

[0065] Determine the initial contamination status value based on the initial cleaning records of the production equipment, and process each production order sequentially according to the time order of the historical production sequence.

[0066] For each order in the sequence, the production quantity of the order is multiplied by the unit pollution coefficient of the corresponding color raw material to obtain the pollution increment of the order; the calculated pollution increment is added to the equipment pollution status value at the current moment to obtain the latest equipment pollution status value after the order is completed.

[0067] When a device cleaning event occurs in the production sequence record, the device contamination status value is reset to the initial contamination status value after the cleaning event is completed. The above steps of contamination increment calculation and status value update are repeated until all orders in the historical production sequence are processed, and finally a complete historical device contamination status value sequence is output.

[0068] First, it's necessary to determine the initial state of the production equipment. After thorough cleaning, the equipment's contamination level is set as the initial baseline value, representing a completely clean state. For example, when new equipment is put into use or after deep cleaning, the system records the equipment's contamination level as zero, serving as the starting point for subsequent calculations.

[0069] Next, each production record is processed sequentially according to the historical production orders. Taking a specific production sequence as an example, assuming the initial contamination state of the equipment is zero, order A is processed first. This order produced four thousand pairs of shoe soles using black raw materials. Based on the pre-determined unit contamination coefficient of the black raw materials, the production quantity is multiplied by the unit contamination coefficient to obtain the contamination increment caused by this order. For example, a higher unit contamination coefficient of the black raw materials will result in a larger contamination increment when the production quantity is larger. Then, this contamination increment is added to the current contamination state value of the equipment to obtain the new equipment contamination state value after the order is completed.

[0070] When processing the next order, the same calculation process is repeated based on the equipment contamination status value after the previous order was completed. For example, when processing order B, which produced 3,000 pairs of shoe soles using white raw materials, although the unit contamination coefficient of white raw materials is low, the contamination increment of this order will be further added to the equipment contamination status value because the equipment already has a certain contamination status value.

[0071] During the calculation process, if the production records show that equipment cleaning was performed, the system will reset the equipment contamination status value to the initial baseline value after cleaning is completed. For example, if a thorough cleaning was performed after completing a dark-colored order, the calculation of subsequent orders will start from zero and accumulate the contamination status value again. The above calculation steps are repeated until all order records in the historical production sequence have been processed, ultimately obtaining a complete historical equipment contamination status value sequence. This sequence accurately reflects the dynamic changes in equipment contamination status during the production process.

[0072] By tracking the accumulation of contamination in equipment during the production process, the previously invisible state of equipment contamination is transformed into a quantifiable numerical sequence. This allows the production system to understand the impact of each production stage on equipment status, providing a data foundation for establishing the correlation between equipment status and product quality. Accurate calculation of equipment contamination status values ​​is a crucial aspect of the entire quality management system. It provides a reliable basis for subsequent yield prediction and production scheduling optimization, thereby helping the production system achieve rational allocation and continuous improvement in the efficiency of production resource utilization while ensuring product quality.

[0073] In another preferred embodiment of the present invention, the specific construction process of the pass rate mapping table in step S3 is as follows:

[0074] Collect the pollution status values ​​of each piece of equipment recorded during historical production processes and the corresponding production quality data of subsequent orders; group the equipment pollution status values ​​into numerical intervals, and calculate the actual pass rate of each color raw material in each interval as the actual pass rate in subsequent production; calculate the arithmetic mean of the pass rate data in each equipment pollution status value interval to obtain the standard pass rate corresponding to that interval; establish a mapping relationship between the equipment pollution status value intervals and the corresponding standard pass rates to form a pass rate mapping relationship table with the equipment pollution status value intervals as the index and the standard pass rates of each color raw material as the values.

[0075] First, the system needs to collect the contamination status values ​​of each piece of equipment recorded during historical production processes, along with the corresponding production quality data for subsequent orders. For example, this can be achieved by retrieving all continuous production records from the past year, including the equipment contamination status values ​​at the time of each order's completion, as well as the actual quality inspection results of immediately following orders. This data comes from production logs and quality inspection reports, ensuring the authenticity and completeness of the data.

[0076] Next, the collected equipment contamination status values ​​are divided into several consecutive intervals based on their numerical values. For example, contamination status values ​​of 0 to 10 are divided into the first interval, 10 to 20 into the second interval, and so on. For the data within each interval, the actual pass rate of each color raw material is calculated as follows: For example, within the contamination status value interval of 5 to 10, the pass rate data of white raw materials in subsequent production is calculated, revealing 20 production batches with pass rates ranging from 95% to 98%.

[0077] Then, the arithmetic mean of the pass rate data within each equipment contamination state value range is calculated. For example, within the contamination state value range of 5 to 10, the average of the 20 pass rate data for white raw materials is calculated to be 96.5%, and this value is used as the standard pass rate for that range. Using the arithmetic mean calculation method can effectively balance the impact of individual abnormal data and obtain a more representative pass rate level.

[0078] Finally, a mapping relationship is established between the equipment contamination status value ranges and the corresponding standard pass rates. For example, the standard pass rate for white raw materials is determined to be 99% for the contamination status value range of 0 to 10, 96% for the range of 10 to 20, and 92% for the range of 20 to 30. This forms a complete pass rate mapping table. This mapping table uses the equipment contamination status value range as an index and the standard pass rate of each color raw material as a specific numerical value, forming a complete query reference system.

[0079] In another preferred embodiment of the present invention, the specific process of solving the predicted pass rate distribution in step S4 is as follows:

[0080] Obtain the raw material color attribute and production quantity of each order in the set of orders to be scheduled for production. Use an exhaustive method to perform all permutations and combinations on the set of orders to be scheduled for production to generate all possible candidate order scheduling sequences.

[0081] For each candidate order production sequence, initialize the simulated contamination status value of the production equipment, and process each order in the order of the sequence; multiply the production quantity of the current order by the corresponding unit contamination coefficient to obtain the contamination increment, and add the contamination increment to the current simulated contamination status value of the equipment to obtain the updated simulated contamination status value of the equipment; query the pass rate mapping table based on the updated simulated contamination status value of the equipment to obtain the predicted pass rate of the current order;

[0082] Repeat the above steps of updating the simulated pollution status value of the equipment and querying the predicted pass rate until all orders in the candidate order scheduling sequence have been processed, and obtain the predicted pass rate distribution of the candidate order scheduling sequence; perform the same processing procedure on all candidate order scheduling sequences to obtain the predicted pass rate distribution of all candidate order scheduling sequences.

[0083] First, it's necessary to obtain basic information about the orders to be scheduled, including the color characteristics of the raw materials used in each order and the planned production quantity. For example, there are three orders to be scheduled: Order A uses black raw materials and needs to produce 4,000 pairs, Order B uses white raw materials and needs to produce 3,000 pairs, and Order C uses dark gray raw materials and needs to produce 2,000 pairs. The system uses an exhaustive approach to perform all permutations and combinations of these orders, generating all possible production sequence schemes, such as ABC, ACB, BAC, etc., six different production sequence sequences. This exhaustive approach ensures that no possible production sequence combination is missed.

[0084] For each candidate production sequence, the contamination status of the production equipment is first initialized to a baseline value. When processing each order sequentially, the production quantity of the current order is multiplied by the unit contamination coefficient corresponding to its raw material to calculate the contamination increment generated during the production process of that order. For example, black raw materials have a higher unit contamination coefficient, and a larger production quantity will result in a significant contamination increment. This contamination increment is then added to the current simulated contamination status value of the equipment to obtain the updated equipment contamination status value. This update process reflects the actual situation of contamination accumulation during production.

[0085] Based on the updated equipment contamination status value, the system queries a pre-established pass rate mapping table. This table records historical pass rate data for each color raw material under different contamination statuses, and the predicted pass rate for the order under the current status can be obtained by querying it. The steps of updating the contamination status and querying the pass rate are repeated until all orders in the sequence have been processed, ultimately yielding a complete predicted pass rate distribution. The same processing flow is executed for all candidate production scheduling sequences to obtain the predicted pass rate distribution data corresponding to all possible production scheduling schemes.

[0086] By assessing the potential quality outcomes of different production scheduling plans in advance, and replacing traditional manual judgment with systematic simulation calculations, the system comprehensively considers the impact of each order's production sequence on equipment contamination status, as well as the cascading effects of contamination status on the quality of subsequent orders, making quality predictions more accurate and reliable. This predictive mechanism provides crucial data support for subsequent production scheduling decisions, enabling production planners to anticipate the quality performance of different plans before actual production begins. This allows them to select the optimal production scheduling plan, effectively avoiding quality problems caused by improper scheduling and improving the scientific rigor and predictability of production management.

[0087] In another preferred embodiment of the present invention, the specific process of performing comprehensive scoring in step S5 is as follows:

[0088] Iterate through each candidate order scheduling sequence. When the predicted pass rate of any order to be scheduled in the sequence is lower than the preset pass rate threshold, insert a machine cleaning mark before the order to be scheduled.

[0089] The total number of equipment cleaning marks in each candidate sequence is counted, and a cleaning cost scoring item is established with the objective function of minimizing the total number of cleaning marks; the difference between the planned completion time and delivery time of each pending production order is calculated, and a delivery timeliness scoring item is established with the objective function of minimizing the total delivery delay time of all pending production orders.

[0090] The cleaning cost score item and the delivery timeliness score item are multiplied by their respective weighting coefficients and then summed to obtain the comprehensive score value of each candidate sequence.

[0091] First, each candidate production sequence is checked one by one. When an order's predicted pass rate is found to be lower than the preset pass rate standard, a machine cleaning mark is added before that order. For example, in a production sequence containing black and white orders, if the system detects that the equipment is highly contaminated after completing the black order, the predicted pass rate of the subsequent white order will not meet the quality requirements, and a mark indicating that the white order needs machine cleaning will be added before the white order. By counting the total number of cleaning marks in each sequence, the required cleaning frequency for different production plans can be evaluated. Fewer cleaning times mean lower production costs and higher production efficiency. At the same time, the system calculates the difference between the planned completion time and the required delivery time for each order, and sums up the delay times of all orders to obtain the total delay time. The shorter the delay time, the more timely the order delivery. Then, the system weights the cleaning times and delay times according to preset importance to obtain a comprehensive score for each production sequence. For example, one sequence may require two cleanings but ensure on-time delivery of all orders, while another sequence may only require one cleaning but cause a two-day delay for one order. The system uses a comprehensive score to weigh these two situations and select the plan with the best overall performance.

[0092] Understandably, this approach considers both production quality and delivery efficiency, two key dimensions. Setting cleaning markers ensures product quality at crucial stages, preventing quality issues caused by equipment contamination. Controlling the number of cleaning cycles reduces production costs and improves equipment utilization. Considering delivery delays maintains corporate reputation and customer satisfaction. This multi-dimensional evaluation system overcomes the limitations of traditional scheduling methods that often focus on a single objective. It prevents resource waste caused by excessive cleaning in pursuit of quality, and avoids the risks of neglecting quality to meet deadlines. By balancing the relationships between different objectives, the final selected scheduling plan achieves an optimal balance between quality, cost, and delivery, providing a comprehensive and reliable basis for production decisions and ultimately improving overall production efficiency.

[0093] In another preferred embodiment of the present invention, the specific process for obtaining the planned completion time of each order to be scheduled for production is as follows:

[0094] Set the initial time of the production plan as the current system time, process each order sequentially according to the order order order sequence, and calculate the theoretical production time of the order based on the ratio of the production quantity of the order to the equipment's unit time capacity;

[0095] Add the current system time to the theoretical production time to obtain the preliminary completion time of the order; if the order is marked as requiring equipment cleaning in the production sequence, add the standard time required for equipment cleaning to the preliminary completion time to obtain the final planned completion time of the order; set the final planned completion time as the starting system time of the next order; repeat the above steps until the planned completion time calculation of all orders in the production sequence is completed.

[0096] In another preferred embodiment of the present invention, in step S2, the order product qualification rate is the ratio of the number of qualified products determined based on the color inspection results to the total production quantity.

[0097] After each batch of orders is completed, quality inspectors conduct a comprehensive color quality inspection, comparing each product to a standard color chart to identify defective products with color discrepancies. For example, if a white order is produced immediately after a black order, inspectors might find gray or black spots on some white products. These products with color contamination are recorded as defective. Subsequently, the total production quantity of this batch is tallied, and the number of defective products due to color issues is removed. The remaining number of products without color defects is divided by the total production quantity to obtain the color pass rate for that order. For instance, if a white order produces 5,000 pairs of shoes, and 100 pairs are found to be defective due to color contamination, then the color pass rate for that order is the ratio of 4,900 acceptable pairs to the total production quantity of 5,000 pairs.

[0098] By differentiating the color compliance rate from compliance rates caused by other factors, the accuracy of subsequent contamination coefficient calculations can be ensured, avoiding the misattribution of other quality issues to color contamination. This targeted compliance rate assessment method provides a reliable data foundation for establishing an accurate contamination propagation model, making the correlation between equipment contamination status and product quality clearer and more explicit. This provides a more precise basis for subsequent intelligent production scheduling decisions, ensuring that production plans can effectively control color contamination risks while rationally allocating production resources.

[0099] See Figure 2 The present invention also includes a big data-based foam shoe production management system for implementing the above-described big data-based foam shoe production management method, comprising:

[0100] The data acquisition module is used to extract raw material composition records, production volume records, and quality inspection results from historical production data based on foam shoe production orders, and to establish an order contamination feature database.

[0101] The data analysis module is used to perform pollution load quantification analysis on the order pollution feature library, statistically analyze the correlation between the production quantity of color raw materials of previous orders and the product qualification rate of subsequent orders in the historical production sequence, and calculate the unit pollution coefficient of each color raw material when it is used as the previous production.

[0102] The data optimization module is used to calculate the equipment pollution status value after the completion of each preceding order based on the unit pollution coefficient and the production quantity of each preceding order in the historical production sequence, and to establish a mapping relationship table between the equipment pollution status value and the pass rate of each color raw material when it is used for subsequent production.

[0103] The qualified prediction module is used to obtain the raw material color attributes and production quantity of each order in the set of orders to be scheduled for production, and to obtain the predicted qualified rate distribution of different candidate order scheduling sequences based on the unit pollution coefficient and the qualified rate mapping relationship table.

[0104] The result generation module is used to establish a multi-objective optimization function based on the predicted pass rate distribution and order delivery time, and to give a comprehensive score to each candidate order scheduling sequence; select the candidate order scheduling sequence with the smallest comprehensive score as the final scheduling scheme, output the production execution instruction of the scheme and drive the production system to execute the order production operation.

[0105] The foregoing has provided a detailed description of one embodiment of the present invention, but this description is merely a preferred embodiment and should not be construed as limiting the scope of the invention. All equivalent variations and modifications made within the scope of the claims of this invention should still fall within the patent coverage of this invention.

Claims

1. A big data-based foaming shoe production management method, characterized by, The method comprises the following steps: S1, based on the production order of foaming shoes, extracting the raw material composition record, production quantity record and quality inspection result in the historical production data, establishing an order pollution characteristic library; S2, performing pollution load quantitative analysis on the order pollution characteristic library, and statistically analyzing the correlation between the color raw material production quantity of the previous order and the product pass rate of the subsequent order in the historical production sequence, to calculate the unit pollution coefficient of each color raw material as the production of the previous order; S3, based on the unit pollution coefficient and the production quantity of each previous order in the historical production sequence, calculating the equipment pollution state value after completing the production of each previous order, and establishing a mapping relationship table between the equipment pollution state value and the pass rate of each color raw material as the subsequent production; S4, obtaining the raw material color attribute and production quantity of each order in the to-be-scheduled order set, and based on the unit pollution coefficient and the pass rate mapping relationship table, obtaining the predicted pass rate distribution of different candidate order scheduling sequences; S5, based on the predicted pass rate distribution and the order delivery time, a multi-objective optimization function is established, when the predicted pass rate of any to-be-scheduled order in the sequence is lower than the preset pass rate threshold, a device cleaning mark is inserted before the to-be-scheduled order; the total number of device cleaning marks in each candidate sequence is counted, and a cleaning cost scoring item is established with the minimum cleaning mark total number as the objective function; the difference between the planned completion time and the delivery time of each to-be-scheduled order is calculated, and a delivery time limit scoring item is established with the minimum total sum of delivery delay time of all to-be-scheduled orders as the objective function; the cleaning cost scoring item and the delivery time limit scoring item are multiplied by the corresponding weight coefficients and then added, to obtain the comprehensive score value of each candidate sequence; the candidate order scheduling sequence with the minimum comprehensive score value is selected as the final scheduling scheme, and the production execution instruction of the scheme is output and the production system is driven to execute order production work; In S2, the specific process of performing pollution load quantitative analysis on the order pollution characteristic library is as follows: The order sequence pairs with direct production sequence relationship are screened from the historical production data, wherein each order sequence pair contains the color attribute of the previous order and the corresponding production quantity data, and the color attribute of the subsequent order and the corresponding quality inspection pass rate data; a linear regression model is established with the production quantity of the previous order as the independent variable and the pass rate of the subsequent order as the dependent variable; the regression coefficient of the linear regression model is solved by the least square method, and the regression coefficient is defined as the unit pollution coefficient corresponding to each color raw material; In S4, the predicted pass rate is obtained by the following method: For each candidate order scheduling sequence, initialize the simulated pollution state value of the production equipment, and process each order in sequence according to the order sequence in the sequence; multiply the production quantity of the current order by the corresponding unit pollution coefficient to obtain the pollution increment, and add the pollution increment to the current equipment simulated pollution state value to obtain the updated equipment simulated pollution state value; query the pass rate mapping relationship table according to the updated equipment simulated pollution state value to obtain the predicted pass rate of the current order. In S3, the specific calculation process of the equipment pollution value is as follows:

2. The big data-based foamed shoe production management method according to claim 1, characterized in that, ​ An initial pollution state value is determined according to an initial cleaning record of the production equipment, and each production order in a historical production sequence is processed in a time sequence; For each order in the sequence, the production quantity of the order is multiplied by the unit pollution coefficient of the corresponding color raw material to obtain the pollution increment of the order; and the pollution increment calculated is added to the equipment pollution state value at the current time to obtain the latest equipment pollution state value after the order is completed; When a cleaning event of the equipment occurs in the production sequence record, the equipment pollution state value is reset to the initial pollution state value after the cleaning event is completed; the pollution increment calculation and state value updating steps are repeatedly executed until the processing of all orders in the historical production sequence is completed, and a complete historical equipment pollution state value sequence is finally output.

3. The big data-based foamed shoe production management method according to claim 1, characterized in that, In the S3, the specific construction process of the qualified rate mapping relationship table is: Collecting the equipment pollution state values recorded in the historical production process and the corresponding subsequent order production quality data; grouping the equipment pollution state values according to numerical intervals, and counting the actual qualified rates of each color raw material as subsequent production in each interval; calculating the arithmetic mean of the qualified rate data in each equipment pollution state value interval to obtain the standard qualified rate corresponding to the interval; establishing a mapping relationship between the equipment pollution state value interval and the corresponding standard qualified rate to form a qualified rate mapping relationship table with the equipment pollution state value interval as the index and the standard qualified rate of each color raw material as the value.

4. The big data-based foamed shoe production management method according to claim 1, characterized in that, In the S4, obtaining the predicted qualified rate distribution of different candidate order scheduling sequences includes: Repeating the steps of updating the equipment simulation pollution state value and querying the predicted qualified rate until all orders in the candidate order scheduling sequence are processed to obtain the predicted qualified rate distribution of the candidate order scheduling sequence; performing the same processing procedure on all candidate order scheduling sequences to obtain the predicted qualified rate distribution of all candidate order scheduling sequences.

5. The big data-based foamed shoe production management method according to claim 1, characterized in that, The specific acquisition process of the planned completion time of each order to be scheduled is: Setting the production plan initial time as the current system time, and processing each order in the order sequence in turn; calculating the theoretical production time of the order according to the ratio of the production quantity of the order to the unit time capacity of the equipment; Adding the current system time and the theoretical production time to obtain the preliminary completion time of the order; If the order is marked as requiring equipment cleaning in the order scheduling sequence, the standard time required for equipment cleaning is added to the preliminary completion time to obtain the final planned completion time of the order; The final planned completion time is set as the starting system time of the next order; Repeating the above steps until the planned completion time calculation of all orders in the order scheduling sequence is completed.

6. The big data-based foamed shoe production management method according to claim 1, characterized by, In the S2, the order product qualified rate is the ratio of the number of qualified products determined based on the color inspection result to the total production quantity.

7. A big data based foamed shoe production management system for implementing the big data based foamed shoe production management method of any one of claims 1-6, characterized in that, It includes: A data acquisition module for extracting raw material composition records, production quantity records and quality inspection results from historical production data based on foaming shoe production orders, and establishing an order pollution feature library; A data analysis module is configured to perform pollution load quantification analysis on the order pollution feature library, count the correlation between the color raw material production quantity of a previous order and the product pass rate of a subsequent order in a historical production sequence, and calculate the unit pollution coefficient of each color raw material when used for production of the previous order; A data optimization module is configured to calculate the equipment pollution state value after production of each previous order based on the unit pollution coefficient and the production quantity of each previous order in the historical production sequence, and establish a mapping relationship table between the equipment pollution state value and the pass rate of each color raw material when used for subsequent production; A pass rate prediction module is configured to obtain the raw material color attribute and production quantity of each order in a set of to-be-scheduled orders, and obtain the predicted pass rate distribution of different candidate order scheduling sequences based on the unit pollution coefficient and the mapping relationship table of the pass rate; A result generation module is configured to establish a multi-objective optimization function based on the predicted pass rate distribution and the order delivery time, and comprehensively score each candidate order scheduling sequence; select the candidate order scheduling sequence with the minimum comprehensive score value as the final scheduling scheme, output the production execution instruction of the scheme, and drive the production system to perform order production work.

Images