Production task real-time scheduling method and system based on production line process parameter analysis

Abstract

The application relates to the technical field of intelligent manufacturing, and discloses a production task real-time scheduling method and system based on production line process parameter analysis. The method acquires first and second process parameters of a first production line and a second production line through a preset first acquisition module, and acquires third and fourth process parameters of corresponding production lines through a preset second acquisition module. First and second process difference values are respectively calculated based on first and second process libraries, a comprehensive difference value is obtained by fusion, and a task scheduling instruction is automatically sent when the comprehensive difference value is greater than a preset reference difference value. After responding to the instruction, an idle third production line is preferentially searched, if the idle third production line exists, the tasks of the first and second production lines are combined and sent to the production line, and if the idle third production line does not exist, the error correlation of the two production lines is compared, and the task of the production line with strong error correlation is combined to another production line for execution. The application improves the operation efficiency and production stability of the intelligent manufacturing production line.

Description

Technical Field

[0001] This application relates to the technical field of intelligent manufacturing, and in particular to a method and system for real-time scheduling of production tasks based on production line process parameter analysis. Background Technology

[0002] With the deepening of Industry 4.0 and intelligent manufacturing, the manufacturing industry is transforming towards flexibility, digitalization, and intelligence, and multi-variety, small-batch, and highly dynamic production models are becoming the mainstream. Production lines are placing higher demands on process stability, quality consistency, and task response speed, striving to improve the adaptability and overall operational efficiency of production lines.

[0003] Current intelligent manufacturing systems widely integrate the functions of acquiring, analyzing, and issuing early warnings for production process parameters. They can monitor key process parameters such as temperature, pressure, speed, and flow rate in real time, and promptly identify quality fluctuations and equipment anomalies through threshold judgment and trend analysis, enabling early warnings of quality problems. Simultaneously, the central control room is equipped with a production management and scheduling interface, allowing operators to manually modify and issue production tasks based on production plans, order requirements, and on-site conditions, adjusting production line operating cycles, process priorities, and processing parameters to ensure orderly production execution.

[0004] However, existing intelligent manufacturing systems lack an unmanned linkage mechanism for process parameter analysis and production task scheduling. When the system issues a process anomaly warning, manual review of the warning information and assessment of the anomaly's impact are still required, followed by manual adjustment of production tasks and rearrangement of scheduling plans. This results in problems such as delayed response, reliance on experience for decision-making, low scheduling efficiency, and susceptibility to human error. Summary of the Invention

[0005] To achieve an unmanned linkage mechanism between process parameter analysis and production task scheduling, this application provides a real-time production task scheduling method and system based on production line process parameter analysis.

[0006] Firstly, this application provides a real-time production task scheduling method based on production line process parameter analysis, employing the following technical solution:

[0007] A real-time production task scheduling method based on production line process parameter analysis includes the following steps:

[0008] The first process parameters on the first production line are obtained based on the preset first acquisition module, and the second process parameters on the second production line are obtained based on the preset second acquisition module; the third process parameters on the first production line are obtained based on the preset second acquisition module, and the fourth process parameters on the second production line are obtained.

[0009] The first process difference value is calculated by comparing the first process parameter with the third process parameter based on a preset first process library; the second process difference value is calculated by comparing the second process parameter with the fourth process parameter based on a preset second process library; the comprehensive difference value is calculated by combining the first process difference value and the second process difference value; if the comprehensive difference value is greater than the preset reference difference value, a task scheduling instruction is issued.

[0010] In response to the task scheduling instruction, find an idle third production line; if an idle third production line is found, merge the tasks on the first production line and the tasks on the second production line to obtain a merged task, and send the merged task to the third production line; otherwise, compare the first production line with the second production line.

[0011] If the comparison result shows that the first production line has a strong error correlation, then the tasks on the first production line are merged into the tasks on the second production line; otherwise, the tasks on the second production line are merged into the tasks on the first production line.

[0012] By adopting the above technical solution, the first and third process parameters of the first production line and the second and fourth process parameters of the second production line are acquired by the first and second acquisition modules, respectively. Combined with the preset first and second process libraries, the first and second process difference values ​​are calculated and merged to obtain a comprehensive difference value. When the comprehensive difference value is greater than the preset reference difference value, a task scheduling command is automatically issued, establishing an unmanned linkage mechanism between process parameter analysis and production task scheduling. After responding to the task scheduling command, an idle third production line is automatically searched for to achieve dynamic adaptation and allocation of production tasks. If an idle third production line exists... The tasks of the first and second production lines are merged into a single merged task and issued. If no merged task exists, the task of one production line is merged into the other by comparing the error correlation between the first and second production lines, thus achieving precise matching between production tasks and production line status. This improves the real-time performance and automation level of production line scheduling, maintains dynamic coordination between production line process parameters and production tasks, and facilitates the rational allocation of production line resources. At the same time, through the collection of multiple process parameters and the fusion calculation of difference values, task scheduling and abnormal process states are highly adapted, improving the overall operational efficiency of the production line and realizing unmanned linkage between production task scheduling and process parameter analysis, reducing problems caused by human intervention.

[0013] Furthermore, the process of comparing the first production line and the second production line also includes the following steps:

[0014] Obtain the first comparison value and the second comparison value;

[0015] A first deviation value is calculated based on the first comparison value and the first process parameter; a second deviation value is calculated based on the first comparison value and the second process parameter; a third deviation value is calculated based on the second comparison value and the third process parameter; a fourth deviation value is calculated based on the second comparison value and the fourth process parameter; a first production line deviation value is calculated based on the first and second deviation values; a second production line deviation value is calculated based on the third and fourth deviation values; if the first production line deviation value is greater than the second production line deviation value, the comparison result indicates that the first production line has a strong error correlation; otherwise, the comparison result indicates that the second production line has a strong error correlation.

[0016] By adopting the above technical solution and calculating the deviation value in multiple dimensions, the relative error correlation between the first production line and the second production line can be accurately determined. The production line with a larger deviation value has a strong error correlation, while the production line with a smaller deviation value has a weak error correlation.

[0017] Furthermore, the process of comparing the first production line and the second production line also includes the following steps:

[0018] Obtain the first comparison value and the second comparison value;

[0019] A first deviation value is calculated based on a first comparison value and a first process parameter; a second deviation value is calculated based on a first comparison value and a second process parameter; a third deviation value is calculated based on a second comparison value and a third process parameter; a fourth deviation value is calculated based on a second comparison value and a fourth process parameter; a first production line deviation value is calculated based on the first and second deviation values; and a second production line deviation value is calculated based on the third and fourth deviation values.

[0020] The first fluctuation curve is obtained by plotting the first production line deviation value within a preset time period, and the second fluctuation curve is obtained by plotting the second production line deviation value within a preset time period. The first fluctuation curve and the second fluctuation curve are superimposed in the same coordinate system and the area deviation is calculated to obtain the fluctuation area deviation.

[0021] If the fluctuation area deviation is greater than the preset area reference deviation, the comparison result is that the first production line has a strong error correlation; otherwise, the comparison result is that the second production line has a strong error correlation.

[0022] By adopting the above technical solution, through the plotting of fluctuation curves and the calculation of area deviation, the relative error correlation between the first production line and the second production line can be accurately determined based on the concrete graphic data, dynamically adapting to the real-time dynamic changes of the production line deviation value, and maintaining the match between the determination result and the actual state of the production line.

[0023] Furthermore, the task merging process also includes the following sub-steps:

[0024] Tasks on production lines with strong error correlation are tasks to be split. Tasks that have not undergone process processing steps are tasks to be issued. Tasks that have undergone process processing steps are tasks in execution.

[0025] If the workload of a task in progress is less than the preset reference workload, the task will continue to be processed in production lines with strong error correlation; otherwise, the processing of the task in progress will be suspended.

[0026] The tasks to be issued are issued to production lines that do not have strong error correlation. The production lines that do not have strong error correlation receive the tasks to be issued and obtain the unprocessed tasks.

[0027] If the number of unprocessed tasks is greater than the total number of processing tasks for production lines that do not have strong error correlation, then the difference between the number of unprocessed tasks and the total number of processing tasks is calculated, and secondary allocation tasks are split from the unprocessed tasks based on the difference.

[0028] Search for other production lines, calculate their full capacity status, extract those that are not yet at full capacity as production lines to be allocated, and then issue the secondary allocation task to the production lines to be allocated.

[0029] By adopting the above technical solution, through task splitting and hierarchical allocation, tasks to be issued are accurately assigned to designated production lines, dynamically adapting to the differences in processing capacity of other production lines, and maintaining the matching of task allocation with the full capacity status of other production lines; tasks are split and allocated a second time according to the total processing capacity of the production line, and searched for and issued to other production lines that are not at full capacity, which helps to optimize the dynamic allocation of production line task load, improve the balance of task processing, further improve the unmanned linkage of process parameter analysis and production task scheduling, and improve the efficiency of production line task processing.

[0030] Furthermore, the task merging process also includes the following sub-steps:

[0031] The production line to be assigned receives an unprocessed merged task after receiving a secondary assignment task;

[0032] If the number of unprocessed merged tasks is greater than the total number of processing tasks corresponding to the production line to be allocated, the difference between the unprocessed merged tasks and the total number of processing tasks is calculated, and tasks are split off from the unprocessed merged tasks and redistributed based on the difference.

[0033] The task will be redistributed to other production lines awaiting redistribution. The number of times the redistribution task has been generated will be obtained, and the task quantity of the redistribution task will be adjusted according to the negative correlation of the number of times.

[0034] By adopting the above technical solution, tasks are split sequentially according to the processing capacity of the production line to be assigned. If the unprocessed merged tasks exceed the total processing capacity of the production line, the difference is automatically calculated and the tasks are split and redistributed. At the same time, the workload of the redistributed tasks is negatively adjusted based on the number of times the secondary allocation tasks are generated. The actual load of each production line to be assigned is dynamically adapted to maintain a dynamic match between the split task workload and the production line processing capacity.

[0035] Furthermore, the method also includes the following steps:

[0036] During the execution of unprocessed tasks and unprocessed merged tasks, the values ​​of the first acquisition module and the second acquisition module are updated in real time.

[0037] Calculate the growth rates of the first process difference value, the second process difference value, the first production line deviation value, and the second production line deviation value;

[0038] If both the first process difference value and the second process difference value are greater than the preset first difference reference value, it is defined as a continuous process deviation.

[0039] If the deviation values ​​of the first production line and the second production line are both greater than the preset second difference reference value, then it is defined as a continuous production deviation.

[0040] If there is a continuous process deviation or a continuous production deviation, then the first and second production lines will be stopped, a production line status alarm will be generated, and the remaining tasks on the first and second production lines will be designated as tasks to be issued.

[0041] By adopting the above technical solution, during task execution, the task breakdown process and related material information are recorded synchronously, and the data of the first and second acquisition modules are updated in real time. The process difference value, production line deviation value and its growth rate are dynamically updated. By monitoring error changes in real time, the continuous status of process deviation and production deviation is accurately captured. If the error continues to expand and meets the preset conditions, the first and second production lines will be automatically stopped and a production line status alarm will be generated. At the same time, the remaining tasks will be organized to maintain the synchronous matching between error monitoring and production line operation, which is conducive to timely avoidance of the impact of error expansion.

[0042] Furthermore, the method also includes the following steps:

[0043] The strong error correlation is updated in real time. If the production line with the strong error correlation changes, the split tasks are reversed to obtain the restored tasks. Among them, the tasks that have undergone the process processing steps are retained in the current production line.

[0044] The restoration task will be sent to the production lines where tasks were merged before the change with strong error correlation.

[0045] By adopting the above technical solution, during the task splitting and execution process, the task splitting process and related material information are recorded simultaneously, and the error changes are statistically analyzed in real time to dynamically adapt to the real-time fluctuation of the error. When it is detected that the error continues to shrink or the production line with strong error correlation changes, the split tasks can be reversed and restored. The tasks that need to be restored and those that need to be retained are reasonably distinguished. This does not include some tasks that were merged but not split. These tasks are still left on the production line of the merged tasks. The restored tasks are sent to the original production line of the merged tasks before the change, maintaining the dynamic matching between task allocation and error changes. This is conducive to the flexible control of task allocation and improves the unmanned linkage mechanism.

[0046] Furthermore, the step of issuing the restore task to the production line where the task was merged before the strong error correlation change also includes the following sub-steps:

[0047] According to the preset percentage value, some tasks in the restoration task will be merged into the production lines that were merged before the change with strong error correlation.

[0048] The duration after a change occurs in a production line with strong cumulative error correlation is used as a time adjustment value. The ratio of this duration to the preset adoption time is calculated and used as the time adjustment value. The percentage value is then adjusted based on the positive correlation of the time adjustment value.

[0049] By adopting the above technical solution, the task recovery speed is adjusted, the recovery task is allocated by a preset percentage value, and the duration of changes in production lines with strong error correlation is accumulated. Combined with the preset adoption time, the time adjustment value is calculated and the percentage value is adjusted in a positive correlation to dynamically adapt to the task recovery needs, so that the recovery speed matches the production line status, which is conducive to the precise control of the amount of recovery tasks within the set time.

[0050] Furthermore, the step of calculating the difference value of the first process based on the comparison between the first process parameters and the third process parameters using a preset first process library also includes the following sub-steps:

[0051] The first parameter curve is plotted based on the first process parameters, the corresponding first standard curve is obtained from the first process library, and the first difference value is calculated based on the first parameter curve and the first standard curve.

[0052] The third parameter curve is plotted based on the third process parameters, the corresponding third standard curve is obtained from the first process library, and the third difference value is calculated based on the third parameter curve and the third standard curve.

[0053] The first process difference value is calculated based on the first difference value and the third difference value using a preset algorithm;

[0054] The step of calculating the difference value of the second process based on the comparison between the second process parameters and the fourth process parameters using a preset second process library also includes the following sub-steps:

[0055] The second parameter curve is plotted based on the second process parameters, the corresponding second standard curve is obtained from the second process library, and the second difference value is calculated based on the second parameter curve and the second standard curve.

[0056] The fourth parameter curve is plotted based on the fourth process parameter, the corresponding fourth standard curve is obtained from the second process library, and the fourth difference value is calculated based on the fourth parameter curve and the fourth standard curve.

[0057] The second process difference value is calculated based on the second and fourth difference values ​​using a preset algorithm.

[0058] By adopting the above technical solution, the difference value is calculated by drawing parameter curves and comparing graphics, combined with the first process library; the real-time changes of the first process parameter and the second process parameter are dynamically adapted to maintain the synchronous matching between the process difference value calculation and the actual process status of the production line.

[0059] Secondly, this application provides a real-time production task scheduling system based on production line process parameter analysis, which adopts the following technical solution:

[0060] A real-time production task scheduling system based on production line process parameter analysis includes a processor, wherein the processor executes the steps of the real-time production task scheduling method based on production line process parameter analysis as described in any one of the above. Attached Figure Description

[0061] Figure 1 This is a flowchart illustrating the steps of a real-time production task scheduling method based on production line process parameter analysis.

[0062] Figure 2 This is a flowchart comparing the first production line and the second production line in the first implementation method.

[0063] Figure 3 This is a flowchart comparing the second implementation method of the first production line and the second production line. Detailed Implementation

[0064] The embodiments of this application are described in detail below, and examples of the embodiments are shown in the accompanying drawings.

[0065] This application discloses a real-time production task scheduling method based on production line process parameter analysis, specifically designed for the dynamic management of production tasks in large-scale film-making process production line clusters. It is adaptable to continuous production scenarios for various film products, including polyethylene (PE) film, polyethylene terephthalate (PET) film, food packaging film, and industrial films. It can seamlessly integrate with film-making production line process data early warning systems and manufacturing execution management systems, achieving unmanned and intelligent linkage of film-making process parameter acquisition, anomaly analysis, and production task scheduling. The central scheduling server of the film-making production line cluster is the executing entity of this method. This server pre-stores a film-making process standard database, a production line operation status database, and a production task information database. It also establishes wired / wireless bidirectional communication connections with the process parameter acquisition modules, production task execution terminals, and equipment status monitoring modules of each film-making production line, enabling real-time interaction of process data, dynamic perception of production line status, and immediate issuance of scheduling instructions.

[0066] This method is applied to a cluster of multiple film-making production lines of the same type and specifications. The first and second production lines are the core production lines for current process parameter monitoring, while the third production line is a dispatchable line. All three are equipped with identical film-making equipment, process parameter acquisition modules, and material supply systems. The produced film products are of the same type and specifications, differing only in production batches and quantities. The process compatibility is 100%, meeting the basic requirements for merging and distributing production tasks across production lines. The following refers to... Figure 1 This method includes the following steps:

[0067] S1: Real-time acquisition of film-forming process parameters:

[0068] This step uses two pre-set dedicated acquisition modules to simultaneously collect core and auxiliary process parameters from the first and second production lines. The first acquisition module consists of online detection equipment used in the film-making industry, specifically including sensors such as infrared temperature sensors, non-contact laser thickness gauges, or die melt pressure sensors. The acquisition frequency of this module can be set from 500ms / time to 5s / time, with a default of 1s / time, depending on the precision requirements of the film-making process. It supports real-time transmission of parameter data and preliminary filtering of outliers. The second acquisition module consists of environmental monitoring and equipment operation parameter detection equipment, specifically including humidity sensors, speed encoders, or tension sensors. It maintains the same acquisition frequency and data transmission protocol as the first acquisition module to ensure the timing synchronization of the two types of parameters.

[0069] The first acquisition module collects the first process parameters of the first production line and the second process parameters of the second production line, with the second process parameters being of the same type as the first process parameters. The first process parameters can be selected as key film forming temperatures (such as die exit temperature, cooling roller surface temperature, and heat setting zone temperature), average film thickness (average of longitudinal / transverse multi-point measurements), or die melt pressure, etc., which are film-related parameters. In specific embodiments, one type is selected as the acquisition object; for example, in this case, the average film thickness is collected. The second acquisition module collects the third process parameters of the first production line and the fourth process parameters of the second production line, with the fourth process parameters being of the same type as the third process parameters. The third process parameters can be selected as film-related parameters such as ambient humidity in the film-making workshop, traction roller speed, winding tension, and raw material drying moisture content, etc. In specific embodiments, one type of parameter is selected as the acquisition object, and the selected acquisition object corresponds one-to-one with the type of the first process parameter; for example, when the first acquisition module collects film thickness, the second acquisition module collects traction roller speed.

[0070] S2: Calculation of comprehensive process difference value and triggering of scheduling command:

[0071] The central scheduling server pre-stores two process libraries, the first and second, which are standardized benchmark databases for film-making process parameters. The first process library contains the standard thresholds and reasonable fluctuation ranges for the corresponding film-making process parameters, and also stores the standard curve models for the corresponding parameters. This curve represents the optimal dynamic trajectory of the core parameters over time for the corresponding film-making product under stable production conditions. The second process library contains the standard thresholds and reasonable fluctuation ranges for the corresponding film-making process parameters, and also stores the standard curve models for the corresponding parameters. The standard curve model is the optimal dynamic trajectory curve of the corresponding process parameters over time when the film-making production line produces qualified film products. It is generated based on a large amount of stable production test data, combined with film-making process simulation optimization. The various characteristics of the model correspond one-to-one with the film-making product specifications, production equipment models, and process parameter types.

[0072] Specific generation steps:

[0073] (1) Test data acquisition: Select a film production line that matches the target film product (such as 50μm PE film, PET food packaging film), adjust the equipment to the best operating state, set the process parameters according to the standard threshold of the first / second process library, and conduct a continuous stable production test. The test duration is not less than 8 hours (covering the entire stage of start-up, stable production, and minor adjustment of the film production line); through the first / second acquisition module of this application, collect the time sequence data of the corresponding process parameters during the test at a acquisition frequency of 1 second / time to form the original data sample.

[0074] (2) Raw data preprocessing: The raw data samples are denoised, outlier removed and smoothed; among them, the 3σ principle is used to remove abnormal data caused by instantaneous equipment jitter and interference from the acquisition module, the linear interpolation method is used to complete the missing data, and the moving average method is used to smooth the data to eliminate the jagged changes in the data caused by single acquisition fluctuations, so as to obtain standardized data samples.

[0075] (3) Curve fitting and optimization: With time as the horizontal axis and process parameter values ​​as the vertical axis, the standardized data samples are imported into conventional data fitting software in this field such as Origin and Matlab. Polynomial fitting or spline curve fitting is used to fit the data to curves. The fitting parameters are adjusted and optimized with the correlation coefficient R² ≥ 0.98 between the fitted curve and the standardized data samples as the fitting qualification standard, and the initial standard curve model is obtained.

[0076] (4) Process verification and finalization: The initial standard curve model is applied to the actual production of the film production line. Three batches of qualified film products are produced continuously to verify the model's adaptability. If the quality pass rate of the film products is ≥99.5% when the process parameters change according to the model trajectory during the production process, the initial standard curve model is finalized and stored in the first / second process library. If the pass rate does not meet the standard, the model is fine-tuned according to the production deviation until the verification requirements are met.

[0077] Model update rules: When the production equipment of the film production line is upgraded, the specifications of the film products are adjusted, or there are significant changes in the batches of raw materials, data is collected again according to the above steps, a new standard curve model is generated and verified, and the original model in the process library is replaced and updated to ensure dynamic matching between the model and the actual production process.

[0078] Based on the continuous acquisition data of the first process parameter within a preset time period (3min~10min, default 5min), a two-dimensional rectangular coordinate system is constructed. The horizontal axis is the time axis, with the unit being seconds; the vertical axis is the first process parameter value axis, with the unit matching the parameter type. The actual value of the parameter at each time node is used as the vertical axis and the corresponding timestamp is used as the horizontal axis. All data points are connected sequentially, and linear interpolation is used for smoothing to avoid sawtooth changes in the curve caused by fluctuations in a single acquisition, thus obtaining the first parameter curve that reflects the dynamic changes of the first process parameter.

[0079] The central dispatch server automatically retrieves the corresponding first standard curve from the first process library based on the current membrane product specification code and the first process parameter type code. The first parameter curve and the first standard curve are superimposed on the same coordinate system, and the area enclosed by the two curves within a preset time period is calculated using the trapezoidal numerical integration method. This area value is the first difference value, which is a unitless value. The larger the area value, the more significant the dynamic deviation of the first process parameter.

[0080] Similarly, the third parameter curve is plotted based on the continuously collected data of the third process parameter, and the corresponding third standard curve is retrieved from the first process library to calculate the third difference value.

[0081] Based on the weights W1 and W2 of the first process parameter and the third process parameter preset in the first process library, W1 + W2 = 1, with W1 = 0.5 and W2 = 0.5 by default, the first process difference value is obtained by fusion through a weighted summation algorithm. The formula is: First process difference value = First difference value × W1 + Third difference value × W2.

[0082] Similarly, based on the continuously collected data of the second process parameter within a preset time period, a second parameter curve is plotted. The second standard curve corresponding to the second process parameter is retrieved from the second process library, and the second difference value is calculated using the trapezoidal numerical integration method. Based on the continuously collected data of the fourth process parameter, a fourth parameter curve is plotted. The corresponding fourth standard curve is retrieved, and the fourth difference value is calculated. The second process difference value is obtained through a weighted summation algorithm, with the formula: Second process difference value = Second difference value × W1 + Fourth difference value × W2.

[0083] The comprehensive difference value is calculated by combining the difference values ​​of the first process and the difference values ​​of the second process. The difference value of the first process is assigned a preset fusion weight W1', and the difference value of the second process is assigned a preset fusion weight W2', where W1'+W2'=1, and the default value is 0.5 for each. The comprehensive difference value = difference value of the first process × W1' + difference value of the second process × W2'.

[0084] The central dispatch server pre-stores reference difference values, which are the critical thresholds for deviations in the process parameters of the film-making production line. These values ​​are jointly set by the quality tolerance range of the film product, the stability requirements of the film-making process, and the loss control standards for mass production. For example, the reference difference value for 50μm PE film is preset to 120μm·s. If the calculated comprehensive difference value is greater than the reference difference value, it indicates that the dynamic deviation of the process parameters of the first and second production lines has exceeded the reasonable range. Continuing independent production will lead to a decrease in the consistency of film product quality. At this time, the central dispatch server generates and issues a production task scheduling instruction. If the comprehensive difference value is less than or equal to the reference difference value, it indicates that the process status of the two production lines is stable. The central dispatch server returns to step S1 and continues to collect and monitor the process parameters in real time.

[0085] S3: Idle Third Production Line Retrieval and Task Merging / Assignment:

[0086] The central dispatch server retrieves the production line operation status database. If the production line meets the following requirements, it is determined to be the third production line in an idle state: there are no film production tasks being executed, and the production task execution terminal is in a standby ready state; the production equipment (extruder, stretching machine, etc.) has no fault alarms and no unfinished maintenance plans, and the equipment operating parameters can be adjusted to the standard range of the first process library / second process library; the inventory of raw materials and semi-finished products meets the production needs of the merged task, and the material specifications are completely matched with the production tasks of the first and second production lines; the process parameter acquisition module and equipment status monitoring module are operating normally, and can realize real-time data acquisition and abnormal early warning of the production process.

[0087] If the third production line is found to be idle, the central dispatch server performs a production task merging and distribution operation. It extracts complete information on the current production tasks of the first and second production lines from the production task information database, including production order numbers, membrane product production quantities, process parameter requirements (such as thickness, temperature, and speed), delivery milestones, material ratios, and process execution progress. The production tasks of the two production lines are merged to generate a merged task order, specifying the merged total production volume (e.g., 5 tons from the first production line + 8 tons from the second production line = 13 tons), process parameter standards, production progress requirements, and material usage plans. The central dispatch server distributes the merged task order to the production task execution terminal of the third production line via a communication link, while simultaneously issuing production task stop instructions to the first and second production lines. Upon receiving the stop instructions, the first and second production lines immediately complete the equipment slowdown shutdown, raw material / semi-finished product cleaning, and material temporary storage according to the standardized membrane manufacturing process. Based on the requirements of the merged task order, the central dispatch server allocates unused raw materials and transferable semi-finished products (such as unstretched membrane preforms) from the two production lines to the third production line through the material conveying system, achieving rational utilization of production line resources. The third production line starts production according to the merged task order. The central scheduling server returns to step S1 to collect and monitor the process parameters of the third production line in real time.

[0088] If no third production line that meets the idle state criteria is found, it indicates that the film production line cluster has no additional capacity to carry the merged task. At this time, the first production line and the second production line are compared. If the comparison result shows that the first production line has a strong error correlation, the task on the first production line is merged into the task on the second production line. Otherwise, the task on the second production line is merged into the task on the first production line.

[0089] In this embodiment, there are two implementation methods for comparing the first production line and the second production line:

[0090] Implementation Method 1:

[0091] Reference Figure 2The central dispatch server automatically retrieves a first comparison value and a second comparison value from a preset process library based on the current membrane product specification code and process parameter type. The first comparison value, retrieved from the first process library, represents the optimal standard value of the membrane manufacturing process parameter, not a fluctuation range. For example, if the first process parameter is the average membrane thickness, the first comparison value is the standard thickness value of 50μm for that membrane product specification; if it is the die head temperature, the first comparison value is the standard temperature value of 180℃. The second comparison value, retrieved from the second process library, also represents the optimal standard value of the membrane manufacturing process parameter. For example, if the third process parameter is the traction roller speed, the second comparison value is the standard speed value of 15m / min; if the third process parameter is the workshop humidity, the second comparison value is the standard humidity value of 40%RH.

[0092] A first deviation value is calculated based on the first comparison value and the first process parameter: First deviation value = |First process parameter - First comparison value|. A second deviation value is calculated based on the first comparison value and the second process parameter: Second deviation value = |Second process parameter - First comparison value|. A third deviation value is calculated based on the second comparison value and the third process parameter: Third deviation value = |Third process parameter - Second comparison value|. A fourth deviation value is calculated based on the second comparison value and the fourth process parameter: Fourth deviation value = |Fourth process parameter - Second comparison value|.

[0093] The first production line deviation value is calculated using a weighted average algorithm based on the first and second deviation values, with each weighting value being 0.5. Similarly, the second production line deviation value is calculated using a weighted average algorithm based on the third and fourth deviation values, with each weighting value also being 0.5. If the first production line deviation value is greater than the second production line deviation value, the comparison result indicates that the first production line has a strong error correlation; otherwise, the comparison result indicates that the second production line has a strong error correlation.

[0094] The deviation values ​​of the first / second production lines in this application are comprehensive quantitative indicators calculated based on the first comparison value and the second comparison value, and by integrating the multi-dimensional deviation values ​​of the core process parameters (first / second process parameters) and auxiliary process parameters (third / fourth process parameters) of the production line. The magnitude of the value directly reflects the degree of deviation between the overall process state of the production line and the standard process state. The larger the value, the more significant the deviation of the process parameters of the production line from the standard value of the film-making process, and the higher the instability of the production state of the production line.

[0095] Strong error correlation is an attribute definition for the process state of film production lines. It specifically refers to a process state in which the probability of production errors occurring due to deviations of process parameters from standards is higher, the scope of error impact is wider, and the risk of product quality defects caused by continuous production is greater. The production line deviation value, as a comprehensive quantitative indicator of the degree of deviation of the production line process, is positively correlated with the error correlation strength of the production line. That is, the larger the production line deviation value, the stronger the instability of its process state and the stronger the correlation of production errors. Therefore, production lines with larger deviation values ​​are judged to have strong error correlation.

[0096] Film manufacturing is a continuous production process. The quality indicators of film products, such as thickness, flatness, and mechanical properties, have extremely high requirements for the stability of process parameters. The greater the deviation of process parameters from the standard values, the more exponentially the probability of quality defects such as uneven film thickness, stretching cracks, and uneven winding will occur in subsequent production. The production line with larger deviation values ​​is identified as having strong error correlation, and its tasks are merged into the production line with smaller deviation values. This is to avoid the product quality risks caused by the continuous production of high-deviation production lines and to meet the actual production quality control needs of film manufacturing production lines.

[0097] Implementation Method Two:

[0098] Reference Figure 3 The deviation values ​​of the first production line and the second production line are calculated using the same principle as in Implementation Method 1. The difference from Implementation Method 1 is that this implementation method continuously calculates the production line deviation values ​​within a preset time period. The preset time period is set according to the stability requirements of the film-making process (e.g., 3 min, 5 min, default 5 min), and the calculation frequency is consistent with the process parameter acquisition frequency (e.g., 1 s / time). Finally, two sets of continuous "time-production line deviation value" data pairs are obtained (first production line: t_1-Y1_1, t_2-Y1_2……t_n-Y1_n; second production line: t_1-Y2_1, t_2-Y2_2……t_n-Y2_n).

[0099] The central dispatch server constructs a two-dimensional Cartesian coordinate system based on continuous data pairs within a preset time period, and plots the deviation fluctuation curves for the two production lines respectively. The horizontal axis is the time axis, and the scale represents the time nodes of the preset time period (unit: s or min). The deviation values ​​(Y1_1~Y1_n) at each time node of the first production line are used as the vertical axis, and the corresponding timestamps (t_1~t_n) are used as the horizontal axis. All data points are connected sequentially, and linear interpolation is used for smoothing to eliminate the sawtooth changes in the curve caused by single-calculation fluctuations, resulting in the first fluctuation curve. The second fluctuation curve is plotted using the same logic, based on the continuous deviation value data of the second production line.

[0100] The first and second wave curves are superimposed on the same two-dimensional coordinate system. The central dispatch server uses the trapezoidal numerical integration method to calculate the area enclosed by the two wave curves in the same coordinate system. This area is the wave area deviation, which is a unitless value.

[0101] The central dispatch server compares the calculated fluctuation area deviation with the preset area reference deviation to determine the strength of the error correlation between the two production lines. If the fluctuation area deviation > the area reference deviation: this indicates that within the preset time period, the dynamic fluctuation amplitude and overall deviation trend of the deviation value of the first production line are significantly different from those of the second production line, and the fluctuation of the first production line is more severe, thus the first production line is judged to have a strong error correlation. If the fluctuation area deviation ≤ the area reference deviation: this indicates that the dynamic fluctuation of the deviation value of the second production line is more significant, and its process state error correlation is stronger, thus the second production line is judged to have a strong error correlation.

[0102] In this application, the fluctuation area deviation is the positive area enclosed by the two curves, which are calculated using the trapezoidal numerical integration method in the same coordinate system, with the first fluctuation curve as the upper reference curve and the second fluctuation curve as the lower reference curve. If the deviation value of the first fluctuation curve is lower than that of the second fluctuation curve in a preset time period, the area enclosed by the two curves is the negative area. In this application, only the positive area value is taken for the calculation of the fluctuation area deviation, and the negative area value will be inverted. After inversion, it will be determined that the second production line corresponding to the second fluctuation curve has more severe fluctuations.

[0103] The premise for determining "fluctuation area deviation > area reference deviation" in this application is that the first fluctuation curve forms a positive enclosed area relative to the second fluctuation curve. The formation of this positive area is essentially as follows: within a preset time period, the real-time value of the deviation value of the first production line is generally higher than that of the second production line, and the fluctuation amplitude and deviation trend of the deviation value are significantly different from those of the second production line. That is, when the fluctuation area deviation > area reference deviation, it necessarily indicates that the deviation value of the first production line fluctuates more violently and the process state is more unstable, thus it is determined to have a strong error correlation. If the two curves enclose a negative area (i.e., the deviation value of the second fluctuation curve is generally higher than that of the first fluctuation curve), the negative area is inverted and compared with the area reference deviation. If the inverted area value > area reference deviation, the second production line is determined to have a strong error correlation; if the absolute value of the enclosed area ≤ area reference deviation, it indicates that the difference in the fluctuation of the deviation values ​​of the two production lines has not reached a significant level, and the first production line monitored first is determined to have a strong error correlation by default (or determined according to the production line capacity priority).

[0104] Following the step of determining the error correlation between the first and second production lines, the task merging step also includes the following sub-steps:

[0105] The central scheduling server defines all film-making production tasks currently carried on production lines with strong error correlation as tasks to be split. Based on the execution status of the film-making process steps, these tasks are split into tasks in progress and tasks awaiting distribution. Tasks that have not undergone process processing steps are identified as tasks awaiting distribution, while tasks that have undergone process processing steps are identified as tasks in progress. In this embodiment, process processing steps refer to the substantive forming processes in film production, including raw material melting and plasticizing, extrusion film formation, stretching and shaping, cooling and winding, and film surface corona treatment.

[0106] The central dispatch server extracts the process execution logs and material feeding records of the tasks to be split from the production task information database, calculates the task volume of tasks in execution and tasks to be issued, generates a task hierarchical split list, and determines the process parameter requirements, material matching status and delivery nodes of the tasks.

[0107] The central dispatch server pre-stores the reference task volume for execution. This value is the minimum task threshold that the film production line can continuously process in a single run, and can be dynamically adjusted according to the film product specifications and equipment model.

[0108] The server compares the workload of the tasks in progress with the reference workload. If the workload of the tasks in progress is less than the reference workload, it indicates that the workload is small, and the cost of stopping production midway is higher than the cost of continuing production. The central dispatch server instructs the production line to be split to continue completing the tasks in progress on its current production line. If the workload of the tasks in progress is greater than or equal to the reference workload, it indicates that the workload is large, and continuing production on the production line to be split will lead to a greater decline in membrane product quality and high rework costs. The central dispatch server instructs the production line to be split to suspend the tasks in progress.

[0109] The central dispatch server directs the tasks to be assigned to production lines that do not have strong error correlation, i.e., the target receiving production lines. The target receiving production lines receive the tasks to be assigned and integrate them with the existing unexecuted tasks of the production lines to form unprocessed tasks. The unprocessed task quantity is calculated as follows: Unprocessed task quantity = Existing unexecuted task quantity of the target receiving production line + Task quantity to be assigned.

[0110] The central dispatch server retrieves the total number of processing tasks for the target production line from the production line process capacity database. This value represents the maximum processing capacity of the target production line under the current equipment status (no faults), material supply (sufficient inventory), and process stability (parameters meet standards), and is updated dynamically in real time.

[0111] If the number of unprocessed tasks is less than or equal to the total number of processed tasks, it means that the target production line can independently handle all tasks. The server instructs the target production line to execute the task according to the normal process without splitting it.

[0112] If the amount of unprocessed tasks is greater than the total number of processed tasks, calculate the task difference: Task difference = Amount of unprocessed tasks - Total number of tasks processed by the target production line;

[0113] The portion of the unprocessed task with a value equal to the task's difference is defined as a secondary allocation task. Other production lines are searched, their full-capacity status is calculated, and those not yet at full capacity are extracted as production lines to be allocated. The secondary allocation task is then assigned to these production lines.

[0114] The full-capacity status is quantified using the "full-capacity rate," calculated as: Full-capacity rate = (Currently accepted tasks / Maximum total tasks that the production line can handle) × 100%. A higher full-capacity rate indicates a greater production line load and less remaining capacity. A production line not at full capacity is defined as having a full-capacity rate below 1. In practical applications, a full-capacity rate below 0.9 is considered in this context.

[0115] If there is only one production line to be allocated, the secondary allocation task will be issued in full. If there are multiple production lines to be allocated, the task will be split and issued according to the remaining capacity of each production line: the amount of task undertaken by a production line = the amount of secondary allocation task × (the remaining capacity of the production line / the sum of the remaining capacity of all production lines to be allocated). After splitting, the amount of task undertaken by each production line must meet the minimum production batch requirement.

[0116] For scenarios where the production line awaiting allocation is still unable to handle secondary allocation tasks, the system implements task further splitting and dynamic adjustment of task volume, specifically including the following steps:

[0117] After the production line to be assigned receives the secondary allocation task, the central scheduling server integrates the original unprocessed task volume and the newly received secondary allocation task volume of the production line and defines it as an unprocessed merged task.

[0118] Retrieve the total number of tasks to be processed in real time for the production line to be assigned; if the number of unprocessed merged tasks is less than or equal to the total number of processed tasks, it indicates that the production line can accept the task and the instruction is executed according to the normal process; if the number of unprocessed merged tasks is greater than the total number of processed tasks, calculate the redistribution difference: redistribution difference = number of unprocessed merged tasks - total number of processed tasks; split the task portion equal to the difference from the unprocessed merged tasks and define it as the redistribution task.

[0119] The tasks are redistributed to other production lines awaiting allocation. The number of times these tasks are generated is recorded, and the workload of each redistribution task is adjusted based on a negative correlation with the number of reassignments. The definition of the number of reassignment task generation is as follows: the initial redistribution task is used as the counting base, denoted as N=1; the current redistribution task is the Nth allocation, e.g., if the initial redistribution N=2, the second redistribution N=3. The number is extracted by the server from the full task split log and updated in real time. The server pre-stores a basic adjustment coefficient K_0, ranging from 0.8 to 0.95, with a default of 0.9. This coefficient can be adjusted according to the complexity of the film-making process. The adjustment formula is: Adjusted redistribution task workload = Initial redistribution task workload × (K_0)^N; the larger N is, the smaller (K_0)^N is, resulting in a smaller adjusted workload, achieving a negative correlation between workload and allocation frequency.

[0120] The adjusted redistribution tasks are then sent to the subsequent production lines in the list of production lines to be redistributed, sorted by full capacity rate priority. If a single production line still cannot take on the task, this step is repeated until all tasks are redistributed, or a production line capacity warning is triggered when there are no remaining production lines to be redistributed.

[0121] In this embodiment, during the entire process of executing unprocessed tasks and unprocessed merged tasks on the film-forming production line, the central scheduling server performs the following synchronization operations:

[0122] The first and second acquisition modules continuously collect process parameters from the first production line, the second production line, and all production lines undertaking the splitting tasks. After calibration and noise reduction, the data is transmitted to the server in real time to update the process parameter database.

[0123] The central dispatch server updates the first process difference value, the second process difference value, the first production line deviation value, and the second production line deviation value in real time based on the updated process parameters. It also calculates the unit-time growth rate of these four types of deviation values ​​to quantify the dynamic trend of the deviations. The growth rate refers to the rate of change of the deviation value per unit time, and the formula is: Growth rate V = (Current deviation value V_t - Previous deviation value V_{t-1}) / Δt; where Δt is the data acquisition time interval, consistent with the acquisition frequency. If V > 0, the deviation is determined to be in a continuous upward trend; V = 0 indicates stable fluctuation; and V < 0 indicates a gradual decline.

[0124] The central dispatch server presets the first difference reference value, which is the maximum acceptable threshold for deviation of the core process parameters of membrane manufacturing, and is set by the membrane product quality standard.

[0125] If the first process difference value > the first difference reference value, the second process difference value > the first difference reference value, the growth rate V of the two types of difference values ​​> 0, and the above conditions continue to meet the preset stable duration, such as 30s, then it is determined that there is a continuous process deviation.

[0126] The central dispatch server presets a second difference reference value, which is the maximum acceptable threshold for the overall deviation of the film production line, set by the production line stability requirements. If the deviation value of the first production line is greater than the second difference reference value, the deviation value of the second production line is greater than the second difference reference value, the growth rate V of the two types of deviation values ​​is greater than 0, and the above conditions are continuously met for a preset stable duration, such as 30 seconds, then it is determined that there is a continuous production deviation.

[0127] If both process deviation and production deviation exist simultaneously, the first and second production lines will be stopped, a production line status alarm will be generated, and the remaining tasks on the first and second production lines will be designated as tasks to be issued.

[0128] In this embodiment, the central scheduling server continuously monitors the process status of the first and second production lines throughout the entire task execution process and updates the strong error correlation determination results in real time.

[0129] If the new judgment result is inconsistent with the original result, and the state continues to meet the preset stable duration, such as 3min~5min, to avoid misjudgment due to temporary fluctuations, it is judged that a production line with strong error correlation has changed. After the production line with strong error correlation changes, the server performs task reverse restoration based on the full log of task splitting. Tasks that have undergone process processing are retained, and tasks that have not undergone process processing are restored. Among them, only tasks that were split due to strong error correlation in the early stage are reverse restored. Regular merged tasks (batch merged tasks without process deviation) are not involved in the restoration and are still retained in the current receiving production line. Tasks that have undergone core process processing steps are removed from the split tasks and retained in the current production line to avoid changes in semi-finished product quality. The remaining tasks that have not undergone any core process processing are defined as restoration tasks, that is, the unassigned or assigned but unexecuted parts of the original tasks to be issued. The core information of the restoration tasks is extracted, including task volume, product specifications, process parameters, delivery nodes, material matching, etc., and a restoration task list is generated with the original production line merged identifier to ensure the accuracy of targeted issuance.

[0130] After the server completes the definition of the restoration task, it performs a targeted distribution operation: it issues a process parameter calibration instruction to the merged original production line, requiring it to adjust the process parameters to the standard range corresponding to the restoration task, and sends a ready signal after the calibration is qualified; it distributes the restoration task list to the production task execution terminal of the merged original production line, and updates the production task information database and the production line operation status database simultaneously, marking the task status as unprocessed; after the merged original production line starts the restoration task production, the server continuously monitors its process parameters and deviation changes.

[0131] In this embodiment, the step of sending the restoration task to the production line where the tasks were merged before the change with strong error correlation also includes the following sub-steps:

[0132] The central dispatch server pre-stores the initial percentage value of the restoration task. This value represents the percentage of the first task issued, and is set based on the equipment startup characteristics, minimum continuous production batch, and capacity recovery gradient of the merged original production line. The value ranges from 10% to 50%, with a default of 20%. The initial task quantity = total restoration task quantity × initial percentage value. If the calculated result is lower than the minimum production batch of the film-making process, such as 1 ton, the initial task quantity is adjusted according to the minimum batch, and the initial percentage value is adjusted accordingly. For example, if the total restoration task quantity is 4 tons and the initial percentage is 20%, the calculated result is 0.8 tons, which is adjusted to 1 ton, and the initial percentage is adjusted to 25%. The initial task is then issued to the merged original production line, and the server monitors the process parameters (such as film thickness and speed) after receiving the task in real time. If there is no trend of increasing deviation, the subsequent adjustment process begins. If fluctuations occur, the issuance is paused until the parameters stabilize before resuming execution.

[0133] The central dispatch server starts from the time point when a change occurs in a production line with strong error correlation, accumulates the duration, and calculates the time adjustment value to dynamically adjust the distribution percentage.

[0134] 1. Define the cumulative duration T as the time it takes for the merged original production line process to stabilize after the change, in minutes; the preset adoption time T_0 is the critical time at which the load can be gradually increased after the merged original production line process stabilizes, set by the production line stability verification data, such as 10 min to 20 min.

[0135] The time adjustment value K_t = T / T_0, with a range of 0 < K_t ≤ 1. When T ≥ T_0, K_t = 1, reaching the maximum adjustment coefficient, reflecting the maturity of the production line process stability. Based on the time adjustment value, the server dynamically adjusts the percentage of restoration tasks distributed in a positive correlation: the adjusted percentage K = initial percentage value × (1 + K_t), and the upper limit of the adjusted percentage is 100%, i.e., full distribution.

[0136] Update T and K_t at preset time intervals (e.g., 1 min to 3 min), recalculate the adjusted percentage, extract the corresponding proportion of tasks from the remaining restoration tasks, and distribute them to the merged original production line. For example:

[0137] Initial status: Total task volume is 4 tons, initial percentage is 20%, 1 ton is distributed in the first phase (after adjustment), and 3 tons remain.

[0138] T=5min (T_0=10min): K_t=0.5, K=20%×(1+0.5)=30%, this batch will issue 3×30%=0.9 tons (adjusted to 1 ton according to the smallest batch), with 2 tons remaining;

[0139] T=10min: K_t=1, K=20%×(1+1)=40%, this time 2×40%=0.8 tons are issued (adjusted to 1 ton), 1 ton remains;

[0140] Final batch: K_t=1, K=40%, the remaining 1 ton was distributed, and the full amount of restoration was completed.

[0141] 3. Process linkage verification: Before each release, the server verifies and merges the real-time process parameters of the original production line (deviation value ≤ reference value, growth rate V ≤ 0). If the parameters fluctuate, the adjustment is paused and executed again after stabilization to ensure process stability during the recovery process.

[0142] This application also discloses a real-time production task scheduling system based on production line process parameter analysis, including a processor, wherein the processor executes the steps of the real-time production task scheduling method based on production line process parameter analysis as described in any of the above embodiments.

[0143] Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of this application.

Claims

1. A real-time production task scheduling method based on production line process parameter analysis, characterized in that, Includes the following steps: The first process parameters on the first production line are obtained based on the preset first acquisition module, and the second process parameters on the second production line are obtained based on the preset second acquisition module; the third process parameters on the first production line are obtained based on the preset second acquisition module, and the fourth process parameters on the second production line are obtained. The difference value of the first process is calculated by comparing the first process parameter with the third process parameter based on a preset first process library. The second process difference value is calculated by comparing the second process parameter with the fourth process parameter based on the preset second process library. The comprehensive difference value is calculated by combining the first process difference value and the second process difference value. If the comprehensive difference value is greater than the preset reference difference value, a task scheduling instruction is issued. In response to task scheduling instructions, it searches for an available third production line; If an available third production line is found, the tasks on the first and second production lines are merged to form a merged task, which is then sent to the third production line; otherwise, the first and second production lines are compared. If the comparison result shows that the first production line has a strong error correlation, then the tasks on the first production line are merged into the tasks on the second production line; otherwise, the tasks on the second production line are merged into the tasks on the first production line.

2. The real-time production task scheduling method based on production line process parameter analysis according to claim 1, characterized in that, The process of comparing the first production line with the second production line also includes the following steps: Obtain the first comparison value and the second comparison value; A first deviation value is calculated based on the first comparison value and the first process parameter, and a second deviation value is calculated based on the first comparison value and the second process parameter. The third deviation value is calculated based on the second comparison value and the third process parameter, and the fourth deviation value is calculated based on the second comparison value and the fourth process parameter; the first production line deviation value is calculated based on the first deviation value and the second deviation value, and the second production line deviation value is calculated based on the third deviation value and the fourth deviation value. If the deviation value of the first production line is greater than that of the second production line, the comparison result is that the first production line has a strong error correlation; otherwise, the comparison result is that the second production line has a strong error correlation.

3. The real-time production task scheduling method based on production line process parameter analysis according to claim 1, characterized in that, The process of comparing the first production line with the second production line also includes the following steps: Obtain the first comparison value and the second comparison value; A first deviation value is calculated based on the first comparison value and the first process parameter, and a second deviation value is calculated based on the first comparison value and the second process parameter. The third deviation value is calculated based on the second comparison value and the third process parameter, and the fourth deviation value is calculated based on the second comparison value and the fourth process parameter; the first production line deviation value is calculated based on the first deviation value and the second deviation value, and the second production line deviation value is calculated based on the third deviation value and the fourth deviation value. The first fluctuation curve is obtained by plotting the first production line deviation value within a preset time period, and the second fluctuation curve is obtained by plotting the second production line deviation value within a preset time period. The first fluctuation curve and the second fluctuation curve are superimposed in the same coordinate system and the area deviation is calculated to obtain the fluctuation area deviation. If the fluctuation area deviation is greater than the preset area reference deviation, the comparison result is that the first production line has a strong error correlation; otherwise, the comparison result is that the second production line has a strong error correlation.

4. The real-time production task scheduling method based on production line process parameter analysis according to claim 2 or 3, characterized in that, The task merging process also includes the following sub-steps: Tasks on production lines with strong error correlation are tasks to be split. Tasks that have not undergone process processing steps are tasks to be issued. Tasks that have undergone process processing steps are tasks in execution. If the workload of a task in progress is less than the preset reference workload, the task in progress will continue to be processed in the production line with strong error correlation. Otherwise, the processing of ongoing tasks will be suspended; The tasks to be issued are issued to production lines that do not have strong error correlation. The production lines that do not have strong error correlation receive the tasks to be issued and obtain the unprocessed tasks. If the number of unprocessed tasks is greater than the total number of processing tasks for production lines that do not have strong error correlation, then the difference between the number of unprocessed tasks and the total number of processing tasks is calculated, and secondary allocation tasks are split from the unprocessed tasks based on the difference. Search for other production lines, calculate their full capacity status, extract those that are not yet at full capacity as production lines to be allocated, and then distribute the secondary allocation task to the production lines to be allocated.

5. The real-time production task scheduling method based on production line process parameter analysis according to claim 4, characterized in that, The task merging process also includes the following sub-steps: The production line to be assigned receives an unprocessed merged task after receiving a secondary assignment task; If the number of unprocessed merged tasks is greater than the total number of processing tasks corresponding to the production line to be allocated, the difference between the unprocessed merged tasks and the total number of processing tasks is calculated, and tasks are split off from the unprocessed merged tasks and redistributed based on the difference. The task will be redistributed to other production lines awaiting redistribution. The number of times the redistribution task has been generated will be obtained, and the task quantity of the redistribution task will be adjusted according to the negative correlation of the number of times.

6. The real-time production task scheduling method based on production line process parameter analysis according to claim 4, characterized in that, The method also includes the following steps: During the execution of unprocessed tasks and unprocessed merged tasks, the values ​​of the first acquisition module and the second acquisition module are updated in real time. Calculate the growth rates of the first process difference value, the second process difference value, the first production line deviation value, and the second production line deviation value; If both the first process difference value and the second process difference value are greater than the preset first difference reference value, it is defined as a continuous process deviation. If the deviation values ​​of the first production line and the second production line are both greater than the preset second difference reference value, then it is defined as a continuous production deviation. If there is a continuous process deviation or a continuous production deviation, then the first and second production lines will be stopped, a production line status alarm will be generated, and the remaining tasks on the first and second production lines will be designated as tasks to be issued.

7. The real-time production task scheduling method based on production line process parameter analysis according to claim 4, characterized in that, The method also includes the following steps: The strong error correlation is updated in real time. If the production line with the strong error correlation changes, the split tasks are reversed to obtain the restored tasks. Among them, the tasks that have undergone the process processing steps are retained in the current production line. The restoration task will be sent to the production lines where tasks were merged before the change with strong error correlation.

8. The real-time production task scheduling method based on production line process parameter analysis according to claim 7, characterized in that, The process of issuing the restore task to the production line where the task was merged before the strong error correlation change also includes the following sub-steps: According to the preset percentage value, some tasks in the restoration task will be merged into the production lines that were merged before the change with strong error correlation. The duration after a change occurs in a production line with strong cumulative error correlation is used as a time adjustment value. The ratio of this duration to the preset adoption time is calculated and used as the time adjustment value. The percentage value is then adjusted based on the positive correlation of the time adjustment value.

9. The real-time production task scheduling method based on production line process parameter analysis according to claim 1, characterized in that, The step of calculating the difference value of the first process based on the comparison between the first process parameter and the third process parameter using a preset first process library also includes the following sub-steps: A first parameter curve is plotted based on the first process parameters, and a corresponding first standard curve is obtained from the first process library. The first difference value is calculated based on the first parameter curve and the first standard curve. The third parameter curve is plotted based on the third process parameters, and the corresponding third standard curve is obtained from the first process library. The third difference value is calculated based on the third parameter curve and the third standard curve. The first process difference value is calculated based on the first difference value and the third difference value using a preset algorithm; The step of calculating the difference value of the second process based on the comparison between the second process parameters and the fourth process parameters using a preset second process library also includes the following sub-steps: The second parameter curve is plotted based on the second process parameters, the corresponding second standard curve is obtained from the second process library, and the second difference value is calculated based on the second parameter curve and the second standard curve. The fourth parameter curve is plotted based on the fourth process parameter, the corresponding fourth standard curve is obtained from the second process library, and the fourth difference value is calculated based on the fourth parameter curve and the fourth standard curve. The second process difference value is calculated based on the second and fourth difference values ​​using a preset algorithm.

10. A real-time production task scheduling system based on production line process parameter analysis, characterized in that, The system includes a processor that executes the steps of the real-time production task scheduling method based on production line process parameter analysis as described in any one of claims 1-9.

Images