Production scheduling optimization system based on data analysis

Abstract

The present application relates to the technical field of production management, in particular to a production scheduling optimization system based on data analysis, which comprises a process data trend analysis module, a workstation resource real-time acquisition module, a scheduling sequence optimization judgment module, a beat coordination inspection module and a scheduling dynamic optimization module.In the present application, by real-time analysis of process duration trend, accurate identification of continuous growth, dynamic correction of beat parameters, ensuring process rhythm stability, based on workstation task completion time, matching task and workstation idle start time, realizing efficient connection of resources, dynamically adjusting task sequence according to workstation idle state, optimizing task connection, checking adjacent task beat ratio, adjusting task interval, controlling beat difference within a reasonable range, preventing beat mutation from interfering with production, overall enhancing the rapid response capability to production changes, realizing task and resource collaborative optimization, improving overall scheduling efficiency and beat consistency, and improving production continuity and resource utilization.

Description

Technical Field

[0001] This invention relates to the field of production management technology, and in particular to a production scheduling optimization system based on data analysis. Background Technology

[0002] The field of production management technology includes aspects such as production resource allocation, manufacturing process control, production planning and scheduling optimization. Its core lies in the effective coordination and management of each link in production activities to improve production efficiency, reduce costs and ensure product quality. It comprehensively covers the management of elements such as materials, personnel, equipment and time. Through scientific management methods, the entire production process is dynamically monitored and adjusted to achieve optimal allocation of various production resources, thereby meeting market demands and improving the overall operational level.

[0003] Among them, the data-driven production scheduling optimization system refers to the establishment of an optimization strategy system for the production operation sequence by collecting, organizing, and analyzing relevant data in the production process. It addresses technical matters such as the rational arrangement of operation sequences, the balanced allocation of resources, and the dynamic adjustment of production capacity. Specifically, it uses real-time collected production data as a basis, models the time required for each operation and resource usage through statistical analysis methods, and then uses linear programming or integer programming algorithms to optimize the scheduling sequence and formulate adjustment plans. Typically, the scheduling optimization task relies on the collaborative efforts of three stages: production data recording, data analysis, and optimization calculation.

[0004] The existing production scheduling management system lacks dynamic response capabilities to the real-time status of the production process, fails to effectively track the changing trends of process durations, and suffers from lagging adjustments to cycle time parameters. This results in uncorrected cycle time fluctuations during production, leading to frequent backlogs or delays. Regarding resource allocation, workstation utilization relies on fixed configurations without optimization based on real-time task completion. Idle workstations are not promptly identified and utilized, resulting in low resource scheduling efficiency. Task sequencing is based on initial plans, lacking a dynamic adjustment mechanism for task execution timing and actual workstation idleness, leading to frequent task waiting and conflicts. In terms of cycle time coordination, the lack of a proportional verification mechanism prevents timely correction of process interruptions caused by imbalanced cycle time ratios, resulting in disrupted production line rhythms. For example, task interval settings that have not been adjusted for a long time prevent some processes from immediately proceeding to the next process after completion, leading to accumulated waiting time and slowing down the overall production pace. This makes it difficult for the production system to respond efficiently to changing conditions, affecting resource utilization efficiency and task execution continuity. Summary of the Invention

[0005] The purpose of this invention is to address the shortcomings of existing technologies by proposing a production scheduling optimization system based on data analysis.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: a production scheduling optimization system based on data analysis includes:

[0007] The process data trend analysis module acquires continuous process duration records, constructs a duration increase sequence, performs a sign consistency judgment, if all signs are positive, calculates the deviation between the latest process duration and the current schedule cycle number, if it is greater than the cycle time deviation threshold, then records and generates a process cycle adjustment parameter set;

[0008] The real-time workstation resource acquisition module collects the completion time of tasks at each workstation, records the start time of workstation idle time, retrieves the target workstation data of the current scheduled tasks, performs task-workstation-idle time matching, and generates a real-time workstation idle dataset.

[0009] The scheduling order optimization judgment module compares the start time of the target workstation for the current task with the start time of the next task's idle workstation based on the real-time workstation idle dataset, determines whether there is a need to swap task orders, performs adjustment operations to update the scheduled task list, and generates a scheduling order optimization result.

[0010] The cycle time coordination verification module, based on the scheduling order optimization results and the process cycle time adjustment parameter set, judges the ratio of the cycle time of adjacent tasks in the scheduling task list. If the ratio of the cycle time of adjacent tasks is greater than the tolerance of the ratio of adjacent cycle time, the task interval time is shortened and cycle time coordination correction data is generated.

[0011] As a further embodiment of the present invention, the process cycle adjustment parameter set includes a duration increase judgment value, a cycle time deviation amount, and a cycle adjustment suggestion value; the workstation idle real-time dataset includes workstation number, idle start time, and task matching degree; the scheduling order optimization result includes task swap record, updated task order list, and adjusted scheduling priority; and the cycle coordination correction data includes a cycle ratio correction value, task interval adjustment amount, and cycle coordination identifier.

[0012] As a further aspect of the present invention, the process data trend analysis module includes:

[0013] The duration difference extraction submodule, based on the acquired continuous process duration record data, arranges them in chronological order, performs difference calculation on the duration of any adjacent process, obtains the first-order difference sequence of adjacent process times, calculates the absolute value of all differences in the duration difference sequence, and obtains the absolute value set of the duration difference sequence.

[0014] The sign consistency judgment submodule performs sign consistency judgment on any continuous difference in the sequence based on the absolute value set of the duration difference sequence, filters the difference combination interval where all signs are positive, extracts the latest process duration and the current scheduling cycle number based on the filtered interval, and calculates the process duration cycle deviation value.

[0015] The cycle time adjustment parameter generation submodule compares the cycle time deviation value of the process duration with the cycle time deviation threshold. If the deviation is greater than the cycle time deviation threshold, it extracts the current schedule cycle number, the latest process duration, and the cycle time deviation threshold, calculates the cycle time adjustment benchmark value, and establishes a cycle time adjustment parameter set.

[0016] As a further aspect of the present invention, the real-time workstation resource acquisition module includes:

[0017] The task completion collection submodule collects the task completion time of each workstation and records the system clock time when the corresponding task at each workstation ends, generating a set of workstation task completion times.

[0018] The idle time recording submodule, based on the workstation task completion time set, records the system clock time after the task is completed for each workstation as the idle start time, and establishes the workstation idle start time value by corresponding with the workstation identifier.

[0019] The workstation data matching submodule performs task and workstation matching based on the workstation idle start time value and the target workstation data of the current scheduled task. It combines the target workstation number and workstation idle start time of each task, integrates the task number, workstation number, and idle start time, and obtains a real-time dataset of workstation idle time.

[0020] As a further aspect of the present invention, the scheduling order optimization judgment module includes:

[0021] The idle time comparison submodule obtains the start time of the target workstation idle time for the current task and the next task based on the real-time workstation idle data set. It then compares the execution times of the target workstation idle time for the current task and the target workstation idle time for the next task to obtain the difference in workstation idle time.

[0022] The sequence swap determination submodule determines whether the start time of the next task's workstation is earlier than the start time of the current task's workstation based on the difference in the workstation's idle time. If the determination result is yes, it records the sequence swap requirement between the current task and the next task and obtains the task sequence swap flag.

[0023] The scheduling list update submodule performs an order adjustment operation on the scheduled task list according to the task order swap markers, updates the scheduled task list after adjusting the task order, and generates the scheduling order optimization result.

[0024] As a further aspect of the present invention, the rhythm coordination testing module includes:

[0025] The cycle ratio judgment submodule extracts the cycle number of adjacent tasks in the scheduling task list based on the scheduling order optimization result and the process cycle adjustment parameter set, performs a ratio relationship judgment on the cycle number of adjacent tasks, calls the adjacent cycle ratio tolerance, calculates the cycle ratio of adjacent tasks, compares it with the cycle ratio tolerance, and obtains the cycle ratio exceeding the limit flag.

[0026] The task interval adjustment submodule shortens the task interval time between adjacent tasks marked as exceeding the limit based on the beat ratio exceeding the limit. After shortening the task interval time, the task interval adjustment value is recorded.

[0027] The configuration record generation submodule adjusts the task interval value according to the task interval, integrates the adjusted task interval time and corresponding task number, establishes a complete task interval configuration record, and generates beat coordination correction data.

[0028] As a further aspect of the present invention, the system further includes:

[0029] The scheduling dynamic optimization module integrates the scheduling task data based on the cycle time coordination correction data, updates the execution time and corresponding workstation configuration of each production task, constructs a globally optimized scheduling timetable and resource allocation table, and generates dynamic optimization results for production scheduling.

[0030] The results of the dynamic optimization of the production scheduling include the optimized scheduling timetable, resource allocation table, and task execution time configuration.

[0031] As a further aspect of the present invention, the scheduling dynamic optimization module includes:

[0032] The task data integration submodule acquires the takt time coordination correction data, performs an overall integration operation on the scheduled task data, extracts the takt time information, workstation configuration and adjusted task interval corresponding to each task, integrates them into a unified set of scheduled tasks, and generates integrated scheduled task data.

[0033] The execution time period update submodule updates the execution time period of each production task based on the integrated data of the scheduling tasks, adjusts the start and end times of each task according to the task order and rhythm coordination correction data, and obtains the task execution time period configuration data.

[0034] The resource table generation submodule, based on the task execution time period configuration data and combined with the workstation number and resource allocation requirements of each task, establishes a globally optimized scheduling timetable and resource allocation table, generating dynamic optimization results for the production schedule.

[0035] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0036] In this invention, by analyzing the trend of process duration in real time, continuous growth is accurately identified, and cycle time parameters are dynamically adjusted to ensure stable process rhythm and avoid local congestion. Based on the completion time of workstation tasks, the idle start time of tasks and workstations is matched to achieve efficient resource connection and reduce idle time. The task sequence is dynamically adjusted according to the idle status of workstations to optimize task connection and improve scheduling flexibility. The cycle time ratio of adjacent tasks is verified and the task interval is adjusted to keep the cycle time difference within a reasonable range and prevent sudden changes in cycle time from interfering with production. Overall, the ability to respond quickly to production changes is enhanced, the optimal synergy between tasks and resources is achieved, the overall scheduling efficiency and cycle time consistency are improved, and the continuity of production and resource utilization are enhanced. Attached Figure Description

[0037] Figure 1 This is a system flowchart of the present invention;

[0038] Figure 2 This is a flowchart of the process data trend analysis module of the present invention;

[0039] Figure 3 This is a flowchart of the real-time workstation resource acquisition module of the present invention;

[0040] Figure 4 This is a flowchart of the scheduling order optimization judgment module of the present invention;

[0041] Figure 5 This is a flowchart of the beat coordination test module of the present invention;

[0042] Figure 6 This is a flowchart of the scheduling dynamic optimization module of the present invention. Detailed Implementation

[0043] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0044] Please see Figure 1 The data-driven production scheduling optimization system includes:

[0045] The process data trend analysis module acquires continuous process duration records, arranges them in chronological order, calculates the difference between any two adjacent durations to construct a duration increase sequence (i.e., a first-order difference sequence of adjacent process times), performs a sign consistency judgment based on any consecutive differences in the sequence, if all signs are positive, calculates the deviation between the latest process duration and the current schedule cycle number, and compares it with the cycle time deviation threshold (the maximum allowable time deviation of the production cycle). If the deviation is greater than the threshold, it records and generates a process cycle adjustment parameter set.

[0046] The real-time workstation resource acquisition module collects the task completion time of each workstation, records the system clock time as the start time of workstation idle time, retrieves the target workstation data of the current scheduled task, performs task-workstation-idle time matching, and generates a real-time workstation idle dataset.

[0047] The scheduling order optimization judgment module compares the start time of the target workstation for the current task with the start time of the next task's workstation based on the real-time dataset of workstation vacancy. If the start time of the next task's workstation vacancy is earlier than the start time of the current task's workstation vacancy, the module records the need to swap the order of the two tasks, performs the order adjustment operation, updates the scheduled task list, and generates the scheduling order optimization result.

[0048] The cycle time coordination test module, based on the scheduling order optimization results and the process cycle time adjustment parameter set, judges the ratio of the cycle time of adjacent tasks in the scheduling task list. If the ratio of the cycle time of adjacent tasks is greater than the tolerance of the ratio of adjacent cycle time (obtained by calculating the production line balance rate or statistically analyzing historical cycle time synchronization data, representing the maximum allowable ratio of the difference in cycle time between adjacent processes, such as the 20%-30% tolerance range commonly seen in industry standards), then the task interval time is shortened, and the adjusted task interval configuration is recorded to generate cycle time coordination correction data.

[0049] The dynamic scheduling optimization module coordinates and corrects data based on the takt time, integrates the scheduling task data as a whole, updates the execution time and corresponding workstation configuration of each production task, constructs a globally optimized scheduling timetable and resource allocation table, and generates dynamic optimization results for production scheduling.

[0050] The process cycle time adjustment parameter set includes the duration increase judgment value, cycle time deviation, and cycle time adjustment suggestion value. The workstation idle real-time dataset includes workstation number, idle start time, and task matching degree. The scheduling order optimization results include task swap records, updated task order list, and adjusted scheduling priority. The cycle time coordination correction data includes cycle time ratio correction value, task interval adjustment amount, and cycle time coordination identifier. The production scheduling dynamic optimization results include the optimized scheduling time table, resource allocation table, and task execution time configuration.

[0051] Please see Figure 2 The process data trend analysis module includes:

[0052] The duration difference extraction submodule, based on the acquired continuous process duration record data, arranges them in chronological order, performs difference calculation on the duration of any adjacent process, obtains the first-order difference sequence of adjacent process times, calculates the absolute value of all differences in the duration difference sequence, and obtains the absolute value set of the duration difference sequence.

[0053] To obtain continuous process duration records, the original records of process execution times must first be collected. This data can be obtained through the timestamp records of the industrial control system. For example, the start and end times of each process on the production line are 10:00-10:05, 10:06-10:12, 10:13-10:20, etc., with corresponding process durations of 5 minutes, 6 minutes, and 7 minutes, respectively. This data is arranged in chronological order to ensure the consistency of the time series of adjacent processes. Through the above time difference calculation, the duration difference between adjacent processes is 1 minute and 1 minute, respectively. Further, the absolute values ​​of all adjacent process duration differences are processed to obtain the absolute value set of the difference sequence as {1, 1}. In specific execution, the duration of the first and second processes is first extracted by executing |6-5|=1, and the duration of the second and third processes is extracted by executing |7-6|=1, forming a complete duration difference sequence, and then generating the absolute value set of the duration difference sequence.

[0054] The sign consistency judgment submodule performs sign consistency judgment on any continuous difference in the sequence based on the absolute value set of the time difference sequence, filters the difference combination interval where all signs are positive, extracts the latest process duration and the current scheduling cycle number based on the filtered interval, and calculates the process duration cycle number deviation value.

[0055] Based on the absolute value set of the time difference sequence, the sign of continuous differences in the difference sequence is determined. Specifically, the sign of each difference is checked sequentially. If all differences in the sequence are positive, the sign consistency condition is met. For example, if the difference sequence is {1, 1}, all of its signs are positive, which meets the consistency requirement. Based on this result, the latest process duration of 7 minutes and the current schedule cycle number of 6 minutes are extracted, and the difference is calculated. 7-6=1 minute, and the process duration cycle deviation value is obtained as 1 minute.

[0056] The cycle time adjustment parameter generation submodule compares the cycle time deviation value of the process duration with the cycle time deviation threshold. If the deviation is greater than the cycle time deviation threshold, it extracts the current schedule cycle time number, the latest process duration, and the cycle time deviation threshold, using the following formula:

[0057]

[0058] Calculate and obtain the beat adjustment reference value P a Establish a set of beat adjustment parameters, where T n S represents the duration of the latest process. p D represents the current schedule tick count. i represents the i-th difference in the set of absolute values ​​of the duration difference sequence, and N represents the number of differences in the sign-consistent interval of the difference sequence.

[0059] Based on the cycle time deviation value of the process duration, a cycle time deviation threshold of 0.8 minutes is set. This threshold is set according to the standard tolerance range of the production cycle time, usually determined based on the stability of equipment operation and the maximum allowable time fluctuation range between processes. Specifically, it refers to the principle that the fluctuation of the rated cycle time of the production line should not exceed 10%. The current scheduled cycle time is 6 minutes, so the maximum allowable deviation is 6 × 10% = 0.6 minutes. Considering the impact of equipment changeover time between processes and the difference in manual operation, the tolerance is adjusted to be widened to 13%, that is, 6 × 13% = 0.78 minutes, which is rounded up to 0.8 minutes. This threshold changes with the adjustment of the scheduled cycle time; the higher the scheduled cycle time, the greater the allowable time deviation, and vice versa, to ensure the flexibility and accuracy of cycle time control. Comparing 1 minute with 0.8 minutes, 1 > 0.8, which meets the condition that the deviation is greater than the threshold. The current scheduled cycle time S is then extracted. p =6 minutes, latest process duration T n =7 minutes, difference set D1=1, D2=1, substitute into the formula:

[0060]

[0061] Calculate and obtain the beat adjustment reference value P a Approximately 0.832 minutes, combined with the current scheduled beat count of 6 minutes and P a A set of cycle time adjustment parameters is established to dynamically reflect the variation range of the cycle time in the process. As shown in Table 1:

[0062] Table 1: Parameters Related to Beat Adjustment

[0063] Current schedule tick count Latest process duration <![CDATA[Difference D1]]> <![CDATA[Difference D2]]> Beat adjustment reference value 6 minutes 7 minutes 1 minute 1 minute 0.832 minutes

[0064] The formula's operational logic lies in comprehensively reflecting the combined impact of changes in process duration on cycle time adjustment through multi-level numerical calculation relationships. First, T n -S p This part represents the direct deviation between the latest process duration and the current scheduled cycle time. This deviation value is used to measure the initial difference between the actual execution of the process and the planned cycle time, and is subsequently summed with the cumulative difference of continuous differences. The summation reflects the cumulative effect of the increase in time between consecutive historical processes, ensuring that the current deviation not only reflects the differences in individual processes but also takes into account the cumulative effect of trend changes. The overall deviation is then summed with... Normalization is performed by scaling the order of magnitude of the latest process duration and the current schedule cycle count using square root operations to avoid abnormal fluctuations caused by numerical amplification. This also improves the stability of the cycle adjustment benchmark and the consistency of the relative scale. Finally, absolute value operations are used to ensure that the deviation is positive, which is beneficial for subsequent judgment and application. Therefore, this formula, through the summation of the cumulative deviation, square root normalization scaling, and absolute value standardization, enables the cycle adjustment benchmark to have a comprehensive ability to reflect actual deviation, historical trends, and cycle count.

[0065] Please see Figure 3 The real-time workstation resource acquisition module includes:

[0066] The task completion collection submodule collects the task completion time of each workstation and records the system clock time when the corresponding task at each workstation ends, generating a set of workstation task completion times.

[0067] The system collects the task completion time of each workstation and records the system clock time as the start time of workstation idle time. First, for each workstation, based on the workstation number and task assignment information, its task execution status is monitored in real time. When the task status changes from "in execution" to "completed", the current system clock time is immediately read. This time is the task completion time. For example, if workstation A completes its task at 10:15:30, this time is recorded. Next, the completion time is stored in the workstation task record table, along with associated information such as workstation number and task number. At the same time, for other workstations within the same time period, such as workstation B and workstation C, their corresponding task completion times are recorded. For example, if workstation B's completion time is 10:17:45 and workstation C's is 10:17:45, then the completion time is recorded. The time sequence is 10:20:10. The collected completion times are then categorized and summarized, arranged in ascending order of workstation number to form a complete time record sequence. Any missing data is checked; if any is found, it needs to be collected again. For example, the collected data is: Workstation A: 10:15:30, Workstation B: 10:17:45, and Workstation C: 10:20:10, as shown in Table 2. Table 2 lists the workstation task completion time collection results. Furthermore, for workstation A, 10:15:30 is recorded as its idle start time, and so on for other workstations. This process needs to consider that the system clock accuracy is at the millisecond level to avoid recording errors caused by insufficient time precision. Finally, the workstation task completion time set is obtained.

[0068] Table 2: Workstation Task Completion Schedule

[0069] Workstation Number Task Number Completion Time A T001 10：15：30 B T002 10：17：45 C T003 10：20：10

[0070] The idle time recording submodule is based on the workstation task completion time set. It records the system clock time after the task is completed for each workstation as the idle start time and associates it with the workstation identifier to establish the workstation idle start time value.

[0071] Based on the workstation task completion time set, the system clock time after task completion for each workstation is recorded as the idle start time and associated with the workstation identifier. First, the workstation task completion time set is called to extract the completion time data for each workstation. The completion times of workstations a, b, and c are read sequentially according to their workstation numbers: 10:15:30, 10:17:45, and 10:20:10, respectively. These times are then assigned as the idle start time. Next, the idle start time is standardized to ensure it is in 24-hour format (HH:MM:SS). For example, if a certain workstation... If the recorded time is 10:5:3, it is converted to 10:05:03. Then, the idle start time is paired with the workstation number to generate a key-value pair structure, such as {a: 10:15:30}, {b: 10:17:45}, {c: 10:20:10}. A list or dictionary structure is then created to manage this data set. The time data is then logically validated to ensure that the time order is reasonable and to avoid incorrect pairings caused by data anomalies. For example, if the idle time of workstation b is earlier than that of workstation a, it is necessary to re-validate or re-collect the data. After completing the above operations, the idle start time value of the workstation is obtained.

[0072] The workstation data matching submodule performs task and workstation matching based on the workstation idle start time value and the target workstation data of the current scheduled task. It combines the target workstation number and workstation idle start time of each task, integrates the task number, workstation number, and idle start time, and obtains the real-time dataset of workstation idle.

[0073] Based on the workstation idle start time, and according to the target workstation data of the current schedule, task and workstation pairing is performed. Combining the target workstation number and idle start time of each task, first, the current schedule task data is obtained, and the correspondence between task number and target workstation number is read. For example, the target workstation for task T004 is a, the target workstation for task T005 is b, and the target workstation for task T006 is c. Then, the established workstation idle start time values ​​are called, such as workstation a: 10:15:30, workstation b: 10:17:45, and workstation c: 10:20:10, and task T004 is paired with workstation a. The idle start time is assigned to 10:15:30. Task T005 is paired with workstation b and assigned an idle start time of 10:17:45. Similarly, task T006 is paired with workstation c. After pairing, the pairing results are summarized to form a set of triplets of task number, workstation number, and idle start time, such as {T004, a, 10:15:30}, {T005, b, 10:17:45}, {T006, c, 10:20:10}. This set is the data set that reflects the current idle status of each task at each workstation in real time. Finally, the real-time idle dataset of workstations is obtained.

[0074] Please see Figure 4 The scheduling order optimization judgment module includes:

[0075] The idle time comparison submodule obtains the start time of the target workstation idle time for the current task and the next task based on the real-time workstation idle data set. It then compares the execution times of the target workstation idle time for the current task and the target workstation idle time for the next task to obtain the difference in workstation idle time.

[0076] The process involves obtaining the start time of the target workstation's idle time for the current and next tasks. It also retrieves the target workstation numbers W1 and W2 for tasks T1 and T2 in the current schedule list. Further, it extracts the corresponding idle start times S1 and S2 from the generated real-time workstation idle data set. S1 and S2 are then compared to determine their temporal relationship. This is further refined to obtaining the specific time values ​​of S1 and S2, for example, S1 = 10:35:20 and S2 = 10:28:45. These time values ​​are then converted to seconds. That is, S1 = 38120 seconds and S2 = 37725 seconds. By comparing the magnitudes of 38120 and 37725, it is determined that S2 is earlier than S1. During this comparison operation, attention should be paid to the standardization of time. All times should be uniformly converted to seconds to avoid conversion errors between hours, minutes and seconds. Here, the idle time difference ΔT = 38120 - 37725 = 395 seconds is obtained. This difference is the idle time difference of the workstation. As shown in Table 3, workstation W2 is idle 395 seconds earlier than workstation W1, indicating that there is a possibility of optimized sorting.

[0077] Table 3: Workstation Idle Start Time Schedule

[0078] Workstation Number Workstation idle start time Workstation idle start time (seconds) W1 10：35：20 38120 W2 10：28：45 37725

[0079] Table 3 shows the idle times of W1 and W2. After calculation using standardized seconds, the corresponding time difference ΔT = 395 seconds was obtained.

[0080] The sequence swap determination submodule determines whether the start time of the next task's workstation is earlier than the start time of the current task's workstation based on the difference in workstation idle time. If the determination result is yes, it records the sequence swap requirement between the current task and the next task and obtains the task sequence swap flag.

[0081] Based on the difference in idle time at each workstation, it is determined whether the start time of the next task's workstation idle time is earlier than the start time of the current task's workstation idle time. The swapping requirement is determined by comparing the sign of the idle time difference ΔT. If ΔT > 0, it means that the next task's workstation idles earlier than the current task's workstation idles. The task order swapping requirement is recorded, and the current task T1 and the next task T2 are marked as needing to be adjusted in order. Here, taking ΔT = 395 seconds as an example, if ΔT is positive, it is determined that the order of T1 and T2 needs to be adjusted. On this basis, a mark M = 1 is generated, where M is the order swapping mark. M = 1 means that swapping is required, and M = 0 means that swapping is not required. This mark is used as the basis for scheduling adjustment. In the further judgment process, if ΔT is in different value ranges, the reference range can be set to 0-60 seconds, 60-300 seconds, and above 300 seconds, which correspond to no swapping, optional swapping, and mandatory swapping, respectively. In this embodiment, ΔT = 395 seconds, which belongs to the mandatory swapping range, and the task order swapping mark is obtained.

[0082] The scheduling list update submodule swaps the markers according to the task order, adjusts the order of the scheduled tasks, updates the scheduled tasks list, and generates the optimized scheduling order result.

[0083] Based on the task order swapping flag, the scheduled task list is reordered. The original scheduled task list {T1, T2, T3, ..., Tn} is extracted. Based on the swapping flag M=1, T1 and T2 are swapped, updating to {T2, T1, T3, ..., Tn}. During the update process, the task attributes and corresponding target workstations must remain unchanged; only their order in the task list is adjusted. The new task list is then renumbered and sorted to form an updated scheduled task set. The updated list is recorded and stored to obtain the scheduling order optimization result.

[0084] Please see Figure 5 The beat coordination test module includes:

[0085] The cycle time ratio judgment submodule, based on the scheduling order optimization results and the process cycle time adjustment parameter set, extracts the cycle time number of adjacent tasks in the scheduling task list, performs a ratio relationship judgment on the cycle time numbers of adjacent tasks, calls the adjacent cycle time ratio tolerance, and compares whether the cycle time ratio of adjacent tasks is greater than the tolerance, using the formula:

[0086]

[0087] Calculate the ratio of adjacent task cycles R gh , will R gh With the rhythm ratio tolerance T r Compare and determine R gh Is it greater than T? r Obtain the beat ratio exceeding the limit flag, where S gS represents the tick number of the g-th task in the scheduled task list. h This represents the number of beats for the h-th task;

[0088] To obtain the scheduling order optimization results and the process cycle time adjustment parameter set, first extract the cycle time of adjacent tasks from the scheduling task list, and set them as the cycle time S of the g-th task. g The number of beats S for the h-th task h For example, on a production line, task 1 has a cycle time of 50 seconds, and task 2 has a cycle time of 60 seconds. After extraction, these are S1 = 50 and S2 = 60 respectively. Then, the cycle time difference is calculated. First, the absolute value of the difference is calculated: |S1 - S2| = |50 - 60| = 10. Then, this difference is divided by the latter cycle time S2. Converted to a percentage, this yields 16.67%, then the beat rate tolerance T is applied. r In comparison, the cycle time ratio tolerance is set based on the production line balance rate, which is obtained through statistical analysis of the average load rate of each workstation and the stability of the process cycle time. Typically, during the stability assessment of multiple production lines, the distribution range of cycle time differences for each workstation under standard operating conditions is used as the basis to select the maximum allowable cycle time difference ratio range. This range fluctuates non-linearly with the number of processes. When the number of processes is between 10 and 20, the cycle time fluctuation tolerance is usually stable between 20% and 30%. Combining the synchronization efficiency and transmission time matching degree of the production equipment, 25% is selected as the standard setting for the cycle time ratio tolerance of adjacent processes. It is determined whether 16.67% is less than 25%. If it is less, no action is needed; if it is greater, it is marked as a cycle time ratio exceeding the limit. Taking Task 3 and Task 4 as examples, with cycle times of 70 seconds and 50 seconds respectively, |70-50| = 20 is calculated, and the ratio R... gh for That is, 40%, which exceeds the tolerance of 25%, so it is marked as exceeding the limit. This process is applied to all adjacent tasks in turn. By calculating the cycle ratio of each pair of tasks and comparing the tolerance, the cycle ratio exceeding the limit is obtained.

[0089] formula The operational logic lies in normalizing the differences between the number of ticks of adjacent tasks and evaluating the relative magnitude of these differences. First, S... g With S h These represent the number of ticks for two adjacent tasks, where |S g -S h | represents the absolute difference between the two. The use of absolute value calculation eliminates the influence of task order on the magnitude of the difference, ensuring that the difference is positive regardless of whether the former is greater than the latter or vice versa, thus guaranteeing the universality of the calculation results. Secondly, the difference value is related to S. hThe division operation is performed to convert the absolute difference into a proportional value relative to the subsequent beat count, thus standardizing the difference and allowing comparisons between different beat count pairs on a uniform scale. Furthermore, multiplying by 100% converts the standardized difference value into a percentage, facilitating comparison with the preset beat ratio tolerance T. r By making direct comparisons, this formula comprehensively utilizes the logic of eliminating the influence of signs through absolute values, normalizing differences through proportional operations, and facilitating threshold judgment through percentage forms, in order to fully reflect the measurement method of the coordination of adjacent task rhythms.

[0090] Table 4: Beat Ratio and Tolerance Judgment Table

[0091]

[0092] As shown in Table 4, the beat ratio between adjacent tasks was judged by tolerance. Tasks 3-4 were marked as exceeding the limit because the beat ratio exceeded 25%.

[0093] The task interval adjustment submodule shortens the task interval time between adjacent tasks marked as exceeding the limit based on the beat ratio exceeding the limit. After shortening the task interval time, the task interval adjustment value is recorded.

[0094] Based on the clock speed ratio exceeding the limit, the interval adjustment is performed on the marked task pairs. First, the original interval time of task 3 and task 4 is read. Assuming the original interval is 30 seconds, the shortening operation is performed according to the preset shortening ratio. If the shortening ratio is 20%, the adjusted interval time is 30 × (1 - 0.2) = 24 seconds. This result is recorded. The new interval of task 3-4 is 24 seconds. Since task 1-2 and task 5-6 have not exceeded the limit, the interval time remains unchanged. This processing logic is repeated for all marked task pairs to obtain the task interval adjustment value.

[0095] Table 5: Task Interval Adjustment Table

[0096]

[0097] See Table 5. Tasks 3-4 have been shortened and adjusted, with the interval now being 24 seconds.

[0098] The configuration record generation submodule adjusts the values ​​according to the task interval, integrates the adjusted task interval time and corresponding task number, establishes a complete task interval configuration record, and generates beat coordination correction data.

[0099] Based on the task interval adjustment values, all task pairs and their corresponding adjusted intervals are integrated into a configuration record. Tasks 1-2 and 5-6 retain their original intervals, while tasks 3-4 use the adjusted interval of 24 seconds. In this way, a data table is established to correspond to the task number pairs and the interval time, forming a task interval configuration record, and finally generating the beat coordination correction data.

[0100] Please see Figure 6 The scheduling dynamic optimization module includes:

[0101] The task data integration submodule acquires the takt time coordination correction data, performs an overall integration operation on the scheduled task data, extracts the takt time information, workstation configuration and adjusted task interval corresponding to each task, integrates them into a unified set of scheduled tasks, and generates integrated scheduled task data.

[0102] To obtain the takt time coordination correction data, a comprehensive integration operation is performed on the scheduled task data. First, the task interval adjustment records and takt time adjustment values ​​are extracted from the takt time coordination correction data. Each item is matched against the task number and original task interval in the scheduled task data. By comparing the interval adjustment value in the takt time coordination correction data with the time period of the original scheduled task data, the correction magnitude for each task is determined, refined to the task level. The takt time adjustment of the task time periods is then accumulated to form continuous scheduled time periods. Based on the workstation configuration corresponding to the task, the updated task time periods are associated and integrated with the workstation resource information of each task. For example, if the original execution time of task a is 8:00-8:30, and after takt time adjustment it is shortened to 8:00-8:25, then the start time of subsequent task b needs to be synchronously adjusted to 8:25. If the workstation resource for task b is workstation W1, then the resource allocation time period for workstation W1 needs to be re-corresponded to 8:25-8:55. In the first step of data integration for all tasks, for tasks with resource conflicts, the order or time period needs to be adjusted by comparing the takt adjustment range of adjacent tasks and combining it with the minimum task interval threshold allowed by the production takt. For example, if task c and task d overlap due to takt adjustment, and the takt adjustment range is task c minus 5 minutes and task d minus 3 minutes, it is necessary to determine whether the time meets the minimum task interval threshold of 5 minutes. If not, the takt adjustment range of task c is redistributed to minus 3 minutes, while task d remains unchanged, thereby avoiding conflicts. In the actual example, if the takt coordination correction data adjusts the task interval as follows: task 1 interval adjusted by 2 minutes, task 2 interval adjusted by 3 minutes, and task 3 interval adjusted by 1 minute, then the integrated scheduled task data is as follows: task 1 execution time 8:00-8:28, task 2 execution time 8:28-8:57, task 3 execution time 8:57-9:26, generating integrated scheduled task data.

[0103] The execution time update submodule updates the execution time of each production task based on the scheduling task integration data, adjusts the start and end times of each task according to the task order and rhythm coordination correction data, and obtains the task execution time configuration data.

[0104] Based on the integrated scheduling task data, the execution time period for each production task is updated. First, the integrated time period information and workstation number corresponding to each task are obtained from the integrated scheduling task data. Continuity checks are performed on the task time periods to ensure that the end time of each task corresponds to the start time of the next task. Then, the adjustment range in the takt time adjustment data is called to calculate the final start and end times of each task, resulting in the updated time period configuration value. In practical application, if the integrated scheduling task data shows that task 4's time period is 9:00-9:30 and task 5's time period is 9:30-10:00, and the takt time adjustment range for task 4 is -2 minutes, then... After the update, the time period for Task 4 is 9:00-9:28, and the start time for Task 5 is updated to 9:28. Further, the updated execution time period configuration values ​​are aggregated to generate a set of four-tuples containing the task number, workstation number, start time, and end time. This set can be used for subsequent resource scheduling and task scheduling. In a specific example, if a production task was originally scheduled for 10:00-10:30, with workstation W2, and after the cycle time adjustment is corrected to 9:55-10:25, then the task execution time period configuration value is: Task Number: z1, Workstation Number: W2, Start Time: 9:55, End Time: 10:25. The task execution time period configuration data is then obtained.

[0105] The resource table generation submodule, based on the task execution time period configuration data, combined with the workstation number and resource allocation requirements of each task, establishes a globally optimized scheduling timetable and resource allocation table, generating dynamic optimization results for the production schedule;

[0106] Based on the task execution time period configuration data, combined with the workstation number and resource allocation requirements of each task, the task entries in the task execution time period configuration value are first sorted according to the task order. The workstation number, start time, and end time of each task are extracted in sequence. By referring to the workstation resource list, the available time period and resource occupancy of each workstation are determined. The task execution time period configuration value is then filled into the scheduling table, and the resource allocation table is filled in simultaneously. In the case of resource conflicts, it is determined whether there is an intersection between the task interval and the available time of the workstation resource. If there is an intersection, the conflict is marked in the resource allocation table. In a real scenario, if workstation W3 has been assigned task z2 from 10:00 to 11:00, and a new task z3 is assigned to W3 with an execution time of 10:30 to 11:00, then a resource conflict needs to be marked. The solution is to adjust the execution time of task z3 or change the workstation. After completing the resource scheduling, a scheduling table and a resource allocation table containing task number, workstation number, execution time period, and resource status are generated, and the dynamic optimization result of the production scheduling is generated.

[0107] Table 6: Production Task Scheduling Time Configuration Table

[0108] Task Number Workstation Number Start time End time Tempo Adjustment (minutes) z1 W1 8：00 8：25 -5 z2 W2 8：25 8：50 -5 z3 W3 8：50 9：20 -10

[0109] Table 6 lists the start and end times of production tasks and the corresponding cycle time adjustment ranges. All tasks are arranged in the order of scheduling optimization.

[0110] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments that can be applied to other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A production scheduling optimization system based on data analysis, characterized in that, The system includes: The process data trend analysis module obtains continuous process duration records, constructs a duration increase sequence, and performs a sign consistency judgment. If all signs are positive, it calculates the deviation between the latest process duration and the current schedule cycle number. If the deviation is greater than the cycle time deviation threshold, it records and generates a process cycle adjustment parameter set. The real-time workstation resource acquisition module collects the task completion time of each workstation, records the start time of workstation idle time, retrieves the target workstation data of the current scheduled task, performs task-workstation-idle time matching, and generates a real-time workstation idle dataset. The scheduling order optimization judgment module compares the start time of the target workstations for the current task and the next task with the real-time workstation idle time based on the workstation idle data set, determines whether there is a need to swap task order, performs adjustment operations to update the scheduling task list, and generates scheduling order optimization results. The cycle time coordination test module, based on the scheduling order optimization results and the process cycle time adjustment parameter set, judges the ratio of the cycle time of adjacent tasks in the scheduling task list. If the ratio of adjacent task cycle time is greater than the tolerance of adjacent cycle time ratio, the task interval time is shortened and cycle time coordination correction data is generated. The process data trend analysis module includes: The duration difference extraction submodule is based on the acquisition of continuous process duration record data, arranged in chronological order, and performs difference calculation on the duration of any adjacent process to obtain the first-order difference sequence of adjacent process times. It then calculates the absolute value of all differences in the duration difference sequence to obtain the set of absolute values ​​of the duration difference sequence. The sign consistency judgment submodule performs sign consistency judgment on any continuous difference in the sequence based on the absolute value set of the duration difference sequence, filters the difference combination interval where all signs are positive, extracts the latest process duration and the current scheduling cycle number based on the filtered interval, and calculates the process duration cycle deviation value. The cycle time adjustment parameter generation submodule compares the cycle time deviation value of the process duration with the cycle time deviation threshold. If the deviation is greater than the cycle time deviation threshold, it extracts the current schedule cycle number, the latest process duration, and the cycle time deviation threshold, calculates the cycle time adjustment benchmark value, and establishes a process cycle time adjustment parameter set.

2. The production scheduling optimization system based on data analysis according to claim 1, characterized in that, The process cycle time adjustment parameter set includes a duration increase judgment value, cycle time deviation, and cycle time adjustment suggestion value. The workstation idle real-time dataset includes workstation number, idle start time, and task matching degree. The scheduling order optimization result includes task swap record, updated task order list, and adjusted scheduling priority. The cycle time coordination correction data includes cycle time ratio correction value, task interval adjustment amount, and cycle time coordination identifier.

3. The production scheduling optimization system based on data analysis according to claim 1, characterized in that, The real-time workstation resource acquisition module includes: The task completion collection submodule collects the task completion time of each workstation and records the system clock time when the corresponding task at each workstation ends, generating a set of workstation task completion times. The idle time recording submodule, based on the workstation task completion time set, records the system clock time after the task is completed for each workstation as the idle start time, and establishes the workstation idle start time value by corresponding it with the workstation identifier. The workstation data matching submodule performs task and workstation matching based on the workstation idle start time value and the target workstation data of the current schedule task. It combines the target workstation number and workstation idle start time of each task, integrates the task number, workstation number, and idle start time, and obtains a real-time dataset of workstation idle time.

4. The production scheduling optimization system based on data analysis according to claim 1, characterized in that, The scheduling order optimization judgment module includes: The idle time comparison submodule obtains the start time of the target workstation idle time for the current task and the next task based on the real-time workstation idle data set. It then compares the execution times of the target workstation idle time for the current task and the target workstation idle time for the next task to obtain the difference in workstation idle time. The sequence swap determination submodule determines whether the start time of the next task's workstation is earlier than the start time of the current task's workstation based on the difference in the workstation's idle time. If the determination result is yes, the sequence swap requirement between the current task and the next task is recorded, and the task sequence swap flag is obtained. The scheduling list update submodule performs an order adjustment operation on the scheduled task list according to the task order swap markers, updates the scheduled task list after adjusting the task order, and generates the scheduling order optimization result.

5. The production scheduling optimization system based on data analysis according to claim 1, characterized in that, The rhythm coordination test module includes: The cycle ratio judgment submodule extracts the cycle number of adjacent tasks in the scheduling task list based on the scheduling order optimization result and the process cycle adjustment parameter set, performs a ratio relationship judgment on the cycle number of adjacent tasks, calls the adjacent cycle ratio tolerance, calculates the cycle ratio of adjacent tasks, compares it with the cycle ratio tolerance, and obtains the cycle ratio exceeding the limit flag. The task interval adjustment submodule shortens the task interval time between adjacent tasks marked as exceeding the limit based on the beat ratio exceeding the limit, and records the task interval adjustment value after shortening the task interval time. The configuration record generation submodule adjusts the value according to the task interval, integrates the adjusted task interval time and corresponding task number, establishes a complete task interval configuration record, and generates beat coordination correction data.

6. The production scheduling optimization system based on data analysis according to claim 1, characterized in that, The system also includes: The scheduling dynamic optimization module integrates the scheduling task data based on the cycle time coordination correction data, updates the execution time and corresponding workstation configuration of each production task, constructs a globally optimized scheduling timetable and resource allocation table, and generates dynamic optimization results for production scheduling. The results of the dynamic optimization of the production scheduling include the optimized scheduling timetable, resource allocation table, and task execution time configuration.

7. The production scheduling optimization system based on data analysis according to claim 6, characterized in that, The scheduling dynamic optimization module includes: The task data integration submodule obtains the takt time coordination correction data, performs an overall integration operation on the scheduled task data, extracts the takt time information, workstation configuration and adjusted task interval corresponding to each task, integrates them into a unified set of scheduled tasks, and generates integrated scheduled task data. The execution time update submodule updates the execution time of each production task based on the integrated data of the scheduling tasks, adjusts the start and end times of each task according to the task order and rhythm coordination correction data, and obtains the task execution time configuration data. The resource table generation submodule, based on the task execution time period configuration data and combined with the workstation number and resource allocation requirements of each task, establishes a globally optimized scheduling timetable and resource allocation table, generating dynamic optimization results for the production schedule.

Images