Batch start and end time correction method, device and storage medium based on yield data

Abstract

The application discloses a batch start and end time correction method and device based on yield data and a storage medium, relates to the technical field of industrial data processing correction, and solves the problem of batch data analysis distortion caused by the advanced or lagged batch start and end time in the production and manufacturing process of an enterprise.

Description

Technical Field

[0001] This invention relates to the field of industrial data processing and correction technology, specifically to a method, device, and storage medium for correcting batch start and end times based on production data. Background Technology

[0002] With the increasing automation of discrete manufacturing sites, industrial data processing technology is gradually being applied to production cycle monitoring, batch traceability management, and process control, involving high-speed assembly equipment, continuous processing lines, and intelligent packaging units. In industrial data processing, batch data and process data are typically correlated based on batch start and end times. However, when the actual production status of the equipment changes asynchronously with the recorded batch time, discrepancies may arise in their time correspondence. Manufacturing Execution Systems (MES) typically record the batch start time based on the operator's execution of the batch start operation and the batch end time based on the execution of the batch end operation, then aggregate the process parameter data generated during equipment operation according to the corresponding time range. In actual production, the equipment may undergo preheating, loading, and acceleration processes before the batch begins, and may undergo production stoppage, deceleration, and cleaning processes before and after the batch ends. The batch start and end times recorded by the MES are not necessarily consistent with the actual start and end times of product production. This may result in the process data aggregated according to the batch time range including data from non-actual production stages or omitting data from actual production stages. In existing MES production execution processes, batch work order issuance and status updates typically require integration with operator operation records. After an operator executes the batch start operation, the equipment may still be in preheating, loading, or operational preparation states; even after the equipment enters actual production, the batch start information may be later than the actual production start time due to differences in operation time. Similarly, at the batch end stage, there may be discrepancies between the MES recorded time and the actual equipment stop production time. Therefore, the batch start and end times stored in the industrial data processing system may have a time offset from the actual equipment production cycle. Furthermore, during continuous equipment operation, the entry time of material batch information, process parameter information, and quality inspection information may also be asynchronous with the occurrence time of the corresponding production events, causing time correspondence discrepancies when associating industrial data from different sources according to batch time, thus affecting production cycle analysis and production process traceability based on batch time ranges.

[0003] For example, the invention patent with announcement number CN110046777B provides a method and apparatus for continuous reconfiguration scheduling in flexible workshops. The method includes: determining multiple process-level batches of all orders within the flexible workshop; selecting currently process-level batches that can be processed from the determined multiple process-level batches; using heuristic rules based on equipment bottleneck levels to sequentially arrange equipment for the currently processable multiple process-level batches to obtain a preliminary production schedule; dividing the preliminary production schedule into multiple time intervals and using a clustering algorithm to sequentially obtain virtual unit candidate schemes for each time interval; using an applicability genetic algorithm to optimize and obtain a virtual unit construction scheme for each time interval; and revising the preliminary production schedule based on the virtual unit construction scheme to obtain a final production schedule. This scheme can assist schedulers in continuously reconfiguring and scheduling flexible workshops with multiple product types and small batches, ensuring that the flexible workshop simultaneously considers production efficiency and production flexibility.

[0004] For example, the invention patent with announcement number CN116128221B discloses a scheduling method for aero-engine blade remanufacturing production line based on digital twins, including: constructing a digital twin model of the remanufacturing workshop and historical twin data based on physical remanufacturing workshop information, and obtaining several sets of damaged blade information; constructing an optimization model for cross-batch workshop scheduling based on the profit margin of damaged blade remanufacturing based on the several sets of damaged blade information, and performing optimization and solving to obtain a cross-batch scheduling scheme based on the profit margin of damaged blade remanufacturing; constructing a workshop dynamic scheduling model based on real-time process completion time based on real-time sensors, and regenerating a workshop dynamic scheduling scheme based on the profit margin model and real-time process completion time; correcting the digital twin simulation working time of the processing trajectory of several processes based on the workshop dynamic scheduling model of real-time process completion time, performing workshop scheduling planning for the next batch of damaged blades, and completing the scheduling of the damaged blade remanufacturing production line.

[0005] Therefore, there is an urgent need for batch start and end time correction methods, equipment, and storage media based on production data, capable of automatically and accurately correcting batch start and end times to correlate batches with equipment process parameters. This method should effectively eliminate timing errors introduced by human operation, ensuring a high degree of consistency between production data and the physical production process, thereby laying a solid data foundation for building a transparent, reliable, and efficient digital workshop. Summary of the Invention

[0006] Technical problems to be solved

[0007] To address the shortcomings of existing technologies, this invention provides a method, device, and storage medium for correcting batch start and end times based on production data, which solves the problem of batch data distortion caused by batch start and end times being ahead or behind during enterprise production and manufacturing processes.

[0008] Technical solution

[0009] To achieve the above objectives, the present invention provides the following technical solution: a batch start and end time correction method based on production data, comprising: S1, collecting production data of the target production equipment based on a preset sampling period and storing the production data in a production data storage table; S2, constructing a production change comparison table based on the production data storage table, dividing production batches, and determining the batch start time and batch end time based on the production data records of the same batch, and excluding invalid candidate batches; S3, removing test batches based on the minimum production time of the batch, determining the best start candidate and the best end candidate through time deviation, and obtaining the corrected batch start time and the corrected batch end time.

[0010] Further, the specific steps for collecting output data of the target production equipment based on a preset sampling period and storing the output data in the output data storage table are as follows: Output data of the target production equipment, including machine tools, assembly equipment, and packaging machinery in the production line, is collected through a preset sampling period; the output data is read from real-time output records by interfacing with the production management platform of the target production equipment, or automatically collected by counting sensors installed on the production line, including photoelectric sensors and encoders; the output data is the cumulative output value of the target production equipment at the sampling time, representing the number of products completed in the current production batch; the sampling period is used as the time interval for writing output data into the output data storage table, with each sampling time corresponding to one output data record; the output data storage table includes a timestamp column, an output value column, a description column, and a quality column, with the timestamp column set as the primary key, and the quality column used to mark normal collection, supplementary collection, missing, or abnormal collection status.

[0011] Further, the specific steps for constructing a production change comparison table based on the production data storage table are as follows: Using the sampling period as a unit, traverse the production data in ascending order by timestamp, associate the production value of the current sampling period, the production value of the previous sampling period, and the production value of the next sampling period to form an adjacent production change relationship based on the production data record corresponding to the current sampling time, and construct a production change comparison table based on the adjacent production change relationship; the production change comparison table includes the sampling period, the production of the current sampling period, the production of the previous sampling period, and the production of the next sampling period, used to identify the time positions of production clearing, production initiation, and production reaching the batch production output; for the first production data record, the production of the previous sampling period is recorded as null, and the first production data record does not participate in the production clearing determination; for the last production data record, the production of the next sampling period is recorded as null, and the last production data record does not participate in the production initiation determination; adjacent production change relationships with null values ​​are not used as the basis for determining the corresponding time position.

[0012] Further, the specific steps for dividing production batches are as follows: A zeroing marker column and a batch number column are added to the production change comparison table to form a production batch division table; when the production at the current sampling time is zero, and the production at the previous sampling time is not zero, the current sampling time is marked as a zeroing point, and the zeroing marker column is set to one; otherwise, the zeroing marker column is set to zero; the sampling time corresponding to the zeroing point is recorded as the production zeroing time; the current batch number is initialized to one, and starting from the first production data record, batch numbers are assigned to each production data record according to the order of sampling times. Production data records before the first zeroing point are all assigned to the initial batch number; when a zeroing point record is identified, the current batch number is incremented first, and then the zeroing point record and the production data records after the zeroing point record and before the next zeroing point record are assigned to the incremented batch number; when the production value at two or more consecutive sampling times is zero, the first zeroing point in the continuous zero-value segment is used as the boundary record of the next production batch, and the remaining zero-value records in the continuous zero-value segment are assigned to the next production batch number with the same number as the first zeroing point.

[0013] Further, the specific steps for determining the batch start time and batch end time based on the production data records of the same batch, and excluding invalid candidate batches, are as follows: In the production data records corresponding to the same batch number, if the production value at the current sampling time is zero, and the production value at the adjacent subsequent sampling time is not zero, mark the subsequent sampling time as the batch start time; obtain the maximum production value in the production data records corresponding to the same batch number, and use the maximum production value as the production output of the current batch; in the production data records corresponding to the same batch number, perform missing value imputation, abnormal jump removal, and adjacent sampling smoothing processing on the production value to obtain the corrected production value; when the previous sampling value appears for the first time... If the corrected output value at the current sampling time is less than the batch production output, and the absolute value of the difference between the corrected output value at the current sampling time and the batch production output is less than or equal to the output tolerance, and remains within the output tolerance range for the next N consecutive sampling windows, then the current sampling time is marked as the batch end time. The batch production duration is obtained by subtracting the batch start time from the batch end time. If the batch start time or batch end time is not identified in the output data record corresponding to the same batch number, or if the batch end time is earlier than the batch start time, then the corresponding batch number is marked as an invalid candidate batch. Invalid candidate batches are not included in the subsequent start and end time deviation comparison.

[0014] Further, the specific steps for eliminating test batches based on the minimum batch production duration are as follows: Read the original batch start time and original batch end time of the batch to be corrected and obtain the original batch duration. The original batch start time and original batch end time are obtained from the production management platform, work order records, or batch record table. The original batch duration is the duration obtained by subtracting the original batch start time from the original batch end time. Summarize the output zeroing time, batch start time, batch end time, batch production duration, and batch production output of each non-invalid candidate batch into a batch candidate record. Set a minimum batch production duration. Determine the batch candidate records whose batch production duration is lower than the minimum batch production duration as test batches in the production process and eliminate them from the batch candidate records to obtain normal production batch candidate records. The minimum batch production duration is determined based on the product process cycle time of the target production equipment, and the unit of the minimum batch production duration is consistent with the timestamp unit.

[0015] Further, the specific steps for determining the optimal start candidate and optimal end candidate based on time deviation are as follows: The normal production batch candidate record includes the candidate batch number, candidate start time, candidate end time, candidate batch production output, and candidate batch production duration; based on the original batch start time and original batch end time, calculate the start time deviation and end time deviation between each normal production batch candidate record and the batch requiring correction; the start time deviation is the absolute value of the difference between the candidate start time and the original batch start time, and the end time deviation is the absolute value of the difference between the candidate end time and the original batch end time; the normal production batch candidate record with the smallest start time deviation is determined as the optimal start candidate. The candidate record of the normal production batch with the smallest deviation from the end time is determined as the best end candidate; when there are two or more candidate records of normal production batches with the same minimum start time deviation, the candidate record of the normal production batch with the smallest absolute difference between the batch production duration and the original batch duration of the batch that needs to be corrected is selected as the best start candidate; when there are two or more candidate records of normal production batches with the same minimum end time deviation, the candidate record of the normal production batch with the smallest absolute difference between the batch production duration and the original batch duration of the batch that needs to be corrected is selected as the best end candidate; if it is still impossible to distinguish, the candidate record of the normal production batch with the earliest start time is selected.

[0016] Further, the specific steps for obtaining the corrected batch start time and corrected batch end time are as follows: when the best start candidate and the best end candidate point to the same candidate batch number, the normal production batch candidate record corresponding to the candidate batch number is determined as the target corrected batch, and the candidate start time and candidate end time of the target corrected batch are used as the corrected batch start time and corrected batch end time of the batch to be corrected, respectively; when the best start candidate and the best end candidate point to different candidate batch numbers, a batch matching anomaly is determined to have occurred, a batch matching anomaly record is generated, and an alarm prompt is output; the batch matching anomaly record includes the batch number to be corrected, the candidate batch number corresponding to the best start candidate, the candidate batch number corresponding to the best end candidate, the start time deviation, the end time deviation, and the alarm generation time.

[0017] A second aspect of the present invention provides a batch start and end time correction device based on production data, comprising: at least one processor and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform any of the batch start and end time correction methods based on production data.

[0018] A third aspect of the present invention provides a batch start and end time correction storage medium based on production data, comprising: a storage medium storing computer-executable instructions, wherein the computer-executable instructions are configured to implement the batch start and end time correction method based on production data as described in any one of the claims.

[0019] Beneficial effects

[0020] The present invention has the following beneficial effects:

[0021] (1) The batch start and end time correction method, equipment and storage medium based on production data corrects the original batch start time and original batch end time by using equipment production data as the basis for batch time correction, thereby reducing the impact of the inconsistency between manual operation time and actual equipment production time on batch time recording and improving the accuracy of the correspondence between batch start and end time and actual production process.

[0022] (2) The batch start and end time correction method, equipment and storage medium based on production data, by associating process parameters with the corrected time window that accurately reflects the actual production cycle, ensures that the traceable data is the actual process data experienced by the batch, and provides a reliable basis for quality traceability.

[0023] (3) The batch start and end time correction method, equipment and storage medium based on production data can correct the time and associate the data without human intervention, which improves efficiency, ensures data consistency and provides a data foundation for realizing "lights-out factories" and unmanned decision-making.

[0024] (4) The batch start and end time correction method, equipment and storage medium based on production data do not require modification of existing hardware equipment, make full use of the existing production and process parameter data in the workshop, realize value through software logic and algorithm innovation, and are easy to promote and apply.

[0025] Of course, any product implementing this invention does not necessarily need to achieve all of the advantages described above at the same time. Attached Figure Description

[0026] Figure 1 This is a flowchart of the batch start and end time correction method based on production data according to the present invention;

[0027] Figure 2 This is a batch matching consistency determination diagram related to the present invention;

[0028] Figure 3 This is a schematic diagram of the batch start and end time correction method based on production data involved in this invention. Detailed Implementation

[0029] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0030] Please see Figures 1-3 This invention provides a technical solution: a batch start and end time correction method, device, and storage medium based on production data, comprising: S1, collecting production data of the target production equipment based on a preset sampling period and storing the production data in a production data storage table; S2, constructing a production change comparison table based on the production data storage table, dividing production batches, and determining the batch start time and batch end time based on the production data records of the same batch, excluding invalid candidate batches; S3, removing test batches based on the minimum production time of the batch, determining the best start candidate and the best end candidate through time deviation, and obtaining the corrected batch start time and the corrected batch end time.

[0031] Specifically, the steps for collecting output data of the target production equipment based on a preset sampling period and storing the output data in the output data storage table are as follows: The output data of the target production equipment is collected through a preset sampling period. The target production equipment refers to specific production equipment that needs to be corrected for batch start time and end time calculation, including processing machine tools, assembly equipment and packaging machinery in the production line; the sampling period is kept at no less than 1 second as the time base for continuous output recording. Each time a sampling moment is reached, an output reading is triggered, and the sampling moment, output value and collection status are synchronously written into the subsequent output data storage table.

[0032] Production data is obtained by connecting to the production management platform of the target production equipment to read real-time production records, or by automatically collecting data through counting sensors installed on the production line, including photoelectric sensors and encoders. The production management platform includes a manufacturing execution platform, an equipment data acquisition platform, a production work order record platform, or a data platform that can provide real-time production records of the equipment. The connection methods include database view reading, data interface calling, equipment communication gateway forwarding, or platform log table reading. The real-time production record includes at least the equipment identifier, sampling time, cumulative production value, work order reference information or batch reference information, and acquisition status. The sampling time is mapped to the timestamp column of the production data storage table, the cumulative production value is mapped to the production value column, the equipment identifier, work order reference information, batch reference information, and data source are written into the description column, and the acquisition status is written into the quality column.

[0033] The production data is the cumulative production value of the target production equipment at the sampling time. The cumulative production value represents the number of products completed in the current production batch. When the production data is provided by the production management platform, the cumulative production field in the platform record is read directly. When the production data is provided by the counting sensor, the cumulative production value is formed by accumulating the number of times the sensor is triggered within the sampling period. After the production is identified as zero, the accumulation starts from zero again to ensure that the production data can reflect the actual number completed in the current production batch.

[0034] The sampling period is used as the time interval for writing production data into the production data storage table. Each sampling time corresponds to one production data record. The production data storage table includes a timestamp column, a production value column, a description column, and a quality column. The timestamp column is set as the primary key, and the quality column is used to mark the normal collection, supplementary collection, missing, or abnormal collection status. Normal acquisition status indicates that the output value is obtained on time within the current sampling period, and the output value is neither empty nor negative, the sampling time is continuous, and the output change conforms to the cumulative output recording rules. Supplementary acquisition status indicates that the output value at the current sampling time was not written in real time within the sampling period, but was supplemented after a traceable output record at the same or adjacent sampling time was read. The supplementary acquisition record retains the original sampling time and is marked as supplementary acquisition in the quality column. Missing status indicates that the output value was not read at the current sampling time, and the corresponding output record cannot be obtained within the allowed supplementary acquisition time range. The output value corresponding to this sampling time is recorded as empty and marked as missing in the quality column. Abnormal acquisition status indicates that the output value is not empty but does not meet the cumulative output recording rules, including output value is negative, output suddenly increases beyond the theoretical maximum output range of the equipment at adjacent sampling times, output decreases in the opposite direction without zeroing, or duplicate output records exist at the same timestamp.

[0035] For production data records marked as "normally collected" in the quality column, they directly participate in subsequent production change comparisons, production batch division, and batch start and end time determination. For production data records marked as "supplemented" in the quality column, they participate in subsequent calculations when the supplemented sampling timestamp can be aligned with the sampling period and the production value meets the cumulative production record rules. For production data records marked as "missing" in the quality column, the time position is first retained when constructing the production change comparison table, and then supplemented based on the normal production records of adjacent sampling times. If supplementation is not possible, it will not be used as the basis for determining production zeroing, production start, and batch end. For production data records marked as "abnormally collected" in the quality column, abnormal jump rejection and adjacent sampling smoothing are first performed. If they can be restored to meet the cumulative production record rules, the corrected production value is used to participate in subsequent batch determination. Records that cannot be restored are only retained in the production data storage table and do not participate in the batch start time and batch end time determination.

[0036] Table 1 Production Data Storage Table

[0037]

[0038] This implementation plan clarifies the data collection sources, storage formats, and quality status processing rules for production data, ensuring that the target production equipment generates traceable, calculable, and verifiable production records at each sampling moment. By defining the production management platform interface, counting sensor acquisition methods, timestamp columns, production value columns, description columns, and quality columns, it guarantees that real-time production data from different sources can be uniformly written into the production data storage table. By distinguishing between normal acquisition, supplementary acquisition, missing data, and abnormal acquisition states, and specifying the processing methods for participating in calculations, supplementing, removing, or retaining data respectively, it avoids erroneous production records from affecting subsequent production change comparisons, production batch division, batch start time determination, and batch end time determination, thereby providing a stable and reliable data foundation for subsequent start and end time corrections.

[0039] Specifically, the steps for constructing a production change comparison table based on the production data storage table are as follows: Using the sampling period as a unit, traverse the production data in ascending order of timestamps, treating each production data record as the current sampling time record, and reading the production value corresponding to the current sampling time, the production value corresponding to the previous sampling time, and the production value corresponding to the next sampling time respectively; use the adjacency relationship of timestamps to stagger the association of the three types of production values, forming an adjacent production change relationship based on the production data record corresponding to the current sampling time, and constructing a production change comparison table based on the adjacent production change relationship. The production change comparison table is used to simultaneously store the current production status, the previous sampling production status, and the next sampling production status under the same time base, enabling subsequent determination of whether the production has been reset to zero, whether it has entered the production state from zero, and whether it has reached the maximum production value corresponding to the same batch number by analyzing the changes in production before and after.

[0040] The output change comparison table includes the sampling period, the output of the current sampling period, the output of the previous sampling period, and the output of the next sampling period. The output of the current sampling period represents the cumulative output value at the current sampling moment. The output of the previous sampling period is compared with the output of the current sampling period to identify the output reset point. The output of the next sampling period is compared with the output of the current sampling period to identify the output start point. After the production batches are subsequently divided, the maximum output value corresponding to the same batch number is defined as the batch production output. Based on the output change comparison table, the time position when the output reaches the batch production output is identified, thus providing a data basis for determining the batch end time.

[0041] For the first production data record, since there is no production data record corresponding to the previous sampling time, the production of the previous sampling period is recorded as null, and the first production data record does not participate in the production zeroing determination. For the last production data record, since there is no production data record corresponding to the next sampling time, the production of the next sampling period is recorded as null, and the last production data record does not participate in the production start determination. The adjacent production change relationship with null values ​​is only retained in the production change comparison table to maintain the integrity of the time series, and is not used as the basis for determining production zeroing, production start, or production reaching the batch production output. For production data records marked as missing in the quality column and unable to be filled, they are used as placeholders in the time of production change comparison table, but are not used as the basis for determining the corresponding time position. For production data records marked as abnormal acquisition and unrecoverable in the quality column, the current sampling time position is retained, and they are not used in the effective determination of adjacent production change relationships. For normally acquired records and records that can be filled or smoothed, the corresponding production value or the corrected production value is used to participate in the construction of the production change comparison table.

[0042] Table 2 Comparison of Production Changes

[0043]

[0044] In this implementation plan, single-point production records are converted into production change comparison data with preceding and following reference relationships. This ensures that each sampling moment not only retains the current production value but also reflects the production changes at the previous and subsequent sampling moments, thus providing a unified data foundation for subsequently identifying the production zeroing position, production start position, and batch end position. At the same time, by setting rules to exclude the first record, last record, missing record, and abnormal record from corresponding judgments, the plan avoids boundary data or abnormal data from accidentally triggering zeroing, start, and end judgments, thereby improving the accuracy and stability of batch division and start / end time correction.

[0045] Specifically, the steps for production batch division are as follows: A zeroing marker column and a batch number column are added to the production change comparison table to form a production batch division table. The production batch division table uses the sampling period, current sampling period output, previous sampling period output, and next sampling period output from the production change comparison table. The zeroing marker column is used as the batch boundary identification field, and the batch number column is used as the attribution field for production data records within the same production batch. Production data records in the production change comparison table are read one by one in ascending order of timestamp. When the output at the current sampling time is zero, and the output at the previous sampling time is not zero, it indicates that the output has dropped from non-zero to zero at the current sampling time. The current sampling time is marked as the zeroing point, and the zeroing marker column is set to one. When the current sampling time does not meet the aforementioned conditions, the zeroing marker column is set to zero. The sampling time corresponding to the zeroing point is recorded as the production zeroing time, used to characterize the time position when the cumulative output after the end of the previous production batch is returned to zero, and serves as the boundary basis for the division of the next production batch.

[0046] The current batch number is initialized to one. Starting from the first production data record, batch numbers are assigned to each production data record according to the sampling time sequence. Before the first zeroing point is identified, all production data records are assigned to the initial batch number to avoid production data records before the first zeroing point being unable to be assigned due to the lack of a preceding batch boundary. When a zeroing point record is identified, the current batch number is incremented first, and then the zeroing point record is assigned to the incremented batch number. From the zeroing point record onwards, until the next zeroing point record, all production data records continue to use the incremented batch number. In this way, the zeroing point record is no longer assigned to the previous completed production batch, but serves as the boundary record of the next production batch, maintaining the same batch number as the production start record, production growth record, and production completion record in the next production batch. This ensures that the batch boundary, batch start time, and batch end time can be calculated subsequently under the same batch number.

[0047] When the output value is zero at two or more consecutive sampling times, boundary assignment processing is performed on the continuous zero-value segment. If the first zero-value record in the continuous zero-value segment meets the zeroing point judgment condition that the output at the current sampling time is zero and the output at the previous sampling time is not zero, then the first zero-value record is regarded as the boundary record of the next production batch, and its zeroing mark is set to one. The remaining zero-value records in the continuous zero-value segment are not marked as new zeroing points, their zeroing marks are set to zero, and they are assigned to the next production batch number with the same number as the first zeroing point, indicating that this segment of zero values ​​belongs to the waiting period after zeroing and before production has started. This processing method can avoid the continuous zero-value segment being incorrectly split into multiple production batches, and can also ensure that the next production batch has a clear batch number assignment before the output actually jumps from zero to non-zero.

[0048] Table 3 Production Batch Classification Table

[0049]

[0050] In this implementation plan, the batch boundaries and data attribution relationships are further clarified in the production change comparison table. The batch switching position where the production returns from non-zero to zero is identified by the zeroing mark. The production records after the same zeroing point and before the next zeroing point are grouped into the same production batch by the batch number, thereby avoiding the repeated division of continuous zero value segments and providing a clear data range for determining the batch start time, batch end time, batch production output and batch production duration within the same batch number.

[0051] Specifically, the steps for determining the batch start and end times based on the production data records of the same batch and excluding invalid candidate batches are as follows: In the production data records corresponding to the same batch number, read the production value at the current sampling time and the production value at the next adjacent sampling time in chronological order. When the production value at the current sampling time is zero and the production value at the next adjacent sampling time is not zero, it indicates that the production has transitioned from zero to production between the current sampling time and the next sampling time. The next sampling time is then marked as the batch start time obtained based on the corrected sampling data. If the counting sensor trigger record, the production management platform production change record, or the equipment communication gateway log contains the first increase time of production between the current sampling time and the next sampling time, then the next sampling time is corrected using this first increase time of production as the batch start time. If there is no more granular first increase time of production, then the next sampling time is used as the batch start time, with the time resolution of the batch start time limited by the sampling period.

[0052] Before obtaining the production output of the current batch, a quality check is performed on the production data records corresponding to the same batch number. The quality check includes missing value imputation, duplicate sampling merging, abnormal jump rejection, production rollback correction, and adjacent sampling smoothing. Among them, missing value imputation is used to fill in the short-term missing sampling positions according to the adjacent valid production records. Duplicate sampling merging is used to retain the records with normal quality status and the writing time closest to the sampling time among multiple production records under the same timestamp. Abnormal jump rejection is used to exclude records with sudden increases in output exceeding the theoretical output capacity of the target production equipment. Production rollback correction is used to identify the production decrease records without a zeroing point as abnormal rollback and correct them according to the previous valid production status. Adjacent sampling smoothing is used to reduce the impact of supplementary sampling, jitter, or sensor errors on the cumulative production curve, resulting in a corrected production value sequence.

[0053] Provided that the cumulative output within the current batch does not decrease after the corrected output value sequence, the maximum corrected output value corresponding to the same batch number is obtained, and this maximum corrected output value is used as the production output of the current batch. If there are unrecoverable missing records, abnormal jump records, or production rollback records within the same batch number, the corresponding records are not included in the determination of the production output of the current batch, and a quality status marker is retained in the candidate batch records. The production output of the current batch is used to represent the number of products completed within the current batch and serves as the target output benchmark for determining the end time of subsequent batches.

[0054] In the production data records corresponding to the same batch number, when the corrected production value at the previous sampling time is less than the current batch production output for the first time, and the absolute value of the difference between the corrected production value at the current sampling time and the current batch production output is less than or equal to the production tolerance, and remains within the production tolerance range for the subsequent N consecutive sampling windows, the current sampling time is marked as the batch end time obtained based on the sampled data correction. Here, N represents the number of consecutive sampling windows used to confirm that the production output has stably reached the current batch production output. N is a positive integer, determined according to the product process cycle of the target production equipment and the sampling period of the production data, ensuring that the observation duration corresponding to the N consecutive sampling windows is not less than one product process cycle. If the counting sensor trigger record, the production management platform production change record, or the equipment communication gateway log contains a time when the current batch production output is first reached between the previous sampling time and the current sampling time, then the current sampling time is corrected using this first time to reach the current batch production output as the batch end time. If no finer-grained first time is found, the current sampling time is used as the batch end time. Subtracting the batch start time from the batch end time yields the batch production duration.

[0055] If the production data record corresponding to the same batch number does not identify the batch start time, or does not identify the batch end time, or the batch end time is earlier than the batch start time, or the valid production record used to determine the production output of the current batch is insufficient to support the batch end time determination, the corresponding batch number will be marked as an invalid candidate batch. Invalid candidate batches will only retain the batch number, abnormal reason and quality status mark, and will not be included in the subsequent start and end time deviation comparison.

[0056] In this implementation plan, within the same batch number range, the batch start time, batch end time, batch production output, and batch production duration are determined from the production data after quality verification. By correcting or eliminating missing, duplicate, abnormal jump, and production rollback data, it is ensured that the batch production output comes from stable and reliable cumulative production records. By using the time when the output enters a non-zero state from zero as the start criterion and the time when the output first stably reaches the batch production output as the end criterion, misjudgment of start and end times due to single-point errors is avoided. At the same time, batches with missing start and end times, abnormal time sequences, or insufficient effective production records are marked as invalid candidates, preventing abnormal batches from entering subsequent time deviation comparisons, thereby improving the accuracy and reliability of batch start and end time correction.

[0057] Specifically, the steps for eliminating test batches based on the minimum production duration are as follows: First, read the original batch start time and end time of the batch to be corrected, and obtain the original batch duration. The original batch start time and end time are obtained from the production management platform, work order records, or batch record table. During reading, match the batch number to be corrected, the target production equipment number, and the batch record time range to ensure that the read original batch start time and end time correspond to the same target production equipment and the same production batch. When batch time records exist simultaneously in the production management platform, work order records, and batch record table, verify the consistency of the equipment number, batch number, and timestamp format in each source record, and uniformly convert the timestamps to the same time unit and time base as the production data storage table. The original batch duration is the duration obtained by subtracting the original batch start time from the original batch end time, and is used for subsequent comparisons of the similarity between the production duration of the batch and that of candidate batches.

[0058] The zeroing time, start time, end time, duration, and output of each non-invalid candidate batch are summarized into a batch candidate record. Non-invalid candidate batches are those for which the start and end times have been identified, and whose end time is no earlier than the start time. Batch candidate records are generated sequentially by batch number, retaining corresponding output quality status markers to facilitate differentiation of whether a candidate batch includes supplementary, missing, or corrected output records. After generation, batch candidate records are sorted in ascending order of start time, serving as the data basis for subsequent batch elimination and time deviation comparison.

[0059] A minimum batch production duration is set. Batch candidate records with a production duration shorter than the minimum duration are identified as test batches and removed from the batch candidate records, resulting in normal production batch candidate records. Test batches include segments of batches that have not formed a complete production process after trial operation, jogging, debugging, short-time counting verification, or zeroing. Although these batches may have output variations, their production duration is insufficient to represent the target production equipment completing a normal batch production process, and therefore they are not included in subsequent batch start and end time corrections. The minimum batch production duration is determined based on the product process cycle time of the target production equipment. The product process cycle time can be derived from the target production equipment's process parameter table, work order process route, standard cycle time record corresponding to the product model, or the equipment cycle time field in the production management platform. The unit of the minimum batch production duration is consistent with the timestamp unit, ensuring that the batch production duration and the minimum batch production duration can be directly compared. The normal production batch candidate records retained after removing test batches are used for subsequent start time deviation and end time deviation calculations.

[0060] In this implementation plan, from the candidate batches whose start and end times have been identified, short-term abnormal production segments such as trial runs, jogging, and debugging are further eliminated, and only normal production batch candidate records that can represent the complete production process are retained. At the same time, by reading the start time, end time, and duration of the original batch, a reference basis is provided for comparing the deviation of the start time, the deviation of the end time, and the similarity of the duration, thereby avoiding test batches or abnormal short segments being mistakenly selected as correction targets and improving the accuracy and reliability of batch start and end time correction results.

[0061] Specifically, the steps for determining the optimal start and end candidates based on time deviation are as follows: The normal production batch candidate record is the candidate record retained after eliminating invalid candidate batches and test batches. Before generating the normal production batch candidate record, the product process cycle time, equipment rated cycle time, and minimum qualified output corresponding to the target production equipment are read, and the product process cycle time and equipment rated cycle time are converted into time units consistent with the timestamp. The minimum batch production time is determined based on the cycle time required to complete a single product and the minimum qualified output. When the product process cycle time exists, the first minimum production time is obtained by multiplying the product process cycle time by the minimum qualified output. When the equipment rated cycle time exists, the second minimum production time is obtained by multiplying the equipment rated cycle time by the minimum qualified output. When both the first and second minimum production times exist, the larger of the two is selected as the minimum batch production time. When only one exists, the existing minimum production time is used as the minimum batch production time. Batch candidate records whose batch production duration is less than the minimum batch production duration will not be included in the normal production batch candidate records, thereby avoiding trial runs, jogging, debugging, or short-time counting verification segments from participating in subsequent time deviation comparisons.

[0062] A normal production batch candidate record includes a candidate batch number, a candidate start time, a candidate end time, a candidate batch production output, and a candidate batch production duration. The candidate batch number is the batch number corresponding to this normal production batch candidate record; the candidate start time is the batch start time corresponding to this batch number; the candidate end time is the batch end time corresponding to this batch number; the candidate batch production output is the batch production output corresponding to this batch number; and the candidate batch production duration is the batch production duration corresponding to this batch number. This definition ensures that the candidate start time and candidate end time in the normal production batch candidate record are consistent with the previously identified batch start time and batch end time, avoiding terminological breaks in different processing stages due to the same time meaning.

[0063] Based on the start and end times of the original batch, the start time deviation and end time deviation between each candidate normal production batch and the batch requiring correction are calculated. The start time deviation is the absolute value of the difference between the candidate start time and the start time of the original batch, and the end time deviation is the absolute value of the difference between the candidate end time and the end time of the original batch. The candidate normal production batch with the smallest start time deviation is determined as the optimal start candidate, and the candidate normal production batch with the smallest end time deviation is determined as the optimal end candidate. The optimal start candidate represents the candidate normal production batch that is closest to the batch requiring correction in terms of start time, and the optimal end candidate represents the candidate normal production batch that is closest to the batch requiring correction in terms of end time.

[0064] When two or more candidate records of normal production batches have the same minimum start time deviation, calculate the absolute difference between the production duration of the candidate batch and the original batch duration of the batch to be corrected for each candidate record of normal production batches, and prioritize the candidate record of normal production batches with the smallest absolute difference as the best start candidate; when two or more candidate records of normal production batches have the same minimum end time deviation, calculate the absolute difference between the production duration of the candidate batch and the original batch duration of the batch to be corrected for each candidate record of normal production batches, and prioritize the candidate record of normal production batches with the smallest absolute difference as the best end candidate; when the absolute difference of start time deviation, end time deviation, and production duration still cannot distinguish candidate records, select the candidate record of normal production batches with the earliest start time to ensure that the best start candidate and the best end candidate have definite selection results.

[0065] In this implementation plan, candidate records for normal production batches are uniformly defined and precisely matched. By clearly defining the candidate start time and candidate end time as corresponding to the batch start time and batch end time identified above, consistent terminology is ensured. The minimum production duration for a batch is determined by combining process cycle time, equipment rated cycle time, and minimum qualified output, excluding short periods such as trial runs, jogging, and debugging. Then, the best start candidate and best end candidate are determined based on the deviation of the start time and the deviation of the end time, respectively. When the deviations are the same, disambiguation is performed by the similarity of production duration and the order of candidate start times, thereby ensuring that the final candidate batch selection result is clear and stable, and avoiding matching ambiguities when multiple candidate batches are close.

[0066] Specifically, the steps to obtain the corrected batch start time and corrected batch end time are as follows: When the best start candidate and the best end candidate point to the same candidate batch number, it means that the start time dimension and the end time dimension match the same normal production batch candidate record. The normal production batch candidate record corresponding to the candidate batch number is determined as the target corrected batch. The candidate start time of the target corrected batch is used as the corrected batch start time of the batch that needs to be corrected, and the candidate end time of the target corrected batch is used as the corrected batch end time of the batch that needs to be corrected. The correction result is written to the batch record table and synchronously written back to the batch time field corresponding to the batch number that needs to be corrected in the production management platform. The original batch start time, original batch end time, corrected batch start time, corrected batch end time and target corrected batch number are retained for subsequent traceability of the correction basis.

[0067] When the best start candidate and the best end candidate point to different candidate batch numbers, it indicates that the start time and end time dimensions match different normal production batch candidate records, and a batch matching anomaly is determined. In this case, the time of either candidate batch is not automatically used as the correction result; instead, the original batch start time and original batch end time of the batch that needs correction are retained, and a batch matching anomaly record is generated. The batch matching anomaly record includes the batch number that needs correction, the candidate batch number corresponding to the best start candidate, the candidate batch number corresponding to the best end candidate, the start time deviation, the end time deviation, and the alarm generation time; wherein, the start time deviation is taken from the start time deviation corresponding to the best start candidate, and the end time deviation is taken from the end time deviation corresponding to the best end candidate.

[0068] The alarm generation time uses the server's current time when the batch matching anomaly record is generated, and is written into the batch matching anomaly record in a time format consistent with the timestamp of the production data storage table; when the production management platform has a unified platform time service, the alarm generation time uses the current time returned by the platform time service; when the platform time service cannot be obtained, the local time of the server executing the batch correction program is used, and the time source mark is retained in the batch matching anomaly record.

[0069] When an alarm is output, the batch matching anomaly record is sent to the batch anomaly message queue of the production management platform, or written to the anomaly status field of the batch record table, and the alarm information is displayed on the terminal. The alarm information includes at least the batch number that needs to be corrected, the candidate batch number corresponding to the best start candidate, the candidate batch number corresponding to the best end candidate, and the reason for the anomaly. After the alarm is triggered, the batch that needs to be corrected enters the manual review process or the batch anomaly review process. After the review is completed, it is confirmed whether to use a certain candidate batch time for correction, thereby avoiding incorrect automatic correction results when the candidate batch numbers are inconsistent.

[0070] like Figure 2 The diagram shown illustrates the batch matching consistency determination logic provided in this application example, used to explain the final determination logic for the corrected batch start and end times. The left side of the diagram represents the optimal start candidate, indicating candidate batches selected based on start time deviation, with their candidate start time derived from the production start position. The right side represents the optimal end candidate, indicating candidate batches selected based on end time deviation, with their candidate end time derived from the batch completion position. Both candidate results simultaneously enter the intermediate consistency comparison stage, where the core judgment object is whether their corresponding candidate batch numbers are the same. When the optimal start candidate and optimal end candidate point to the same candidate batch number, the diagram enters the lower left green branch, determining that the numbers are consistent, and identifying this candidate batch as the target correction batch. Finally, the candidate start time of this candidate batch is used as the corrected batch start time, and the candidate end time of this candidate batch is used as the corrected batch end time. When they point to different candidate batch numbers, the diagram enters the lower right red branch, determining a batch matching anomaly, and generating an anomaly record. The record includes the batch number to be corrected, the two candidate batch numbers, the start time deviation, the end time deviation, and the alarm generation time. The batch matching consistency judgment diagram highlights that candidate start points and candidate end points must undergo consistency verification before final correction. It intuitively demonstrates the judgment mechanism that only candidates from the same batch are allowed to output correction results, while different candidate batches will trigger an anomaly alarm.

[0071] like Figure 3The diagram illustrates the principle of the batch start and end time correction method based on production data provided in this application, used to explain the principle of using production volume to correct the batch start and end times. During the production process, the cumulative production volume is reset to zero when switching production batches. The cumulative production volume increases from zero to non-zero after the equipment produces actual products. In the batch on the left side of the diagram, T2 is the start determination position corresponding to the production volume entering the non-zero state. When the production volume value corresponding to T2 is zero and the production volume value at the next sampling time adjacent to T2 is not zero, the next sampling time is determined as the batch start time; T14 is the end determination position corresponding to the first time the production volume of the current batch is reached, used to determine the batch end time of this batch.

[0072] Subsequently, T18 in the diagram represents the output zeroing time, used to characterize the output zeroing position after the end of the previous production batch and serving as the boundary for the next production batch. The output remains zero from T18 to T21, indicating that the equipment has not yet produced the actual output for the next production batch. T21 is the starting determination position corresponding to the output transitioning from zero to a non-zero state within this continuous zero-value segment. When the output value corresponding to T21 is zero and the output value at the next sampling time adjacent to T21 is not zero, that next sampling time is determined as the batch start time for the next production batch. As production continues, the cumulative output gradually increases, and T31 is the ending determination position corresponding to the first time the output reaches the output of the next production batch, used to determine the batch end time of the next production batch.

[0073] Therefore, T2 and T21 in the figure are used to mark the starting judgment position before the output changes from zero to non-zero, T18 is used to mark the output zeroing position between production batches, and T14 and T31 are used to mark the ending judgment position when the output reaches the corresponding batch production output. The batch start time and batch end time are determined by the adjacent output change relationship corresponding to each judgment position.

[0074] In this implementation plan, consistency checks are performed on the best start candidate and the best end candidate to ensure that the batch start time and batch end time come from the same normal production batch candidate record. When both point to the same candidate batch number, the corrected batch start time and corrected batch end time are directly generated, and the basis for the correction is written into the batch record to ensure the traceability of the correction result. When both point to different candidate batch numbers, automatic correction is stopped, a batch matching anomaly record is generated, and an alarm prompt is output, causing the batch to enter the manual review or anomaly review process. This avoids incorrect correction caused by the start and end points coming from different production batches. At the same time, by unifying the alarm generation time source and record format, the consistency and traceability of the anomaly record are ensured.

[0075] A second aspect of the present invention provides a batch start and end time correction device based on production data, comprising: at least one processor and a memory communicatively connected to the at least one processor; wherein the processor is used to call and execute calculation instructions required for batch start and end time correction, and the memory is used to store instructions executable by the processor, a production data storage table, a production change comparison table, a production batch division table, batch candidate records, batch matching anomaly records, and intermediate data related to batch start and end time correction. The processor and the memory interact via a data bus, a device communication interface, or an industrial control communication interface, enabling the processor to read the production data of the target production equipment, the original batch start time, the original batch end time, and the process cycle time data, and write the corrected batch start time, the corrected batch end time, or the batch matching anomaly record back to the corresponding data table. The memory stores instructions that can be executed by at least one processor. When executed by at least one processor, the instructions enable the at least one processor to perform any batch start and end time correction method based on production data, including collecting and storing production data, constructing a production change comparison table, generating a production batch division table, determining the batch start time and batch end time, eliminating invalid candidate batches and test batches, calculating the time deviation, determining the target correction batch, and generating a batch matching anomaly record and outputting an alarm prompt when candidate batches are inconsistent.

[0076] In this implementation scheme, a batch start and end time correction method based on production data is implemented. The processor reads the production data, original batch start time, original batch end time, and process cycle time data of the target production equipment, and calls the execution instructions in the memory to complete the production data storage, production change comparison, production batch division, batch start time and batch end time identification, invalid candidate batch rejection, test batch rejection, time deviation calculation, and target correction batch determination. At the same time, the equipment can write the corrected batch start time and corrected batch end time back to the batch record. When the best start candidate and the best end candidate are inconsistent, a batch matching anomaly record is generated and an alarm prompt is output, thereby realizing automatic correction of batch start and end times, anomaly tracking, and result write-back.

[0077] A third aspect of the present invention provides a batch start and end time correction storage medium based on production data, comprising: a storage medium storing computer-executable instructions, wherein the computer-executable instructions, when read and executed by a processor, implement any batch start and end time correction method based on production data. The computer-executable instructions stored in the storage medium include production data reading instructions, production data writing instructions, production change comparison table generation instructions, production batch division table generation instructions, batch start time identification instructions, batch end time identification instructions, invalid candidate batch rejection instructions, test batch rejection instructions, time deviation calculation instructions, target correction batch determination instructions, correction result write-back instructions, and batch matching anomaly alarm instructions. When the processor executes computer-executable instructions, it reads the output data of the target production equipment according to the sampling period, writes the output data into the output data storage table, constructs an output change comparison table based on the output change relationship between adjacent sampling periods, and forms an output batch division table based on the output change comparison table. Subsequently, within the same batch number range, it identifies the batch start time, batch end time, batch production output, and batch production duration, excluding invalid candidate batches and test batches. Based on the original batch start time and original batch end time, it calculates the time deviation, determines the target correction batch, and obtains the corrected batch start time and corrected batch end time. If the best start candidate and best end candidate point to different candidate batch numbers, the processor generates a batch matching anomaly record when executing the computer-executable instructions and outputs an alarm, causing the abnormal batch to enter the review process. The storage medium includes non-volatile storage media capable of storing computer-executable instructions, including read-only memory, flash memory, solid-state drive, hard disk, memory card, or other data carriers capable of persistently storing computer-executable instructions.

[0078] In this implementation scheme, the computer-executable instructions used to perform batch start and end time correction are persistently stored. This enables the processor to automatically complete processes such as production data reading, production data storage, production change comparison, production batch division, batch start time identification, batch end time identification, invalid candidate batch removal, test batch removal, time deviation calculation, target correction batch determination, correction result write-back, and batch matching anomaly alarm after reading and running these instructions. By embedding each processing logic in a non-volatile data carrier, the batch start and end time correction method can be repeatedly called and stably executed in different devices or different production environments, thereby providing a deployable software carrier for automatic batch time correction, anomaly tracking, and production data traceability.

[0079] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0080] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to the specific implementations described. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims

1. A batch start and end time correction method based on production data, characterized in that: Includes the following steps: S1, collect output data of the target production equipment based on a preset sampling period, and store the output data in the output data storage table; S2, construct a production change comparison table based on the production data storage table, divide the production batches, and determine the batch start time and batch end time based on the production data records of the same batch, and exclude invalid candidate batches; S3, based on the minimum production time of the batch, eliminates test batches, determines the best start candidate and the best end candidate through the time deviation, and obtains the corrected batch start time and the corrected batch end time.

2. The batch start and end time correction method based on production data according to claim 1, characterized in that: The specific steps for collecting output data of the target production equipment based on a preset sampling period and storing the output data in the output data storage table are as follows: Production data of the target production equipment is collected through a preset sampling period. The target production equipment includes machine tools, assembly equipment and packaging machinery in the production line. Production data is read by connecting to the production management platform of the target production equipment to read real-time production records, or by automatically collecting data through counting sensors installed on the production line, including photoelectric sensors and encoders; The production data is the cumulative production value of the target production equipment at the sampling time. The cumulative production value represents the number of products that have been completed in the current production batch. The sampling period is used as the time interval for writing production data into the production data storage table. Each sampling time corresponds to one production data record. The production data storage table includes a timestamp column, a production value column, a description column, and a quality column. The timestamp column is set as the primary key, and the quality column is used to mark the normal collection, supplementary collection, missing, or abnormal collection status.

3. The batch start and end time correction method based on production data according to claim 2, characterized in that: The specific steps for constructing the production change comparison table based on the production data storage table are as follows: Using the sampling period as a unit, the production data is traversed in ascending order by timestamp. The production value of the current sampling period, the production value of the previous sampling period, and the production value of the next sampling period are associated to form the adjacent production change relationship based on the production data record corresponding to the current sampling time. A production change comparison table is constructed based on the adjacent production change relationship. The output change comparison table includes the sampling period, the current sampling period output, the previous sampling period output, and the next sampling period output, which is used to identify the time position when output is reset to zero, output is started, and output reaches the batch production output. For the first production data record, the production in the previous sampling period is recorded as null, and the first production data record is not included in the production zeroing determination; for the last production data record, the production in the next sampling period is recorded as null, and the last production data record is not included in the production start determination; the relationship between adjacent production changes with null values ​​is not used as the basis for determining the corresponding time position.

4. The batch start and end time correction method based on production data according to claim 3, characterized in that: The specific steps for dividing production batches are as follows: Based on the production change comparison table, add a zeroing mark column and a batch number column to form a production batch division table; when the production at the current sampling time is zero and the production at the previous sampling time is not zero, mark the current sampling time as the zeroing point and set the zeroing mark column to one; otherwise, set the zeroing mark column to zero; the sampling time corresponding to the zeroing point is recorded as the production zeroing time. The current batch number is initialized to one. Starting from the first production data record, batch numbers are assigned to each production data record according to the sampling time sequence. Production data records before the first zeroing point are all assigned to the initial batch number. When a zeroing point record is identified, the current batch number is incremented first, and then the zeroing point record and the production data records after the zeroing point record and before the next zeroing point record are assigned to the incremented batch number. When the output value is zero at two or more consecutive sampling times, the first zeroing point in the continuous zero value segment is used as the boundary record of the next production batch, and the remaining zero value records in the continuous zero value segment are assigned to the next production batch number with the same number as the first zeroing point.

5. The batch start and end time correction method based on production data according to claim 4, characterized in that: The specific steps for determining the batch start time and batch end time based on the production data records of the same batch, and excluding invalid candidate batches, are as follows: In the production data record corresponding to the same batch number, if the production value at the current sampling time is zero and the production value at the next adjacent sampling time is not zero, the next sampling time is marked as the batch start time; the maximum production value in the production data record corresponding to the same batch number is obtained, and the maximum production value is used as the production output of the current batch; in the production data record corresponding to the same batch number, missing value filling, abnormal jump removal, and adjacent sampling smoothing are performed on the production value to obtain the corrected production value; When the corrected output value at the previous sampling time is less than the batch production output for the first time, and the absolute value of the difference between the corrected output value at the current sampling time and the batch production output is less than or equal to the output tolerance, and remains within the output tolerance range for the next N consecutive sampling windows, the current sampling time is marked as the batch end time; the batch production duration is obtained by subtracting the batch start time from the batch end time. If the production data record corresponding to the same batch number does not identify the batch start time or the batch end time, or if the batch end time is earlier than the batch start time, the corresponding batch number will be marked as an invalid candidate batch. Invalid candidate batches will not be included in subsequent start and end time deviation comparisons.

6. The batch start and end time correction method based on production data according to claim 5, characterized in that: The specific steps for eliminating test batches based on the minimum batch production time are as follows: Read the original batch start time and original batch end time of the batch that needs to be corrected, and obtain the original batch duration. The original batch start time and original batch end time are obtained from the production management platform, work order records, or batch record table. The original batch duration is the duration obtained by subtracting the original batch start time from the original batch end time. Summarize the production zeroing time, batch start time, batch end time, batch production duration, and batch production output of each non-invalid candidate batch into a batch candidate record. Set a minimum batch production duration. Batch candidate records with a batch production duration lower than the minimum batch production duration are judged as test batches in the production process and removed from the batch candidate record to obtain normal production batch candidate records. The minimum production time for batch production is determined based on the product process cycle time of the target production equipment, and the unit of the minimum production time for batch production is consistent with the timestamp unit.

7. The batch start and end time correction method based on production data according to claim 6, characterized in that: The specific steps for determining the optimal start and end candidates based on time deviation are as follows: The candidate record for normal production batches includes the candidate batch number, candidate start time, candidate end time, candidate batch production output, and candidate batch production duration. Based on the start and end times of the original batch, calculate the start time deviation and end time deviation between each candidate record of normal production batch and the batch that needs correction; the start time deviation is the absolute value of the difference between the candidate start time and the start time of the original batch, and the end time deviation is the absolute value of the difference between the candidate end time and the end time of the original batch; the candidate record of normal production batch with the smallest start time deviation is determined as the best start candidate, and the candidate record of normal production batch with the smallest end time deviation is determined as the best end candidate. When two or more candidate records of normal production batches have the same minimum start time deviation, the candidate record of normal production batch with the smallest absolute difference between the batch production duration and the original batch duration of the batch that needs to be corrected is selected as the best start candidate; when two or more candidate records of normal production batches have the same minimum end time deviation, the candidate record of normal production batch with the smallest absolute difference between the batch production duration and the original batch duration of the batch that needs to be corrected is selected as the best end candidate; if they still cannot be distinguished, the candidate record of normal production batch with the earliest start time is selected.

8. The batch start and end time correction method based on production data according to claim 7, characterized in that: The specific steps for obtaining the corrected batch start time and corrected batch end time are as follows: When the best start candidate and the best end candidate point to the same candidate batch number, the normal production batch candidate record corresponding to the candidate batch number is determined as the target correction batch, and the candidate start time and candidate end time of the target correction batch are used as the corrected batch start time and corrected batch end time, respectively. When the best start candidate and the best end candidate point to different candidate batch numbers, a batch matching anomaly is determined to have occurred, a batch matching anomaly record is generated, and an alarm is output. The batch matching anomaly record includes the batch number to be corrected, the candidate batch number corresponding to the best start candidate, the candidate batch number corresponding to the best end candidate, the start time deviation, the end time deviation, and the alarm generation time.

9. A batch start and end time correction device based on production data, employing the batch start and end time correction method based on production data as described in any one of claims 1-8, characterized in that: include: At least one processor and a memory communicatively connected to the at least one processor; The memory stores instructions executable by at least one processor, which, when executed, enables the at least one processor to perform the batch start and end time correction method based on production data as described in any one of claims 1 to 8.

10. A batch start and end time correction storage medium based on production data, wherein a computer program is stored thereon, and the computer program, when executed by a processor, implements the batch start and end time correction method based on production data according to any one of claims 1 to 8, characterized in that: include: A storage medium storing computer-executable instructions, wherein the computer-executable instructions are configured to implement the batch start and end time correction method based on production data as described in any one of claims 1 to 8.

Images