A method, system and device for material demand forecasting of a tea production plant

Abstract

The application discloses a kind of tea production workshop material demand prediction method, system and equipment, it is related to tea leaf production process intelligent management field, including the following steps: obtaining the production state data of each production batch in tea production workshop;Production state data is preprocessed;Corresponding semi-finished product grade split unit set of each production batch is constructed;Grade split proportion, cross-grade flow relationship and grade residence quantity are extracted, and semi-finished product grade split representation sequence is formed;Semi-finished product grade split representation sequence is spliced and associated and packing path is associated;The corresponding packing bag demand sequence, inner bag demand sequence, label demand sequence, gift box demand sequence and splicing auxiliary material demand sequence of each finished product reconstruction unit are determined reversely;It is summarized to obtain the material demand prediction sequence of tea production workshop.The application adopts grade split modeling method, realizes tea workshop material accurate prediction, with the advantages of high accuracy, strong adaptability.

Description

Technical Field

[0001] This invention relates to the field of intelligent management of tea production processes, and in particular to a method, system and equipment for predicting material demand in a tea production workshop. Background Technology

[0002] Tea production workshops typically involve multiple continuous processes such as screening, selection, grading, blending, and packaging. Different production batches of tea vary in appearance, color, particle size, moisture content, and subsequent grading destinations, which in turn affects the amount of packaging materials and blending additives used. Currently, material preparation in tea production workshops often relies on production plans, historical statistical data, or manual experience for estimation, and materials are prepared and mixed according to packaging specifications or feeding rules.

[0003] However, existing technologies typically treat material requirements as direct conversion results under a fixed plan, lacking a comprehensive analysis of the grade division state of tea after grading, the cross-grade flow relationship between different grading destinations, and the blending and reconstruction process. This makes it difficult to accurately reflect the actual changes in the demand for packaging bags, inner bags, labels, gift boxes, and blending auxiliary materials for different production batches, which can easily lead to insufficient material preparation, excessive material preparation, or a mismatch between material configuration and the actual finished product structure. Summary of the Invention

[0004] One objective of this invention is to propose a method, system, and equipment for predicting material demand in tea production workshops. This invention adopts a hierarchical split modeling method to achieve accurate prediction of materials in tea workshops, and has the advantages of high accuracy and strong adaptability.

[0005] A method for predicting material requirements in a tea production workshop according to an embodiment of the present invention includes the following steps:

[0006] Obtain production status data for each production batch in the tea production workshop;

[0007] Preprocess the production status data to generate a standardized production status sequence;

[0008] Based on the standardized production status sequence, construct a set of semi-finished product grade splitting units corresponding to each production batch;

[0009] For the set of semi-finished product grade splitting units, the grade splitting ratio, cross-grade flow relationship and grade retention are extracted to form a semi-finished product grade splitting characterization sequence;

[0010] By combining the material feeding data and packaging specification mapping data in the standardized production state sequence, the semi-finished product grade splitting characterization sequence is associated with the material feeding data and the packaging path, and a finished product reconstruction characterization sequence is generated.

[0011] Based on the finished product reconstruction characterization sequence, the packaging bag demand sequence, inner bag demand sequence, label demand sequence, gift box demand sequence, and compounding auxiliary material demand sequence corresponding to each finished product reconstruction unit are determined in reverse.

[0012] By summarizing the demand sequences for packaging bags, inner bags, labels, gift boxes, and blending auxiliary materials, a material demand forecast sequence for the tea production workshop is obtained.

[0013] Optionally, the production status data includes batch identification, semi-finished tea weight data, sieved particle size distribution data, appearance grade data, color grade data, moisture content data, grading destination identification data, blending and feeding data, and packaging specification mapping data.

[0014] Optionally, the preprocessing includes time synchronization, batch association, process association, anomaly removal, and missing information completion.

[0015] Optionally, the generation of the semi-finished product level splitting unit set specifically includes:

[0016] Extract the weight data, sieve particle size distribution data, appearance grade data, color grade data, moisture content data, and grading destination identification data of the semi-finished tea products corresponding to the same production batch from the standardized production status sequence according to the batch identifier;

[0017] Based on the hierarchical destination identification data, the data extracted from the same production batch is divided into hierarchical destinations to form multiple data subsequences corresponding to the hierarchical destinations.

[0018] In the data subsequence corresponding to each grade destination, the corresponding semi-finished tea weight data is correlated with the sieve particle size distribution data, shape grade data, color grade data and moisture content data to obtain the grade split data group corresponding to each grade destination.

[0019] The data sets of each level of splitting are combined with the corresponding batch identifier and grade destination identifier data to form the corresponding semi-finished product level splitting unit.

[0020] All semi-finished product grade splitting units corresponding to each grade destination under the same production batch are aggregated to construct the semi-finished product grade splitting unit set corresponding to that production batch.

[0021] Optionally, the generation of the semi-finished product level splitting characterization sequence specifically includes:

[0022] Extract the batch identifier, grading destination identifier, semi-finished tea weight data, sieve particle size distribution data, appearance grade data, color grade data and moisture content data corresponding to each semi-finished product grade splitting unit set corresponding to each production batch.

[0023] Using the same production batch as a unit, the semi-finished product grade splitting units are classified and statistically analyzed according to the grade destination identifier. The proportion of the weight data of the semi-finished tea corresponding to each grade destination in the total weight data of the semi-finished tea in the production batch is determined to obtain the grade splitting ratio.

[0024] Within the same production batch, the semi-finished product grade splitting units are sorted according to appearance grade data, color grade data, and sieve particle size distribution data.

[0025] Based on the grade sorting results and grade destination identifiers, the corresponding semi-finished product grade splitting units for different grade destinations are compared and the cross-grade flow relationship between different grade destinations is extracted.

[0026] By combining the weight and moisture content data of the semi-finished tea leaves corresponding to each semi-finished product grade division unit, the grade retention amount corresponding to each grade is determined.

[0027] The grade splitting ratio, cross-grade transfer relationship, and grade retention volume are arranged sequentially according to batch identifier and grade destination identifier to form a semi-finished product grade splitting characterization sequence.

[0028] Optionally, the generation of the finished product reconstruction representation sequence specifically includes:

[0029] Extract the assembly and feeding data and packaging specification mapping data corresponding to each production batch from the standardized production status sequence;

[0030] Based on batch identifiers and grade destination identifiers, the blending and feeding data are correlated with the semi-finished product grade splitting characterization sequence to form the blending correlation results corresponding to each production batch.

[0031] Based on the assembly correlation results, determine the finished product combination relationship corresponding to different grade destinations under each production batch;

[0032] Based on the packaging specification mapping data, the packaging path is matched to the finished product combination relationship to form the packaging path association result;

[0033] The assembly association results, finished product combination relationships, and packaging path association results are correlated according to the production batch sequence to generate a finished product reconstruction representation sequence.

[0034] Optionally, the reverse determination specifically includes:

[0035] Extract the assembly association results, finished product combination relationships, and packaging path association results for each production batch from the finished product reconstruction characterization sequence;

[0036] Extract feeding targets, graded destination identifiers and feeding quantities from the assembly association results; extract the corrected finished product composition results from the finished product combination relationship; and extract packaging specifications, packaging levels and packaging paths from the packaging path association results.

[0037] Based on the packaging specifications, packaging levels, and the revised finished product composition, determine the packaging bag configuration, inner bag configuration, label configuration, and gift box configuration for each production batch to form a packaging configuration group;

[0038] Based on the packaging configuration group and the revised finished product composition, a reverse conversion is performed to form a packaging material requirement group;

[0039] Based on the material feeding target, graded destination identification and packaging path, the packaging material demand groups are classified and merged into packaging bag demand, inner bag demand, label demand and gift box demand to form classified packaging demand groups.

[0040] Based on the material input quantity, material input target, and graded destination identifier in the blending association results, the blending auxiliary material requirements corresponding to each production batch are determined, forming a blending auxiliary material requirement group;

[0041] The categorized packaging demand group and the blending auxiliary material demand group are aggregated according to production batches to obtain the corresponding packaging bag demand sequence, inner bag demand sequence, label demand sequence, gift box demand sequence and blending auxiliary material demand sequence for each production batch.

[0042] Optionally, the generation of the material demand forecast sequence specifically includes:

[0043] Extract the demand sequence for packaging bags, inner bags, labels, gift boxes, and auxiliary materials for each production batch;

[0044] The demand sequences for packaging bags, inner bags, labels, and gift boxes are classified and grouped according to production batch, packaging specifications, and packaging route.

[0045] The packaging bag requirements, inner bag requirements, label requirements, and gift box requirements in each category group are summed to obtain the category packaging requirements for each production batch.

[0046] The demand sequence of blended auxiliary materials is aggregated according to production batches to obtain the classification of blended auxiliary material demand results for each production batch.

[0047] By combining the classification and packaging requirements and the classification and blending requirements of each production batch, a material requirement prediction sequence for the tea production workshop is obtained.

[0048] A material demand forecasting system for a tea production workshop according to an embodiment of the present invention includes:

[0049] The data acquisition module is used to acquire production status data of each production batch in the tea production workshop during the screening, selection, grading, blending and packaging stages;

[0050] The preprocessing module is used to preprocess production status data and generate a standardized production status sequence.

[0051] The grade splitting construction module is used to construct a set of semi-finished product grade splitting units corresponding to each production batch based on the standardized production state sequence.

[0052] The splitting characterization module is used to extract the splitting ratio, cross-level flow relationship and level retention of the semi-finished product grade splitting unit set, and form a semi-finished product grade splitting characterization sequence.

[0053] The finished product reconstruction module is used to combine the assembly and feeding data and packaging specification mapping data in the standardized production state sequence to perform assembly association and packaging path association on the semi-finished product grade splitting characterization sequence, and generate the finished product reconstruction characterization sequence.

[0054] The demand reverse deduction module is used to reverse-determine the demand sequence for packaging bags, inner bags, labels, gift boxes, and auxiliary materials for each production batch based on the finished product reconstruction representation sequence.

[0055] The summary output module is used to summarize the demand sequences for packaging bags, inner bags, labels, gift boxes, and blending auxiliary materials to obtain the material demand forecast sequence for the tea production workshop.

[0056] A material demand forecasting device for a tea production workshop according to an embodiment of the present invention includes: a processor and a memory; the memory is used to store a computer program, and the processor calls the computer program stored in the memory to execute a material demand forecasting method for a tea production workshop.

[0057] The beneficial effects of this invention are:

[0058] Compared with existing technologies, this invention is based on the production status data of each production batch in the tea production workshop during the screening, selection, grading, blending and packaging stages. It first constructs a set of semi-finished product grade splitting units, then extracts the grade splitting ratio, cross-grade flow relationship and grade retention amount to form a semi-finished product grade splitting characterization sequence. Then, it combines blending material input data and packaging specification mapping data to generate a finished product reconstruction characterization sequence. On this basis, it reversely determines the demand sequence of packaging bags, inner bags, labels, gift boxes and blending auxiliary materials. Thus, the material preparation process that originally relied on experience judgment is transformed into a structured prediction process based on grading results, blending relationships and packaging paths, which can more accurately reflect the actual material demand changes of different production batches of tea in the subsequent finished product organization.

[0059] This invention not only classifies, predicts, and summarizes the demand for packaging materials and blending auxiliary materials based on the different grading destinations, cross-grade circulation relationships, and blending and reconstruction relationships formed after tea grading, but also avoids the deviations caused by estimations based solely on production plans or historical average consumption in existing technologies. This helps improve the accuracy of the configuration of packaging bags, inner bags, labels, gift boxes, and blending auxiliary materials, reduces the risks of insufficient or excessive material preparation, and mismatch between material configuration and actual finished product structure, thereby enhancing the targeted nature of material organization in tea production workshops, the smoothness of production connections, and the stability of overall production management. Attached Figure Description

[0060] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0061] Figure 1 This is a flowchart of a material demand forecasting method for a tea production workshop proposed in this invention;

[0062] Figure 2 This is a schematic diagram illustrating the construction of the finished product reconstruction characterization sequence for a material demand prediction method for a tea production workshop proposed in this invention. Detailed Implementation

[0063] The present invention will now be described in further detail with reference to the accompanying drawings. These drawings are simplified schematic diagrams, illustrating only the basic structure of the invention, and therefore only show the components relevant to the invention.

[0064] refer to Figure 1 and Figure 2 A method for predicting material requirements in a tea production workshop includes the following steps:

[0065] To obtain production status data for each production batch in the tea production workshop during the screening, selection, grading, blending and packaging stages;

[0066] Preprocess the production status data to generate a standardized production status sequence;

[0067] Based on the standardized production status sequence, construct a set of semi-finished product grade splitting units corresponding to each production batch;

[0068] For the set of semi-finished product grade splitting units, the grade splitting ratio, cross-grade flow relationship and grade retention are extracted to form a semi-finished product grade splitting characterization sequence;

[0069] By combining the material feeding data and packaging specification mapping data in the standardized production state sequence, the semi-finished product grade splitting characterization sequence is associated with the material feeding data and the packaging path, and a finished product reconstruction characterization sequence is generated.

[0070] Based on the finished product reconstruction characterization sequence, the packaging bag demand sequence, inner bag demand sequence, label demand sequence, gift box demand sequence, and compounding auxiliary material demand sequence corresponding to each finished product reconstruction unit are determined in reverse.

[0071] By summarizing the demand sequences for packaging bags, inner bags, labels, gift boxes, and blending auxiliary materials, a material demand forecast sequence for the tea production workshop is obtained.

[0072] In this embodiment, the production status data includes batch identifier, semi-finished tea weight data, sieved particle size distribution data, appearance grade data, color grade data, moisture content data, grading destination identifier data, blending feed data, and packaging specification mapping data. The batch identifier is a unique identification information used to distinguish different production batches of tea and establish the corresponding relationship between data at each stage. The semi-finished tea weight data is the weight of the intermediate tea material corresponding to each production batch of tea. The sieved particle size distribution data is the distribution of the same production batch of tea in different particle size ranges after sieving. The appearance grade data is the data formed by classifying the tea according to its appearance characteristics. The color grade data is the data formed by classifying the tea according to its color performance. The moisture content data is the data of the moisture content of the corresponding production batch of tea at the current stage. The semi-finished product is the intermediate tea material that has completed the previous stage of processing but has not yet formed the final packaged product and still needs to participate in grading, blending, or packaging.

[0073] In this embodiment, preprocessing includes time synchronization, batch association, process association, anomaly removal, and missing data completion.

[0074] In this embodiment, the generation of the semi-finished product grade splitting unit set specifically includes:

[0075] Extract the weight data, sieve particle size distribution data, appearance grade data, color grade data, moisture content data, and grading destination identification data of the semi-finished tea products corresponding to the same production batch from the standardized production status sequence according to the batch identifier;

[0076] Based on the hierarchical destination identification data, the data extracted from the same production batch is divided into hierarchical destinations to form multiple data subsequences corresponding to the hierarchical destinations.

[0077] The grading destination refers to the subsequent circulation path of tea leaves in the same production batch after grading. The grading destination identification data is used to characterize the circulation category and processing direction of each semi-finished tea leaf in the subsequent production organization. The grading destination is set according to the actual production in the tea production workshop, specifically corresponding to one or more of the following: high-grade finished product packaging destination, medium-grade finished product packaging destination, ordinary finished product packaging destination, blending raw material destination, re-screening destination, re-selection destination, temporary storage destination, rework destination, and destination for specific packaging specifications. Different production batches form different grading destination configurations based on the corresponding tea leaf's sieve particle size distribution, appearance grade, color grade, moisture content, and finished product organization requirements. The grading destination identification data corresponds to the semi-finished tea leaf weight data, sieve particle size distribution data, appearance grade data, color grade data, and moisture content data, and is used to characterize the actual grade splitting state of the semi-finished tea leaves under each grading destination.

[0078] In the data subsequence corresponding to each grade destination, the corresponding semi-finished tea weight data is correlated with the sieve particle size distribution data, shape grade data, color grade data and moisture content data to obtain the grade split data group corresponding to each grade destination.

[0079] The grade split data group is a set of grade split descriptive data consisting of the weight data, sieve particle size distribution data, appearance grade data, color grade data and moisture content data of semi-finished tea leaves corresponding to each other under the same grade.

[0080] The data sets of each level of splitting are combined with the corresponding batch identifier and grade destination identifier data to form the corresponding semi-finished product level splitting unit.

[0081] All semi-finished product grade splitting units corresponding to each grade destination under the same production batch are aggregated to construct the semi-finished product grade splitting unit set corresponding to that production batch.

[0082] In this embodiment, the generation of the semi-finished product grade splitting characterization sequence specifically includes:

[0083] Extract the batch identifier, grading destination identifier, semi-finished tea weight data, sieve particle size distribution data, appearance grade data, color grade data and moisture content data corresponding to each semi-finished product grade splitting unit set corresponding to each production batch.

[0084] Using the same production batch as a unit, the semi-finished product grade splitting units are classified and statistically analyzed according to the grade destination identifier. The proportion of the weight data of the semi-finished tea corresponding to each grade destination in the total weight data of the semi-finished tea in the production batch is determined to obtain the grade splitting ratio.

[0085] The grade split ratio is used to characterize the degree to which a production batch of tea is split and distributed into different grades and different distribution paths;

[0086] Within the same production batch, the semi-finished product grade splitting units are sorted according to appearance grade data, color grade data, and sieve particle size distribution data.

[0087] The process of ranking is as follows: Within the same production batch, firstly, extract the shape grade data, color grade data, and sieve particle size distribution data corresponding to each semi-finished product grade splitting unit, and generate shape grade values, color grade values, and particle size grade values ​​respectively; then, normalize the shape grade values, color grade values, and particle size grade values ​​to obtain standard grade values; subsequently, perform a weighted summation of the standard grade values ​​corresponding to each semi-finished product grade splitting unit to generate a comprehensive grade value; sort each semi-finished product grade splitting unit according to the comprehensive grade value from high to low, and output the grade ranking results;

[0088] The shape grade data is formed by scoring the tightness, uniformity, and fragmentation of the strands. The color grade data is formed by scoring the consistency of color, the overall color level, and the smoothness. The tightness, uniformity, and smoothness of the strands are scored separately, and the scores of each sub-item are unified to the same numerical range before being weighted and summed to output the shape grade value. The consistency of color, the overall color level, and the smoothness are scored separately, and the scores of each sub-item are unified to the same numerical range before being weighted and summed to output the color grade value. The weight percentage of the same semi-finished product grade splitting unit in each particle size range is calculated. The range grade coefficients are assigned according to the grade order corresponding to the particle size range. The weight percentage of each particle size range is multiplied by the corresponding range grade coefficient and then summed to output the particle size grade value.

[0089] The tightness of the tea leaves refers to the degree to which the tea leaves are curled, tightened, and firm, indicating whether the tea leaves are tightly formed. The uniformity refers to the consistency of the shape, length, thickness, and overall neatness of the same batch of tea leaves, indicating whether the distribution of the leaves is even. The degree of wholeness refers to the state of the tea leaves and broken pieces, indicating the level of integrity of the tea leaves. The color consistency refers to the degree of uniformity in the color of the same batch of tea leaves, indicating whether the color distribution is harmonious. The overall color grade is the quality grade corresponding to the overall color of the tea leaves, indicating what level the overall color belongs to. The degree of luster refers to the degree of gloss and oiliness on the surface of the tea leaves, indicating whether the color is fresh and bright.

[0090] Based on the grade sorting results and grade destination identifiers, the corresponding semi-finished product grade splitting units for different grade destinations are compared and the cross-grade flow relationship between different grade destinations is extracted.

[0091] The generation of cross-level transfer relationships specifically includes: numbering each semi-finished product grade splitting unit according to the grade sorting result within the same production batch, and outputting the sorting number sequence; then grouping the sorting number sequence according to the grade destination identifier, and outputting the sorting group corresponding to each grade destination; using the absolute values ​​of the difference in sorting number, the difference in appearance grade value, the difference in color grade value, and the difference in particle size grade value as comparison items, pairing and calculating semi-finished product grade splitting units with different grade destinations one by one, and weighting and summing each comparison item to output the unit pairing difference value; for each semi-finished product grade splitting unit, screening unit pairing difference... Candidate pairings with outliers less than or equal to a preset difference threshold are selected, and the candidate pairing with the smallest unit pairing difference value is selected from the corresponding candidate pairings to form the corresponding cross-level corresponding unit group. When multiple semi-finished product level splitting units correspond to the same candidate pairing result, only the candidate pairing result with the smallest unit pairing difference value is retained. All cross-level corresponding unit groups are aggregated and output as a cross-level corresponding unit group set. Based on the sorting number ascending and descending direction and the change direction of the grade destination identifier of each group in the cross-level corresponding unit group set, the cross-level flow relationship between different grade destinations is determined.

[0092] The pairing difference values ​​of semi-finished product grade splitting units that have been confirmed to have cross-level transfer relationships in historical production batches are sorted from smallest to largest, and the 90th percentile value is taken as the preset difference threshold. The direction of the sorting number is the direction of change of the numbers of the two semi-finished product grade splitting units corresponding to the cross-level in the grade sorting result from high to low or from low to high. The direction of change of the grade destination identifier is the direction of change of the grade destination corresponding to the two semi-finished product grade splitting units from which grade destination to another grade destination.

[0093] By combining the weight data and moisture content data of the semi-finished tea leaves corresponding to each semi-finished product grade division unit, the grade retention amount corresponding to each grade destination is determined, which is used to characterize the comprehensive quantitative value of the distribution of semi-finished tea leaves in the moisture content state under the corresponding grade destination.

[0094] When generating grade retention, the semi-finished product grade splitting units in the same production batch are grouped according to the grade destination identifier, and the unit group corresponding to each grade destination is output. Then, the weight data and moisture content data of the semi-finished tea leaves corresponding to each semi-finished product grade splitting unit in each unit group are extracted, and the weight data of the semi-finished tea leaves is multiplied by the moisture content data to obtain the moisture retention of each semi-finished product grade splitting unit. The moisture retention of each semi-finished product grade splitting unit corresponding to the same grade destination is accumulated, and the total moisture retention of the grade destination is output. Then, the total moisture retention of the grade destination is divided by the accumulated value of the weight data of the semi-finished tea leaves in the grade destination to obtain the retention moisture ratio of the grade destination. The total moisture retention of the grade destination and the retention moisture ratio are normalized and then weighted and summed to output the grade retention of the grade destination.

[0095] The grade splitting ratio, cross-grade transfer relationship, and grade retention volume are arranged sequentially according to batch identifier and grade destination identifier to form a semi-finished product grade splitting characterization sequence.

[0096] In this embodiment, the generation of the finished product reconstruction characterization sequence specifically includes:

[0097] Extract the assembly and feeding data and packaging specification mapping data corresponding to each production batch from the standardized production status sequence;

[0098] Based on batch identifiers and grade destination identifiers, the blending and feeding data are correlated with the semi-finished product grade splitting characterization sequence to form the blending correlation results corresponding to each production batch.

[0099] When generating the blending association results, the batch identifier, grade destination identifier, feeding quantity, and feeding target corresponding to each feeding record are extracted from the blending feeding data, and the blending feeding subset is output. Then, the corresponding batch identifier, grade destination identifier, grade split ratio, cross-grade flow relationship, and grade retention quantity are extracted from the semi-finished product grade splitting representation sequence, and the representation matching subset is output. Using the batch identifier and grade destination identifier as dual indexes, the blending feeding subset and the representation matching subset are matched accordingly, and the effective association groups are output. The effective association groups are merged according to the feeding target to form the blending association results corresponding to each production batch.

[0100] Based on the assembly correlation results, determine the finished product combination relationship corresponding to different grade destinations under each production batch;

[0101] When determining the finished product combination relationship, the material input target, grade destination identifier, material input quantity, and corresponding grade splitting ratio, cross-grade flow relationship, and grade retention quantity for each production batch are extracted from the blending association results, and a blending characterization data set is output. Then, the blending characterization data set is classified according to the material input target, and a blending unit group corresponding to each material input target is output. Subsequently, the material input quantities corresponding to each grade destination within the same blending unit group are accumulated to obtain the total material input quantity corresponding to each material input target, and the proportion of the material input quantity corresponding to each grade destination in the total material input quantity is calculated, and a finished product composition proportion group is output. Then, the finished product composition proportion group is corrected by combining the grade splitting ratio, cross-grade flow relationship, and grade retention quantity corresponding to each grade destination, and a finished product composition correction group is output. The material input target, the total material input quantity corresponding to each grade destination, and the finished product composition correction group are combined to form the finished product combination relationship corresponding to different grade destinations under each production batch.

[0102] When correcting the finished product composition percentage group, the finished product composition percentage, grade splitting ratio, cross-grade flow relationship, and grade retention amount corresponding to each grade destination under the same feeding target are extracted. Then, the grade splitting ratio corresponding to each grade destination is multiplied by the finished product composition percentage to obtain the initial splitting correction value. Subsequently, the flow correction direction corresponding to each grade destination is determined according to the cross-grade flow relationship. For grade destinations with inflow relationship, the corresponding inflow correction amount is added to the initial splitting correction value, and for grade destinations with outflow relationship, the corresponding outflow correction amount is subtracted from the initial splitting correction value, and the flow correction value is output. Then, the flow correction value corresponding to each grade destination is normalized and weighted and summed to obtain the combined correction value corresponding to each grade destination, and the finished product composition correction group is output. The finished product composition correction group is arranged according to the grade destination identifier to obtain the corrected finished product composition results corresponding to different grade destinations under each feeding target.

[0103] The inflow correction is the sum of the products of the grade splitting ratio and the grade retention amount corresponding to each source grade destination that forms an inflow relationship with the current grade destination. The outflow correction is the product of the grade splitting ratio and the grade retention amount corresponding to the current grade destination. The inflow relationship indicates that the semi-finished product grade splitting unit corresponding to other grade destinations flows to the semi-finished product grade splitting unit corresponding to the current grade destination. The outflow relationship indicates that the semi-finished product grade splitting unit corresponding to the current grade destination flows to the semi-finished product grade splitting unit corresponding to other grade destinations.

[0104] Based on the packaging specification mapping data, the packaging path is matched to the finished product combination relationship to form the packaging path association result;

[0105] When generating packaging path association results, the finished product combination relationship and packaging specification mapping data corresponding to each production batch are extracted, and the matching input group is output. Then, the matching input group is searched according to the material feeding target and grade destination identifier to determine the packaging specifications, packaging level and packaging path corresponding to each finished product combination relationship. Subsequently, the corrected finished product composition results corresponding to the same packaging path are merged, and the path merging group is output. The packaging specifications, packaging level, packaging path and path merging group are combined to form the packaging path association results.

[0106] The assembly association results, finished product combination relationships, and packaging path association results are correlated according to the production batch sequence to generate a finished product reconstruction representation sequence.

[0107] In this embodiment, the reverse determination specifically includes:

[0108] Extract the assembly association results, finished product combination relationships, and packaging path association results for each production batch from the finished product reconstruction characterization sequence;

[0109] Extract feeding targets, graded destination identifiers and feeding quantities from the assembly association results; extract the corrected finished product composition results from the finished product combination relationship; and extract packaging specifications, packaging levels and packaging paths from the packaging path association results.

[0110] The packaging path association result is formed based on the packaging specification mapping data. In specific implementation, the corresponding packaging specification, packaging level and packaging path are read from the packaging path association result, or the packaging specification mapping data corresponding to the packaging path association result is read.

[0111] Based on the packaging specifications, packaging levels, and the revised finished product composition, determine the packaging bag configuration, inner bag configuration, label configuration, and gift box configuration for each production batch to form a packaging configuration group;

[0112] Based on the packaging configuration group and the revised finished product composition, a reverse conversion is performed to form a packaging material requirement group;

[0113] In specific implementation, in production batch B, the corrected finished product composition results correspond to a finished product weight of 240 kg for high-grade finished product packaging and 160 kg for ordinary finished product packaging. The packaging specification mapping data for high-grade finished product packaging uses 80g per bag, with one inner bag, one label, and one gift box per bag; for ordinary finished product packaging, it uses 200g per bag, with one inner bag, one label, and no gift box. Based on this, the packaging configuration group is determined as follows: for high-grade finished product packaging, the corresponding packaging bag configuration is 80g / bag, one inner bag / bag, one label / bag, and one gift box / bag; for ordinary finished product packaging, the corresponding packaging bag configuration is 200g / bag. The packaging material requirements are as follows: 1 bag / inner bag, 1 label / bag, 0 gift boxes / bag. Then, based on the packaging configuration group and the revised finished product composition, a reverse conversion is performed: 240 kg is converted to 240,000 g and divided by 80 g. This yields a requirement of 3,000 bags, 3,000 inner bags, 3,000 labels, and 3,000 gift boxes for high-grade finished product packaging. 160 kg is converted to 160,000 g and divided by 200 g. This yields a requirement of 800 bags, 800 inner bags, 800 labels, and 0 gift boxes for ordinary finished product packaging. This forms the packaging material requirement group for this production batch.

[0114] Based on the material feeding target, graded destination identification and packaging path, the packaging material demand groups are classified and merged into packaging bag demand, inner bag demand, label demand and gift box demand to form classified packaging demand groups.

[0115] Extract the production batch, material input target, hierarchical destination identifier, packaging path, and packaging bag, inner bag, label, and gift box requirements corresponding to each record from the packaging material requirement group, and output a detailed packaging requirement group. Classify the detailed packaging requirement group according to the material input target, hierarchical destination identifier, and packaging path to form multiple subgroups. Then, sum the packaging bag, inner bag, label, and gift box requirements in each subgroup to obtain the total packaging bag, inner bag, label, and gift box requirements for each subgroup. Combine the material input target, hierarchical destination identifier, packaging path, total packaging bag, inner bag, label, and gift box requirements for each subgroup to form a categorized packaging requirement group.

[0116] Based on the material input quantity, material input target, and graded destination identifier in the blending association results, the blending auxiliary material requirements corresponding to each production batch are determined, forming a blending auxiliary material requirement group;

[0117] In specific implementation, within production batch C, the target material input is finished product C2. The graded destination identifiers correspond to the destination of blended raw materials and the destination of mid-grade finished product packaging. The blending correlation results show that the input quantity corresponding to the destination of blended raw materials is 120 kg, and the input quantity corresponding to the destination of mid-grade finished product packaging is 80 kg. When the blended auxiliary material configuration relationship corresponding to the target material finished product C2 is that 2 kg of auxiliary flavoring and 1 kg of stabilizing auxiliary material need to be added per 100 kg of finished product, the 120 kg and 80 kg are first... Adding them together, we get a total feed amount of 200 kg for finished product C2; then divide 200 kg by 100 kg to get an auxiliary material conversion factor of 2; then multiply the required amount of auxiliary flavoring (2 kg) by 2 to get a required amount of auxiliary flavoring (4 kg); multiply the required amount of stable auxiliary material (1 kg) by 2 to get a required amount of stable auxiliary material (2 kg); combine the production batch, feed target, grade destination identifier, required amount of auxiliary flavoring, and required amount of stable auxiliary material to form the required amount of blended auxiliary materials for this production batch.

[0118] The categorized packaging demand group and the blending auxiliary material demand group are aggregated according to production batches to obtain the corresponding packaging bag demand sequence, inner bag demand sequence, label demand sequence, gift box demand sequence and blending auxiliary material demand sequence for each production batch.

[0119] In this embodiment, the generation of the material demand forecast sequence specifically includes:

[0120] Extract the demand sequence for packaging bags, inner bags, labels, gift boxes, and auxiliary materials for each production batch;

[0121] The demand sequences for packaging bags, inner bags, labels, and gift boxes are classified and grouped according to production batch, packaging specifications, and packaging route.

[0122] The packaging bag requirements, inner bag requirements, label requirements, and gift box requirements in each category group are summed to obtain the category packaging requirements for each production batch.

[0123] The demand sequence of blended auxiliary materials is aggregated according to production batches to obtain the classification of blended auxiliary material demand results for each production batch.

[0124] By combining the classification and packaging requirements and the classification and blending requirements of each production batch, a material requirement prediction sequence for the tea production workshop is obtained.

[0125] A material requirements forecasting system for a tea production workshop includes:

[0126] The data acquisition module is used to acquire production status data of each production batch in the tea production workshop during the screening, selection, grading, blending and packaging stages;

[0127] The preprocessing module is used to preprocess production status data and generate a standardized production status sequence.

[0128] The grade splitting construction module is used to construct a set of semi-finished product grade splitting units corresponding to each production batch based on the standardized production state sequence.

[0129] The splitting characterization module is used to extract the splitting ratio, cross-level flow relationship and level retention of the semi-finished product grade splitting unit set, and form a semi-finished product grade splitting characterization sequence.

[0130] The finished product reconstruction module is used to combine the assembly and feeding data and packaging specification mapping data in the standardized production state sequence to perform assembly association and packaging path association on the semi-finished product grade splitting characterization sequence, and generate the finished product reconstruction characterization sequence.

[0131] The demand reverse deduction module is used to reverse-determine the demand sequence for packaging bags, inner bags, labels, gift boxes, and auxiliary materials for each production batch based on the finished product reconstruction representation sequence.

[0132] The summary output module is used to summarize the demand sequences for packaging bags, inner bags, labels, gift boxes, and blending auxiliary materials to obtain the material demand forecast sequence for the tea production workshop.

[0133] A material demand forecasting device for a tea production workshop includes a processor and a memory; the memory stores a computer program, and the processor calls the computer program stored in the memory to execute a material demand forecasting method for a tea production workshop.

[0134] Example 1: To verify the feasibility of this invention in practice, it was applied to a tea production workshop in a certain location. This tea production workshop continuously undertakes the screening, selection, grading, blending, and packaging of multiple batches of tea. Different batches of tea have significant differences in appearance, color, particle size, moisture content, and subsequent grading direction. In actual operation, the workshop has long adopted a method of preparing materials based on production plans, manual experience, and historical consumption records.

[0135] Because tea leaves do not flow to a fixed single packaging path after grading, but instead go to high-grade finished product packaging, ordinary finished product packaging, blending raw materials, and other subsequent circulation paths according to the grading results, inconsistencies often arise between the initial judgment and the actual use in the later stages of preparing packaging bags, inner bags, labels, gift boxes, and blending auxiliary materials. Specifically, some packaging materials are prepared in excess and accumulate, while other packaging materials are temporarily replenished due to changes in the structure of the finished products. It is difficult for workshop managers to timely determine the actual material requirements of different batches of tea in the subsequent packaging and blending stages.

[0136] In this scenario, production status data of each production batch during the screening, selection, grading, blending and packaging stages are first collected, and time synchronization, batch association, process association, anomaly removal and missing information completion are completed to obtain a standardized production status sequence. Then, based on the standardized production status sequence, a set of semi-finished product grade splitting units corresponding to each production batch is constructed, and the grade splitting ratio, cross-grade flow relationship and grade retention are extracted to form a semi-finished product grade splitting characterization sequence.

[0137] By combining material feeding data and packaging specification mapping data, the semi-finished product grade splitting characterization sequence is associated with the material feeding and packaging path to generate the finished product reconstruction characterization sequence. Based on this, the demand sequence for packaging bags, inner bags, labels, gift boxes, and compounding auxiliary materials is determined in reverse. The above demand sequences are summarized and the corresponding material demand prediction sequence for the production batch is output for workshop material preparation, material requisition, and packaging production scheduling.

[0138] During continuous production periods, the workshop compares and records the material organization before and after implementation. The recording locations are the packaging area and blending area of ​​the tea production workshop. The recording content includes material preparation deviation, temporary material replenishment, backlog of packaging materials, and accuracy of blending auxiliary material preparation.

[0139] To verify the performance of the present invention, a comparison was conducted, and the comparison results are shown in Table 1.

[0140] Table 1. Comparison of the results of this invention with traditional empirical estimation methods.

[0141] Comparison items Traditional empirical estimation methods Method of the present invention Improvement range Average deviation rate of packaging material demand forecast 12.8% 5.1% Reduced by 60.15% Average deviation rate of demand forecast for blended auxiliary materials 10.6% 3.5% Reduced by 66.98% Number of temporary material replenishments 11 times 6 times Reduced by 45.45% Packaging material stockpiling rate 9.3% 3.2% Reduced by 65.59% Average connection time from material preparation to packaging line 46 minutes 29 minutes Reduced by 36.96%

[0142] As shown in Table 1, the method of this invention significantly reduces the average deviation rate of packaging material demand forecasting and blending auxiliary material demand forecasting compared to traditional empirical estimation methods. Specifically, the average deviation rate of packaging material demand forecasting decreased from 12.8% to 5.1%, and the average deviation rate of blending auxiliary material demand forecasting decreased from 10.6% to 3.5%. This indicates that this invention does not rely solely on existing production plans or historical average consumption for static estimation. Instead, it integrates production status data from screening, selection, grading, blending, and packaging into the analysis process. By constructing a set of semi-finished product grade splitting units, it further extracts grade splitting ratios, cross-grade flow relationships, and grade retention amounts. This allows for a more accurate reflection of the actual material consumption trends of different production batches of tea in the subsequent finished product organization, resulting in a smaller deviation between the predicted results and actual usage.

[0143] From the perspective of implementation results, the method of this invention reduced the number of temporary material replenishments from 11 to 6, the backlog rate of packaging materials from 9.3% to 3.2%, and the average connection time from material preparation to packaging line launch from 46 minutes to 29 minutes. The reason for these improvements lies in the fact that after forming the semi-finished product grade splitting characterization sequence, this invention combines blending material input data and packaging specification mapping data to generate a finished product reconstruction characterization sequence. Based on this, it reversely determines the requirements for packaging bags, inner bags, labels, gift boxes, and blending auxiliary materials. This allows the material preparation results to be adjusted synchronously with changes in grading destinations, blending relationships, and packaging paths, thereby reducing material mismatches, duplicate material preparation, and on-site waiting caused by estimation lags. Therefore, this invention not only improves the accuracy of demand forecasting but also enhances the consistency between material organization in the tea production workshop and the actual finished product structure, demonstrating high application value.

[0144] The above are merely preferred embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A method for predicting material requirements in a tea production workshop, characterized in that, Includes the following steps: Obtain production status data for each production batch in the tea production workshop; Preprocess the production status data to generate a standardized production status sequence; Based on the standardized production status sequence, construct a set of semi-finished product grade splitting units corresponding to each production batch; For the set of semi-finished product grade splitting units, the grade splitting ratio, cross-grade flow relationship and grade retention are extracted to form a semi-finished product grade splitting characterization sequence; By combining the material feeding data and packaging specification mapping data in the standardized production state sequence, the semi-finished product grade splitting characterization sequence is associated with the material feeding data and the packaging path, and a finished product reconstruction characterization sequence is generated. Based on the finished product reconstruction characterization sequence, the packaging bag demand sequence, inner bag demand sequence, label demand sequence, gift box demand sequence, and compounding auxiliary material demand sequence corresponding to each finished product reconstruction unit are determined in reverse. The material demand forecast sequence for the tea production workshop is obtained by summarizing the demand sequences for packaging bags, inner bags, labels, gift boxes, and blending auxiliary materials. The reverse determination specifically includes: Extract the assembly association results, finished product combination relationships, and packaging path association results for each production batch from the finished product reconstruction characterization sequence; Extract feeding targets, graded destination identifiers and feeding quantities from the assembly association results; extract the corrected finished product composition results from the finished product combination relationship; and extract packaging specifications, packaging levels and packaging paths from the packaging path association results. Based on the packaging specifications, packaging levels, and the revised finished product composition, determine the packaging bag configuration, inner bag configuration, label configuration, and gift box configuration for each production batch to form a packaging configuration group; Based on the packaging configuration group and the revised finished product composition, a reverse conversion is performed to form a packaging material requirement group; Based on the material feeding target, graded destination identification and packaging path, the packaging material demand groups are classified and merged into packaging bag demand, inner bag demand, label demand and gift box demand to form classified packaging demand groups. Based on the material input quantity, material input target, and graded destination identifier in the blending association results, the blending auxiliary material requirements corresponding to each production batch are determined, forming a blending auxiliary material requirement group; The categorized packaging demand group and the blending auxiliary material demand group are aggregated according to production batches to obtain the corresponding packaging bag demand sequence, inner bag demand sequence, label demand sequence, gift box demand sequence and blending auxiliary material demand sequence for each production batch.

2. The material demand forecasting method for a tea production workshop according to claim 1, characterized in that, The production status data includes batch identification, semi-finished tea weight data, sieved particle size distribution data, appearance grade data, color grade data, moisture content data, grading destination identification data, blending and feeding data, and packaging specification mapping data.

3. The material demand forecasting method for a tea production workshop according to claim 1, characterized in that, The preprocessing includes time synchronization, batch association, process association, anomaly removal, and missing data completion.

4. The material demand forecasting method for a tea production workshop according to claim 1, characterized in that, The generation of the semi-finished product grade splitting unit set specifically includes: Extract the weight data, sieve particle size distribution data, appearance grade data, color grade data, moisture content data, and grading destination identification data of the semi-finished tea products corresponding to the same production batch from the standardized production status sequence according to the batch identifier; Based on the hierarchical destination identification data, the data extracted from the same production batch is divided into hierarchical destinations to form multiple data subsequences corresponding to the hierarchical destinations. In the data subsequence corresponding to each grade destination, the corresponding semi-finished tea weight data is correlated with the sieve particle size distribution data, shape grade data, color grade data and moisture content data to obtain the grade split data group corresponding to each grade destination. The data sets of each level of splitting are combined with the corresponding batch identifier and grade destination identifier data to form the corresponding semi-finished product level splitting unit. All semi-finished product grade splitting units corresponding to each grade destination under the same production batch are aggregated to construct the semi-finished product grade splitting unit set corresponding to that production batch.

5. The material demand forecasting method for a tea production workshop according to claim 1, characterized in that, The generation of the semi-finished product grade splitting characterization sequence specifically includes: Extract the batch identifier, grading destination identifier, semi-finished tea weight data, sieve particle size distribution data, appearance grade data, color grade data and moisture content data corresponding to each semi-finished product grade splitting unit set corresponding to each production batch. Using the same production batch as a unit, the semi-finished product grade splitting units are classified and statistically analyzed according to the grade destination identifier. The proportion of the weight data of the semi-finished tea corresponding to each grade destination in the total weight data of the semi-finished tea in the production batch is determined to obtain the grade splitting ratio. Within the same production batch, the semi-finished product grade splitting units are sorted according to appearance grade data, color grade data, and sieve particle size distribution data. Based on the grade sorting results and grade destination identifiers, the corresponding semi-finished product grade splitting units for different grade destinations are compared and the cross-grade flow relationship between different grade destinations is extracted. By combining the weight and moisture content data of the semi-finished tea leaves corresponding to each semi-finished product grade division unit, the grade retention amount corresponding to each grade is determined. The grade splitting ratio, cross-grade transfer relationship, and grade retention volume are arranged sequentially according to batch identifier and grade destination identifier to form a semi-finished product grade splitting characterization sequence.

6. The material demand forecasting method for a tea production workshop according to claim 1, characterized in that, The generation of the finished product reconstruction characterization sequence specifically includes: Extract the assembly and feeding data and packaging specification mapping data corresponding to each production batch from the standardized production status sequence; Based on batch identifiers and grade destination identifiers, the blending and feeding data are correlated with the semi-finished product grade splitting characterization sequence to form the blending correlation results corresponding to each production batch. Based on the assembly correlation results, determine the finished product combination relationship corresponding to different grade destinations under each production batch; Based on the packaging specification mapping data, the packaging path is matched to the finished product combination relationship to form the packaging path association result; The assembly association results, finished product combination relationships, and packaging path association results are correlated according to the production batch sequence to generate a finished product reconstruction representation sequence.

7. The material demand forecasting method for a tea production workshop according to claim 1, characterized in that, The generation of the material demand forecast sequence specifically includes: Extract the demand sequence for packaging bags, inner bags, labels, gift boxes, and auxiliary materials for each production batch; The demand sequences for packaging bags, inner bags, labels, and gift boxes are classified and grouped according to production batch, packaging specifications, and packaging route. The packaging bag requirements, inner bag requirements, label requirements, and gift box requirements in each category group are summed to obtain the category packaging requirements for each production batch. The demand sequence of blended auxiliary materials is aggregated according to production batches to obtain the classification of blended auxiliary material demand results for each production batch. By combining the classification and packaging requirements and the classification and blending requirements of each production batch, a material requirement prediction sequence for the tea production workshop is obtained.

8. A material demand forecasting system for a tea production workshop, comprising executing the material demand forecasting method for a tea production workshop as described in any one of claims 1 to 7, characterized in that, include: The data acquisition module is used to acquire production status data of each production batch in the tea production workshop during the screening, selection, grading, blending and packaging stages; The preprocessing module is used to preprocess production status data and generate a standardized production status sequence. The grade splitting construction module is used to construct a set of semi-finished product grade splitting units corresponding to each production batch based on the standardized production state sequence. The splitting characterization module is used to extract the splitting ratio, cross-level flow relationship and level retention of the semi-finished product grade splitting unit set, and form a semi-finished product grade splitting characterization sequence. The finished product reconstruction module is used to combine the assembly and feeding data and packaging specification mapping data in the standardized production state sequence to perform assembly association and packaging path association on the semi-finished product grade splitting characterization sequence, and generate the finished product reconstruction characterization sequence. The demand reverse deduction module is used to reverse-determine the demand sequence for packaging bags, inner bags, labels, gift boxes, and auxiliary materials for each production batch based on the finished product reconstruction representation sequence. The summary output module is used to summarize the demand sequences for packaging bags, inner bags, labels, gift boxes, and blending auxiliary materials to obtain the material demand forecast sequence for the tea production workshop.

9. A material demand forecasting device for a tea production workshop, characterized in that, include: A processor and a memory; the memory is used to store a computer program, and the processor calls the computer program stored in the memory to execute a material demand forecasting method for a tea production workshop as described in any one of claims 1 to 7.

Images