Batch circulation method for tobacco primary processing workshop

Abstract

The invention discloses a batch circulation method for a tobacco primary processing workshop, and the method comprises the steps: generating a unique batch code based on raw material warehousing information, binding a two-dimensional code with an RFID dual identifier, and enabling the batch code to comprise a date, a production line code, a raw material batch number and a sequence number; the tobacco shred making process is divided into the working procedures of raw material moisture regaining, slicing / tip cutting, feeding and leaf moistening, shredding, tobacco shred drying, blending and perfuming and tobacco shred storing, and each working procedure is provided with an independent operation unit and an electronic lock. Intelligent temporary storage bins are arranged at a cut tobacco drying outlet and a blending and perfuming inlet, inter-process residence time is monitored through an MES system, early warning is triggered when time is out, and batches are frozen; when the batch enters each process, the preset process parameter range is automatically loaded, and the equipment is only allowed to operate in the parameter compliance range; process data and quality inspection results of all procedures are collected in real time, and raw material batches and finished silk boxes are associated to form a forward / reverse tracing chain; and starting freezing, reviewing and reworking / scrapping processes for the batches which are detected to be abnormal, and synchronously triggering an equipment self-cleaning program.

Description

Technical Field

[0001] The present invention relates to the technical field of batch circulation, and more particularly to a batch circulation method in a tobacco silk-making workshop. Background Art

[0002] Traditional batch management methods in tobacco silk production suffer from numerous technical flaws. First, the batch traceability system suffers from significant shortcomings, primarily relying on single identification technologies such as paper labels or barcodes. These labels are prone to damage, detachment, or ambiguity in the harsh, high-temperature, and high-humidity environments of the silk workshop, resulting in a break in the batch information chain. Data from various links is stored in isolated systems such as ERP and SCADA, lacking effective linkage and integration mechanisms. When quality traceability is required, manual data retrieval across multiple systems is often required, resulting in inefficient reverse tracing.

[0003] Secondly, the efficiency of inter-process coordination is poor. When the production capacity of upstream and downstream process equipment is mismatched—for example, when the processing speed of the tomato drying machine is lower than the capacity of the tomato cutting machine—manual intervention is often required for material handling and temporary storage. This not only increases labor intensity but also easily leads to material backlogs between processes or equipment idling, seriously affecting the overall utilization rate of equipment. During batch switching, equipment cleaning mainly relies on manual operation, making it difficult to completely remove residual materials and posing a significant risk of cross-batch contamination.

[0004] Quality control also suffers from significant lags. Key process parameters, such as material moisture content and flavoring accuracy, rely primarily on post-production sampling testing. This leads to significant delays in feedback, making real-time quality monitoring during production impossible. Manual recording of process parameters is prone to errors and struggles to meet the stringent data accuracy requirements of modern pharmaceutical manufacturing quality management practices.

[0005] Furthermore, the exception handling mechanism was inadequate. When anomalies such as metallic foreign matter or excessive moisture content occurred, troubleshooting and resolution relied primarily on the operator's experience, a time-consuming and unstandardized process. The lack of systematic control measures for rework led to the recurrence of similar quality issues, severely impacting production efficiency and product quality stability.

[0006] Therefore, how to provide a batch flow method for a tobacco silk workshop is an urgent problem that needs to be solved by those skilled in the art. Summary of the Invention

[0007] In view of this, the present invention provides a batch flow method for a tobacco silk-making workshop, aiming to solve the above-mentioned technical problems.

[0008] In order to achieve the above object, the present invention adopts the following technical solutions:

[0009] A batch flow method for a tobacco silk-making workshop comprises the following steps:

[0010] S1. Generate a unique batch code based on the raw material warehousing information, using a QR code and RFID dual identification binding. The batch code includes the date, production line code, raw material batch number and sequence number;

[0011] S2. Divide the silk-making process into the following steps: raw material rehydration, slicing / cutting, adding ingredients to moisten leaves, shredding, drying, blending and flavoring, and storing. Each step is equipped with an independent operation unit and electronic lock.

[0012] S3. Set up intelligent temporary storage warehouses at the drying outlet and the blending and flavoring entrance. Use the MES system to monitor the retention time between processes. If the timeout exceeds the limit, trigger an early warning and freeze the batch.

[0013] S4. When a batch enters each process, the preset process parameter range is automatically loaded, and the equipment is only allowed to operate within the parameter compliance range;

[0014] S5. Real-time collection of process data and quality inspection results of each process, linking raw material batches with finished silk boxes, forming a forward / reverse traceability chain;

[0015] S6. Initiate the freezing, review, and rework / scrap processes for batches with abnormal detection, and simultaneously trigger the equipment self-cleaning program.

[0016] Furthermore, the S1 is specifically:

[0017] Conduct moisture content, impurity content and grade testing on tobacco raw materials;

[0018] Bind the test data with the raw material origin information, store it in the ERP system, and store it in zones according to origin and grade. The temperature and humidity in the storage area are controlled at 25℃±2℃ and 60%±5%;

[0019] Generate a unique code according to the rule of YYMMDD-production line code-raw material batch number-sequence number;

[0020] Triggering conditions for code generation: Scan the code when raw materials are shipped to confirm that they meet the production formula requirements;

[0021] A high-temperature-resistant and waterproof QR code label is affixed to the material transfer box, and an RFID chip is embedded in the pallet;

[0022] The batch code is linked to the following data through the MES system: raw material quality inspection report, preset process parameters and target finished product specifications;

[0023] AGVs are dispatched to extract raw materials based on the first-in-first-out principle. If the current batch of raw material grade deviates from the formula requirement by more than ±1 level, the system will lock the delivery instruction and issue an alarm.

[0024] After the code is scanned when the goods are shipped out, ERP automatically generates a production batch task order and simultaneously sends it to the control system of each process equipment.

[0025] Furthermore, the electronic lock control logic of each process in step S2 includes:

[0026] After the raw material rehumidification process is completed, when the infrared moisture meter detects that the moisture content reaches 19-21%, the MES system unlocks the entry to the slicing / pointing process;

[0027] When the moisture content at the outlet of the drying process is within the range of 11.5-13.5%, the electronic lock of the blending and flavoring process is released.

[0028] Furthermore, the method for dynamically adjusting the capacity of the intelligent temporary storage warehouse in step S3 is:

[0029] According to the difference in production capacity between the previous and next processes:

[0030] ΔQ = output rate of the previous process - receiving rate of the next process

[0031] Calculate temporary storage demand. When ΔQ is greater than 10% for 5 minutes, activate the backup storage location.

[0032] The AGV transport path is dynamically planned based on batch priority, and the weight of the rework batch path is higher than that of the regular batch.

[0033] Furthermore, the process parameter binding in step S4 includes:

[0034] The error of liquid application in the leaf moistening process is controlled within ±0.5%. If the error does not meet the standard, the shredding process is prohibited.

[0035] The drying process uses PID to control the hot air flow. When the outlet moisture content exceeds the standard for three consecutive times, the batch flow is forcibly interrupted.

[0036] Furthermore, the quality traceability data in step S5 includes:

[0037] Positive traceability: query raw material grade, slice thickness, drying temperature curve, flavoring accuracy and storage environment data through production batch number;

[0038] Reverse tracing: Locate abnormal processes through the traceability code of the finished silk box, and link the operation logs and operator information of all equipment in the batch.

[0039] Furthermore, the equipment self-cleaning procedure in step S6 includes:

[0040] When switching batches, a mixture of compressed air and food-grade alcohol is sprayed into the pipeline. After the residue test meets the standards, the new batch processing is unlocked;

[0041] Cleaning process data is uploaded to the MES system in real time, including cleaning time, pressure value and residue weight.

[0042] Furthermore, it also includes performance optimization modules:

[0043] A tow-bread steam consumption model is established based on historical data. When the ambient humidity is greater than 70%, the preheating temperature is automatically increased by 3-5°C to reduce energy consumption.

[0044] An infrared thermal imager is deployed at the outlet of the wire storage cabinet to detect temperature uniformity and provide feedback to adjust the operating parameters of the constant humidity unit.

[0045] The present invention discloses a batch flow method for a tobacco silk-making workshop. The present invention can achieve full traceability of batch information (forward / reverse traceability time is shortened to within 5 minutes) through dual identification (high-temperature resistant QR code + embedded RFID) and MES data penetration technology, and solve the problem of traceability failure caused by easy damage of traditional single identification and information islands. Dynamic grade verification (error ≤ 1 level) is performed when raw materials are shipped out to eliminate accidents of mismatch between formula and raw material grade from the source. Intelligent temporary storage capacity adaptive adjustment (dynamic expansion based on ΔQ = output rate of the previous process - receiving rate of the next process) reduces the equipment idling rate from 15% to below 5%, and improves the overall equipment efficiency (OEE) to more than 82%. The AGV path dynamic avoidance algorithm, combined with batch priority (rework batch > expiring batch > regular batch), reduces handling conflicts and shortens batch change time by 40%. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are merely embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0047] Figure 1 The present invention is a schematic flow chart of a batch flow method for a tobacco shred workshop. DETAILED DESCRIPTION

[0048] The following is a clear and complete description of the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making any creative efforts are within the scope of protection of the present invention.

[0049] In the traditional production system of tobacco silk-making workshops, batch traceability relies on a single identification system, which is prone to physical damage in high-temperature and high-humidity environments, resulting in a disconnect between raw material properties and process parameters. Distributed storage of multi-source data creates information silos, and reverse tracing requires manual data retrieval across systems, with response cycles exceeding the process control window. Material flow between processes relies on manual scheduling and mechanical buffering, and equipment operating parameters lack coordinated control. Fluctuations in production capacity cause process blockages or idling, and overall equipment efficiency falls below industry benchmarks. There is a time lag between quality inspection data and the production process, making offline spot checks unable to correct process deviations in real time. Manual record-keeping results in insufficient data credibility, making it difficult to meet quality compliance audit requirements. Abnormal event handling relies on manual experience and judgment, lacking standardized handling procedures and equipment linkage mechanisms. This leads to extended production line downtime and an increased risk of batch cross-contamination.

[0050] For example, in the silk-making workshop, during the drying process, when the capacity of the shredder exceeded the drying machine's processing rate, semi-finished products accumulated in the temporary storage area, exceeding the physical capacity. Operators were required to manually adjust the AGV path and shut down upstream equipment. At this point, moisture content test data at the drying outlet had to be transmitted to the central control room via paper documents. Manual data entry delays meant that the parameter exceeding the standard was not discovered until three batches had been processed in the blending and flavoring process. When the metal detector alarm was detected, maintenance personnel had to review the operation logs of six machines in sequence. During this downtime, five subsequent batches were delayed, and production was restarted without performing equipment cleaning procedures, resulting in residual tobacco residue contaminating the new batch of raw materials.

[0051] If these issues are not addressed, a broken batch information chain will render quality traceability ineffective, making it impossible to accurately pinpoint the source of process deviations and increasing the risk of product recalls. Equipment idling and manual scheduling errors will lead to energy waste and lost production capacity, increasing unit production costs. Lagging quality control will result in substandard products continuing to enter subsequent processes, resulting in wasteful raw materials and increased rework costs. Lack of standardization in exception handling procedures will prolong production line recovery time, and residual contamination from equipment can trigger a chain reaction of quality incidents, impacting the timely execution of production plans.

[0052] When faced with the above problems, this application first considers how to establish a full-process data association mechanism to ensure the reliability of batch traceability. The traditional single identification is easily damaged in a complex environment, resulting in data loss. For this, a multi-modal identification binding solution is explored. In response to the problem of low process collaboration efficiency, the material accumulation and idling phenomena caused by production capacity fluctuations between equipment are analyzed, and attempts are made to achieve production capacity balance through dynamic buffering and logical interlocking mechanisms. In terms of quality control, traditional offline sampling has time delays that make it impossible to intercept abnormal batches in time, and it is necessary to build a real-time data collection and parameter boundary control system. For situations where exception handling relies on manual experience, research is conducted on how to standardize the disposal process and link it with the equipment status to avoid the risk of cross-contamination.

[0053] In this regard, the present application proposes: S1. Generate a unique batch code based on the raw material warehousing information, and use a QR code and RFID dual identification to bind. The batch code includes the date, production line code, raw material batch number and serial number; S2. Divide the silk making process into raw material rehydration, slicing / tip cutting, adding materials to moisten leaves, shredding, drying silk, blending and flavoring, and silk storage. Each process is equipped with an independent work unit and electronic lock; S3. Set up an intelligent temporary storage warehouse at the silk drying outlet and the blending and flavoring entrance, monitor the residence time between processes through the MES system, trigger an early warning and freeze the batch when the timeout expires; S4. When the batch enters each process, the preset process parameter range is automatically loaded, and the equipment is only allowed to operate within the parameter compliance range; S5. Collect the process data and quality inspection results of each process in real time, associate the raw material batch with the finished silk box, and form a forward / reverse traceability chain; S6. Initiate the freezing, review and rework / scrap process for batches with abnormal detection, and simultaneously trigger the equipment self-cleaning program.

[0054] The unique batch code is a non-repeatable identifier generated based on the date, production line code, raw material batch number, and sequence number. This can be achieved using a hash algorithm that combines a timestamp with the production line number, ensuring that each production batch has a unique identity. This code, as a data carrier, addresses the issue of information loss caused by the fragility of traditional single identifiers.

[0055] QR code and RFID dual-identification refers to the simultaneous use of optical recognition and radio frequency identification technologies for data storage. This is achieved by combining a QR code tag made of heat-resistant and waterproof material with an embedded RFID chip. The two identifiers serve as a backup for enhanced data reading reliability. This technology overcomes the problem of tag shedding or damage in high-humidity workshop environments.

[0056] Independent work cells and electronic locks establish physically isolated areas for each process, equipped with access control systems. These can be achieved through PLC-controlled electromagnetic locks linked to infrared sensors. Interlocking mechanisms between processes ensure the controllable production sequence. This eliminates the risk of process disruptions caused by manual handling.

[0057] Smart temporary storage devices are buffer storage devices with dynamic capacity adjustment capabilities. They use laser ranging sensors to monitor material accumulation in real time and, combined with AGV scheduling algorithms, automatically expand storage locations. This device balances capacity differences between upstream and downstream processes, reducing equipment idling and material backlogs.

[0058] The preset process parameter range refers to the permissible operating range of the equipment. Specifically, standard process parameters can be stored in a database and loaded through the boundary condition setting module of the equipment control system. This setting prevents manual operation from causing parameter deviations outside the permissible range.

[0059] The forward / reverse traceability chain is a bidirectional link linking raw material and finished product quality data. Specifically, blockchain technology can be used to record the timestamp and hash value of each process data, achieving tamper-proof full-process tracking. This mechanism breaks through the traceability efficiency bottleneck of traditional decentralized data management.

[0060] The equipment's self-cleaning program is an automated cleaning process that occurs during batch changes. This process utilizes a mixed injection system of compressed air and food-grade alcohol, coupled with a residual weight sensor to monitor cleaning effectiveness. This program eliminates the risk of cross-contamination caused by incomplete manual cleaning.

[0061] Furthermore, the silk-making process is divided into raw material conditioning, slicing / tip cutting, leaf conditioning, shredding, drying, blending and flavoring, and storage. Each process is equipped with an independent work unit and electronic lock. The electronic lock triggers the unlocking mechanism based on preset conditions, achieving physical isolation and logical interlocking between processes, preventing process disruptions caused by manual intervention.

[0062] Smart temporary storage bins are installed at the tofu drying exit and the blending and flavoring entrance. The MES system monitors the retention time between processes. When the retention time exceeds a preset threshold, the system triggers an alert and freezes the batch to prevent material accumulation or equipment idling. The capacity of the smart temporary storage bins can be dynamically adjusted based on the production capacity difference between the previous and next processes, and AGV transport routes are planned based on batch priority.

[0063] When a batch enters each process, the system automatically loads the preset process parameter range, allowing the equipment to operate only within the parameter compliance range. For example, the amount of liquid applied during the leaf moistening process is controlled within a specific range, and the hot air flow rate in the tofu drying process is controlled using PID. This ensures standardized process execution and prevents the accumulation of parameter deviations.

[0064] As a result, process data and quality inspection results for each process are collected in real time and linked to raw material batches and finished yarn boxes, forming a forward and reverse traceability chain. Forward traceability allows for querying information such as raw material grade and processing parameters, while reverse traceability allows for locating abnormal processes through the finished yarn box traceability code.

[0065] For batches detected as abnormal, the system initiates a freezing, review, and rework / scrap process, simultaneously triggering the equipment's self-cleaning routine. This involves injecting a mixture of compressed air and food-grade alcohol into the pipes, with cleaning process data uploaded to the MES system in real time. This mechanism prevents the spread of problems while eliminating cross-contamination risks, forming a proactive defense system.

[0066] When raw materials enter the warehouse, the system generates a batch code in the format of YYMMDD-production line code-raw material batch number-sequence number. High-temperature and waterproof QR code labels are affixed to the material transfer boxes, and RFID chips are embedded in the pallets. The MES system links the batch code with data such as raw material quality inspection reports, preset process parameters, and target finished product specifications.

[0067] The silk-making process is divided into seven steps: raw material conditioning, slicing / tip cutting, adding ingredients to moisten leaves, shredding, drying, blending and flavoring, and storing. Each step is equipped with an independent work unit and electronic lock. For example, after the raw material conditioning step is completed, the MES system unlocks the entry to the slicing / tip cutting step when the infrared moisture meter detects that the moisture content reaches the preset range.

[0068] Smart temporary storage bins are installed at the tofu drying exit and the flavoring entrance. The MES system monitors the dwell time between processes in real time. When the dwell time exceeds a preset threshold, the system triggers an alert and freezes the batch. The capacity of the smart temporary storage bins is dynamically adjusted based on the production capacity difference between the previous and next processes. If the capacity difference exceeds a specific percentage for a preset period of time, the system activates a backup storage location.

[0069] As a batch enters each process, the system automatically loads pre-set process parameters. For example, the liquid dosage during the leaf moistening process is controlled within a specific tolerance; if this tolerance is not met, the system will not proceed to the shredding process. The shredded rice drying process uses PID control for hot air flow. If the outlet moisture content exceeds the standard multiple times in a row, the system will forcibly interrupt the batch flow.

[0070] Process data and quality inspection results from each process are collected in real time and uploaded to the MES system. Forward traceability allows users to query raw material grade, slice thickness, drying temperature curve, flavoring accuracy, and storage environment data using the production batch number. Reverse traceability uses the traceability code on the finished silk box to locate abnormal processes and link the operation logs and operator information of all equipment in that batch.

[0071] When an abnormal batch is detected, the system automatically initiates the freezing, review, and rework / scrap processes. Simultaneously, the equipment's self-cleaning process is triggered, injecting a mixture of compressed air and food-grade alcohol into the pipes. Cleaning process data, including cleaning time, pressure, and residue weight, is uploaded to the MES in real time. Once residue testing meets standards, the system unlocks a new batch for processing.

[0072] Tobacco leaf raw materials are tested for moisture content, impurity content, and grade. Moisture content is measured using an infrared moisture meter with an accuracy of ±0.1%. Impurity content is measured using a near-infrared spectrometer, capable of simultaneously analyzing multiple impurity components. Grade testing is performed by professional graders based on indicators such as leaf color and leaf integrity.

[0073] Test data is linked to raw material origin information and stored in the ERP system, where it is partitioned by origin and grade. A raw material information database is established within the ERP system, including fields such as origin, grade, moisture content, and impurity content. Warehouses are organized into dedicated areas based on origin and grade, each equipped with independent temperature and humidity control equipment. The temperature and humidity in the warehouse are maintained at 25°C ± 2°C and 60% ± 5%.

[0074] A unique code is generated using the format YYMMDD-production line code-raw material batch number-sequence number. For example, 230615-A01-20230501001-001 indicates the 001st box of finished product produced on June 15, 2023, by production line A01 using raw materials from batch 001 received on May 1, 2023.

[0075] The trigger for code generation is scanning the barcode on the raw material packaging to confirm that it meets the production formula requirements when the raw material is shipped out. Warehouse managers use handheld terminals to scan the barcode on the raw material packaging. The system automatically compares the raw material information with the production formula requirements and generates a new batch code after confirming a match.

[0076] A heat-resistant and waterproof QR code label is affixed to the material transfer box, and an RFID chip is embedded in the pallet. The QR code label is made of heat-resistant polyester material and can withstand temperatures of 200°C. The RFID chip uses the UHF frequency band and has a read range of up to 5 meters.

[0077] The MES system links the batch code with the raw material quality inspection report, preset process parameters, and target finished product specifications. The MES system establishes a batch information association table containing fields such as the batch code, raw material quality inspection report number, process parameter number, and finished product specification number.

[0078] AGVs are dispatched to extract raw materials based on the first-in, first-out principle. The AGV dispatch system prioritizes raw materials that arrived the earliest. If the raw material grade of the current batch deviates from the recipe by more than ±1 level, the system locks the delivery instruction and issues an alarm.

[0079] After the product is shipped out of the warehouse and scanned, the ERP automatically generates a production batch order and simultaneously sends it to the equipment control systems of each process. The order includes the batch code, raw material information, process parameters, and other information, and is sent to each process PLC in real time via industrial Ethernet.

[0080] Temporary storage demand is calculated based on the capacity difference between the previous and next processes: ΔQ = previous process output rate - next process receiving rate. When ΔQ remains > 10% for 5 minutes, the backup storage location is activated. The AGV transport path is dynamically planned based on batch priority, and the rework batch path weight is higher than the regular batch path.

[0081] The capacity difference calculation model is configured to acquire real-time data on the material transfer rates of the preceding and following processes. The sensor network collects the instantaneous flow rates at the drying outlet and the blending and flavoring inlet, updating the ΔQ value every minute. If the output rate of the preceding process exceeds the receiving rate of the subsequent process by more than 10% for five consecutive minutes, the temporary storage expansion command is triggered, and the foldable storage space is deployed via a hydraulic lifting device, increasing the available storage area by 20%-30%. The AGV scheduling system has a built-in path optimization algorithm. When generating transport routes, the priority coefficient of rework batches is set to 1.5-2 times that of regular batches. This shortens the travel interval between path nodes and prioritizes the allocation of barrier-free passages.

[0082] Specifically, a laser flowmeter installed at the exit of the tobacco drying process monitors the tobacco output rate in real time, and this data is transmitted to the MES system via Industrial Ethernet. A weighing sensor at the entrance of the blending and flavoring process simultaneously collects the material receipt rate, and the system performs a ΔQ difference calculation every 30 seconds. When ΔQ is detected to consistently exceed a set threshold, the central controller sends a capacity expansion signal to the temporary storage warehouse, automatically extending the slide rail mechanism of the spare storage space to create additional buffer space. Simultaneously, the AGV scheduling module dynamically adjusts path weights based on batch type, assigning rework batches higher access rights in path planning. For example, when a regular batch and a rework batch simultaneously request the same channel, the system prioritizes a straight path for the rework batch, while assigning a detour to the regular batch. This collaborative control mechanism links the physical expansion of the temporary storage warehouse with logistics scheduling, effectively alleviating material backlogs caused by production capacity fluctuations while ensuring that rework batches can quickly enter the processing flow, avoiding production delays caused by confused priorities.

[0083] The method for dynamically adjusting the capacity of an intelligent temporary storage warehouse can include the following steps: First, real-time monitoring of the output rate of the upstream process and the receiving rate of the downstream process. For example, the output rate of the upstream process's tofu drying machine is 1000 kg / h, while the receiving rate of the downstream process's blending and flavoring machine is 900 kg / h. Then, the capacity difference ΔQ is calculated: ΔQ = 1000 - 900 = 100 kg / h. If ΔQ remains greater than 10% for 5 minutes, the system automatically activates the backup storage location. In this example, ΔQ = 100 kg / h, representing 10% of the upstream process's output rate. If this condition persists for 5 minutes, activation of the backup storage location is triggered. The backup storage location can be a pre-configured additional temporary storage space connected to the main temporary storage warehouse via a conveyor or automated guided vehicle (AGV). AGV transport routes are dynamically planned based on batch priority, with rework batches given a higher path weight. In specific implementations, regular batches can be given a weight of 1, while rework batches can be given a weight of 2. The path planning algorithm prioritizes batches with higher weights for transport, ensuring timely processing of rework batches.

[0084] This application further proposes a dynamic capacity adjustment method based on the capacity difference between the previous and next processes, as well as a priority planning strategy for AGV transport paths.

[0085] Dynamic capacity adjustment calculates the difference between the output rate of the preceding process and the receiving rate of the following process. It activates the backup storage location when the capacity deviation exceeds 10% and persists for 5 minutes. This threshold is based on experimental data from the silk-making workshop's production cycle and effectively filters out transient fluctuations. When activated, the backup storage location is physically connected to the main temporary storage warehouse via a telescopic conveyor, expanding the storage area by 35%-40%.

[0086] AGV path planning uses a weighted approach, assigning rework batches two to three times the weight of regular batches. Path weighting is implemented through the AGV scheduling system's priority queue. When multiple batches of transport requests are received concurrently, the system automatically calculates the expected time cost for each batch to reach the target workstation, prioritizing the most heavily weighted tasks.

[0087] Specifically, during the material flow between the tofu drying outlet and the flavoring and blending entrance, the MES system continuously collects operational data from the equipment in the preceding and following processes. If the system detects that the output per unit time of the preceding process consistently exceeds the output received by the subsequent process by 10%, it identifies a structural capacity imbalance and triggers a pneumatic damper to open the entrance to the backup storage location. Once the backup storage location is activated, the effective capacity of the main temporary storage warehouse is expanded from the standard 12 pallets to 16, and the buffer time is extended by approximately 25 minutes until the capacity difference returns to within the threshold range.

[0088] When the AGV performs a handling task, the system identifies rework batches by parsing the status identifier in the batch code. Rework batches are assigned the highest access rights in path planning, and their transportation routes avoid conventional material transportation routes and instead go directly to the target process via dedicated channels. This path planning uses a laser navigation system to update map data in real time, ensuring that the average transit time of rework batches is reduced to less than 60% of that of regular batches. Furthermore, the physical storage area of ​​rework batches within the temporary storage warehouse is designated as an independent isolation zone, and the storage time is strictly limited to 30 minutes to prevent the risk of cross-contamination.

[0089] Through this approach, the intelligent temporary storage warehouse can flexibly expand capacity based on actual production fluctuations, increasing storage utilization to over 92%. It also boosted rework batch transfer efficiency by 40%, reducing the standard deviation of inter-process hold times from ±18 minutes to ±6 minutes. This technical solution integrates closed-loop control between capacity matching detection and logistics scheduling, creating a dynamic adjustment mechanism that adapts to production fluctuations.

[0090] The method for dynamically adjusting the capacity of the intelligent temporary storage warehouse includes the following steps:

[0091] First, the output rate of the previous process and the reception rate of the subsequent process are monitored in real time. The output rate of the previous process can be obtained by a counting sensor at the exit of the tofu drying machine, while the reception rate of the subsequent process is measured by a weighing device at the entrance of the blending and flavoring process.

[0092] Next, calculate the capacity difference ΔQ between the previous and next processes. The system collects data once a minute and calculates ΔQ = the output rate of the previous process - the receiving rate of the next process.

[0093] Next, the system determines whether ΔQ continues to exceed 10%. The system continuously records the ΔQ value for 5 minutes. If all 5 sampling points are greater than 10%, the temporary storage capacity adjustment mechanism is triggered.

[0094] Finally, the backup storage space is activated. The temporary storage warehouse is equipped with retractable storage spaces, which are normally retracted. When expansion conditions are met, the control system issues a command, driving the motor to expand the storage space, increasing temporary storage capacity by 50%.

[0095] AGV transport routes are dynamically planned based on batch priority. The system assigns a priority weight to each batch, with regular batches receiving a weight of 1 and rework batches receiving a weight of 2. The AGV scheduling algorithm prioritizes batches with higher weights when planning routes. For example, if both regular and rework batches are being transported, the system assigns the AGV to the rework batch first, ensuring faster movement.

[0096] The error of the amount of liquid applied in the leaf moistening process is controlled at ±0.5%. If the standard is not met, the shredding process is prohibited. The shredded leaves process uses PID to control the hot air flow. When the outlet moisture content exceeds the standard for three consecutive times, the batch flow is forcibly interrupted.

[0097] The error control of liquid application is achieved through the linkage of a mass flowmeter and a servo valve. The flowmeter monitors the deviation between the actual application amount and the set value in real time. When the absolute value of the deviation exceeds 0.5%, the servo valve is locked in the closed position. The PID control parameters for the tofu drying process include a proportional band set to 5%-8%, an integral time set to 120-180 seconds, and a differential time set to 30-60 seconds. The opening of the hot air flow regulating valve is dynamically adjusted based on the outlet moisture content. The logic for three consecutive detections exceeding the standard is implemented through a timer module. The detection interval is set to 30 seconds as required by the process standard, and the three violations must occur within 90 consecutive seconds.

[0098] Specifically, during the leaf-feeding and moistening process, a mass flow meter collects the instantaneous flow rate of the feed liquid every 5 seconds. The accumulated flow error is calculated in real time and compared against a threshold of ±0.5%. If the accumulated error exceeds the allowable range, the power to the drive motor of the inter-process conveyor belt is cut off, and the damper at the inlet of the shredder is simultaneously driven to a closed position by a pneumatic cylinder. During the shredded tofu drying process, an online moisture content detector outputs a measurement value every 30 seconds. If three consecutive measured values ​​exceed the range of 11.5%-13.5%, the pneumatic shut-off valve on the batch flow path immediately closes, isolating the current batch in the shredded tofu outlet buffer. This control logic is linked to the electronic lock module of the MES system. When an interrupt command is triggered, the batch task order is frozen synchronously, preventing equipment preparation for subsequent processes. By improving the flow control accuracy to ±0.5%, the standard deviation of the tobacco absorption rate was reduced from 1.2% of the traditional method to 0.3%; after adopting the three-time continuous detection mechanism, the misjudgment rate of moisture content control was reduced from 15% of a single test to below 2%, and the interception time of unqualified batches was shortened from 30 minutes of traditional manual sampling to real-time interception.

[0099] During the leaf-feeding and moistening process, the amount of liquid applied is monitored in real time by a high-precision flow meter. Utilizing the Coriolis principle, the flow meter boasts a measurement accuracy of ±0.1%. The control system calculates the target application rate based on material weight and a pre-set recipe, achieving precise control through variable-frequency pump speed regulation. If the actual application rate deviates from the target by more than ±0.5%, the system automatically triggers an alarm and stops the conveyor belt, preventing unqualified material from entering the shredding process.

[0100] During the tofu drying process, a PID control algorithm dynamically adjusts the hot air flow rate. The outlet moisture content is sampled every 30 seconds by an online near-infrared moisture meter with a measurement accuracy of ±0.2%. The control system compares the measured value with the set target, calculates the deviation, and outputs an adjustment signal based on the PID parameters to drive the variable frequency fan to adjust the hot air flow rate. If the moisture content test results exceed the range of 11.5%-13.5% for three consecutive times, the system automatically closes the tofu drying machine feed gate and notifies the MES system to interrupt the batch flow.

[0101] Forward traceability uses the production batch number to query the raw material grade, slice thickness, drying temperature curve, flavoring accuracy and storage environment data; reverse traceability uses the finished silk box traceability code to locate abnormal processes and link the operation logs and operator information of all equipment in the batch.

[0102] Among them, the production batch number serves as a unique identifier throughout the entire process data chain. The association between raw material grade and slice thickness is realized through the raw material quality inspection report stored in the batch code. The wire drying temperature curve stores the temperature value collected every 30 seconds with the timestamp as the index. The traceability code of the finished wire box is generated by laser engraving technology, and a two-way mapping relationship is established with the production batch number. The equipment operation log records include motor current fluctuation values, valve opening change curves and sensor calibration records. The operator information is bound to the fingerprint recognition system login record and the parameter modification operation log. When it is detected that the humidity of the wire storage environment exceeds the standard, the hot air valve opening curve of the batch of wire drying machines and the operator's parameter adjustment time node can be retrieved through reverse tracing.

[0103] Specifically, during forward tracing, raw material grade data and slice thickness parameters are linked using the batch number field in the batch code. When a slice thickness deviation exceeding ±0.1mm is detected, the system automatically compares the raw material grade within the same batch to ensure it meets the preset processing precision requirements. The drying temperature curve is stored as time series data, with each temperature sampling point associated with the drying drum speed and hot air flow parameters, forming a multidimensional parameter matrix. When reverse tracing is triggered, the finished yarn box traceability code is parsed into the original production batch number, and equipment logs for all processes in that batch are automatically retrieved. This includes records of cutter tool wear, drying heat exchanger efficiency curves, and the operating status of the flavoring machine metering pump. Operator information is linked to the operation logs of the equipment control terminal through the MES system, allowing the relationship between parameter modification records and equipment status changes at specific points in time to be traced. If anomalies in flavoring precision are detected, reverse tracing can simultaneously retrieve the work order receipt time, equipment parameter adjustment records, and concurrent equipment maintenance logs of the flavoring process operator for that batch, enabling cross-analysis of multidimensional factors across the human-machine-material-method-environmental chain. This technical solution uses a two-way data association mechanism to shorten the time for locating abnormal processes from 2 hours with traditional methods to within 10 minutes, and improves the completeness of equipment status parameter traceability to 98%.

[0104] A bidirectional traceability mechanism is employed in quality traceability data management. Forward traceability uses the production batch number as a unique identifier, correlating raw material grade, slice thickness, drying temperature profile, flavoring accuracy, and storage environment data. For example, for batch number "20230501-A01-123456-001," the query indicates that this batch used B2F grade tobacco leaves, a slice thickness of 0.35mm, a drying temperature profile fluctuating between 80-120°C, a flavoring accuracy of ±0.02%, a storage temperature of 22°C, and a relative humidity of 55%.

[0105] Reverse tracing uses the traceability code of the finished silk box to locate abnormal processes. When an abnormality is detected in the finished silk box number "SX20230501-001," the system automatically links the operation logs and operator information of all equipment in that batch. Specifically, the system traces back to the speed fluctuations of cutter #3 between 9:15 and 9:30 AM, which caused operator Zhang San to adjust the parameters at 9:25 AM. Simultaneously, the temperature sensor of dryer #2 displayed abnormal readings between 9:40 and 10:00 AM, which were subsequently calibrated by maintenance technician Li Si.

[0106] This two-way traceability mechanism provides data penetration throughout the entire process, from raw materials to finished product, and can accurately pinpoint the specific causes of process anomalies. By linking equipment operation logs with operator information, equipment failures or human errors can be quickly identified, providing accurate and timely data support for quality anomaly analysis.

[0107] The equipment's self-cleaning procedure includes: when switching batches, a mixture of compressed air and food-grade alcohol is sprayed into the pipeline, and a new batch of processing is unlocked after the residue test meets the standards; the cleaning process data is uploaded to the MES system in real time, including cleaning time, pressure value and residue weight.

[0108] The injection pressure of compressed air and food-grade alcohol is controlled within the range of 0.3-0.6 MPa, and the alcohol concentration is 75% ± 5% food-grade ethanol solution. The mixed gas forms a turbulent flow through a porous nozzle to cover the inner wall of the pipe. Residue detection uses a near-infrared spectroscopy online analyzer, and the detection threshold is set to a residue weight of ≤ 0.1g / m 2 The test results are transmitted to the PLC controller via the industrial bus. Cleaning time is recorded accurately to the second level, pressure values ​​are collected once per second, and the residue weight data is automatically uploaded after calibration using an electronic balance.

[0109] Specifically, when the batch switching signal is triggered, the injection system of the compressed air and alcohol mixture is started, and the pressure of the mixed gas is controlled by adjusting the opening of the solenoid valve to form an effective airflow flushing effect in the pipeline. The volatility of alcohol causes it to evaporate quickly after cleaning, avoiding residual liquid from causing microbial growth. During the cleaning process, the pressure sensor monitors the pressure fluctuations inside the pipeline in real time. If the pressure value is detected to be lower than the preset threshold of 0.3MPa, the system automatically replenishes the pressure to ensure the cleaning effectiveness. The weight of the residue is collected by the weighing sensor. When the detection values ​​are all lower than 0.1g / m3 for three consecutive times, the pressure is automatically replenished to ensure the cleaning effectiveness. 2When the batch lock is released, the batch transfer lock is automatically unlocked. All cleaning parameters are uploaded to the MES system via the OPC protocol, creating a triple-verified record with timestamp, equipment number, and operator, ensuring data cannot be tampered with. This solution achieves 360-degree cleaning of the pipeline interior through the synergistic effect of physical cleaning and chemical disinfection, while also establishing a complete electronic cleaning log to meet the mandatory data integrity requirements of production audits.

[0110] The equipment self-cleaning procedure includes the following steps: First, when the batch is switched, a mixture of compressed air and food-grade alcohol is sprayed into the pipeline. Among them, the compressed air pressure can be set to 0.6-0.8MPa, and the concentration of food-grade alcohol is 75%. The spraying time lasts 120 seconds to ensure that the mixture fully contacts the inner wall of the pipeline. Next, a photoelectric sensor is used to detect the residue content at the pipeline outlet. When the residue concentration is lower than 10ppm, the cleaning standard is determined to be met. After meeting the standard, the system automatically unlocks the new batch processing flow. Furthermore, key data of the cleaning process is uploaded to the MES system in real time. Specifically, the uploaded data includes the start and end time of cleaning, the spray pressure value (recorded every 5 seconds), and the weight of the residue (measured by a high-precision electronic scale). Thus, a complete electronic record of the cleaning process is formed, which is convenient for subsequent audits and analysis.

[0111] The efficiency optimization module includes: establishing a silk-drying steam consumption model based on historical data, automatically increasing the preheating temperature by 3-5°C to reduce energy consumption when the ambient humidity is greater than 70%; deploying an infrared thermal imager at the outlet of the silk storage cabinet to detect temperature uniformity and provide feedback to adjust the operating parameters of the constant humidity unit.

[0112] The towheat steam consumption model analyzes historical operating data from at least three production quarters to establish a correlation curve between ambient humidity and steam consumption. When the humidity sensor detects that the current ambient humidity exceeds a set threshold, the temperature control instructions for the preheating phase are automatically adjusted. For example, when the humidity reaches 75%, the preheating temperature can be raised to the upper limit of the reference temperature range, reducing the amount of steam required to maintain the target temperature in the subsequent phases.

[0113] At the wire storage cabinet outlet, an infrared thermal imager scans twice per second to collect temperature distribution data. The detected temperature data is filtered to generate a temperature standard deviation indicator. When this indicator exceeds 0.8°C, the constant humidity unit's circulating air volume is increased by 10%-15%. Furthermore, the coordinates of the temperature abnormality area are converted into opening adjustment instructions for the constant humidity unit's air outlets. Specifically, the opening of the corresponding air outlet in the area with detected high temperatures is increased by 3-5%, while the opening of the air outlet in the low-temperature area is correspondingly reduced.

[0114] Specifically, the implementation process of the tow-bread drying steam consumption model includes: installing a temperature and humidity sensor at the tow-bread drying machine feed end to collect real-time ambient humidity data; inputting the collected data into a trained regression model to output a preheat temperature compensation value; and finally, a PLC controller adjusts the heater power setpoint based on the compensation value. When the ambient humidity exceeds 70%, every 1°C increase in preheat temperature reduces subsequent steam consumption by 2.3%-4.1%. This adjustment process is linked to the closed-loop moisture content control of the tow-bread drying process as specified in the preceding claims. After the preheat temperature is adjusted, the moisture content detection frequency at the tow-bread drying outlet is automatically increased to once per minute, ensuring that process parameters remain within the regulatory compliance range.

[0115] During the wire storage process, infrared thermal imagers are deployed 0.5-1.2 meters from the discharge conveyor, achieving an imaging resolution of 0.1°C. Temperature data is transmitted via Industrial Ethernet to the constant humidity control unit's control cabinet. A control algorithm generates three-dimensional adjustment parameters based on the temperature distribution thermodynamic map. This adjustment process forms a closed-loop data loop with the electronic lock control of the wire storage process. The electronic lock at the discharge outlet is released only when temperature uniformity reaches the set standard, ensuring that batches under abnormal temperature control conditions do not flow into the next process.

[0116] An LSTM neural network algorithm is integrated into the tow-fiber drying process control system to construct a steam consumption prediction model. This model uses ambient humidity, initial moisture content of raw materials, and preheating temperature as input nodes. By training on 100,000 sets of process parameters from five years of historical production data, it establishes a nonlinear mapping relationship between humidity and steam consumption. When the online ambient humidity exceeds 70%, the preheating temperature is raised to the set value of +4°C by adjusting the PID controller of the preheating heater. In the tow-fiber storage process, a dual-band infrared thermal imager with operating wavelengths of 8-14μm and 3-5μm is used to scan the temperature field at the discharge port at a sampling rate of 30 frames per second. The temperature standard deviation is calculated using an image processing algorithm. If the standard deviation exceeds 0.8°C, the temperature distribution data is fed into the fuzzy control system of the constant humidity unit, which dynamically adjusts the air guide vane opening and humidifier spray volume to maintain temperature uniformity within the cabinet within ±0.5°C.

[0117] The efficiency optimization module collects data from environmental sensors in the tobacco drying process. When the humidity exceeds the 70% threshold, it uses a steam consumption model to calculate a preheat temperature compensation value. This preheat temperature increase is achieved by adjusting the heater power using a PID controller, with the compensation range maintained between 3 and 5°C to prevent sudden temperature rises that could cause tobacco charring. Simultaneously, an infrared thermal imager in the tobacco storage process scans the entire area every five minutes, identifying areas with abnormal temperature standard deviations exceeding ±1.2°C and generating thermal map coordinate data. This coordinate data is transmitted to the constant humidity unit PLC via the OPC protocol, dynamically adjusting the air outlet guide vane angle and humidifier spray volume to improve temperature uniformity to within ±0.8°C. The preheat temperature compensation in the tobacco drying process and the temperature control in the tobacco storage process form a closed-loop linkage. If the moisture content of the tobacco at the drying outlet fluctuates by more than ±0.3%, the tobacco storage constant humidity unit is triggered to enter a standby mode to prevent sudden temperature and humidity fluctuations from affecting tobacco quality. By combining historical data modeling with a real-time feedback mechanism, steam consumption can be reduced by 12-15% when humidity exceeds the standard, and the temperature uniformity of the wire storage cabinet can be improved by more than 40%.

[0118] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0119] The above description of the disclosed embodiments is intended to enable one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be readily apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention is not limited to the embodiments shown herein but is intended to conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A batch flow method for a tobacco silk workshop, characterized in that: The following steps are involved: S1. Generate a unique batch code based on the raw material warehousing information, using a QR code and RFID dual identification binding. The batch code includes the date, production line code, raw material batch number and sequence number; S2. Divide the silk-making process into the following steps: raw material rehydration, slicing / cutting, adding ingredients to moisten leaves, shredding, drying, blending and flavoring, and storing. Each step is equipped with an independent operation unit and electronic lock. S3. Set up intelligent temporary storage warehouses at the drying outlet and the blending and flavoring entrance. Use the MES system to monitor the retention time between processes. If the timeout exceeds the limit, trigger an early warning and freeze the batch. S4. When a batch enters each process, the preset process parameter range is automatically loaded, and the equipment is only allowed to operate within the parameter compliance range; S5. Real-time collection of process data and quality inspection results of each process, linking raw material batches with finished silk boxes, forming a forward / reverse traceability chain; S6. Initiate the freezing, review, and rework / scrap processes for batches with abnormal detection, and simultaneously trigger the equipment self-cleaning program.

2. The batch flow method of a tobacco silk workshop according to claim 1, characterized in that: The S1 is specifically: Conduct moisture content, impurity content and grade testing on tobacco raw materials; Bind the test data with the raw material origin information, store it in the ERP system, and store it in zones according to origin and grade. The temperature and humidity in the storage area are controlled at 25℃±2℃ and 60%±5%; Generate a unique code according to the rule of YYMMDD-production line code-raw material batch number-sequence number; Triggering conditions for code generation: Scan the code when raw materials are shipped to confirm that they meet the production formula requirements; A high-temperature-resistant and waterproof QR code label is affixed to the material transfer box, and an RFID chip is embedded in the pallet; The batch code is linked to the following data through the MES system: raw material quality inspection report, preset process parameters and target finished product specifications; AGVs are dispatched to extract raw materials based on the first-in-first-out principle. If the current batch of raw material grade deviates from the formula requirement by more than ±1 level, the system will lock the delivery instruction and issue an alarm. After the code is scanned when the goods are shipped out, ERP automatically generates a production batch task order and simultaneously sends it to the control system of each process equipment.

3. The batch flow method of a tobacco silk workshop according to claim 1, characterized in that: The electronic lock control logic of each process in step S2 includes: After the raw material rehumidification process is completed, when the infrared moisture meter detects that the moisture content reaches 19-21%, the MES system unlocks the entry to the slicing / pointing process; When the moisture content at the outlet of the drying process is within the range of 11.5-13.5%, the electronic lock of the blending and flavoring process is released.

4. The batch flow method of a tobacco silk workshop according to claim 1, characterized in that: The method for dynamically adjusting the capacity of the intelligent temporary storage warehouse in step S3 is: According to the difference in production capacity between the previous and next processes: ΔQ = output rate of the previous process - receiving rate of the next process Calculate temporary storage demand. When ΔQ is greater than 10% for 5 minutes, activate the backup storage location. The AGV transport path is dynamically planned based on batch priority, and the weight of the rework batch path is higher than that of the regular batch.

5. The batch flow method of a tobacco silk workshop according to claim 1, characterized in that: The process parameter binding in step S4 includes: The error of liquid application in the leaf moistening process is controlled within ±0.5%. If the error does not meet the standard, the shredding process is prohibited. The drying process uses PID to control the hot air flow. When the outlet moisture content exceeds the standard for three consecutive times, the batch flow is forcibly interrupted.

6. The batch flow method of a tobacco silk-making workshop according to claim 1, characterized in that: The quality traceability data in step S5 includes: Positive traceability: query raw material grade, slice thickness, drying temperature curve, flavoring accuracy and storage environment data through production batch number; Reverse tracing: Locate abnormal processes through the traceability code of the finished silk box, and link the operation logs and operator information of all equipment in the batch.

7. The batch flow method of a tobacco silk-making workshop according to claim 1, characterized in that: The equipment self-cleaning procedure in step S6 includes: When switching batches, a mixture of compressed air and food-grade alcohol is sprayed into the pipeline. After the residue test meets the standards, the new batch processing is unlocked; Cleaning process data is uploaded to the MES system in real time, including cleaning time, pressure value and residue weight.

8. The batch flow method of a tobacco silk-making workshop according to claim 1, characterized in that: Also includes performance optimization modules: A tow-bread steam consumption model is established based on historical data. When the ambient humidity is greater than 70%, the preheating temperature is automatically increased by 3-5°C to reduce energy consumption. An infrared thermal imager is deployed at the outlet of the wire storage cabinet to detect temperature uniformity and provide feedback to adjust the operating parameters of the constant humidity unit.

Images