Tobacco leaf bundling data networking management system and method based on 5G cloud platform

Abstract

The invention relates to the technical field of tobacco leaf bundling management, and discloses a tobacco leaf bundling data networking management system and method based on a 5G cloud platform. The system comprises a cigarette packet density portrait generation module, a real-time compression portrait generation module, a compression deviation analysis module, a fluctuation threshold judgment module and a cigarette packet data cleaning module. A cigarette packet density portrait generation module obtains a density backtracking portrait of a target batch cigarette packet compression dimension acquisition moment in a preset period, wherein the density backtracking portrait comprises an initial compression duration and a pressure parameter; a real-time compressed portrait generation module obtains a real-time compressed portrait containing the current compression duration and the pressure parameter; the compression deviation analysis module generates a deviation vector based on the parameters; a fluctuation threshold value judgment module retrieves a matched deviation module value set, and when the variance exceeds a preset threshold value, the cigarette packet data deletion proportion is calculated; and the cigarette packet data cleaning module deletes related data from the storage starting moment according to the proportion.

Description

Technical Field

[0001] The present invention relates to the technical field of tobacco leaf baling management, and specifically to a tobacco leaf baling data networking management system and method based on a 5G cloud platform. Background Art

[0002] In the tobacco production process, quality control in the baling stage has a direct impact on subsequent storage, transportation, and processing. Currently, data management in the tobacco baling process relies heavily on traditional local recording methods, which limits the scope of data collection and makes it difficult to achieve information exchange between different links and equipment.

[0003] Under existing tobacco baling data management models, monitoring of key parameters during the tobacco bale compression process, such as compression duration and pressure, is often incomplete and incomplete. In most cases, only partial parameter data is recorded, failing to form a complete picture of the compression process. This significantly limits the ability to trace and analyze the compression status of tobacco bales. Due to the lack of a networked management mechanism, data generated by various bundling machines is stored in separate locations, making centralized integration and comparative analysis difficult. When evaluating the compression quality of a particular batch of cigarette packs, manual data retrieval from each device is required, which is not only time-consuming and labor-intensive, but also prone to data omissions and errors. As bundled data accumulates, storage capacity gradually increases, but a large amount of redundant or invalid data is also present. Existing data management methods lack an effective data cleanup mechanism to filter and delete unnecessary data based on the actual conditions of the cigarette pack compression process. This not only consumes a large amount of storage resources but also affects data processing efficiency. When the compression quality of cigarette packs fluctuates, traditional management methods are unable to quickly locate the source of the deviation and cannot adjust management strategies in time according to changes in compression parameters. This may cause unqualified cigarette packs to enter subsequent links and affect the overall quality of tobacco leaf products. Summary of the Invention

[0004] The purpose of the present invention is to provide a tobacco baling data networking management system and method based on a 5G cloud platform to solve the problems raised in the above background technology.

[0005] To achieve the above objectives, the present invention provides a tobacco leaf baling data network management system based on a 5G cloud platform, the system comprising: A cigarette pack density profile generation module is used to obtain a density retrospective profile of the target cigarette pack batch corresponding to the tobacco leaf compression dimension at the collection moment when the preset cigarette pack management cycle is met. The density retrospective profile includes the initial compression duration and initial pressure parameters; A real-time compression profile generation module is used to obtain a real-time compression profile of a target cigarette pack batch, wherein the real-time compression profile includes current compression duration and current pressure parameters; A compression deviation analysis module, configured to generate a deviation vector of the tobacco leaf compression image based on the initial compression duration, the initial pressure parameter, the current compression duration, and the current pressure parameter; a fluctuation threshold determination module, configured to retrieve a set of tobacco leaf compression dimension deviation modulus values ​​that matches the deviation vector, and when the variance of the tobacco leaf compression dimension deviation modulus value set exceeds a preset variance threshold, calculate the ratio of the length of a first time period from the storage start time to the collection time to the length of a second time period from the storage start time to the current time as the cigarette pack data deletion ratio; The cigarette pack data cleaning module is used to delete the tobacco leaf compressed dimension data starting from the start time of compressed data storage of the target cigarette pack batch based on the cigarette pack data deletion ratio.

[0006] Preferably, the cigarette pack density portrait generation module includes: A compression parameter acquisition unit, configured to obtain a first preset pressure parameter up to an Nth preset pressure parameter of a target cigarette pack batch; a compression duration statistics unit, configured to use the first preset pressure parameter as a constraint condition and to calculate a first alienated compression duration of the tobacco leaf compression dimension based on a historical cigarette pack sample library; A multi-parameter compression analysis unit, configured to calculate the Nth alienated compression duration of the tobacco leaf compression dimension based on the Nth preset pressure parameter as a constraint condition; The period dynamic calculation unit is used to extract 0.5 times the minimum value of the first alienation compression time length to the Nth alienation compression time length, and set it as the preset cigarette pack management period.

[0007] Preferably, the cigarette pack density portrait generation module further includes: A cloud storage capacity monitoring unit, used to obtain the upper threshold of the 5G cloud platform storage capacity for the compressed data of the target cigarette pack batch; The portrait backtracking trigger unit is used to trigger the acquisition of the density backtracking portrait of the target cigarette pack batch at the collection moment of the tobacco leaf compression dimension corresponding to the target cigarette pack batch when the compressed data volume of the target cigarette pack batch reaches the upper limit threshold of the storage capacity of the 5G cloud platform.

[0008] Preferably, the compression duration statistics unit includes: a tobacco leaf physical property retrieval subunit, configured to retrieve, from the historical cigarette pack sample library, a set of physical property record feature values ​​and corresponding initial compression duration record values ​​that match the first preset pressure parameter when the tobacco leaf compression dimension is the physical compression property; A physical property deviation calculation subunit, configured to perform paired deviation analysis on the physical property record characteristic value set and the initial compression duration record value, to generate a physical property characteristic deviation modulus value set and an initial compression duration deviation modulus value set; A compression feature screening subunit is configured to extract, from the initial compression duration deviation modulus value set, a compression duration selected deviation modulus value set whose physical property feature deviation modulus exceeds a preset feature deviation threshold; The compression duration mode evaluation subunit is configured to perform frequency mode analysis on the selected deviation modulus value set of the compression duration and output the first alienated compression duration.

[0009] Preferably, the compression duration statistics unit includes: a bale type mapper scheduling subunit, configured to, when the tobacco leaf compression dimension is a bale type derived attribute, call a compression dimension mapper bound to the bale type derived attribute and a first preset pressure parameter, wherein the compression dimension mapper requires input of a physical compression attribute; a derived association retrieval subunit, configured to retrieve a set of physical attribute record feature values ​​and a derived compression duration record value matching the first preset pressure parameter from the historical cigarette pack sample library; a bundle-type feature conversion subunit, configured to convert a set of physical attribute record feature values ​​through the compressed dimension mapper to generate a set of bundle-type derived attribute feature values; The alienated duration derivation subunit is configured to calculate the first alienated compressed duration based on the bundle-type derived attribute feature value set and the derived compressed duration record value.

[0010] Preferably, the system further comprises: A derived attribute configuration module, configured to configure a preset physical compression input attribute of a preset bundle-type derived attribute through a user terminal; A mapper training module, configured to collect multiple sets of tobacco leaf bundle shape data and train the compressed dimension mapper using a preset cigarette package model and a preset pressure parameter as constraints; A mapping relationship binding module, configured to bind the compression dimension mapper to the preset bundle type derived attribute and the preset pressure parameter; The multiple groups of tobacco leaf bundle type data include preset physical compression input attribute record values ​​and preset bundle type derived attribute record values.

[0011] Preferably, the fluctuation threshold determination module includes: a deviation decomposition unit, configured to decompose the deviation vector into a compression duration deviation vector and a pressure parameter deviation vector; The cloud-based ciphertext retrieval unit is used to send an encrypted retrieval request to the 5G cloud platform, obtain the key, encrypt the tobacco leaf compression dimension, cigarette package model, compression time deviation vector, and pressure parameter deviation vector to generate a retrieval ciphertext, and then receive the initial tobacco leaf compression dimension deviation modulus value set after uploading it to the cloud; The outlier data filtering unit is used to perform outlier elimination on the initial tobacco leaf compression dimension deviation module value set to generate a target tobacco leaf compression dimension deviation module value set.

[0012] Preferably, the cloud-based ciphertext retrieval unit includes: A cloud-based decryption subunit is used to decrypt the search ciphertext through the 5G cloud platform, extract the tobacco leaf compression dimension, cigarette package model, compression time deviation vector, and pressure parameter deviation vector, and construct distributed search task constraints; The node ciphertext request subunit is used to initiate a retrieval request to the edge node through the 5G cloud platform to obtain the edge node key; The edge encryption retrieval subunit is used to encrypt the retrieval task constraints and retrieval key based on the edge node key, generate the edge task ciphertext and send it to the edge node, and receive the retrieval feedback information including the initial tobacco leaf compression dimension deviation modulus value set.

[0013] Preferably, the system further comprises: A compression fluctuation monitoring module is used to obtain the adjustment signal and adjustment process data of the cigarette package compression parameters, analyze and generate the compression fluctuation amplitude, and generate the compression delay coefficient based on the compression fluctuation amplitude; A rate dynamic correction module, configured to adjust the pressure application rate of the cigarette pack compression device according to the compression delay coefficient, and optimize the compression process based on the compression efficiency parameter; The compression delay coefficient is generated by calculating the minimum difference between the cigarette pack compression parameter and the preset compression range boundary.

[0014] Preferably, the present invention further includes a tobacco leaf baling data networking management method based on a 5G cloud platform, which is applied to the above-mentioned tobacco leaf baling data networking management system based on a 5G cloud platform, and the method includes: When the preset cigarette pack management cycle is met, a density retrospective portrait of the target cigarette pack batch corresponding to the tobacco leaf compression dimension at the collection moment is obtained, wherein the density retrospective portrait includes the initial compression time and initial pressure parameters; Obtaining a real-time compression profile of the target cigarette pack batch, wherein the real-time compression profile includes a current compression duration and a current pressure parameter; generating a compression image deviation vector based on the initial compression duration, the initial pressure parameter, the current compression duration, and the current pressure parameter; Retrieving a set of tobacco leaf compression dimension deviation modulus values ​​that matches the deviation vector; when the variance of the tobacco leaf compression dimension deviation modulus value set exceeds a preset variance threshold, calculating a ratio of a first time period length from the storage start time to the collection time to a second time period length from the storage start time to the current time, and setting the ratio as the cigarette pack data deletion ratio; According to the cigarette pack data deletion ratio, the tobacco leaf compressed dimension data is deleted starting from the start time of compressed data storage of the target cigarette pack batch.

[0015] Compared with the prior art, the present invention has the following beneficial effects: Through the cigarette pack density portrait generation module, when the cigarette pack management cycle is preset, the density retrospective portrait of the target cigarette pack batch corresponding to the tobacco leaf compression dimension at the collection moment can be obtained, including the initial compression duration and initial pressure parameters, which can fully present the state characteristics of the initial stage of cigarette pack compression, facilitating subsequent comparison with real-time data. The real-time compression portrait generated by the real-time compression portrait generation module includes the current compression duration and current pressure parameters, which can reflect the current state of cigarette package compression in real time, allowing managers to promptly grasp the dynamic changes of the compression process. The compression deviation analysis module generates a deviation vector based on the initial and real-time compression duration and pressure parameters, which can clearly show the changes in parameters during the cigarette pack compression process and provide intuitive information for understanding the deviation of the compression state. The fluctuation threshold judgment module retrieves the set of tobacco leaf compression dimension deviation modulus values ​​that match the deviation vector. When the variance exceeds the preset threshold, it calculates the ratio of cigarette pack data deletion. It can determine the scale of data cleaning based on the fluctuation of the compression parameters, so that data deletion is more in line with the needs of the actual compression process. The cigarette pack data cleaning module deletes tobacco leaf compression dimension data from the storage start time according to the data deletion ratio, which can reduce the storage of redundant data, avoid invalid data occupying too many resources, and make the stored information more targeted and effective. The networked management of tobacco baling data is achieved through the 5G cloud platform, breaking the geographical limitations of traditional data management and enabling the centralized aggregation of baling equipment data in different locations for unified monitoring and analysis. The system generates a retrospective density profile and real-time compression profile of the cigarette pack, fully demonstrating the entire compression process and providing rich information for analyzing state changes during the compression process. The generation of deviation vectors quantifies parameter deviations during the compression process, providing a deeper understanding of compression quality trends. The determination and execution of data deletion ratios ensures that data storage always revolves around valuable compressed information, improving the accuracy of data management while also reducing the burden of data processing and making data processing more efficient. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 This is a working principle diagram of the tobacco baling data network management system based on the 5G cloud platform according to the present invention; Figure 2 Flowchart of the module for generating density profiles of cigarette packs; Figure 3 Flowchart for compressing duration statistics for physical attributes; Figure 4 Flowchart for compression duration statistics of bundle derived attributes; Figure 5This is a flow chart of the fluctuation threshold determination module. DETAILED DESCRIPTION

[0017] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. 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 creative efforts are within the scope of protection of the present invention.

[0018] See Figure 1 The present invention provides a tobacco leaf baling data network management system based on a 5G cloud platform. The system includes: a tobacco package density profile generation module, a real-time compression profile generation module, a compression deviation analysis module, a fluctuation threshold determination module, and a tobacco package data cleaning module. The specific implementation steps are as follows: When the preset cigarette pack management cycle is met, the cigarette pack density profile generation module obtains the density retrospective profile of the target cigarette pack batch corresponding to the tobacco leaf compression dimension at the collection moment. The density retrospective profile includes the initial compression time and initial pressure parameters.

[0019] The real-time compression profile generation module obtains a real-time compression profile of a target cigarette pack batch, and the real-time compression profile includes current compression duration and current pressure parameters.

[0020] The compression deviation analysis module generates a deviation vector of the tobacco leaf compression image based on the initial compression duration, initial pressure parameters, current compression duration and current pressure parameters.

[0021] The fluctuation threshold determination module retrieves the tobacco leaf compression dimension deviation modulus value set that matches the deviation vector. When the variance of the tobacco leaf compression dimension deviation modulus value set exceeds the preset variance threshold, the ratio of the length of the first time period from the storage start time to the collection time to the length of the second time period from the storage start time to the current time is calculated as the cigarette pack data deletion ratio.

[0022] The cigarette pack data cleaning module deletes the tobacco leaf compression dimension data starting from the start time of compressed data storage of the target cigarette pack batch based on the cigarette pack data deletion ratio.

[0023] Example 1: See Figure 2The cigarette pack density profile generation module involves a series of units working together to achieve the acquisition and cycle management of tobacco leaf compression data in a 5G cloud platform environment. The module as a whole consists of a compression parameter acquisition unit, a compression duration statistics unit, a multi-parameter compression analysis unit, a cycle dynamic calculation unit, a cloud storage capacity monitoring unit, and a profile backtracking trigger unit. The compression parameter acquisition unit is responsible for accessing the relevant data of the target cigarette pack batch. The unit is connected to the batch data storage area through the application program interface (API) of the 5G cloud platform and automatically parses the preset pressure parameter sequence; the sequence structure includes an order from the first preset pressure parameter to the Nth preset pressure parameter, and each parameter represents the control value of a different stage in the tobacco leaf compression process. For example, the first parameter corresponds to the initial pressure setting, and the Nth parameter corresponds to the final stable stage. The parsing process uses a data formatting mechanism to convert the read pressure value into a processable numerical array. The compression time statistics unit is executed immediately, and the unit establishes a data link with the historical cigarette pack sample library. The sample library stores the complete compression records of past batches, including pressure parameter values ​​and corresponding compression times. The compression time statistics unit uses the first preset pressure parameter as the screening condition and uses the structured query language (SQL) to retrieve matching records in the sample library. The search query specifies the pressure parameter range and returns all record instances that meet the conditions. The average duration value is then calculated. This duration is the first alienated compression duration of the tobacco leaf compression dimension, which reflects the historical average level of compression time under a specific pressure.

[0024] The multi-parameter compression analysis unit operates in parallel, and its function is to perform similar duration statistics for the Nth preset pressure parameter. The unit uses a distributed parallel computing framework to process constraint queries to avoid single-point bottlenecks. The framework supports multi-threaded operations, takes the Nth parameter as input, extracts a subset of matching records from the historical sample library, uses an aggregation algorithm to calculate the average duration, and generates the Nth alienated compression duration as output. Data transmission uses an internal message queue to ensure timing consistency. The periodic dynamic calculation unit receives the duration results from the first two units and performs sequence analysis. The unit logic involves extracting the minimum value from the first alienated compression duration to the Nth alienated compression duration sequence, using the array minimum algorithm to traverse and compare, identifying the minimum element and multiplying it by a fixed coefficient of 0.5, and finally outputting the preset cigarette pack management cycle value. This cycle value serves as the reference duration interval for the system to manage cigarette pack data and is used to trigger the acquisition of density profiling. During operation, the module also includes a cloud storage capacity monitoring unit, which continuously monitors the storage occupancy of the compressed data of the target cigarette pack batch on the 5G cloud platform. The monitoring mechanism involves reading the real-time capacity indicators of the cloud storage area, including the current occupied bytes and the total upper limit threshold. For example, the capacity threshold is set to a fixed value or a dynamic value based on the system configuration. When the data volume growth reaches the preset upper limit threshold, the unit sends a trigger signal. The image backtracking trigger unit responds to this signal and calls the data collection service module. This service retrieves a complete data snapshot at the time of collection. The snapshot generation involves data warehousing technology to ensure that the density backtracking image contains a complete data set of initial compression duration and initial pressure parameters. The unit logic includes verifying data integrity and preventing omissions through a verification mechanism in the cloud environment.

[0025] The entire module relies on the distributed database of the 5G cloud platform to store historical sample libraries and real-time data. The database supports partitioning to accelerate query efficiency. At the same time, a real-time monitoring service runs on the cloud to detect data flow anomalies and system load, ensuring module consistency. The specific implementation of the compression parameter acquisition unit also includes error handling logic. For example, when the API responds abnormally, a retry mechanism is activated to prevent data loss. The unit caches the parsing results internally so that subsequent units can directly access them. The implementation details of the compression duration statistics unit include optimizing query efficiency, such as indexing the pressure parameter column of historical records to shorten retrieval time. When calculating the first alienation compression duration, the algorithm considers the size of the data sample and automatically adjusts the average weighting method to avoid small sample bias. The output of this unit is passed as the base value. The distributed framework of the multi-parameter compression analysis unit achieves load balancing. When processing the Nth parameter constraint, the task is decomposed into multiple computing nodes. The node results are aggregated to generate the Nth alienation compression duration. The framework's fault tolerance mechanism ensures that tasks are rescheduled in the event of calculation failure. The sequence processing logic of the period dynamic calculation unit is expanded to dynamic array input. The duration sequence is stored in the unit memory. The minimum value algorithm traverses all elements and outputs the period value after multiplying by the coefficient. The output value is automatically written to the system configuration file.

[0026] The cloud storage capacity monitoring unit continuously polls storage status, with the polling interval adaptively adjusted based on system load. When capacity exceeds a threshold, a monitoring signal is broadcast via the event bus. The implementation of the profile backtracking trigger unit involves asynchronously invoking the data collection service. The service generates a data snapshot in the cloud, containing all relevant attributes of the compression dimension. The service also restores compressed data from a historical point, using timestamps to match the exact acquisition moment. The module is fully integrated within the 5G network architecture, ensuring low-latency transmission. The distributed database of the historical sample library uses sharding technology to support massive data storage. The real-time monitoring service is embedded in the alarm system to notify maintenance personnel in the event of module component failures. The compression duration statistics unit further refines the historical record retrieval process, for example, by optimizing query performance through parameter binding and avoiding full table scans. The unit interface design allows for dynamic input of pressure parameter values. The parallel processing logic of the multi-parameter compression analysis unit can be expanded to a task scheduler to coordinate computing resources. The unit output format is standardized for easy downstream use. The cloud storage capacity monitoring unit's internal algorithm monitors occupancy trends, predicts threshold triggering times, and preloads resources in advance. The profile backtracking trigger unit verifies data integrity during execution, ensuring the accuracy of initial parameters through checksum comparison. Communication between module components utilizes cloud messaging middleware to ensure data consistency; implementation processes are seamless and avoid interruptions. Module environment deployment incorporates elastic scalability to adapt to resource changes on the 5G cloud platform. The database access layer isolates physical storage details to simplify data operations. Error handling mechanisms are implemented throughout each unit, for example, rewriting logic to execute a fallback solution when SQL queries fail. Unit status logs are stored in the cloud to aid debugging. The entire implementation emphasizes modular design, with each unit independently developed and tested before integration. Parameters are uniformly deployed through configuration management tools. Implementation results generate preset management cycles and density profiles to support cycle-optimized management of cigarette pack data. The compression parameter collection unit processes pressure parameter sequences, including data cleansing to remove invalid or noisy records. Parameter order is loaded from a configuration file, and the unit parsing logic supports multiple data formats. The compression duration statistics unit's calculation algorithm optimizes memory usage to avoid overflows when processing large record sets. Unit output undergoes format validation before being passed to subsequent units. The periodic dynamic calculation unit uses floating-point arithmetic for coefficient multiplication to ensure output accuracy. System parameter registers are updated after unit execution. The data retrieval of the profile backtracking trigger unit covers all state variables at the time of collection, ensuring comprehensiveness of the profile. The threshold settings for the cloud storage capacity monitoring unit are configured through the system administrator interface, offering flexible adjustments. The module operates efficiently across the 5G network, with real-time data streams buffered by message queues to maintain system stability. The unit interaction protocol defines a clear interface, with each unit's output serving as the input for the next. This interlocking logic ensures reliable and systematic cigarette pack data processing. During module initialization, the module loads metadata from the historical sample library to pre-calibrate query performance. During operation, the module automatically adapts to changes in data size and balances computational overhead through a resource allocation mechanism.Implementation details also include data encapsulation standards between units, such as using JSON or binary format to transmit results; error logs are reported to the cloud monitoring dashboard for easy operation and maintenance. The entire process has strict timing control, and the compression duration statistics unit and the multi-parameter compression analysis unit can be executed in parallel to shorten the overall delay; after outputting the preset management cycle value, it is automatically applied to the subsequent cigarette pack cleaning process. The module components are deployed on cloud virtual machine instances, and the instance resources are automatically expanded to cope with peak loads; the indexing strategy of the historical sample library optimizes query performance, such as building a hash index of pressure parameters. Single points of failure are avoided during implementation, and key units such as the image backtracking trigger unit are designed with redundant backups; unit communication encryption ensures data security; the overall implementation is verified for feasibility and compatibility in a standard industrial environment.

[0027] Example 2: See Figure 3 When a tobacco leaf compression dimension is identified as a physical compression attribute, the compression duration statistics unit initiates the physical attribute processing process, which involves the coordinated operation of four core components: the tobacco leaf physical attribute retrieval subunit, the physical attribute deviation calculation subunit, the compression feature screening subunit, and the compression duration mode evaluation subunit. The tobacco leaf physical attribute retrieval subunit sends a data query request to the historical tobacco pack sample library, which carries the first preset pressure parameter as a key constraint. The historical tobacco pack sample library is deployed in a distributed database cluster and uses a sharded storage mechanism to manage the massive amount of historical compression records. The query request is parsed by the database routing layer to locate the physical node storing the matching data. The query logic has two goals: to retrieve the set of characteristic values ​​of all tobacco leaf physical attribute records that meet the first preset pressure parameter threshold and simultaneously extract the corresponding initial compression duration record value. The set of characteristic values ​​of the physical attribute records covers key indicators of the tobacco leaf raw material, including but not limited to structured values ​​such as moisture content percentage, leaf thickness in millimeters, and fiber toughness index. The initial compression duration record value is stored with millisecond precision and associated with the unique identifier of the corresponding tobacco pack. The database performs parallel scanning operations, accelerates the matching process through pre-built pressure parameter indexes, and encapsulates the returned result set into data packets for transmission to the sub-unit memory buffer.

[0028] The physical property deviation calculation subunit receives the above data packet and starts the paired deviation analysis program. The program traverses each associated record of the physical property record characteristic value set and the initial compression duration record value. The analysis process calculates the absolute deviation of the physical property characteristic value and the average value of the same group one by one to generate a physical property characteristic deviation modulus value set. The absolute deviation of the initial compression duration record value and the average value of the same group duration is synchronously calculated to form an initial compression duration deviation modulus value set. The calculation process adopts the memory stream processing mode to avoid disk read and write delays. The physical property characteristic deviation modulus value set and the initial compression duration deviation modulus value set are stored in a matrix structure, and the row identifier corresponds to the original record index to facilitate subsequent association operations. The deviation calculation results are written to the distributed cache system in real time for downstream units to call on demand.

[0029] The compression feature screening subunit loads the physical attribute feature deviation modulus value set and the initial compression duration deviation modulus value set from the cache. The unit internally presets a feature deviation threshold, which is dynamically obtained from the system configuration center. The screening logic traverses each entry in the physical attribute feature deviation modulus value set and performs a scalar comparison of its modulus value with the preset feature deviation threshold. When the modulus value exceeds the threshold, the system automatically marks the entry index. The duration deviation modulus value at the corresponding position is then extracted from the initial compression duration deviation modulus value set based on the index list. All extracted duration deviation modulus values ​​constitute the set of selected compression duration deviation modulus values. The screening process applies a batch processing optimization algorithm, employing a segmented filtering strategy for large datasets. The threshold comparison operation is executed using a parallel thread pool, with each thread processing a subset of the data and ultimately merging the screening results. The set of selected compression duration deviation modulus values ​​is transferred to a persistent storage area and affixed with a timestamp and batch tag.

[0030] The compression duration mode evaluation subunit accesses persistent storage and reads the set of selected compression duration deviation moduli. This unit initiates the frequency mode analysis process, which consists of three phases: data discretization, frequency statistics, and mode determination. The data discretization phase divides the continuous duration deviation moduli into equal-width intervals, with the interval widths automatically adjusted based on the data distribution density. The frequency statistics phase iterates over all data points, assigns them to corresponding intervals, and increments a counter. The results generate a frequency distribution histogram, recording the number of data points in each interval. The mode determination phase scans the histogram to identify the peak interval containing the most data points. The median value of this interval is determined as the first dissimilarized compression duration output value. The analysis process utilizes a streaming statistics algorithm and supports incremental updates. As new data arrives, the system dynamically adjusts the interval divisions and updates the frequency distribution. The mode evaluation results are output to the message-based middleware for invocation by the system's core modules. The entire process runs in a containerized environment on the 5G cloud platform, with resource allocation dynamically scaling based on load.

[0031] At the implementation level, the tobacco leaf physical property retrieval subunit integrates a query retry mechanism. When the database response times out, the subunit automatically resets the connection and resubmits the query request. Query statements are parameterized to prevent injection attacks. The physical property deviation calculation subunit implements data validation, detects null-value records, and performs interpolation or elimination operations. The core deviation calculation logic is encapsulated as a microservice, exposing the computational interface through an API. The compression feature screening subunit's mechanism for obtaining preset feature deviation thresholds incorporates a multi-level fallback strategy: prioritizing parameters from the real-time configuration center; loading a local cached copy when the configuration is unreachable; and ultimately, using the system default value. The threshold comparison operation incorporates a fuzzy matching factor to handle boundary condition data. The interval partitioning algorithm of the compression duration mode evaluation subunit incorporates an outlier isolation mechanism to prevent extreme values ​​from disrupting the main data distribution. Frequency statistics utilize a distributed counter architecture, aggregating statistical results across nodes. The mode determination logic includes a tie-breaking rule: when multiple interval frequencies are identical, the minimum median value is used as the output. The system operation log records the operation trajectory of each subunit, including input parameters, processing time, and output results, and the log is synchronized in real time to the cloud log analysis system. The error handling process covers scenarios such as network outages, data anomalies, and computational overflows, triggering alarms to notify the operation and maintenance node. Data exchange between sub-units uses a serialization protocol to ensure transmission efficiency and compatibility. The unit status monitor periodically checks the health of components and automatically restarts or migrates abnormal instances. The entire implementation architecture is deployed in a hybrid environment of 5G edge computing nodes and the cloud, with critical path data transmitted through dedicated channels to reduce latency. The access interface to the historical cigarette pack sample library implements rate limiting and permission verification to ensure data security. The execution results of the physical property processing process serve as the basic parameters for tobacco leaf compression analysis and participate in subsequent compression cycle calculations and data optimization management. The system maintenance module provides a configuration interface that allows administrators to adjust operating parameters such as preset feature deviation thresholds and retry times. Component version updates use a blue-green deployment strategy to ensure service continuity. The implementation process strictly adheres to data processing specifications, and audit trails are retained for both original records and intermediate results.

[0032] Example 3: See Figure 4When the tobacco leaf compression dimension is identified as a bundle-type derived attribute, the compression duration statistics unit initiates the derived attribute processing process, which consists of a bundle-type mapper scheduling subunit, a derived association retrieval subunit, a bundle-type feature conversion subunit, and an alienated duration derivation subunit. The bundle-type mapper scheduling subunit first accesses the system registry, which stores the binding relationship records between the compression dimension mapper and the bundle-type derived attributes and pressure parameters. The registry query uses the current bundle-type derived attribute identifier and the first preset pressure parameter as the joint key to locate the target compression dimension mapper instance. The mapper instance contains a pre-trained machine learning model, and its input interface requires receiving a set of physical compression attributes. After the mapper is loaded into memory, the derived association retrieval subunit initiates a search request to the historical cigarette pack sample library, carrying the first preset pressure parameter as the core filtering condition. The distributed nodes of the sample library perform a joint index scan and return the set of physical attribute record feature values ​​and the associated derived compression duration record values ​​of the matching records. The set of physical property record feature values ​​includes original measurement values ​​such as tobacco leaf quality and moisture level, and the derived compression duration record value corresponds to the actual compression time in the historical operation log.

[0033] The bundle feature conversion subunit receives a set of physical property record feature values ​​and formats it into a tensor input structure required by the compressed dimension mapper. The mapper model performs forward reasoning calculations. The model can be a deep neural network or an integrated learning architecture, and its hidden layer performs nonlinear transformations on the input features. The model output layer generates a set of bundle-derived attribute feature values, which characterizes the abstract characteristics of the tobacco leaf bundle type, such as the structural tightness index or the shape stability score. The feature value set is stored in the cache in the form of a floating-point vector. The alienated duration derivation subunit synchronously obtains the derived compressed duration record value and aligns it with the bundle-derived attribute feature value set according to the record index. The unit executes a regression analysis algorithm inside to fit the mapping relationship between the feature vector and the duration value. The algorithm minimizes the error function between the predicted duration and the actual duration, and finally outputs the first alienated compressed duration. The duration is calculated using the following formula:

[0034] in Indicates the derived compression duration, For the The weight coefficient of the derived attribute feature, Indicates the derived attribute eigenvalues, is the characteristic nonlinear transformation function, is the total number of derived attribute features.

[0035] The system simultaneously runs the derived attribute configuration module, which provides a graphical user interface (GUI) allowing administrators to define metadata for predefined bundle type derived attributes through a web terminal. This metadata includes a list of predefined physical compression input attributes, feature normalization rules, and mapper version requirements. This configuration information is stored in the cloud configuration center and synchronized to all nodes in real time. The mapper training module runs independently on the training cluster. Its data pipeline collects multiple sets of tobacco bundle type data samples from the data lake. The sample structure includes predefined cigarette pack model labels, predefined pressure parameter values, predefined physical compression input attribute records, and predefined bundle type derived attribute records. The training process uses cigarette pack model and pressure parameter as grouping constraints. Within each group, the following operations are performed: the data cleaning module removes anomalous samples; the feature engineering component constructs high-order cross features; the model selector evaluates candidate architectures based on predefined metrics; and finally, the compression dimension mapper model is trained and saved to the model repository. The mapping binding module monitors model repository update events. When a new model version is released, it automatically extracts the model meta-descriptor and establishes a binding relationship between it and the predefined bundle type derived attribute identifier and the predefined pressure parameter value. The binding records are written to the distributed key-value storage system, and the registry is regularly updated.

[0036] During implementation, the bundled mapper scheduling subunit implements a hot model reload mechanism to avoid service interruptions. The derived association retrieval subunit utilizes paging query optimization to handle extremely large historical records. The bundled feature conversion subunit includes dimension checking during tensor formatting to prevent model input errors. The regression algorithm in the alienation duration derivation subunit supports incremental training, automatically updating weight coefficients as new data arrives. The derived attribute configuration module's interface includes dynamic validation logic to ensure input attributes are compatible with the sample library structure. The mapper training module's pipeline is designed as a directed acyclic graph, with each processing stage running in a separate container. The training cluster resource scheduler dynamically allocates GPU resources based on data size. The mapping relationship binding module implements a versioning strategy and retains historical binding records for rollback operations. All subunits are interconnected via a service mesh, using Protobuf serialization for communication. The system is deployed in a Kubernetes cluster on a 5G cloud platform, with compute nodes equipped with dedicated AI accelerator cards. The historical cigarette packet sample library implements a data partitioning strategy, dividing physical storage by pressure parameter ranges. The model inference service exposes a gRPC interface, supporting millisecond response times. The configuration center implements access control policies, ensuring that only authorized terminals can modify derived attribute definitions. The training dataset collection process includes desensitization to remove sensitive business information. The storage system for binding records uses a multi-copy mechanism to prevent data loss. The entire process execution log is integrated into a distributed tracing system, supporting cross-unit call chain analysis. An anomaly monitoring agent detects component failures and automatically triggers a switch to a backup instance. Data transmission is encrypted throughout the process, complying with 5G network security specifications.

[0037] Example 4: See Figure 5The Fluctuation Threshold Determination Module receives an input vector from the Compression Deviation Analysis Module, which contains a continuous record of deviations in the tobacco leaf compression profile. The Deviation Decomposition Unit initiates a structured parsing program to identify the data type and dimensions in the vector. The program splits the input vector into two independent data sequences: a compression duration deviation sequence that stores the time dimension differences, and a pressure parameter deviation sequence that records the changes in pressure readings. For example, for tobacco batch QY2023-06, the system processes deviation data from eight consecutive inspection points. The specific values ​​are shown in the following table:

[0038] The cloud-based ciphertext retrieval unit activates the encryption process and encapsulates the data structure using the system's pre-set key. The encapsulated object contains the tobacco leaf compression dimension identifier "Compression_Dimension_V2," the cigarette package model "QY2023-06," and the duration and pressure deviation sequences listed above. The encryption engine performs AES-GCM conversion, generating the retrieval ciphertext string "8E2F9A...C54D1B." This ciphertext is uploaded to cloud storage node CN-07 via a dedicated 5G network channel, with a recorded transmission time of 47 milliseconds.

[0039] The cloud decryption subunit performs decryption operations on the cloud server node SVR-09. After decryption, the core parameters are extracted: cigarette pack model, compression dimension identifier, and deviation sequence data. The system constructs distributed retrieval task constraints: Data time range: June 1 to 15, 2023 Deviation matching accuracy: ±12% Maximum number of returned records: 500 The node ciphertext request subunit broadcasts the task metadata to the edge computing group. Edge node NJ-12 responds to the request and returns the node session key "F7B302...A891D4". The edge encryption retrieval subunit uses this key to re-encrypt the task constraints, generating the edge task ciphertext "3C8E...D209" and sending it to node NJ-12 via the 5G network (with a latency of 13 milliseconds). This node performs a query in its local history database and returns a set of initial tobacco leaf compression dimension deviation moduli: [22.5,8.7,36.2,12.4,9.8,44.3,7.1,27.6,18.9,31.4,5.3,39.8] The outlier filtering unit loads the dataset. The processing flow includes the following steps: Sort the collection in ascending order as: [5.3,7.1,8.7,9.8,12.4,18.9,22.5,27.6,31.4,36.2,39.8,44.3] Quartile calculation: Determine the first quartile value as 10.1 and the third quartile value as 33.8.

[0040] Boundary setting: Calculate the interquartile range 23.7, the lower boundary is 10.1-1.5×23.7=-25.45, and the upper boundary is 33.8+1.5×23.7=69.35; Outlier filtering: All data points are checked to be within the range of [-25.45, 69.35], retaining a complete set of 12 records. This implementation involves multiple layers of technical detail. Queries on edge node NJ-12 utilize a local cache acceleration mechanism, directly returning cached result sets when query conditions match recent tasks. A dual-channel verification mechanism is implemented during encrypted transmission, with data packets appended with sequence numbers and CRC32 checksums. The cloud-based decryption subunit is deployed in an isolated container environment, with key storage utilizing a hardware security module. The distributed task constraint generator implements an intelligent sharding strategy, automatically partitioning historical data into multiple subtasks for parallel processing when the volume exceeds a threshold.

[0041] The edge node search service includes data clipping, automatically truncating the returned result set to the required upper limit of 500 records. The outlier filtering unit implements dynamic algorithm switching, using the kernel density estimation algorithm when the data distribution exhibits multimodal characteristics. The filtering operation log records the complete mapping between the input set, boundary parameters, and output results. The system resource monitor adjusts computing resource allocation in real time, automatically enabling streaming mode when the data size exceeds 1,000 records.

[0042] Security protection mechanisms are implemented throughout the entire process: two-way certificate authentication is enabled for cloud transmission, device fingerprints are bound to edge task ciphertexts, and digital signatures are attached to filtering results. Dedicated 5G slicing is implemented at the network layer, ensuring a secure data transmission path with controlled latency. Component status information, including node response time, data processing volume, and exception event codes, is reported to the monitoring center in real time. If an edge node goes offline, the system automatically switches to the backup node, NJ-14, to continue service. An unauthorized access control mechanism is implemented within the historical data query range, prohibiting access to records within unauthorized time intervals. Temporary data generated during the implementation process is immediately deleted upon task completion, and key material is validated once it is valid for a single use.

[0043] Example 5: The compression fluctuation monitoring module runs on the real-time data processing engine of the 5G cloud platform. The module is connected to the sensor network of the cigarette pack compression device through the device gateway interface. The sensor network includes a pressure transmitter, a displacement meter and a control signal receiver, which collects the raw data stream of the compression parameters at a frequency of milliseconds. The core functions of the module include three consecutive stages: adjustment signal analysis, fluctuation amplitude calculation, and delay coefficient generation. The adjustment signal analysis stage monitors the device control bus and captures the compression parameter adjustment instructions and their timestamp sequence. In the example scenario, the target device triggers an adjustment event during the compression process of batch TJ2024, and the system records the event sequence: Always receive pressure increase instructions (+5kPa), Always receive pressure maintenance instructions, The system constantly receives rate reduction commands (-2 kPa / s). Data from the adjustment process is collected synchronously, including pressure fluctuation curves (0.5-second interval sampling values: 152.3 kPa, 157.1 kPa, 155.8 kPa, 149.6 kPa) and hydraulic cylinder displacement trajectories (3.2 mm, 3.5 mm, 3.7 mm, 3.4 mm).

[0044] The time series analysis starts at the stage of calculating the fluctuation range. The algorithm loads and adjusts the time window. For the process data set within, the following operations are performed: the pressure data flow is smoothed through a sliding window (window width = 500ms), and the calculated pressure standard deviation is 4.82kPa; the calculated range of the displacement data is 0.5mm; the control signal frequency statistics show an adjustment frequency of 3 times / minute. The weighted fusion of the three indicators generates a compression fluctuation amplitude value AW=6.37 (dimensionless), and the weight coefficient is dynamically loaded from the configuration library according to the device model. The boundary comparator is activated in the delay coefficient generation stage. This component pre-stores the current batch compression range boundary value: pressure lower limit 148kPa, pressure upper limit 160kPa, displacement tolerance ±0.3mm. The system calculates the minimum difference between the acquisition parameters and the boundary in real time: the current pressure value 149.6kPa differs from the lower limit by 1.6kPa, and the displacement 3.4mm differs from the upper limit by 0.1mm. Through the nonlinear conversion function Output compression delay factor .

[0045] The dynamic rate correction module connects to the compression equipment actuator via an industrial control protocol. Upon receiving the compression delay coefficient (CD) of 0.909, the module activates the dual-track correction mechanism. The pressure application rate adjustment channel calculates the base rate setpoint (120 kPa / s) multiplied by CD to obtain a corrected rate of 109.08 kPa / s. This value is converted to a 4-20 mA current signal via a digital-to-analog converter and transmitted to the hydraulic servo valve via the control bus. The compression process optimization channel operates simultaneously, extracting compression efficiency parameters from the equipment log: the current cycle time is 42 seconds and the energy consumption index is 78 kW·h. The optimization algorithm compares this to the historical best record (cycle time of 38 seconds and energy consumption of 70 kW·h) and generates a process adjustment plan: shortening the pressure holding phase by 3 seconds and advancing the pressure relief trigger point. The optimization instructions are encapsulated as Modbus TCP protocol frames and transmitted to the PLC controller.

[0046] The implementation details include multi-layer security design. The device gateway interface implements protocol adaptive conversion and supports multiple industrial bus standards such as Profinet and EtherCAT. The sensor data stream adopts a frame check sequence mechanism, and the erroneous data packets are automatically retransmitted. The adjustment signal parser is equipped with an instruction semantic analyzer to identify the parameter adjustment intention in the control signal. The time series analysis module deploys a streaming computing framework, and the window width is dynamically adjusted according to the degree of data jitter. The boundary comparator implements a dynamic update strategy for the boundary value, and synchronizes the new threshold from the cloud configuration center when the process parameters change. The nonlinear conversion function is defined in sections. An exponential decay response curve is enabled when the module is in operation. Double checks are implemented during the transmission of rate adjustment commands: packet checksum verification and actuator feedback confirmation. The process optimization algorithm includes rollback protection logic, which automatically restores the original parameters when the efficiency of the first cycle after optimization drops by more than 5%. Historical versions of control commands are stored in a ring buffer, supporting 10 operation backtrackings. The module's operating status is reported to the cloud monitoring center in real time via the 5G network. Key indicators include command latency (average 23ms), data processing throughput (850 items / second), and abnormal event counts. The fault recovery mechanism includes automatic reconnection, parameter cache continuation, and safe state switching functions. The feedback data from the compression device actuator is closed-loop verified to verify the adjustment effect, forming a continuous optimization loop. All operation records are written to the blockchain evidence storage system to meet industrial audit requirements.

[0047] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus.

[0048] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A tobacco leaf baling data network management system based on a 5G cloud platform, characterized in that: include: A cigarette pack density profile generation module is used to obtain a density retrospective profile of the target cigarette pack batch corresponding to the tobacco leaf compression dimension at the collection moment when the preset cigarette pack management cycle is met. The density retrospective profile includes the initial compression duration and initial pressure parameters; A real-time compression profile generation module is used to obtain a real-time compression profile of a target cigarette pack batch, wherein the real-time compression profile includes current compression duration and current pressure parameters; A compression deviation analysis module, configured to generate a deviation vector of the tobacco leaf compression image based on the initial compression duration, the initial pressure parameter, the current compression duration, and the current pressure parameter; a fluctuation threshold determination module, configured to retrieve a set of tobacco leaf compression dimension deviation modulus values ​​that matches the deviation vector, and when the variance of the tobacco leaf compression dimension deviation modulus value set exceeds a preset variance threshold, calculate the ratio of the length of a first time period from the storage start time to the collection time to the length of a second time period from the storage start time to the current time as the cigarette pack data deletion ratio; The cigarette pack data cleaning module is used to delete the tobacco leaf compressed dimension data starting from the start time of compressed data storage of the target cigarette pack batch based on the cigarette pack data deletion ratio.

2. The tobacco leaf bundling data network management system based on the 5G cloud platform according to claim 1 is characterized in that: The cigarette pack density profile generation module includes: A compression parameter acquisition unit, configured to obtain a first preset pressure parameter up to an Nth preset pressure parameter of a target cigarette pack batch; a compression duration statistics unit, configured to use the first preset pressure parameter as a constraint condition and to calculate a first alienated compression duration of the tobacco leaf compression dimension based on a historical cigarette pack sample library; A multi-parameter compression analysis unit, configured to calculate the Nth alienated compression duration of the tobacco leaf compression dimension based on the Nth preset pressure parameter as a constraint condition; The period dynamic calculation unit is used to extract 0.5 times the minimum value of the first alienation compression time length to the Nth alienation compression time length, and set it as the preset cigarette pack management period.

3. The tobacco leaf bundling data network management system based on the 5G cloud platform according to claim 1 is characterized in that: The cigarette pack density portrait generation module further includes: A cloud storage capacity monitoring unit, used to obtain the upper threshold of the 5G cloud platform storage capacity for the compressed data of the target cigarette pack batch; The portrait backtracking trigger unit is used to trigger the acquisition of the density backtracking portrait of the target cigarette pack batch at the collection moment of the tobacco leaf compression dimension corresponding to the target cigarette pack batch when the compressed data volume of the target cigarette pack batch reaches the upper limit threshold of the storage capacity of the 5G cloud platform.

4. The tobacco leaf bundling data network management system based on the 5G cloud platform according to claim 2 is characterized in that: The compression duration statistics unit includes: a tobacco leaf physical property retrieval subunit, configured to retrieve, from the historical cigarette pack sample library, a set of physical property record feature values ​​and corresponding initial compression duration record values ​​that match the first preset pressure parameter when the tobacco leaf compression dimension is the physical compression property; A physical property deviation calculation subunit, configured to perform paired deviation analysis on the physical property record characteristic value set and the initial compression duration record value, to generate a physical property characteristic deviation modulus value set and an initial compression duration deviation modulus value set; A compression feature screening subunit is configured to extract, from the initial compression duration deviation modulus value set, a compression duration selected deviation modulus value set whose physical property feature deviation modulus exceeds a preset feature deviation threshold; The compression duration mode evaluation subunit is configured to perform frequency mode analysis on the selected deviation modulus value set of the compression duration and output the first alienated compression duration.

5. The tobacco leaf bundling data network management system based on the 5G cloud platform according to claim 2 is characterized in that: The compression duration statistics unit includes: a bale type mapper scheduling subunit, configured to, when the tobacco leaf compression dimension is a bale type derived attribute, call a compression dimension mapper bound to the bale type derived attribute and a first preset pressure parameter, wherein the compression dimension mapper requires input of a physical compression attribute; a derived association retrieval subunit, configured to retrieve a set of physical attribute record feature values ​​and a derived compression duration record value matching the first preset pressure parameter from the historical cigarette pack sample library; a bundle-type feature conversion subunit, configured to convert a set of physical attribute record feature values ​​through the compressed dimension mapper to generate a set of bundle-type derived attribute feature values; The alienated duration derivation subunit is configured to calculate the first alienated compressed duration based on the bundle-type derived attribute feature value set and the derived compressed duration record value.

6. The tobacco leaf bundling data network management system based on the 5G cloud platform according to claim 5 is characterized in that: Also includes: A derived attribute configuration module, configured to configure a preset physical compression input attribute of a preset bundle-type derived attribute through a user terminal; A mapper training module, configured to collect multiple sets of tobacco leaf bundle shape data and train the compressed dimension mapper using a preset cigarette package model and a preset pressure parameter as constraints; A mapping relationship binding module, configured to bind the compression dimension mapper to the preset bundle type derived attribute and the preset pressure parameter; The multiple groups of tobacco leaf bundle type data include preset physical compression input attribute record values ​​and preset bundle type derived attribute record values.

7. The tobacco leaf bundling data network management system based on the 5G cloud platform according to claim 1 is characterized in that: The fluctuation threshold determination module includes: a deviation decomposition unit, configured to decompose the deviation vector into a compression duration deviation vector and a pressure parameter deviation vector; The cloud-based ciphertext retrieval unit is used to send an encrypted retrieval request to the 5G cloud platform, obtain the key, encrypt the tobacco leaf compression dimension, cigarette package model, compression time deviation vector, and pressure parameter deviation vector to generate a retrieval ciphertext, and then receive the initial tobacco leaf compression dimension deviation modulus value set after uploading it to the cloud; The outlier data filtering unit is used to perform outlier elimination on the initial tobacco leaf compression dimension deviation module value set to generate a target tobacco leaf compression dimension deviation module value set.

8. The tobacco leaf bundling data network management system based on the 5G cloud platform according to claim 7 is characterized in that: The cloud-based ciphertext retrieval unit includes: A cloud-based decryption subunit is used to decrypt the search ciphertext through the 5G cloud platform, extract the tobacco leaf compression dimension, cigarette package model, compression time deviation vector, and pressure parameter deviation vector, and construct distributed search task constraints; The node ciphertext request subunit is used to initiate a retrieval request to the edge node through the 5G cloud platform to obtain the edge node key; The edge encryption retrieval subunit is used to encrypt the retrieval task constraints and retrieval key based on the edge node key, generate the edge task ciphertext and send it to the edge node, and receive the retrieval feedback information including the initial tobacco leaf compression dimension deviation modulus value set.

9. The tobacco leaf bundling data network management system based on the 5G cloud platform according to claim 1 is characterized in that: Also includes: A compression fluctuation monitoring module is used to obtain the adjustment signal and adjustment process data of the cigarette package compression parameters, analyze and generate the compression fluctuation amplitude, and generate the compression delay coefficient based on the compression fluctuation amplitude; A rate dynamic correction module, configured to adjust the pressure application rate of the cigarette pack compression device according to the compression delay coefficient, and optimize the compression process based on the compression efficiency parameter; The compression delay coefficient is generated by calculating the minimum difference between the cigarette pack compression parameter and the preset compression range boundary.

10. A tobacco leaf baling data networking management method based on a 5G cloud platform, applied to the tobacco leaf baling data networking management system based on a 5G cloud platform according to any one of claims 1 to 9, characterized in that: The method comprises: When the preset cigarette pack management cycle is met, a density retrospective portrait of the target cigarette pack batch corresponding to the tobacco leaf compression dimension at the collection moment is obtained, wherein the density retrospective portrait includes the initial compression time and initial pressure parameters; Obtaining a real-time compression profile of the target cigarette pack batch, wherein the real-time compression profile includes a current compression duration and a current pressure parameter; generating a compression image deviation vector based on the initial compression duration, the initial pressure parameter, the current compression duration, and the current pressure parameter; Retrieving a set of tobacco leaf compression dimension deviation modulus values ​​that matches the deviation vector; when the variance of the tobacco leaf compression dimension deviation modulus value set exceeds a preset variance threshold, calculating a ratio of a first time period length from the storage start time to the collection time to a second time period length from the storage start time to the current time, and setting the ratio as the cigarette pack data deletion ratio; According to the cigarette pack data deletion ratio, the tobacco leaf compressed dimension data is deleted starting from the start time of compressed data storage of the target cigarette pack batch.

Images