Iot-based shell production parameter storage method and system

Abstract

The application relates to the technical field of production parameter storage methods, and discloses a shell production parameter storage method and system based on an Internet of Things, which comprises the following steps: determining a sub-storage path of a production data set of each shell according to a bidirectional storage path determined by the data amount of the production data set of each shell and the Internet of Things, a storage node of the production data set and a storage congestion coefficient of the Internet of Things; determining a plurality of abnormal storage parameters according to each sub-storage path; determining an abnormal storage area according to the plurality of abnormal storage parameters and corresponding production procedures; determining a storage abnormality processing item based on the abnormal storage area and a storage progress of the production data set of each shell; determining a plurality of sub-processing events according to the identification of the storage abnormality processing item; and determining a production parameter storage system based on the plurality of sub-processing events, each sub-storage path and a production state of each shell. The method can realize the synchronous storage of production parameters of shells in a production process.

Description

Technical Field

[0001] This invention relates to the field of production parameter storage methods, and in particular to a method and system for storing shell production parameters based on the Internet of Things. Background Technology

[0002] With the development of technology, the casing, as part of electronic devices and exposed to the external environment, has seen a gradual maturation of automated production in the manufacturing of electronic devices. In existing technologies, the casing outputs corresponding production parameters during the production process and stores these parameters accordingly. However, the production parameters are stored along a single path, which can easily lead to storage congestion during the production of multiple casings. This affects the accuracy of the sub-storage paths of the production data sets of each casing and makes it impossible to achieve synchronous storage of production parameters during the production process. Summary of the Invention

[0003] The purpose of this invention is to overcome the shortcomings of the prior art. This invention provides a method and system for storing shell production parameters based on the Internet of Things.

[0004] This invention provides a method for storing shell production parameters based on the Internet of Things, including:

[0005] The combination of production parameters is determined by real-time detection of multiple production processes of the shell, and the production status of each production process is determined based on the identification of the combination of production parameters.

[0006] A production dynamic diagram of the shell is constructed based on multiple production images of the shell in each production process. The production data set of the shell is determined based on each production dynamic diagram, the production progress of the shell, and the production status of each production process.

[0007] When multiple shells are in simultaneous production, a bidirectional storage path is determined based on the data volume of each shell's production data set and the Internet of Things (IoT). A sub-storage path for each shell's production data set is determined based on the bidirectional storage path, the storage nodes of each shell's production data set, and the IoT's storage congestion coefficient.

[0008] Multiple abnormal storage parameters are determined based on each sub-storage path. Abnormal storage areas are determined based on the multiple abnormal storage parameters and the corresponding production processes. Storage anomaly handling projects are determined based on the abnormal storage areas and the storage progress of the production data sets of each shell.

[0009] Based on the identification of storage exception handling projects, multiple sub-processing events are determined, and the production parameter storage system is determined based on the multiple sub-processing events, each sub-storage path, and the production status of each shell.

[0010] Optionally, determining the combination of production parameters based on real-time detection of multiple production processes of the casing, and determining the production status of each production process based on the identification of the combination of production parameters, includes:

[0011] The model number of the outer casing is collected. Based on the matching of the outer casing model and the outer casing production database, multiple production processes of the outer casing are determined. Based on the real-time detection of multiple production processes of the outer casing, multiple production parameters are determined, and the production process and time node corresponding to each production parameter are marked. The combination of each production parameter is determined according to the part of the outer casing corresponding to each production parameter, the production process and time node corresponding to each production parameter.

[0012] In each production process, multiple sub-production features are determined based on the identification of the corresponding combination of production parameters. The production status of the production process is determined based on the multiple sub-production features, the shape of the shell in the production process, and the part of the shell that is affected by the production process, so as to collect the production status of each production process.

[0013] Optionally, the step of constructing a production dynamic map of the casing based on multiple production images of the casing at each production process, and determining the production data set of the casing based on each production dynamic map, the production progress of the casing, and the production status of each production process, includes:

[0014] The system captures the position of the outer shell at each production stage, triggering multiple cameras around it based on that position. These cameras then capture images of the outer shell's production process from multiple directions, resulting in multiple production images of the outer shell at each stage. Each production image shows different aspects of the outer shell's production and records the corresponding posture of the outer shell.

[0015] Multiple pose features are determined based on the recognition of each production image. Multiple production behaviors of the shell are determined based on the multiple pose features, the shape of the shell, and the time sequence corresponding to each production image. A production dynamic map of the shell is constructed based on the synthesis of the multiple production behaviors of the shell.

[0016] The production progress of the shell is marked in the production dynamic diagram. The production events of the shell are determined based on each production dynamic diagram, the corresponding production process, and the production progress of the shell. The shell production data set is determined based on the production events of the shell and the production status corresponding to each production process.

[0017] Optionally, when multiple casings are in simultaneous production, determining a bidirectional storage path based on the data volume of each casing's production data set and the Internet of Things (IoT), and determining a sub-storage path for each casing's production data set based on the bidirectional storage path, the storage nodes of each casing's production data set, and the IoT's storage congestion coefficient, includes:

[0018] Multiple shells are produced simultaneously, and the shell production data sets are stored simultaneously. At this time, the data volume of each shell production data set is collected, and the first data storage event is determined based on the data volume of each shell production data set and the shape of each shell.

[0019] The second data storage event is determined based on the data volume of each shell's production data set and the Internet of Things, and a bidirectional storage path is determined based on the first and second data storage events.

[0020] Optionally, when multiple casings are in simultaneous production, determining a bidirectional storage path based on the data volume of each casing's production data set and the Internet of Things (IoT), and determining a sub-storage path for each casing's production data set based on the bidirectional storage path, the storage nodes of each casing's production data set, and the IoT's storage congestion coefficient, further includes:

[0021] The system collects production data sets from various casings, identifies storage nodes, determines multiple data storage routes based on the location of each storage node and bidirectional storage path, determines corresponding storage processes based on the detection of multiple data storage routes, determines corresponding storage loads based on the identification of these storage processes, and determines sub-storage paths for each casing's production data set based on the storage load of each data storage route, bidirectional storage paths, and the storage congestion coefficient of the Internet of Things.

[0022] Optionally, the step of determining multiple abnormal storage parameters based on each sub-storage path, determining abnormal storage areas based on the multiple abnormal storage parameters and corresponding production processes, and determining storage anomaly handling items based on the abnormal storage areas and the storage progress of each shell's production data set includes:

[0023] Collect each sub-storage path, determine the storage parameter set based on the synchronous detection of each sub-storage path, determine multiple storage parameter combinations based on the detection of the storage parameter set, and determine multiple abnormal storage parameters based on the identification of multiple storage parameter combinations.

[0024] The first region is determined based on multiple abnormal storage parameters and corresponding sub-storage paths, the second region is determined based on multiple abnormal storage parameters and corresponding production processes, and the abnormal storage region is determined based on the cross-matching of the first region and the second region.

[0025] Optionally, the step of determining multiple abnormal storage parameters based on each sub-storage path, determining abnormal storage areas based on the multiple abnormal storage parameters and corresponding production processes, and determining storage anomaly handling items based on the abnormal storage areas and the storage progress of each shell's production data set, further includes:

[0026] Collect the current production events of the shell, determine the current production content of the shell based on the detection of the current production events, and determine the storage anomaly handling items based on the current production content of the shell, the regional location of the abnormal storage area, and the storage progress of the production data set of each shell.

[0027] Optionally, the step of determining multiple sub-processing events based on the identification of storage exception handling items, and determining the production parameter storage system based on the multiple sub-processing events, each sub-storage path, and the production status of each casing, includes:

[0028] Collect storage exception handling items, determine multiple sub-storage exception handling lists based on the identification of storage exception handling items, and determine multiple sub-processing events based on the parsing of multiple sub-storage exception handling lists.

[0029] Optionally, the step of determining multiple sub-processing events based on the identification of storage exception handling items, and determining the production parameter storage system based on the multiple sub-processing events, each sub-storage path, and the production status of each casing, further includes:

[0030] Collect data from each sub-storage path, determine the first sub-storage system based on multiple sub-processing events and each sub-storage path, and simultaneously determine the second sub-storage system based on multiple sub-processing events and the production status of each casing.

[0031] The matching coefficient is determined based on the matching of the first sub-storage system and the second sub-storage system. If the matching coefficient is lower than the preset matching coefficient threshold, the first sub-storage system and the second sub-storage system are traced back in reverse to determine the matching coefficient impact event. The matching coefficient is adjusted according to the optimization of the matching coefficient impact event so that the adjusted matching coefficient reaches the preset matching coefficient threshold. The first sub-storage system or the second sub-storage system is optimized simultaneously. The production parameter storage system is determined based on the optimized first sub-storage system and the second sub-storage system.

[0032] This invention also provides an IoT-based shell production parameter storage system, comprising:

[0033] The production status module is used to determine the combination of production parameters based on the real-time detection of multiple production processes of the shell, and to determine the production status of each production process based on the identification of the combination of production parameters.

[0034] The multimodal data module is used to construct a production dynamic map of the shell based on multiple production images of the shell in each production process, and to determine the production data set of the shell based on each production dynamic map, the production progress of the shell, and the production status of each production process.

[0035] The sub-storage path module is used to determine the bidirectional storage path based on the data volume of the production data set of each shell and the Internet of Things when multiple shells are in synchronous production. It also determines the sub-storage path of the production data set of each shell based on the bidirectional storage path, the storage nodes of the production data set of each shell, and the storage congestion coefficient of the Internet of Things.

[0036] The storage exception handling project module is used to determine multiple exception storage parameters based on each sub-storage path, determine the exception storage area based on the multiple exception storage parameters and the corresponding production process, and determine the storage exception handling project based on the exception storage area and the storage progress of the production data set of each shell.

[0037] The production parameter storage system module is used to identify multiple sub-processing events based on the identification of storage anomaly handling items, and to determine the production parameter storage system based on multiple sub-processing events, each sub-storage path, and the production status of each shell.

[0038] Compared with the prior art, the beneficial effects of the present invention are:

[0039] The IoT-based shell production parameter storage method provided in this embodiment of the invention determines a bidirectional storage path based on the data volume of each shell's production data set and the IoT when multiple shells are in synchronous production. It then determines a sub-storage path for each shell's production data set based on the bidirectional storage path, the storage nodes of each shell's production data set, and the IoT's storage congestion coefficient. This introduces the shell's production data set, controls the synchronous production of multiple shells, and incorporates a holistic consideration of the bidirectional storage path, the storage nodes of each shell's production data set, and the IoT's storage congestion coefficient, thereby improving the accuracy of the sub-storage paths for each shell's production data set.

[0040] Therefore, multiple abnormal storage parameters are determined based on each sub-storage path, and abnormal storage areas are determined based on these parameters and corresponding production processes. Storage anomaly handling items are determined based on the storage progress of the abnormal storage areas and the production data sets of each casing. Multiple sub-processing events are determined based on the identification of these storage anomaly handling items. The production parameter storage system is determined based on these sub-processing events, each sub-storage path, and the production status of each casing. By introducing storage anomaly handling items, further control over these items is achieved, realizing a holistic consideration of multiple sub-processing events, each sub-storage path, and the production status of each casing. This improves the accuracy of the production parameter storage system and enables the casing to synchronously complete the storage of production parameters during the production process. Attached Figure Description

[0041] Figure 1 This is a flowchart illustrating the method for storing shell production parameters based on the Internet of Things in an embodiment of the present invention.

[0042] Figure 2 This is a flowchart illustrating step S11 in the IoT-based shell production parameter storage method of this invention.

[0043] Figure 3 This is a flowchart illustrating step S12 in the IoT-based shell production parameter storage method in this embodiment of the invention.

[0044] Figure 4 This is a flowchart illustrating step S13 in the IoT-based shell production parameter storage method in this embodiment of the invention.

[0045] Figure 5 This is a flowchart illustrating step S14 in the IoT-based shell production parameter storage method in this embodiment of the invention.

[0046] Figure 6 This is a flowchart illustrating step S15 of the IoT-based shell production parameter storage method in this embodiment of the invention.

[0047] Figure 7 This is a schematic diagram of the structural composition of the Internet of Things-based shell production parameter storage system in an embodiment of the present invention. Detailed Implementation

[0048] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.

[0049] Please see Figures 1 to 7 A method for storing shell production parameters based on the Internet of Things (IoT), applied to production parameter storage scenarios; the method for storing shell production parameters based on the IoT includes:

[0050] Step S11: Determine the combination of production parameters based on the real-time detection of multiple production processes of the shell, and determine the production status of each production process based on the identification of the combination of production parameters.

[0051] Step S12: Construct a production dynamic diagram of the shell based on multiple production images of the shell in each production process, and determine the production data set of the shell based on each production dynamic diagram, the production progress of the shell, and the production status of each production process.

[0052] Step S13: When multiple shells are in synchronous production, determine the bidirectional storage path based on the data volume of the production data set of each shell and the Internet of Things, and determine the sub-storage path of the production data set of each shell based on the bidirectional storage path, the storage node of the production data set of each shell, and the storage congestion coefficient of the Internet of Things.

[0053] Step S14: Determine multiple abnormal storage parameters based on each sub-storage path, determine abnormal storage areas based on multiple abnormal storage parameters and corresponding production processes, and determine storage abnormality handling items based on the abnormal storage areas and the storage progress of each shell's production data set.

[0054] Step S15: Based on the identification of storage exception handling items, determine multiple sub-processing events, and determine the production parameter storage system based on the multiple sub-processing events, each sub-storage path, and the production status of each shell.

[0055] refer to Figure 2 In step S11, the specific steps are as follows:

[0056] S111: Collect the model number of the casing, determine multiple production processes of the casing based on the matching of the casing model and the casing production database, determine multiple production parameters based on the real-time detection of multiple production processes of the casing, and mark the production process and time node corresponding to each production parameter; determine the combination of each production parameter based on the part of the casing corresponding to each production parameter, the production process and time node corresponding to each production parameter.

[0057] S112: In each production process, multiple sub-production features are determined based on the identification of the corresponding combination of production parameters. The production status of the production process is determined based on the multiple sub-production features, the shape of the shell in the production process, and the part of the shell that the production process acts on, so as to collect the production status of each production process.

[0058] In the embodiments of this application, the shell model information is obtained through barcode scanning, RFID reading or visual recognition system; the model information is input into the system to trigger the subsequent process matching process; the model information includes key identifiers such as product code, version number, batch number; the system matches the identified model with the process flow table in the production database; the database stores the standard process sequence, logical relationship between processes and process parameter range corresponding to each model.

[0059] Deploy corresponding sensors and detection equipment in each process; collect physical, chemical, and geometric parameters related to the process in real time; dynamically adjust the parameter acquisition frequency according to process characteristics and requirements; add process identifiers and timestamps to each acquired parameter; the timestamps are accurate to the millisecond level to ensure the timing accuracy of the parameters; establish a three-dimensional correlation between parameters, processes, and time.

[0060] Based on the corresponding parts of the casing, production processes, and time points for each production parameter, the combination of production parameters is determined. The correlation between parameters and specific parts of the casing is analyzed. Parameters at the same time point, in the same process, and at the same location are combined to establish a weighted relationship between parameter combinations, reflecting the degree of influence of each parameter on the final quality. This refined parameter combination management ensures that every detail in the production process can be accurately recorded and analyzed, providing strong data support for the quality control of e-cigarette casings. When quality problems occur, the specific combination of production parameters can be quickly traced back to find the root cause of the problem.

[0061] Furthermore, the system analyzes the parameter combinations of each process to extract key features that reflect production quality. Sub-production features include multiple dimensions such as physical characteristics, geometric characteristics, and surface characteristics. The feature extraction process considers the interaction and comprehensive effect between parameters. A threshold range is set for each sub-feature for subsequent status evaluation. The system integrates all sub-production feature data and combines it with the current shell morphology information. It analyzes the feature changes of the parts affected by the process and evaluates the quality status of those parts. The system comprehensively evaluates the production status through algorithmic models (such as neural networks, decision trees, etc.). The production status is divided into multiple levels: excellent, good, qualified, warning, abnormal, etc.

[0062] The production status assessment results of each process are recorded in the database. The status data includes status level, confidence level, key indicators, and anomaly details. A status history record is established for trend analysis and prediction. The status data is bound to the product ID to achieve full-process traceability. This refined production status collection mechanism enables the production process of each e-cigarette shell to be fully monitored and evaluated. When the system detects a "warning" or "abnormal" status, it automatically triggers the corresponding processing procedure to ensure that the product quality meets the requirements. At the same time, this status data also provides valuable reference information for subsequent process optimization and quality improvement.

[0063] refer to Figure 3 In step S12, the specific steps are as follows:

[0064] S121: Collect the position of the outer shell in each production process, and trigger multiple cameras around it according to the position. The multiple cameras take pictures of the outer shell in multiple directions to collect multiple production images of the outer shell in each production process. Each production image presents different production content of the outer shell and records the corresponding posture of the outer shell.

[0065] S122: Based on the recognition of each production image, determine multiple pose features, determine multiple production behaviors of the shell based on the multiple pose features, the shape of the shell and the time sequence corresponding to each production image, and construct the production dynamic map of the shell based on the synthesis of the multiple production behaviors of the shell.

[0066] S123: Mark the production progress of the shell in the production dynamic diagram of the shell, determine the production events of the shell according to each production dynamic diagram, the corresponding production process and the production progress of the shell; determine the shell production data set according to the production events of the shell and the production status corresponding to each production process.

[0067] In the embodiments of this application, the precise position of the casing on the production line is monitored in real time through a position sensor network; the position acquisition system includes various technologies such as RFID tags, photoelectric sensors, and visual positioning systems; the position data includes three-dimensional coordinate information (X, Y, Z axes) and production line segment identification; the position update frequency is dynamically adjusted according to the production line speed (usually 10-100Hz); the system establishes a real-time mapping relationship between the casing ID and the position; the system determines which process area it enters based on the casing position information; each process area is pre-set with multiple cameras to form a 360° monitoring network without blind spots; camera triggering adopts an event-driven mechanism to avoid invalid shooting and data redundancy; triggering conditions include: position arrival, process start, specific action detection, etc.; the system dynamically adjusts the number of cameras and shooting strategy according to the importance of the process.

[0068] The system receives and stores image data from multiple cameras in real time; the image acquisition process includes preprocessing (noise reduction, enhancement, normalization, etc.); metadata tags are added to each image to facilitate subsequent retrieval and analysis; image storage adopts a hierarchical strategy (important images are stored at high resolution, while general images are stored in compressed form); an image indexing system is established to support fast querying and backtracking; at the same time, each image focuses on specific production content, reflecting the key quality characteristics of the process; attitude recording includes six degrees of freedom information such as the spatial position, orientation, and angle of the shell; attitude data is acquired through image analysis or additional sensors; attitude information is associated with image content, supporting 3D reconstruction and motion analysis; attitude change records reflect the dynamic characteristics of the production process.

[0069] Furthermore, computer vision algorithms (such as CNN, YOLO, key point detection, etc.) are used to process the images captured by each camera to identify key parts and feature points of the shell; features such as position, angle, rotation, displacement, and surface state of the shell are extracted from the images to form "pose features"; image features from different angles are unified into the same coordinate system to facilitate subsequent fusion analysis; and the pose features of multiple images at the same time point are correlated to form a complete pose description of the shell at that moment.

[0070] Arrange the attitude features at different time points in chronological order to analyze the movement and change trends of the shell; based on the changes in attitude features (such as displacement, rotation, and surface condition changes), identify the "production behavior" of the shell (such as movement, rotation, processing, and inspection); combine the CAD model or design specifications of the shell to verify whether the attitude features meet the expected shape; classify the identified behaviors as "normal", "warning", or "abnormal" and record the duration of the behavior.

[0071] Therefore, in the constructed production dynamic diagram, each shell is marked with its current production stage, such as: completed process, currently in progress process, and process not started; the production completion rate of the shell is represented by a percentage or stage number (such as "3 / 10"); each progress node is associated with a timestamp for easy timeline tracking; progress information comes from process sensors, PLC signals, image recognition results, etc.; different states are marked with different colors or symbols in the dynamic diagram (such as green for completed, yellow for in progress, and gray for not started).

[0072] Production events for the casing are determined based on the various production dynamic diagrams, corresponding production processes, and the production progress of the casing. A production event refers to a specific point in time or a change in state that has specific significance during the production process, such as: process start, process completion, equipment alarm, quality anomaly, etc. Events are recorded based on conditions such as progress changes, process state transitions, and sensor signal changes in the dynamic diagrams. Events are categorized into normal events (such as process completion), warning events (such as parameters approaching thresholds), and abnormal events (such as equipment failure). Each event includes information such as event type, occurrence time, associated processes, relevant parameters, and processing status. Events are associated with the corresponding production dynamic diagrams to form a complete event chain.

[0073] All relevant data (including parameters, images, events, status, etc.) of the shell throughout the entire production process are integrated into a complete dataset; production events are associated with the production status of the corresponding process (from S112) to form a complete production record; the dataset is typically stored in a structured format (such as JSON, XML) or database table format, containing the following information: basic information: shell ID, model, batch, etc.; production parameters: process parameters of each process; production images: image data of key processes; production events: event records; production status: status evaluation results of each process; quality data: inspection results, quality scores, etc.; appropriate storage methods (such as relational database, file storage, object storage, etc.) are selected according to data type and access requirements.

[0074] refer to Figure 4 In step S13, the specific steps are as follows:

[0075] S131: Multiple shells are being produced simultaneously, and the shell production data sets are being stored simultaneously. At this time, the data volume of each shell's production data set is collected, and the first data storage event is determined based on the data volume of each shell's production data set and the shape of each shell.

[0076] S132: Determine the second data storage event based on the data volume of the production data set of each shell and the Internet of Things, and determine the bidirectional storage path based on the first data storage event and the second data storage event;

[0077] S133: Collect the storage nodes of the production data sets of each casing, determine multiple data storage routes based on the location of each storage node and the bidirectional storage path, determine the corresponding storage process based on the detection of multiple data storage routes, determine the corresponding storage load based on the identification of the storage process, and determine the sub-storage path of the production data sets of each casing based on the storage load of each data storage route, the bidirectional storage path and the storage congestion coefficient of the Internet of Things.

[0078] In the embodiments of this application, multiple shells are produced synchronously, and the shell production data sets are stored synchronously. At this time, the data volume of each shell production data set is collected, and a first data storage event is determined based on the data volume of each shell production data set and the shape of each shell, thus introducing a first data storage event.

[0079] At this time, multiple shells are being produced simultaneously, and the production data sets of the shells are being stored synchronously. Meanwhile, multiple e-cigarette shells are in different production processes at the same time, such as injection molding, surface coating, assembly, and quality inspection. These shells may belong to the same batch or different models, but they are all processed within the same time period. The system collects and stores the production data sets of these shells in parallel (including images, parameters, status, timestamps, etc.). The storage system needs to support high-concurrency writing, such as a distributed storage architecture (such as HDFS, MinIO, AWS S3). Asynchronous writing, batch submission, and data sharding technologies are used to ensure that stability and consistency are maintained even when the data volume surges.

[0080] Specifically, assuming there are currently 3 e-cigarette casings (ID: EC001, EC002, EC003) being produced simultaneously, the data status is illustrated in Table 1:

[0081] Table 1: Data Status Diagram

[0082]

[0083] The system writes this data to the distributed storage nodes simultaneously.

[0084] For example: EC001 → node1.storage.local;

[0085] EC002 → node2.storage.local;

[0086] EC003 → node3.storage.local;

[0087] The system calculates the size of each shell production data set in real time, including: image data (such as high-definition photos taken from multiple angles); process parameters (temperature, pressure, speed, etc.); status logs (qualified, warning, abnormal, etc.); timestamps and metadata; and dynamically adjusts the data according to the complexity of the process, such as collecting data once per second for critical processes (such as spraying) and once every 5 seconds for ordinary processes; the data volume is usually in MB or GB units, which facilitates the formulation of subsequent storage strategies.

[0088] The first data storage event is determined based on the data volume of each shell's production data set and the shape of each shell. Shell shape analysis includes the shell's model, structural complexity, number of processes, and material properties. For example, complex models (such as those with textures or curved surfaces) typically have larger data volumes and higher storage requirements. The system determines whether to trigger a specific storage mechanism based on the data volume and shape characteristics, such as: compressed storage: lossless compression of shells with large data volumes (e.g., PNG, ZIP); block storage: splitting large datasets into multiple smaller blocks for distributed storage; priority queue: prioritizing the storage of high-priority data (e.g., critical processes); caching mechanism: temporarily storing frequently accessed data to improve read efficiency. The triggering conditions for the first storage event are: data volume exceeding a threshold (e.g., >20MB); high shell shape complexity (e.g., multiple curved surfaces, multiple processes); and high storage system load (e.g., CPU / memory utilization >80%).

[0089] Specifically, the data collected by the system is illustrated in Table 2:

[0090] Table 2: Data Composition Diagram

[0091]

[0092] Assume the system is configured with the following rules: Data size > 20MB → trigger block storage; Complex data type (e.g., ECPro model) → trigger compressed storage. A table illustrating the first trigger event is provided, as shown in Table 3.

[0093] Table 3. Illustration of the First Triggering Event

[0094]

[0095] Specific storage event: EC002: Block splitting: Split 22.8MB of data into 3 blocks (7.6MB / block); Compression: Compress each block to 6MB using a lossless compression algorithm (such as ZIP); Storage: Write the compressed blocks to node1, node2, and node3 respectively.

[0096] Furthermore, based on the data volume of the production data sets of each shell and the Internet of Things, a second data storage event is determined. A bidirectional storage path is determined based on the first and second data storage events, which takes into account the overall consideration of the first and second data storage events and ensures the accuracy of the bidirectional storage path.

[0097] At this time, the system continuously monitors the size of the production data sets of each shell and makes a comprehensive judgment based on the real-time status of the Internet of Things (IoT) platform (such as network bandwidth, storage node load, cloud response latency, etc.). The IoT system provides indicators such as storage node health, network transmission quality, and storage space utilization to assess whether the current storage environment is suitable for direct storage or needs optimization. Second data storage event determination: When the data volume is large or the IoT environment is poor (such as high latency, node congestion), the system triggers a second data storage event, such as: data fragmentation and uploading; data caching and temporary storage; data degradation storage (such as transferring from the cloud to the edge node); data priority adjustment (high-priority data is uploaded first).

[0098] Specifically, assuming the system is currently monitoring an IoT status indicator table, as shown in Table 4:

[0099] Table 4. IoT Status Diagram

[0100]

[0101] The system determines the status based on the IoT status: EC001: Moderate data volume, good IoT status, no need to trigger the second event; EC002: Large data volume and poor IoT status, trigger the second event: data fragmentation and upload + priority increase; EC003: Small data volume, good IoT status, no need to trigger the second event.

[0102] A bidirectional storage path is determined based on the first and second data storage events. The first data storage event is determined by the data volume and shell form, such as direct storage, block storage, and compressed storage. The second data storage event is determined by the IoT status, such as caching, fragmented uploading, and degraded storage. Combining the first and second events, the system constructs a bidirectional path of "upload + backup" or "local + cloud" to ensure that data can be quickly responded to locally and securely backed up to the cloud. Path selection strategy: If the IoT status is good: directly upload to the cloud while keeping a copy locally; if the IoT status is poor: cache locally first, and synchronize to the cloud after the network recovers; if the data volume is extremely large: fragmented and uploaded to multiple nodes in parallel to improve transmission efficiency.

[0103] Specifically, EC001: the first event is "direct storage", and the second event is "none"; EC002: the first event is "block compression", and the second event is "fragmented upload"; EC003: the first event is "direct storage", and the second event is "none"; the system constructs a bidirectional storage path, and the schematic diagram of this bidirectional storage path is shown in Table 5:

[0104] Table 5. Schematic Diagram of Bidirectional Storage Path

[0105]

[0106] EC001: Data is written directly to local node1 and simultaneously uploaded asynchronously to cloud backup1; EC002: Data is compressed into 3 blocks; the blocks are uploaded in parallel to cloud nodes3 and 4; local node2 retains a complete copy; EC003: Same as EC001, data is written directly to local node1 and cloud backup1.

[0107] Therefore, by collecting the storage nodes of the production data sets of each casing, determining multiple data storage routes based on the location of each storage node and the bidirectional storage path, determining the corresponding storage process based on the detection of multiple data storage routes, determining the corresponding storage load based on the identification of the storage process, and determining the sub-storage path of the production data set of each casing based on the storage load of each data storage route, the bidirectional storage path, and the storage congestion coefficient of the Internet of Things, the accuracy of the sub-storage path of the production data set of each casing is improved.

[0108] At this time, the system collects the production data sets of each shell and monitors the currently available storage nodes in real time, including local nodes (such as edge servers and local databases) and remote nodes (such as cloud storage and distributed storage systems). Each node has attributes such as location (such as local data center, regional cloud, and overseas cloud), available storage space, current load, network bandwidth, and latency. The system allocates one or more candidate storage nodes for each shell's production data set.

[0109] In S132, the path strategy for each shell, such as "local-cloud" or "shard-backup", has been determined. Combining the node location and path strategy, one or more storage routes are generated for each shell. For example: local priority route: write to the local node first, and then synchronize to the cloud; cloud priority route: upload directly to the cloud, suitable for local resources that are scarce; hybrid route: store some data locally and some data in the cloud.

[0110] Specifically, assuming there are multiple storage nodes in the current system, a storage node illustration table is collected, as shown in Table 6:

[0111] Table 6. Storage Node Diagram

[0112]

[0113] For the three electronic cigarette casings EC001, EC002, and EC003, the storage nodes collected by the system are as follows: the candidate storage nodes for EC001 are node1 and node3; the candidate storage nodes for EC002 are node2, node3, and node4; and the candidate storage nodes for EC003 are node1 and node3.

[0114] Assume the bidirectional storage paths are as follows: EC001 - local + cloud; EC002 - sharding + multi-node backup; EC003 - local + cloud; based on the node locations, determine the data storage routes: the data storage route for EC001 is node1 → node3; the data storage route for EC002 is node2 → node3, node2 → node4; the data storage route for EC003 is node1 → node3.

[0115] The system checks each storage route, including: reachability (whether the node is online); transmission rate (whether the bandwidth is sufficient); and whether there is packet loss or latency. Based on the detection results, the system starts the corresponding storage process for each route. For example: synchronous process: real-time writing, suitable for critical data; asynchronous process: delayed writing, suitable for non-critical data; sharding process: parallel writing to multiple nodes to improve efficiency.

[0116] Storage load refers to the current storage process's utilization of node resources, including: CPU utilization; memory usage; disk I / O; network bandwidth usage. The system monitors the load of each storage process in real time to ensure that the node is not overloaded. Specifically, a status illustration table of each route is collected, as shown in Table 7.

[0117] Table 7: Route Status Diagram

[0118]

[0119] Therefore, the storage processes for each shell are: EC001 - synchronous process (node1 → node3); EC002 - asynchronous process (node2 → node3) + sharding process (node2 → node4); EC003 - synchronous process (node1 → node3); the load diagram of each process is shown in Table 8.

[0120] Table 8: Load Diagram of Each Process

[0121]

[0122] The sub-storage paths for the production data sets of each enclosure are determined based on the storage load of each data storage route, the bidirectional storage path, and the storage congestion coefficient of the Internet of Things (IoT). The storage congestion coefficient is a congestion index (0~1) calculated by the IoT system based on the current overall storage environment, with higher values ​​indicating greater congestion. The system combines storage load, path strategy, and congestion coefficient to ultimately determine the sub-storage path for each enclosure. The rules include: high load + high congestion coefficient → select low load route or delayed storage; low load + low congestion coefficient → prioritize high performance route; critical data → prioritize high reliability route.

[0123] Specifically, assuming the current IoT storage congestion coefficient is 0.7 (high congestion); this congestion coefficient is illustrated in Table 9:

[0124] Table 9: Congestion Coefficient Diagram

[0125]

[0126] The final sub-storage paths are as follows: EC001's sub-storage path is node1 → node3 (low-priority synchronization); EC002's sub-storage path is node2 → node4 (sharded upload); EC003's sub-storage path is node1 → node3 (low-priority synchronization). Meanwhile, EC001 and EC003, due to high congestion, use low-priority synchronous storage; EC002, due to high load, only retains sharded uploads to node4. This method ensures that data storage remains efficient, reliable, and resource-efficiently utilized even in high-concurrency, high-dynamic production environments.

[0127] refer to Figure 5 In step S14, the specific steps are as follows:

[0128] S141: Collect each sub-storage path, determine the storage parameter set based on the synchronous detection of each sub-storage path, determine multiple storage parameter combinations based on the detection of the storage parameter set, and determine multiple abnormal storage parameters based on the identification of multiple storage parameter combinations.

[0129] S142: Determine the first region based on multiple abnormal storage parameters and corresponding sub-storage paths, determine the second region based on multiple abnormal storage parameters and corresponding production processes, and determine the abnormal storage region based on the cross-matching of the first region and the second region;

[0130] S143: Collect the current production events of the shell, determine the current production content of the shell based on the detection of the current production events, and determine the storage anomaly handling items based on the current production content of the shell, the location of the abnormal storage area, and the storage progress of the production data set of each shell.

[0131] In the embodiments of this application, each sub-storage path is collected. The system performs real-time synchronous detection on each sub-storage path, collecting multiple key indicators to form a storage parameter set. Typical storage parameters include: transmission rate (MB / s); packet loss rate (%); transmission latency (ms); storage completion rate (%); node response time (ms); retransmission count; bandwidth utilization rate (%). The system combines the storage parameter set into multiple storage parameter combinations according to logical correlation or business rules. Each parameter combination represents a specific storage behavior pattern, such as: high-efficiency transmission combination: high transmission rate + low packet loss rate + low latency; congestion risk combination: high bandwidth utilization + high retransmission count + high latency; unstable connection combination: high packet loss rate + high retransmission count + low completion rate.

[0132] The system evaluates the combination of storage parameters through preset thresholds or intelligent algorithms (such as machine learning models) and identifies abnormal parameters. Abnormal parameters refer to parameter values ​​that deviate significantly from the normal range or indicate potential problems. Common abnormal parameters include: transmission rate below the threshold; packet loss rate exceeding the upper limit; excessive transmission latency; abnormally increased retransmission count; and stagnant storage completion.

[0133] Specifically, assuming there are three e-cigarette casings EC001, EC002, and EC003, they use the following sub-storage paths: EC001: Local edge node → Cloud main storage (path A); EC002: Local edge node → Regional cloud node (path B); EC003: Local edge node → Local backup storage (path C). The system detects that all three paths are currently active and data is being transmitted. After performing synchronization detection on the above three paths, the system obtains a schematic table of storage parameter sets, as shown in Table 10.

[0134] Table 10: A schematic diagram of the storage parameter set.

[0135]

[0136] The system identified the following combinations of storage parameters: High-efficiency transmission combination (path A): high transmission rate (50 MB / s); low packet loss rate (0.1%); low latency (30 ms); high completion rate (80%); Congestion risk combination (path B): high bandwidth utilization (90%); high retransmission frequency (8 times); high latency (200 ms); low completion rate (30%); General performance combination (path C): all parameters are at a moderate level; no obvious abnormalities but not particularly efficient.

[0137] After evaluating the above parameter combinations, the system identified the following abnormal storage parameters: Abnormal parameters for path B: abnormally low transmission rate (10 MB / s, below the normal threshold of 25 MB / s); abnormally high packet loss rate (5%, exceeding the upper limit of 1%); abnormally high transmission latency (200 ms, far exceeding the normal value of 50 ms); abnormally high number of retransmissions (8 times, normally less than 3 times); excessively low storage completion rate (30%, expected to reach at least 60%); Slight abnormalities for path C: slightly high packet loss rate (0.5%, close to the upper limit); excessive number of retransmissions (3 times, close to the upper limit); No abnormalities for path A: all parameters are within the normal range. This refined anomaly detection mechanism provides an accurate data foundation for subsequent storage path optimization and anomaly handling, ensuring that electronic cigarette casing production data can be stored reliably and efficiently.

[0138] Furthermore, a first region is determined based on multiple abnormal storage parameters and corresponding sub-storage paths, and a second region is determined based on multiple abnormal storage parameters and corresponding production processes. An abnormal storage region is determined based on the cross-matching of the first and second regions, which takes into account the overall consideration of the cross-matching of the first and second regions and ensures the accuracy of the abnormal storage region.

[0139] At this point, the first region is determined based on multiple abnormal storage parameters and corresponding sub-storage paths. The system first analyzes the abnormal storage parameters identified in S141 and associates the sub-storage paths where these parameters are located. The first region refers to the problem area in the storage system, which may be a storage node, network link, storage service, or storage policy module. The abnormal parameters are grouped by type (such as network type, node type, policy type), the paths in the abnormal parameter set are analyzed, commonalities are found (such as shared nodes, shared network segments), and the paths are mapped to physical or logical storage areas (such as "cloud main storage area", "edge node cluster", "regional cloud network segment", etc.).

[0140] Specifically, the system identified the following: Path B (local edge node → regional cloud node) has a serious anomaly; Path C (local edge node → local backup storage) has a minor anomaly. The system then determined the first region: Anomalous parameters for Path B: low transmission rate, high packet loss rate, high latency, and high retransmission frequency; Anomalous parameters for Path C: slightly higher packet loss rate and higher retransmission frequency; Both Path B and C pass through "local edge node"; Path B uniquely contains "regional cloud node"; Path C uniquely contains "local backup storage"; Region mapping: Common region: "local edge node cluster" (because both paths pass through this area); Specific region: "regional cloud network link" (unique to Path B); Specific region: "local backup storage network" (unique to Path C); The first region determined by the system is: Primary anomaly region: "local edge node cluster" (because both paths have anomalies); Secondary anomaly region: "regional cloud network link" (Path B has a serious anomaly).

[0141] The second region is determined based on multiple abnormal storage parameters and corresponding production processes. The system analyzes the production processes corresponding to the abnormal storage parameters to identify the problem areas in the production system. The second region refers to the problematic link in the production process, which may be a certain station, equipment, process stage, or production batch. The abnormal parameters are associated with the production processes that generated these data, and the correspondence between the time of occurrence of the abnormal parameters and the production processes is analyzed to map the problem to specific production links (such as "injection molding station", "spraying section", "quality inspection link", etc.).

[0142] Specifically, the system analyzes the production processes corresponding to the abnormal stored parameters: data for path B mainly comes from the "injection molding" and "surface coating" processes; data for path C mainly comes from the "quality inspection" process; the anomaly in path B is most obvious during the peak injection molding period (14:00-16:00); the anomaly in path C occurs during batch processing of quality inspection (16:30-17:30); the anomaly in path B corresponds to the "injection molding station" and "surface coating section"; the anomaly in path C corresponds to the "quality inspection link"; the second area determined by the system is: the main anomaly area: "injection molding station" (because path B has a serious anomaly and a large amount of data); the secondary anomaly area: "quality inspection link" (path C has a slight anomaly).

[0143] The abnormal storage area is determined based on the cross-matching of the first and second regions. The system performs cross-matching analysis on the first region (the storage system problem area) and the second region (the production process problem area). The purpose of cross-matching is to find the correlation between storage anomalies and production anomalies and to determine the root cause of the problem. The system constructs a correlation matrix between storage areas and production processes, calculates the correlation coefficients of different region combinations, and infers the most likely abnormal storage area based on the correlation.

[0144] Specifically, the system performs cross-matching analysis and collects an association matrix diagram, as shown in Table 11:

[0145] Table 11. Illustration of the Association Matrix

[0146]

[0147] "Local edge node cluster" and "injection molding station": correlation coefficient 0.85 (high correlation); "Regional cloud network link" and "injection molding station": correlation coefficient 0.78 (high correlation); "Local edge node cluster" and "quality inspection link": correlation coefficient 0.25 (low correlation); the highest correlation occurs in the combination of "local edge node cluster" and "injection molding station"; the second highest correlation occurs in the combination of "regional cloud network link" and "injection molding station".

[0148] The abnormal storage areas identified by the system are as follows: Primary abnormal storage area: "Local edge node cluster - injection molding station" interface area; Explanation: The injection molding station generates a large amount of data, and the local edge node cluster encounters a bottleneck when processing this data; Secondary abnormal storage area: "Regional cloud network link - injection molding station" transmission channel; Explanation: The transmission link from the edge node to the regional cloud is congested during peak injection molding periods.

[0149] Therefore, the system collects production events of each enclosure in real time through IoT sensors, PLC controllers, MES systems, etc. Production events include: process start / end, equipment status changes, quality inspection results, material changes, manual intervention, etc. Each event contains attributes such as timestamp, event type, related equipment, operator, and event parameters. The system will establish an event queue and process these events in chronological order.

[0150] The system analyzes the latest collected production events to determine the current production activities of each shell. The production content includes: process type, process progress, equipment status, process parameters, quality status, etc. The system will establish a production content model and convert the events into structured production status information. For complex processes, the system will infer detailed sub-steps and completion rates based on the event sequence.

[0151] Based on the current production content of the casing, the location of the abnormal storage area, and the storage progress of the production data sets of each casing, the storage anomaly handling items are determined. The system comprehensively analyzes three key factors: Current production content: understanding the production status and urgency of each casing; Abnormal storage area: information on the problem area obtained from S142; Storage progress: the storage completion status of the data sets of each casing. The system establishes a processing decision matrix, considering: Production urgency: prioritizing critical processes and high-value products; Storage impact: prioritizing those with high risk of data loss; Resource availability: currently available storage resources and network bandwidth. The determined processing items include: Processing type: data migration, storage path switching, compression strategy adjustment, cache cleanup, etc.; Processing priority: high / medium / low; Processing time: immediate execution, delayed execution, scheduled execution; Resource allocation: which storage resources need to be mobilized.

[0152] Specifically, the system analyzes the following information: Current production content: EC001: Quality inspection in progress (critical process); EC002: Spraying completed, awaiting the next process (non-urgent); EC003: Injection molding in progress (important process); Abnormal storage area (obtained from S142): Major anomaly: "Local edge node cluster - Injection molding station" interface area; Minor anomaly: "Regional cloud network link - Injection molding station" transmission channel; Storage progress: EC001: 85% stored (remaining quality inspection data to be stored); EC002: 100% stored (completed); EC003: 30% stored (a large amount of injection molding data to be stored).

[0153] System-identified storage anomaly handling items: Item 1: Emergency optimization of EC003 data storage: Processing type: storage path switching + data compression; Specific measures: Immediately switch the storage path of EC003 from "local edge node → regional cloud" to "local edge node → local backup → regional cloud"; enable high compression ratio algorithm (increase from 5:1 to 8:1); reserve dedicated bandwidth to ensure data transmission; Priority: High; Execution time: Immediate execution; Resource allocation: allocate 50% of the local backup server resources, reserve 100Mbps network bandwidth; Item 2: EC001 quality inspection data storage guarantee: Processing type: storage path optimization + priority increase; Specific measures: increase the priority of EC001 quality inspection data storage to the highest level; bypass the abnormal "local edge node cluster" and directly transmit to the regional cloud; enable dual-write mechanism (simultaneous writing to local and cloud); Priority: High; Execution time: Immediate execution; Resource allocation: reserve cloud storage space, enable dedicated transmission channel; Item 3: Edge node cluster load balancing: Processing type: system optimization + Resource scheduling; Specific measures: reallocate storage tasks for edge node clusters; clean up temporary cache files to release space; optimize data sharding strategy; Priority: Medium; Execution time: to be executed in 30 minutes (after emergency processing is completed) Resource allocation: system maintenance window, call cluster management resources.

[0154] refer to Figure 6 In step S15, the specific steps are as follows:

[0155] S151: Collect storage exception handling items, determine multiple sub-storage exception handling lists based on the identification of storage exception handling items, and determine multiple sub-processing events based on the parsing of multiple sub-storage exception handling lists;

[0156] S152: Collect each sub-storage path, determine the first sub-storage system based on multiple sub-processing events and each sub-storage path, and determine the second sub-storage system based on multiple sub-processing events and the production status of each casing.

[0157] S153: Based on the matching of the first sub-storage system and the second sub-storage system, the corresponding matching coefficient is determined. If the matching coefficient is lower than the preset matching coefficient threshold, the first sub-storage system and the second sub-storage system are traced back in reverse to determine the matching coefficient impact event. The matching coefficient is adjusted according to the optimization of the matching coefficient impact event so that the adjusted matching coefficient reaches the preset matching coefficient threshold. The first sub-storage system or the second sub-storage system is optimized simultaneously. The production parameter storage system is determined based on the optimized first sub-storage system and the second sub-storage system.

[0158] In the embodiments of this application, the system first collects the "storage anomaly handling items" already determined in S143. These items are processing strategies generated based on the abnormal storage area and the current production event. Each processing item includes: anomaly type (e.g., path congestion, node overload, data loss, etc.); scope of impact (e.g., shell number, production process, storage path, etc.); processing suggestions (e.g., switching paths, compressing data, retransmission mechanism, etc.); priority (e.g., high, medium, low); and execution conditions (e.g., triggered when packet loss rate > 5%).

[0159] The system categorizes the collected storage anomaly handling items according to their handling type, forming multiple "sub-storage anomaly handling lists." Common classification methods include: by handling mechanism: path optimization, data compression, retransmission recovery, and resource scheduling; by impact scope: shell-level anomalies, process-level anomalies, and system-level anomalies; and by urgency: emergency handling list, routine handling list, and preventative handling list.

[0160] Specifically, assuming the system identifies storage anomaly handling items in S143, the system categorizes these anomaly handling items into the following sub-storage anomaly handling lists: Path optimization list: P1: EC003 injection molding data path switching (switching from path 1 to path 3); Data compression list: P2: EC001 quality inspection data compressed storage (compression rate 50%); Retransmission recovery list: P3: EC002 packaging data retransmission mechanism enabled (maximum 3 retries); Resource scheduling list: P4: EC005 assembly data historical data cleanup (retaining data from the last 7 days).

[0161] The system performs in-depth analysis of each sub-storage exception handling list, identifying specific execution actions, triggering conditions, and affected objects, forming multiple "sub-processing events." Each sub-processing event contains the following key information: Event ID: a unique identifier; Event type: such as path switching, data compression, retransmission, etc.; Triggering condition: when the event is executed; Execution action: what exactly needs to be done; Affected objects: which shells or processes are affected; Expected effect: the desired state after execution; Execution priority: the relative importance to other events.

[0162] Specifically, the system parses the above sub-storage exception handling list and generates the following sub-processing events: Path Optimization Event: Event ID: EVT-P1-001; Event Type: Path Switching; Trigger Condition: Path 1 latency > 500ms; Action: Switch EC003 injection molding data from path 1 to path 3; Affected Object: Injection molding data of EC003 shell; Expected Effect: Latency reduced to <200ms; Execution Priority: High; Data Compression Event: Event ID: EVT-P2-001; Event Type: Data Compression; Trigger Condition: Edge node CPU utilization > 90%; Action: Compress EC001 quality inspection data by 50%; Affected Object: Quality inspection data of EC001 shell; Expected Effect: Reduce storage space usage by 50%; Execution Priority: Medium.

[0163] Retransmission Recovery Event: Event ID: EVT-P3-001; Event Type: Data Retransmission; Trigger Condition: Packet loss rate of Path 2 > 5%; Execution Action: Enable EC002 packaging data retransmission mechanism (maximum 3 times); Affected Object: Packaging data of EC002 shell; Expected Effect: Ensure data integrity > 99.9%; Execution Priority: High; Resource Scheduling Event: Event ID: EVT-P4-001; Event Type: Resource Cleanup; Trigger Condition: Remaining space of edge nodes < 10%; Execution Action: Clean up historical data of EC005 assembly data from 7 days ago; Affected Object: Assembly data of EC005 shell; Expected Effect: Release > 20% of storage space; Execution Priority: Low. This structured processing method ensures that storage anomalies can be handled systematically and in a standardized manner, laying a solid foundation for subsequent storage system optimization (S152, S153); It is particularly suitable for complex scenarios such as multi-process and multi-shell synchronous production of e-cigarette shells, and can effectively ensure the integrity of production data and the efficient operation of the system.

[0164] Furthermore, each sub-storage path is collected, and the first sub-storage system is determined based on multiple sub-processing events and each sub-storage path. At the same time, the second sub-storage system is determined based on multiple sub-processing events and the production status of each shell. This comprehensive consideration of multiple sub-processing events and the production status of each shell ensures the accuracy of the second sub-storage system.

[0165] At this point, each sub-storage path is collected, and the first sub-storage system is determined based on multiple sub-processing events and each sub-storage path. The system analyzes the multiple sub-processing events generated in S151 and, in conjunction with the current state of the sub-storage paths, constructs the first sub-storage system. The first sub-storage system refers to the storage strategy system built based on storage paths and processing events, including: path allocation strategy (which event uses which path); load balancing strategy (how to allocate data to different paths); fault tolerance strategy (backup plan in case of path failure); and performance optimization strategy (how to improve transmission efficiency).

[0166] Specifically, based on the sub-processing events generated by S151 (path switching, data compression, retransmission mechanism, resource cleanup), the system constructs the first sub-storage system: Path allocation strategy: EVT-P1-001 (path switching): EC003 injection molding data is switched from Path-B to Path-A; EVT-P2-001 (data compression): EC001 quality inspection data continues to use Path-A, but compression is enabled; EVT-P3-001 (retransmission mechanism): EC002 packaging data uses Path-C, and retransmission is enabled; EVT-P4-001 (resource cleanup): Historical data on Path-B is cleaned up; Load balancing strategy: Path-A: handles 60% of injection molding and quality inspection data; Path-C: handles 40% of packaging and assembly data; Path-B: used only for low-priority data transmission; Fault tolerance strategy: When Path-D fails, data is automatically switched to Path-C; When Path-A is congested, some data is offloaded to Path-C; Performance optimization strategy: For Path-A... Enable data compression (50% compression rate); enable incremental synchronization for Path-C (transfer only the changed parts).

[0167] The second sub-storage system is determined based on multiple sub-processing events and the production status of each shell. The system simultaneously analyzes the current production status of each shell (such as production progress, process status, and data generation rate) and constructs the second sub-storage system.

[0168] The second sub-storage system refers to the storage requirement system built based on production status and processing events, including: production priority strategy (which shell's data is stored first); data timeliness strategy (which data needs to be stored in real time); storage capacity strategy (estimate the required storage space); and data integrity strategy (how to ensure that data is not lost).

[0169] Specifically, based on the current shell production status (EC001-EC005), the system constructs a second sub-storage system: Production Priority Strategy: EC001: Injection molding completed, entering quality inspection → High priority; EC002: Packaging in progress → Medium priority; EC003: Injection molding in progress → High priority (real-time data); EC004: Awaiting assembly → Low priority; EC005: Assembly completed → Medium priority; Data Timeliness Strategy: EC003 Injection molding data: Real-time storage (latency <100ms); EC001 Quality inspection data: Near real-time storage (latency <1s); EC002 Packaging data: Batch storage (latency <5s); EC004 / EC005: Delayed storage (latency <10s); Storage Capacity Strategy: EC001: Requires 1GB storage space; EC002: Requires 500MB storage space; EC003: Requires 2GB storage space (large real-time data volume); EC004: Requires 300MB Storage space; EC005: Requires 800MB of storage space; Data integrity policy: EC003: Enable multi-replica storage (3 replicas); EC001: Enable checksum verification; EC002: Enable periodic integrity checks; EC004 / EC005: Basic storage policy.

[0170] Therefore, a matching coefficient is determined based on the matching of the first and second sub-storage systems. If the matching coefficient is lower than the preset matching coefficient threshold, the first and second sub-storage systems are traced back to determine the events affecting the matching coefficient. The matching coefficient is adjusted according to the optimization of the events affecting the matching coefficient, so that the adjusted matching coefficient reaches the preset matching coefficient threshold. The first or second sub-storage system is optimized simultaneously. The production parameter storage system is determined based on the optimized first and second sub-storage systems. At the same time, a storage anomaly handling project is introduced to further control the storage anomaly handling project. This achieves a holistic consideration of multiple sub-processing events, each sub-storage path, and the production status of each shell, improving the accuracy of the production parameter storage system. As a result, the shell can synchronously complete the storage of production parameters during the production process.

[0171] At this point, a matching coefficient is determined based on the matching of the first and second sub-storage systems. The matching coefficient is a quantitative indicator used to measure the degree of fit between the first sub-storage system (the technical system based on storage paths and events) and the second sub-storage system (the business system based on production status). The system calculates the matching coefficient through the following dimensions: path allocation consistency: whether the technical path meets business needs; load balancing rationality: whether the storage resource allocation is balanced; data timeliness guarantee: whether the real-time requirements of production data are met; fault tolerance coverage: whether the reliability requirements of critical data are met; performance indicator compliance: whether the performance requirements such as transmission speed and latency are met.

[0172] The system performs reverse tracing of the first and second sub-storage systems to determine the events affecting the matching coefficient. For dimensions with matching coefficients below the threshold, the system performs reverse tracing to identify specific events affecting the matching: inconsistent path allocation: checking which production demands are not being met; unbalanced load: identifying overloaded or idle storage resources; insufficient timeliness: identifying data transmission paths with excessively high latency; lack of fault tolerance: identifying critical data that has not been backed up; substandard performance: locating performance bottlenecks.

[0173] Specifically, the system traced and found the following impact events: Inconsistent path allocation: EC003 (injection molding station) requires real-time data transmission, but was assigned to the high-latency Path-B; EC005 (quality inspection station) requires high reliability, but multi-replica storage was not enabled; Uneven load: Path-A load was only 40%, but Path-B load reached 95%; all 2GB of data from EC003 was loaded onto Path-B; Insufficient timeliness: Path-B latency of 800ms could not meet the 100ms real-time requirement of the injection molding station; Lack of fault tolerance: The quality inspection data of EC005 was not backed up; Performance failure: Insufficient bandwidth of Path-B caused slow data transmission in EC003.

[0174] The matching coefficient is adjusted based on the optimization of events affected by the matching coefficient. For each event, the system formulates optimization measures: path reallocation: migrate critical data to more suitable paths; load rebalancing: adjust data allocation strategies; performance optimization: enable compression, caching, and other technologies; fault tolerance enhancement: add backup strategies; resource expansion: add storage resources when necessary. Based on the optimization measures, the system synchronously updates two sub-storage systems: the first sub-storage system: updates path allocation, load strategies, and fault tolerance mechanisms; the second sub-storage system: adjusts the priority of production needs and data timeliness requirements.

[0175] Specifically, the system implements the following optimization measures: Path reallocation: migrate EC003 from Path-B to Path-A (10ms latency); enable multi-replica storage for EC005 on Path-C; Load rebalancing: split 2GB of data from EC003: 1.5GB stored on Path-A and 0.5GB on Path-C; migrate some data from EC002 from Path-B to Path-A; Performance optimization: enable data compression for Path-B to reduce transmission volume; enable edge caching for EC003 to accelerate access; Fault tolerance enhancement: enable a 3-replica storage strategy for EC005; add checksums and verification for critical quality inspection data; Resource expansion: increase bandwidth resources for Path-B by 50%.

[0176] Optimized system update: First sub-storage system update: Path allocation: EC003→Path-A, EC005→Path-C (multiple replicas); Load strategy: Path-A 60%, Path-B 70%, Path-C 65%; Fault tolerance mechanism: EC005 enables 3 replicas, critical data enables checksums; Performance optimization: Path-B enables compression, EC003 enables edge caching; Second sub-storage system update: Production priority: EC003 is promoted to the highest priority; Data timeliness: Injection molding data requirements are relaxed from 100ms to 200ms; Storage capacity: EC003 reserves 2.5GB of space; Data integrity: Integrity checks are enabled for all quality inspection data.

[0177] Based on the optimized first and second sub-storage systems, the production parameter storage system is determined. The system integrates the two optimized sub-storage systems to form the final production parameter storage system: clarifying the storage location and method of each type of data; determining the data transmission path and priority; establishing a complete data protection system; and ensuring that the performance indicators meet production requirements.

[0178] Specifically, the final production parameter storage system is as follows: Storage Architecture: Injection Molding Data: Path-A (primary) + Path-C (backup); Quality Inspection Data: Path-C (3 replicas); Packaging Data: Path-B (compressed transmission); Assembly Data: Path-A + Path-C (load balancing); Transmission Strategy: Real-time Data: Prioritize Path-A (latency <20ms); Batch Data: Use Path-B (compressed transmission); Critical Data: Dual-path transmission (Path-A + Path-C); Fault Tolerance Mechanism: Critical Data: 3 replicas + checksum verification; Ordinary Data: 2 replicas; Anomaly Recovery: Automatic retransmission + local backup; Performance Guarantee: Latency Requirement: <200ms (real-time data); Bandwidth Guarantee: ≥100Mbps after Path-B expansion; Availability: 99.9% (critical data).

[0179] Please see Figure 7 , Figure 7 This is a schematic diagram of the structural composition of the IoT-based shell production parameter storage system according to an embodiment of the present invention; the IoT-based shell production parameter storage system includes:

[0180] The production status module 21 is used to determine the combination of production parameters based on the real-time detection of multiple production processes of the shell, and to determine the production status of each production process based on the identification of the combination of production parameters.

[0181] The multimodal data module 22 is used to construct a production dynamic diagram of the shell based on multiple production images of the shell in each production process, and to determine the production data set of the shell based on each production dynamic diagram, the production progress of the shell and the production status of each production process.

[0182] The sub-storage path module 23 is used to determine the bidirectional storage path based on the data volume of the production data set of each shell and the Internet of Things when multiple shells are in synchronous production, and to determine the sub-storage path of the production data set of each shell based on the bidirectional storage path, the storage nodes of the production data set of each shell and the storage congestion coefficient of the Internet of Things.

[0183] The storage exception handling project module 24 is used to determine multiple exception storage parameters based on each sub-storage path, determine the exception storage area based on the multiple exception storage parameters and the corresponding production process, and determine the storage exception handling project based on the exception storage area and the storage progress of the production data set of each shell.

[0184] The production parameter storage system module 25 is used to determine multiple sub-processing events based on the identification of storage anomaly handling items, and to determine the production parameter storage system based on multiple sub-processing events, each sub-storage path and the production status of each shell.

[0185] The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity, not all combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

Claims

1. A method for storing shell production parameters based on the Internet of Things, characterized in that, include: The combination of production parameters is determined by real-time detection of multiple production processes of the shell, and the production status of each production process is determined based on the identification of the combination of production parameters. A production dynamic diagram of the shell is constructed based on multiple production images of the shell in each production process. The production data set of the shell is determined based on each production dynamic diagram, the production progress of the shell, and the production status of each production process. When multiple shells are in simultaneous production, the shell production data set is stored synchronously. At this time, the data volume of each shell production data set is collected, and the first data storage event is determined based on the data volume and the shape of each shell. The first data storage event includes compressed storage, block storage, high priority, and caching mechanism. The second data storage event is determined based on the data volume of each shell's production data set and the real-time status of the IoT platform; the real-time status of the IoT platform includes network bandwidth, storage node load, and cloud response latency, while the second data storage event includes data fragment uploading, data caching, data degradation storage, and data priority adjustment. By combining the first and second data storage events, a bidirectional path is constructed: "upload and backup" or "local and cloud". Path selection strategy: If the IoT status is good: upload directly to the cloud while keeping a copy locally; if the IoT status is poor: cache locally first and synchronize to the cloud after the network recovers. The storage nodes collect production data sets from each casing. Based on the location of each storage node and the bidirectional storage path, multiple data storage routes are determined. The corresponding storage process is determined based on the detection of multiple data storage routes. At this time, the storage nodes include local nodes and remote nodes. Each node has attributes including location, available storage space, current load, network bandwidth, and latency. The storage process includes synchronous process, asynchronous process, and sharding process. Based on the identification of the storage process, the corresponding storage load is determined. Sub-storage paths for the production data sets of each shell are determined according to the storage load of each data storage route, bidirectional storage paths, and the storage congestion coefficient of the Internet of Things (IoT). The storage congestion coefficient is a congestion index calculated by the IoT based on the current overall storage environment. The rules for sub-storage paths include: if the load is high and the congestion coefficient is high, then a low-load route or delayed storage is selected; if the load is low and the congestion coefficient is low, then a high-performance route is selected first; and for critical data, a high-reliability route is selected first. Multiple abnormal storage parameters are determined based on each sub-storage path. An abnormal storage region is determined based on these parameters and the corresponding production process. Storage anomaly handling items are determined based on the abnormal storage region and the storage progress of each shell's production data set. This includes: collecting data from each sub-storage path; determining a set of storage parameters based on the synchronous detection of each sub-storage path; determining multiple combinations of storage parameters based on the detection of the storage parameter set; determining multiple abnormal storage parameters based on the identification of these combinations; determining a first region based on the multiple abnormal storage parameters and the corresponding sub-storage path; determining a second region based on the multiple abnormal storage parameters and the corresponding production process; and determining an abnormal storage region based on the cross-matching of the first and second regions. Based on the identification of storage anomaly handling projects, multiple sub-processing events are determined, and the production parameter storage system is determined based on the multiple sub-processing events, each sub-storage path, and the production status of each shell.

2. The method for storing shell production parameters based on the Internet of Things according to claim 1, characterized in that, The process of determining various production parameter combinations based on real-time monitoring of multiple production processes of the outer casing, and determining the production status of each production process based on the identification of various production parameter combinations, includes: The model number of the outer casing is collected. Based on the matching of the outer casing model and the outer casing production database, multiple production processes of the outer casing are determined. Based on the real-time detection of multiple production processes of the outer casing, multiple production parameters are determined, and the production process and time node corresponding to each production parameter are marked. The combination of each production parameter is determined according to the part of the outer casing corresponding to each production parameter, the production process and time node corresponding to each production parameter. In each production process, multiple sub-production features are determined based on the identification of the corresponding combination of production parameters. The production status of the production process is determined based on the multiple sub-production features, the shape of the shell in the production process, and the part of the shell that is affected by the production process, so as to collect the production status of each production process.

3. The method for storing shell production parameters based on the Internet of Things according to claim 1, characterized in that, The process involves constructing a production dynamic map of the outer shell based on multiple production images at each production stage, and determining the production data set of the outer shell based on each production dynamic map, the production progress of the outer shell, and the production status of each production stage, including: The system captures the position of the outer shell at each production stage, triggering multiple cameras around it based on that position. These cameras then capture images of the outer shell's production process from multiple directions, resulting in multiple production images of the outer shell at each stage. Each production image shows different aspects of the outer shell's production and records the corresponding posture of the outer shell. Multiple pose features are determined based on the recognition of each production image. Multiple production behaviors of the shell are determined based on the multiple pose features, the shape of the shell, and the time sequence corresponding to each production image. A production dynamic map of the shell is constructed based on the synthesis of the multiple production behaviors of the shell. The production progress of the shell is marked in the production dynamic diagram. The production events of the shell are determined according to each production dynamic diagram, the corresponding production process and the production progress of the shell. The shell production data set is determined according to the production events of the shell and the production status corresponding to each production process.

4. The method for storing shell production parameters based on the Internet of Things according to claim 1, characterized in that, The process of determining multiple abnormal storage parameters based on each sub-storage path, determining abnormal storage areas based on the multiple abnormal storage parameters and corresponding production processes, and determining storage anomaly handling items based on the abnormal storage areas and the storage progress of each shell's production data set, further includes: Collect the current production events of the shell, determine the current production content of the shell based on the detection of the current production events, and determine the storage anomaly handling items based on the current production content of the shell, the regional location of the abnormal storage area, and the storage progress of the production data set of each shell.

5. The method for storing shell production parameters based on the Internet of Things according to claim 1, characterized in that, The process of identifying multiple sub-processing events based on the identification of storage anomaly handling items, and determining the production parameter storage system based on these multiple sub-processing events, each sub-storage path, and the production status of each casing, includes: Collect storage exception handling items, determine multiple sub-storage exception handling lists based on the identification of storage exception handling items, and determine multiple sub-processing events based on the parsing of multiple sub-storage exception handling lists.

6. The method for storing shell production parameters based on the Internet of Things according to claim 5, characterized in that, The step of determining multiple sub-processing events based on the identification of storage anomaly handling items, and determining the production parameter storage system based on the multiple sub-processing events, each sub-storage path, and the production status of each casing, further includes: Collect data from each sub-storage path, determine the first sub-storage system based on multiple sub-processing events and each sub-storage path, and simultaneously determine the second sub-storage system based on multiple sub-processing events and the production status of each casing. The matching coefficient is determined based on the matching of the first sub-storage system and the second sub-storage system. If the matching coefficient is lower than the preset matching coefficient threshold, the first sub-storage system and the second sub-storage system are traced back in reverse to determine the matching coefficient impact event. The matching coefficient is adjusted according to the optimization of the matching coefficient impact event so that the adjusted matching coefficient reaches the preset matching coefficient threshold. The first sub-storage system or the second sub-storage system is optimized simultaneously. The production parameter storage system is determined based on the optimized first sub-storage system and the second sub-storage system.

Images