A control system for cylindrical lithium battery production and processing

Abstract

This invention discloses a control system for the production and processing of cylindrical lithium batteries, relating to the field of lithium battery production control technology. The system includes: a local acquisition module for acquiring multi-source production data of cylindrical lithium batteries; a cloud storage module for preprocessing and storing the multi-source production data; a graph database construction module for constructing a graph database based on the data relationships of the preprocessed data; and an execution module for anomaly tracing based on the graph database and real-time multi-source production data. This invention achieves centralized management of multi-source production data by constructing a graph database. By utilizing a quantitative correlation strength approach, combined with real-time acquisition of multi-source production data from different processes in cylindrical lithium battery production, it enables precise location of anomalies, providing a reliable and intuitive tracing path for anomalies.

Description

Technical Field

[0001] This invention relates to the field of lithium battery production control technology, specifically a control system for the production and processing of cylindrical lithium batteries. Background Technology

[0002] Cylindrical lithium batteries, with their advantages of high energy density, long cycle life, compact structure, and good safety, are widely used in consumer electronics, new energy vehicles, and energy storage devices. With increasing demands for performance, quality consistency, and production costs of cylindrical lithium batteries, refined and intelligent control of their production process has become a crucial optimization path. Currently, cylindrical lithium battery production involves multiple continuous processes, including electrode coating, rolling, slitting, winding, assembly, formation, capacity testing, and inspection, with close interrelationships between these processes. Current industry methods for controlling cylindrical lithium battery production mostly rely on single-process anomaly detection, which has the following drawbacks:

[0003] It is impossible to quickly and accurately locate anomalies across processes. Summary of the Invention

[0004] To address the shortcomings of existing technologies, the technical problem solved by this invention is: how to achieve traceability of anomalies in the production process of cylindrical lithium batteries.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a control system for the production and processing of cylindrical lithium batteries, the system comprising:

[0006] A local acquisition module is used to acquire multi-source production data of cylindrical lithium batteries.

[0007] A cloud storage module is used to preprocess multi-source production data and store the preprocessed data.

[0008] A graph database construction module is used to construct a graph database based on the data relationships of the preprocessed data.

[0009] An execution module is used to perform anomaly tracing based on a graph database and real-time multi-source production data.

[0010] The present invention is further configured such that: the local acquisition module includes several sets of data acquisition units, local storage units and off-site storage units, and the data acquisition units are connected to the local storage units;

[0011] The data acquisition unit is used to collect production data from different processes of cylindrical lithium batteries, serving as multi-source production data for cylindrical lithium batteries. This production data includes equipment operating parameters, process control parameters, product quality inspection data, and environmental condition data for the corresponding process.

[0012] The local storage unit is used to store the production data collected by the corresponding data acquisition unit and upload the stored production data to the cloud storage module.

[0013] The off-site storage unit is used to store production data stored in at least one set of non-local storage units.

[0014] The present invention is further configured such that: the cloud storage module includes a data processing unit and a remote allocation unit;

[0015] The data processing unit is used to perform outlier filtering, missing value completion, and standardization on multi-source production data, and to acquire and store preprocessed data.

[0016] The off-site allocation unit is used for the transmission and allocation of off-site storage.

[0017] The present invention is further configured such that the transmission allocation process includes:

[0018] Obtain the spatial coordinates of the local acquisition modules and determine the distance between any two local acquisition modules;

[0019] Choose any one local acquisition module as the information sending point, and select the two local acquisition modules that are furthest away from the information sending point as information backup points;

[0020] Determine whether the local acquisition module corresponding to the information backup point exceeds the preset threshold. If so, remove the local acquisition module corresponding to the current information backup point and re-select the remaining local acquisition module with the furthest distance as the latest information backup point; otherwise, establish communication between the information sending point and the information backup point.

[0021] The present invention is further configured such that: the graph database construction module includes a node construction unit, a data filling unit, and an edge construction unit;

[0022] The node construction unit is used to determine the node hierarchy, node unique identifier, and node attributes;

[0023] The data filling unit is used to extract node-related data based on the preprocessed data and fill the node-related data into the corresponding node as the node's basic data.

[0024] The edge construction unit is used to determine the node association type based on the production logic of cylindrical lithium batteries and the basic data of the nodes, and then use it as the edge type. After quantifying the association strength according to the node association type, it is used as the edge attribute. After completing the mapping between nodes and edges, the graph database is obtained.

[0025] The present invention is further configured such that the type of the edge includes being collected from, affecting, associated with, causing anomalies, and affected by the environment.

[0026] The present invention is further configured such that the calculation method for quantifying the association strength based on the node association type includes:

[0027] When data is collected at the edge, the association strength is quantized to 1;

[0028] When the influence is on an edge, the quantitative calculation formula for its association strength is:

[0029] ;

[0030] In the formula, To affect the strength of the association with the edge, and , When the value is less than 0.8, it is considered a weak association, and no edge of this type is constructed. For the sample size, For the first Data after preprocessing of each process parameter For the first The average value of the data after preprocessing each process parameter For the first The data after preprocessing corresponding quality inspection data. For the first The average value of the preprocessed data for each corresponding quality inspection data;

[0031] When associated with an edge, the quantitative calculation formula for its association strength is:

[0032] ;

[0033] In the formula, The synchronization runtime for two associated nodes. This is the total production time for this batch. Let be the association strength of the edges, and ;

[0034] When an abnormal edge is caused, the quantized association strength is 1;

[0035] When edges are affected by the environment, the quantitative calculation formula for their association strength is as follows:

[0036] ;

[0037] In the formula, The correlation coefficient between environmental data and target data. The correlation coefficient between environmental data and interference factor data. The correlation coefficient between the target data and the data of interfering factors. The correlation strength of edges affected by the environment, and .

[0038] The present invention is further configured such that: the execution module includes an anomaly diagnosis unit and a source tracing unit;

[0039] The anomaly diagnosis unit is used to determine whether real-time multi-source production data is abnormal based on the graph database.

[0040] The tracing unit is used to determine the corresponding node as an abnormal node based on the abnormal data, and to perform tracing based on the abnormal node in conjunction with the graph database.

[0041] The present invention is further configured such that the judgment process for determining whether real-time multi-source production data is abnormal includes:

[0042] The DBSCAN algorithm is used to calculate the deviation between real-time multi-source production data and the corresponding historical normal data clusters. If the deviation exceeds a preset threshold, the data is identified as abnormal.

[0043] The present invention is further configured such that the traceability process includes:

[0044] S1. Using the abnormal node as the center, determine all incoming edges pointing to the abnormal node according to the graph database, and filter out candidate nodes whose association strength exceeds the preset lower limit threshold.

[0045] S2. Determine if the candidate node is an abnormal node. If yes, go to S1; otherwise, go to S3.

[0046] S3. End the tracing process by taking the finally selected abnormal node as the root cause node.

[0047] This invention provides a control system for the production and processing of cylindrical lithium batteries. It has the following beneficial effects:

[0048] By constructing a topological association between nodes and edges with the process as the core, equipment operating parameters, process control parameters, product quality inspection data, and environmental condition data are associated with process nodes to build a graph database, enabling centralized management of multi-source production data. By using a quantitative association strength method, combined with the real-time collection of multi-source production data from different processes of cylindrical lithium batteries, the system can accurately locate abnormal data and provide a reliable and intuitive traceability path for abnormal data. Attached Figure Description

[0049] Figure 1 This is a system principle block diagram of the present invention;

[0050] Figure 2 This is a system principle block diagram of the local acquisition module of the present invention;

[0051] Figure 3 This is a system principle block diagram of the cloud storage module of the present invention;

[0052] Figure 4This is a system principle block diagram of the graph database construction module of the present invention;

[0053] Figure 5 This is a system principle block diagram of the execution module of the present invention. Detailed Implementation

[0054] 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.

[0055] This invention provides a control system for the production and processing of cylindrical lithium batteries, referring to... Figure 1 The system comprises several local acquisition modules, a cloud storage module, a graph database construction module, and an execution module. The local acquisition modules interface with the cloud storage module, the cloud storage module with the graph database construction module, and the graph database construction module with the execution module. The specific number of local acquisition modules corresponds to the cylindrical lithium battery production process. These local acquisition modules are used to collect multi-source production data for cylindrical lithium batteries (equipment operating parameters, process control parameters, product quality inspection data, and environmental condition data). Equipment operating parameters include coating speed, roller pressure, winding tension, formation current / voltage, and capacity-based charge / discharge rate; process control parameters include coating temperature, roller gap, winding speed, and formation temperature curve; product quality inspection data includes electrode thickness deviation, cell internal resistance, battery capacity, and cycle life; and environmental condition data includes workshop temperature and humidity, dust concentration, and equipment vibration values.

[0056] The cloud storage module is used to preprocess multi-source production data and store the preprocessed data; the graph database construction module is used to construct a graph database based on the data relationships of the preprocessed data; and the execution module is used to trace anomalies based on the graph database and real-time multi-source production data.

[0057] In one embodiment, such as Figure 2 As shown, the aforementioned local acquisition module includes several sets of data acquisition units, local storage units, and off-site storage units, with the data acquisition units interfacing with the local storage units. The sets of data acquisition units are used to collect production data (equipment operating parameters, process control parameters, product quality inspection data, and environmental condition data) from different processes of cylindrical lithium batteries (electrode coating, rolling, slitting, winding, assembly, formation, capacity testing, and inspection processes) to obtain multi-source production data for cylindrical lithium batteries.

[0058] The local storage unit is used to store the production data collected by the corresponding data acquisition unit and upload the stored production data to the cloud storage module. The storage format of the production data is as follows: the process is the parent directory, and the equipment operating parameters, process control parameters, product quality inspection data, and environmental condition data are the subdirectories.

[0059] Off-site storage units are used to store production data stored in at least one set of non-local storage units as backup data.

[0060] In this way, the local acquisition module only considers data acquisition and storage, ensuring data acquisition efficiency and long-term stable service life.

[0061] In one embodiment, such as Figure 3 As shown, the aforementioned cloud storage module includes a data processing unit and a remote distribution unit. The data processing unit is used to perform outlier removal, missing value completion, and standardization on multi-source production data, obtaining and storing preprocessed data.

[0062] The above outlier identification uses a time-series outlier detection algorithm to filter out single-point outliers, drift outliers, and mutation outliers; missing value completion is based on the correlation between adjacent process data to complete missing data; standardization processing is used to unify the format, units, and encoding rules of multi-source data.

[0063] The off-site allocation unit is used for the transmission allocation of off-site storage; the transmission allocation process includes:

[0064] Obtain the spatial coordinates of the local acquisition modules and determine the distance between any two local acquisition modules;

[0065] Choose one local acquisition module as the information sending point, and select the two local acquisition modules that are farthest from the information sending point as information backup points (i.e., for off-site storage).

[0066] Determine whether the number of local acquisition modules corresponding to the information backup point exceeds a preset threshold (e.g., 2). If so, remove the local acquisition module corresponding to the current information backup point and re-select the remaining local acquisition module with the furthest distance as the latest information backup point; otherwise, establish communication between the information sending point and the information backup point.

[0067] This avoids data loss caused by the failure of a local acquisition module, and improves the resilience and anti-interference capability of data storage.

[0068] In one embodiment, such as Figure 4 As shown, the graph database construction module includes a node construction unit, a data filling unit, and an edge construction unit.

[0069] The node construction unit is used to determine the node hierarchy (such as the main node of the process node, and the subordinate branch nodes of equipment nodes, process control parameter nodes, product quality inspection data nodes, and environmental condition nodes), the unique identifier of the node, and the node attributes; among which the nodes include process nodes, equipment nodes, process control parameter nodes, product quality inspection data nodes, and environmental condition nodes.

[0070] Node attributes include:

[0071] Process nodes: process number, name, production line, sequence;

[0072] Equipment Nodes: Equipment Number, Name, Model, and Process;

[0073] Process control parameter nodes: parameter number, name, unit, default threshold;

[0074] Product quality inspection data nodes: inspection number, item, value, time;

[0075] Environmental condition node: Environment number, type, value, time.

[0076] The data filling unit is used to extract node-related data from the preprocessed data and fill the node-related data into the corresponding node as the node's basic data.

[0077] The edge construction unit is used to determine the node association relationship (such as the association between equipment under the process and the collection of production data, the association between process control parameters under the process and the influence of product quality inspection data, and the process association between each process) based on the production logic of cylindrical lithium batteries and the basic data of the nodes. This relationship is used as the edge type (including data collected from, data affected by, data associated with, data causing anomalies, and data affected by the environment). After quantifying the association strength according to the node association relationship type, this relationship is used as the edge attribute (such as association strength, timestamp, association basis). After completing the mapping between nodes and edges, the graph database is obtained.

[0078] The above-mentioned calculation methods for quantifying association strength based on node association type include:

[0079] When data is collected at the edge, the node association relationship is deterministic (e.g., data must be collected by the corresponding device), and its association strength is directly quantified as 1;

[0080] When the influence is on an edge, the quantitative calculation formula for its association strength is:

[0081] ;

[0082] In the formula, To affect the strength of the association with the edge, and , When the value is less than 0.8, it is considered a weak association, and no edge of this type is constructed. This refers to the sample size (e.g., the number of samples collected from the corresponding process parameters and product quality inspection data for this process over the past 30 days). For the first Data after preprocessing of process parameters (such as coating thickness and rolling pressure). For the first The average value of the data after preprocessing each process parameter For the first The corresponding preprocessed data of quality inspection (such as cell internal resistance, electrode thickness deviation). For the first The average value of the preprocessed data for each corresponding quality inspection data;

[0083] When associated with an edge, the quantitative calculation formula for its association strength is:

[0084] ;

[0085] In the formula, The synchronous running time of two related nodes (such as coating process and rolling process, coating machine and rolling press). This is the total production time for this batch. Let be the association strength of the edges, and When the synchronous runtime accounts for ≥62.5%, The value is 0.8; when the synchronous runtime accounts for ≤62.5%, Calculated linearly, with a minimum value of 0.5;

[0086] When an abnormal edge is caused, the node association relationship is a deterministic relationship (such as the root cause association confirmed by the anomaly tracing algorithm), and the association strength is directly quantified as 1;

[0087] When edges are affected by the environment, the quantitative calculation formula for their association strength is as follows:

[0088] ;

[0089] In the formula, The correlation coefficient between environmental data (such as workshop temperature and humidity) and target data (such as capacity yield and coating thickness). The correlation coefficient between environmental data and interference factor data (such as equipment operating time). The correlation coefficient between the target data and the data of interfering factors. The correlation strength of edges affected by the environment, and When the above formula is calculated If the result is ≥0.9, take 0.9; if the result is ≤0.6, it is determined to have a weak impact and no edge of this type is constructed.

[0090] In one embodiment, such as Figure 5 As shown, the above execution module includes an anomaly diagnosis unit and a source tracing unit.

[0091] The anomaly diagnosis unit is used to determine whether real-time multi-source production data is abnormal based on the graph database; the anomaly judgment process includes:

[0092] The DBSCAN algorithm is used to calculate the deviation between real-time multi-source production data and the corresponding historical normal data clusters (this deviation is the Euclidean distance between the real-time production data and the center of the corresponding cluster). If the deviation exceeds a preset threshold (this preset threshold is set based on production experience; for example, taking the coating thickness parameter of the coating process as an example, the maximum deviation of the standardized historical data over the past 30 days is 0.15. Based on production experience, it has been found that excessive fluctuations in coating thickness can lead to abnormal internal resistance of the battery cell, affecting the product qualification rate. The preset threshold is set to 1.5 times the historical maximum deviation, i.e., 0.225. If the calculated deviation of a certain real-time coating thickness data from the center of the corresponding cluster is 0.25, and this exceeds 0.225, then the coating thickness data is determined to be abnormal data; when no specific value is specified, the threshold is set to exceed the distribution range of historical normal data), then it is determined to be abnormal data.

[0093] The tracing unit is used to identify abnormal nodes based on abnormal data, and to trace the source of the abnormal nodes using the graph database. This tracing process includes:

[0094] S1. Using the abnormal node as the center, determine all incoming edges pointing to the abnormal node according to the graph database, and filter out candidate nodes whose association strength exceeds the preset lower limit threshold.

[0095] S2. Determine if the candidate node is an abnormal node. If yes, go to S1; otherwise, go to S3.

[0096] S3. End the tracing process by taking the finally selected abnormal node as the root cause node.

[0097] This allows us to obtain a complete chain of abnormal nodes, providing clear guidance for fault diagnosis and subsequent abnormal repair.

[0098] The following is an explanation of the traceability process using an abnormal internal resistance of a battery cell as an example. The process is as follows:

[0099] When a cell internal resistance node is determined to be an abnormal node, it is recorded as Q001.

[0100] Centered on Q001, query all incoming edges in the graph database pointing to that node (i.e., who influences it and who is related to it), and filter out upstream candidate nodes with a correlation strength greater than or equal to a preset lower limit:

[0101] Checking the "Affected by" edge: the correlation strength between Q001 and the coating thickness parameter node (i.e., P001) is 0.92 (≥0.8, which meets the requirements), so P001 is retained as the core candidate node;

[0102] Check the "Affected by the environment" edge: the correlation strength between Q001 and the workshop humidity node (i.e., E001) is 0.55 (<0.6, which does not meet the requirements), so remove E001;

[0103] Check the "associated with" edges: the association strength between Q001 and the capacity-sharing process node (i.e., Pro002) is 0.45 (<0.5, which does not meet the requirements), so remove Pro002;

[0104] This ultimately led to the selection of a single core upstream candidate node: P001;

[0105] When P001 is determined to be an abnormal node, using P001 as the center, query the upstream incoming edges it points to again, filter for nodes with association strength ≥ the lower limit, and determine whether they are abnormal:

[0106] Check the "Collected from" edge: the association strength between P001 and the coating machine equipment node (i.e., D001) is 1 (fixed), so D001 is retained as a candidate node.

[0107] Check the "associated with" edges: the association strength between P001 and the coating process node (i.e., Pro001) is 0.85 (≥0.5, which meets the requirements), so Pro001 is retained as a candidate node;

[0108] If D001 is determined not to be an abnormal node, it is removed directly. If Pro001 is determined not to be an abnormal node, it is removed directly. At this point, there are no abnormal upstream nodes, and the current tracing path is determined to be Q001→P001.

[0109] It should be noted that the sequence numbers of the embodiments in this application are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0110] The terms "comprising" and "having," and any variations thereof, in the specification, claims, and accompanying drawings of this application are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to such process, method, product, or apparatus. The terms "first," "second," and "third," etc., are used to distinguish different objects, etc., and do not indicate a sequence, nor do they limit "first," "second," and "third" to different types.

[0111] In the description of the embodiments of this application, terms such as "exemplary," "for example," or "for instance" are used to indicate examples, illustrations, or explanations. Any embodiment or design described as "exemplary," "for example," or "for instance" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of terms such as "exemplary," "for example," or "for instance" is intended to present the relevant concepts in a concrete manner.

[0112] In the description of the embodiments of this application, unless otherwise stated, " / " means "or". For example, A / B can mean A or B. The "and / or" in the text is merely a description of the relationship between related objects, indicating that there can be three relationships. For example, A and / or B can mean: A exists alone, A and B exist simultaneously, and B exists alone. In addition, in the description of the embodiments of this application, "multiple" means two or more.

[0113] In some processes described in the embodiments of this application, multiple operations or steps are included in a specific order. However, it should be understood that these operations or steps may not be executed in the order they appear in the embodiments of this application, or they may be executed in parallel. The sequence number of the operation is only used to distinguish different operations, and the sequence number itself does not represent any execution order. In addition, these processes may include more or fewer operations, and these operations or steps may be executed sequentially or in parallel, and these operations or steps may be combined.

[0114] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) as described above, and includes several instructions to cause a terminal device to execute the methods described in the various embodiments of this application.

[0115] The above are merely specific embodiments of this application, but the protection scope of this application is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in this application, and these modifications or substitutions should all be covered within the protection scope of this application. Therefore, the protection scope of this application should be determined by the scope of the claims.

Claims

1. A control system for the production and processing of cylindrical lithium batteries, characterized in that, The system includes: A local acquisition module is used to acquire multi-source production data of cylindrical lithium batteries. A cloud storage module is used to preprocess multi-source production data and store the preprocessed data. A graph database construction module is used to construct a graph database based on the data relationships of the preprocessed data. An execution module is used to perform anomaly tracing based on a graph database and real-time multi-source production data.

2. The control system for cylindrical lithium battery production and processing according to claim 1, characterized in that, The local acquisition module includes several sets of data acquisition units, local storage units, and off-site storage units, and the data acquisition units are connected to the local storage units. The data acquisition unit is used to collect production data from different processes of cylindrical lithium batteries, serving as multi-source production data for cylindrical lithium batteries. This production data includes equipment operating parameters, process control parameters, product quality inspection data, and environmental condition data for the corresponding process. The local storage unit is used to store the production data collected by the corresponding data acquisition unit and upload the stored production data to the cloud storage module. The off-site storage unit is used to store production data stored in at least one set of non-local storage units.

3. The control system for cylindrical lithium battery production and processing according to claim 2, characterized in that, The cloud storage module includes a data processing unit and a remote allocation unit; The data processing unit is used to perform outlier filtering, missing value completion, and standardization on multi-source production data, and to acquire and store preprocessed data. The off-site allocation unit is used for the transmission and allocation of off-site storage.

4. The control system for cylindrical lithium battery production and processing according to claim 3, characterized in that, The transmission allocation process includes: Obtain the spatial coordinates of the local acquisition modules and determine the distance between any two local acquisition modules; Choose any one local acquisition module as the information sending point, and select the two local acquisition modules that are furthest away from the information sending point as information backup points; Determine whether the local acquisition module corresponding to the information backup point exceeds the preset threshold. If so, remove the local acquisition module corresponding to the current information backup point and re-select the remaining local acquisition module with the furthest distance as the latest information backup point; otherwise, establish communication between the information sending point and the information backup point.

5. A control system for the production and processing of cylindrical lithium batteries according to claim 4, characterized in that, The graph database construction module includes a node construction unit, a data filling unit, and an edge construction unit; The node construction unit is used to determine the node hierarchy, node unique identifier, and node attributes; The data filling unit is used to extract node-related data based on the preprocessed data and fill the node-related data into the corresponding node as the node's basic data. The edge construction unit is used to determine the node association type based on the production logic of cylindrical lithium batteries and the basic data of the nodes, and then use it as the edge type. After quantifying the association strength according to the node association type, it is used as the edge attribute. After completing the mapping between nodes and edges, the graph database is obtained.

6. A control system for the production and processing of cylindrical lithium batteries according to claim 5, characterized in that, The types of edges include those collected from, those that affect, those that are associated with, those that cause anomalies, and those that are affected by the environment.

7. A control system for the production and processing of cylindrical lithium batteries according to claim 6, characterized in that, The calculation method for quantifying association strength based on node association type includes: When data is collected at the edge, the association strength is quantized to 1; When the influence is on an edge, the quantitative calculation formula for its association strength is: ； In the formula, To affect the strength of the association with the edge, and , When the value is less than 0.8, it is considered a weak association, and no edge of this type is constructed. For the sample size, For the first Data after preprocessing of each process parameter For the first The average value of the data after preprocessing each process parameter For the first The data after preprocessing corresponding quality inspection data. For the first The average value of the preprocessed data for each corresponding quality inspection data; When associated with an edge, the quantitative calculation formula for its association strength is: ； In the formula, The synchronization runtime for two associated nodes. This is the total production time for this batch. Let be the association strength of the edges, and ; When an abnormal edge is caused, the quantized association strength is 1; When edges are affected by the environment, the quantitative calculation formula for their association strength is as follows: ； In the formula, The correlation coefficient between environmental data and target data. The correlation coefficient between environmental data and interference factor data. The correlation coefficient between the target data and the data of interfering factors. The correlation strength of edges affected by the environment, and .

8. A control system for the production and processing of cylindrical lithium batteries according to claim 1, characterized in that, The execution module includes an anomaly diagnosis unit and a source tracing unit; The anomaly diagnosis unit is used to determine whether real-time multi-source production data is abnormal based on the graph database. The tracing unit is used to determine the corresponding node as an abnormal node based on the abnormal data, and to perform tracing based on the abnormal node in conjunction with the graph database.

9. A control system for the production and processing of cylindrical lithium batteries according to claim 8, characterized in that, The process for determining whether real-time multi-source production data is abnormal includes: The DBSCAN algorithm is used to calculate the deviation between real-time multi-source production data and the corresponding historical normal data clusters. If the deviation exceeds a preset threshold, the data is identified as abnormal.

10. A control system for the production and processing of cylindrical lithium batteries according to claim 8, characterized in that, The source tracing process includes: S1. Using the abnormal node as the center, determine all incoming edges pointing to the abnormal node according to the graph database, and filter out candidate nodes whose association strength exceeds the preset lower limit threshold. S2. Determine if the candidate node is an abnormal node. If yes, go to S1; otherwise, go to S3. S3. End the tracing process by taking the finally selected abnormal node as the root cause node.

Images