Movatterモバイル変換


[0]ホーム

URL:


CN118014165B - Traceability management method and traceability management system for lithium ion battery production - Google Patents

Traceability management method and traceability management system for lithium ion battery production
Download PDF

Info

Publication number
CN118014165B
CN118014165BCN202410414673.5ACN202410414673ACN118014165BCN 118014165 BCN118014165 BCN 118014165BCN 202410414673 ACN202410414673 ACN 202410414673ACN 118014165 BCN118014165 BCN 118014165B
Authority
CN
China
Prior art keywords
production
flow
initial
data
optimization
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN202410414673.5A
Other languages
Chinese (zh)
Other versions
CN118014165A (en
Inventor
李彬
李永洪
常伟
曾锦辉
廖辉
张静
戴天童
秦鸿
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Jiade Energy Technology Zhuhai Co ltd
Original Assignee
Jiade Energy Technology Zhuhai Co ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Jiade Energy Technology Zhuhai Co ltdfiledCriticalJiade Energy Technology Zhuhai Co ltd
Priority to CN202410414673.5ApriorityCriticalpatent/CN118014165B/en
Publication of CN118014165ApublicationCriticalpatent/CN118014165A/en
Application grantedgrantedCritical
Publication of CN118014165BpublicationCriticalpatent/CN118014165B/en
Activelegal-statusCriticalCurrent
Anticipated expirationlegal-statusCritical

Links

Classifications

Landscapes

Abstract

The application relates to the technical field of production management, and discloses a traceability management method and system for lithium ion battery production. The method comprises the following steps: constructing a cloud edge cooperative framework, and collecting a production association data set of an initial production flow in a lithium ion battery production line; carrying out risk factor identification and comprehensive weight calculation on the production associated data set to obtain a plurality of production risk factors and comprehensive weight values of each production risk factor; creating a causal traceability map and a stock flow map of the initial production flow according to the plurality of production risk factors and the comprehensive weight value; and carrying out production flow optimization point analysis on the causal tracing graph and the stock flow graph to obtain production flow optimization points and carrying out flow optimization on the initial production flow to obtain a target production flow.

Description

Traceability management method and traceability management system for lithium ion battery production
Technical Field
The application relates to the technical field of production management, in particular to a traceability management method and system for lithium ion battery production.
Background
In the lithium ion battery production industry, along with the continuous growth of market demands and the rapid development of production technologies, the production process becomes more and more complex, and the related risk factors are also increased. These risk factors include fluctuations in raw material quality, stability of production equipment, changes in production environment, consistency in product quality, and the like. The uncertainty of these factors increases the complexity of the production process, posing a potential threat to production efficiency and product quality. Therefore, how to effectively identify and manage these risk factors becomes a key to improving production efficiency and ensuring product quality.
The traditional production management method often depends on experience judgment and linear analysis, and nonlinear relation and dynamic change in the production process are difficult to accurately capture. This results in production risk management that is often reactive rather than preventive, and once a problem occurs, it begins to find reasons and solutions, which are not only inefficient, but also often difficult to fundamentally solve. In addition, the lack of systematic risk assessment and management tools makes the identification of risk factors and weight calculation lack of scientific basis, and thus, the comprehensive and dynamic management of production risks is difficult to realize.
Disclosure of Invention
The application provides a traceability management method and a traceability management system for lithium ion battery production, which are used for improving the production flow optimization accuracy of lithium ion battery production and realizing intelligent traceability management of lithium ion battery production.
In a first aspect, the present application provides a method for tracing management of lithium ion battery production, where the method for tracing management of lithium ion battery production includes:
constructing a cloud edge cooperative framework, and collecting a production association data set of an initial production flow in a lithium ion battery production line;
Carrying out risk factor identification and comprehensive weight calculation on the production associated data set to obtain a plurality of production risk factors and comprehensive weight values of each production risk factor;
Creating a causal traceability map and a stock flow map of the initial production flow according to the plurality of production risk factors and the comprehensive weight value;
and carrying out production flow optimization point analysis on the causal tracing graph and the stock flow graph to obtain production flow optimization points, and carrying out flow optimization on the initial production flow to obtain a target production flow.
In a second aspect, the present application provides a traceability management system for lithium ion battery production, where the traceability management system for lithium ion battery production includes:
The acquisition module is used for constructing a cloud edge cooperative framework and acquiring a production association data set of an initial production flow in a lithium ion battery production line;
The calculation module is used for carrying out risk factor identification and comprehensive weight calculation on the production association data set to obtain a plurality of production risk factors and comprehensive weight values of each production risk factor;
The creation module is used for creating a causal traceability map and a stock flow map of the initial production flow according to the plurality of production risk factors and the comprehensive weight value;
and the optimizing module is used for carrying out production flow optimizing point analysis on the causal tracing graph and the stock flow graph to obtain a production flow optimizing point and carrying out flow optimization on the initial production flow to obtain a target production flow.
According to the technical scheme provided by the application, the cloud edge collaborative architecture is constructed, and the production associated data set of each flow area is collected, so that comprehensive data support is provided for the whole lithium ion battery production flow. The method not only increases the transparency of the production flow, but also greatly improves the traceability of each production stage, and provides a data base for problem diagnosis and quality control in the production process. The production related data set is subjected to deep analysis through a cyclic variation self-encoder, a network analytic hierarchy process and a fuzzy comprehensive evaluation method, so that key risk factors in the production process can be accurately identified, comprehensive weight values of the key risk factors are calculated, and the optimization points of the production process can be subjected to deep analysis through creating a causal tracing graph and a stock flow graph of the initial production process, so that efficiency bottlenecks and quality risk points in the production process can be identified. By formulating and implementing specific flow optimization strategies aiming at the optimization points, the production efficiency can be effectively improved, the product quality is ensured, and the production cost is reduced. Based on the comprehensive weight calculation and the process optimization analysis result, the continuous dynamic adjustment and optimization of the production process can be performed. The causal tracing graph and the stock flow graph are updated regularly, so that the change and the newly-appearing risk in the production process are reflected in time, the continuous monitoring and the self-adaptive optimization of the production process are realized, the production flow optimization accuracy of lithium ion battery production is further improved, and the intelligent traceability management of lithium ion battery production is realized.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for the description of the embodiments will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and other drawings may be obtained based on these drawings without inventive effort for a person skilled in the art.
Fig. 1 is a schematic diagram of an embodiment of a traceability management method for lithium ion battery production according to an embodiment of the present application;
fig. 2 is a schematic diagram of an embodiment of a traceability management system for lithium ion battery production according to an embodiment of the present application.
Detailed Description
The embodiment of the application provides a traceability management method and system for lithium ion battery production. The terms "first," "second," "third," "fourth" and the like in the description and in the claims and in the above drawings, if any, are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used may be interchanged where appropriate such that the embodiments described herein may be implemented in other sequences than those illustrated or otherwise described herein. Furthermore, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed or inherent to such process, method, article, or apparatus.
For easy understanding, the following describes a specific flow of an embodiment of the present application, referring to fig. 1, and an embodiment of a tracing management method for lithium ion battery production in the embodiment of the present application includes:
Step S101, constructing a cloud edge cooperative framework, and collecting a production association data set of an initial production flow in a lithium ion battery production line;
It can be understood that the execution subject of the present application may be a traceability management system for lithium ion battery production, and may also be a terminal or a server, which is not limited herein. The embodiment of the application is described by taking a server as an execution main body as an example.
Specifically, a cloud edge cooperative architecture is constructed, and the data processing capacity of the cloud computing platform and the real-time data processing advantage of the edge computing nodes are fused. By deploying the data analysis and machine learning model on the cloud computing platform and deploying the edge computing nodes at the near end of the production line, the distributed coordination of data processing is realized, the data processing efficiency is ensured, and the real-time requirement is met. And the rough set theory is adopted to divide the production area of the initial production flow of the lithium ion battery production line, the uncertainty and the ambiguity information are effectively processed, the flow areas with similar production characteristics are identified through the analysis of the production flow, and the production line is divided into a plurality of flow areas. After the edge computing nodes are deployed in each process area, the nodes can collect initial raw material data, initial equipment state data, initial production environment data and initial product quality data in the respective areas in real time based on the advantages of the physical positions of the nodes. By implementing data cleaning operation on the edge computing nodes, noise and abnormal values are removed, accuracy and consistency of data are ensured, and a standardized data set is obtained. And carrying out data standardization and integrated conversion on the standard raw material data, the standard equipment state data, the standard production environment data and the standard product quality data of each flow area to form a production association data set with a unified format and standard. The normalized production association data set is transmitted to the cloud computing platform by the edge computing node. On the cloud platform, potential risk factors in the production process are analyzed and evaluated through a data analysis technology and a machine learning algorithm, so that the comprehensive monitoring and optimization of the production process are realized, and the high efficiency, the stability and the traceability of the lithium ion battery production process are ensured.
Step S102, carrying out risk factor identification and comprehensive weight calculation on a production associated data set to obtain a plurality of production risk factors and comprehensive weight values of each production risk factor;
Specifically, the production-related data set is subjected to abnormal feature identification through a cyclic variation self-encoder. The cyclic variation self-encoder effectively identifies abnormal data characteristics which are inconsistent with a normal production mode through learning the distribution characteristics of the data, and a plurality of first abnormal data characteristics are obtained. And screening the first abnormal data features by setting a predefined target value, eliminating anomalies possibly caused by accidental factors, and retaining second abnormal data features directly related to production risk. And (3) carrying out risk factor identification on the screened second abnormal data characteristics by adopting a network analytic hierarchy process, combining quantitative analysis and qualitative analysis, decomposing the complex risk factor identification problem into a plurality of sub-problems by constructing a hierarchical structure model, and then analyzing the plurality of sub-problems one by one. In the process, each abnormal characteristic is regarded as a potential production risk factor, and the core risk factors which truly affect the production safety and quality are identified through comparison and analysis, so that a plurality of production risk factors related to the production flow are obtained. And (3) carrying out comprehensive weight calculation on the production risk factors by a fuzzy comprehensive evaluation method, and solving the evaluation problems of mutual influence and strong uncertainty among the factors. By constructing an evaluation factor set and a weight set and combining the fuzzy relation matrix, the comprehensive weight value of each production risk factor is calculated, and the weight values reflect the relative magnitude of the influence of each risk factor on the production flow.
Further, a network hierarchical analysis method is adopted to construct a relation network structure between the risk criteria and the risk factors. In the relational network structure, risk criteria and risk factors are well defined and interconnected to form an ordered network reflecting the interdependence between the risk factors. And (3) comparing each layer in the relational network structure with the target layer in pairs, and determining the relative importance of different risk factors on the final evaluation target. Subjective assessment is quantified as objective data by constructing a pair-wise comparison matrix. Each matrix element represents the relative importance of one risk factor in achieving the final assessment objective as compared to another. And carrying out feature vector solving on the second abnormal data features in the paired comparison matrixes. The feature vector reflects the relative weight of each risk factor throughout the risk assessment system. A consistency check is performed on each feature vector to verify consistency and reliability of the pair-wise comparison matrix. The consistency check can detect and correct logical contradictions or deviations that may occur during the pair-wise comparison, and only feature vectors that pass the consistency check can be used to generate the final production risk factor. Based on the consistency check result of each feature vector, a plurality of production risk factors associated with the second anomalous data feature are generated.
Further, a fuzzy comprehensive evaluation method is used for determining an evaluation set and a factor set of a plurality of production risk factors, wherein the evaluation set is used for defining risk levels, specifically comprises three levels of low risk, medium risk and high risk, and the factor set comprises all the identified production risk factors. An initial weight value is generated for each production risk factor based on the set of evaluations and the set of factors. The initial weight value reflects a preliminary estimate of the relative importance of each risk factor throughout the production process. And constructing an initial fuzzy relation matrix of each risk factor according to the initial weight value and the evaluation set of each risk factor. And obtaining a comprehensive fuzzy relation matrix by calculating an initial fuzzy relation matrix of each risk factor. The comprehensive fuzzy relation matrix comprehensively integrates the relation between each risk factor and the risk level in the evaluation set, and provides a global view for comprehensive evaluation. And obtaining the comprehensive evaluation result vector of each production risk factor by performing fuzzy synthesis operation on the comprehensive fuzzy relation matrix. The resulting vector quantifies the contribution of each risk factor to overall production risk and reveals its distribution at different risk levels. And (3) recalculating the initial weight value of each production risk factor according to the comprehensive evaluation result vector to obtain the comprehensive weight value of each production risk factor. The comprehensive weight value is processed by the fuzzy logic, so that the real influence of each risk factor in the production process is reflected more accurately.
Step S103, creating a causal traceability map and a stock flow map of the initial production flow according to a plurality of production risk factors and comprehensive weight values;
Specifically, a plurality of relationship nodes of the initial production flow are defined according to a plurality of production risk factors. The relationship nodes represent various production risk factors that may occur during the production process, such as raw material quality, production equipment status, operator skill, and the like. And initializing the tracing graphs of the plurality of relation nodes based on the application of graph theory algorithm, such as minimum spanning tree or network flow algorithm, and constructing an initial tracing graph. And analyzing and weighting the initial traceability map according to the comprehensive weight value of each production risk factor. And distributing a weight matched with the comprehensive weight value of the corresponding risk factor to each node in the initial traceability graph through graph node weight analysis to obtain the causal traceability graph of the initial production flow. And labeling the stock and the flow of the initial production flow based on the causal tracing graph to obtain stock flow graphs of a plurality of production risk factors. The manifestations of each production risk factor in the production process are classified. For example, as a resource inventory (e.g., raw material storage, equipment availability, etc.), or as a flow factor in a production process (e.g., material consumption rate, product production rate, etc.) affects production. The stock flow map can reveal how production risk factors are transferred and transformed between different production nodes.
Step S104, carrying out production flow optimization point analysis on the causal tracing graph and the stock flow graph to obtain production flow optimization points, and carrying out flow optimization on the initial production flow to obtain a target production flow.
Specifically, the causal tracing graph is subjected to production flow optimization point analysis to obtain interaction and influence among production risk factors, a first flow optimization point is determined, and the optimization point mainly focuses on the source of the risk factors and the transmission path of the risk factors in the production process. Meanwhile, the stock flow graph is subjected to production flow optimization point analysis, the configuration and the utilization efficiency of production resources are focused, and a second batch of flow optimization points are obtained, wherein the points are focused on aspects of material consumption, storage management, production efficiency and the like. And comprehensively analyzing and screening the optimizing points of the first flow optimizing point and the second flow optimizing point, and identifying the production flow optimizing point which is most critical to improving the production efficiency and the product quality. The potential influence of each optimizing point on the aspects of reducing production cost, improving production speed, guaranteeing product quality and the like is comprehensively considered, and the screened optimizing points are guaranteed to be capable of comprehensively improving the performance of the production flow. Initializing the flow optimization strategy of the screened production flow optimization points through a genetic algorithm, and generating a plurality of possible flow optimization strategies. By simulating natural selection and genetic mechanisms, genetic algorithms can iteratively search for optimal solutions in multiple flow optimization strategies. To evaluate the effectiveness of the first process optimization strategy, fitness calculations are performed to determine the potential of each strategy in improving the production process. The comprehensive effects of strategies in the aspects of reducing production delay, reducing cost, improving output quality and the like are considered in the fitness calculation, and the screened strategies are ensured to be capable of effectively aiming at key problems in the production process. Based on the fitness value, a second set of process optimization strategies is screened out, which represent solutions that under the current conditions are likely to achieve the best production process improvement. And determining the target flow optimization strategy by carrying out optimization solution on the strategy. And systematically optimizing the initial production flow according to the target flow optimization strategy to obtain the target production flow. The optimization process includes adjustments to various aspects of the line layout, operating procedures, resource allocation, etc., in order to achieve higher production efficiency, lower cost, and better product quality.
In the embodiment of the application, the cloud edge collaborative architecture is constructed and the production associated data set of each flow area is collected, so that comprehensive data support is provided for the whole lithium ion battery production flow. The method not only increases the transparency of the production flow, but also greatly improves the traceability of each production stage, and provides a data base for problem diagnosis and quality control in the production process. The production related data set is subjected to deep analysis through a cyclic variation self-encoder, a network analytic hierarchy process and a fuzzy comprehensive evaluation method, so that key risk factors in the production process can be accurately identified, comprehensive weight values of the key risk factors are calculated, and the optimization points of the production process can be subjected to deep analysis through creating a causal tracing graph and a stock flow graph of the initial production process, so that efficiency bottlenecks and quality risk points in the production process can be identified. By formulating and implementing specific flow optimization strategies aiming at the optimization points, the production efficiency can be effectively improved, the product quality is ensured, and the production cost is reduced. Based on the comprehensive weight calculation and the process optimization analysis result, the continuous dynamic adjustment and optimization of the production process can be performed. The causal tracing graph and the stock flow graph are updated regularly, so that the change and the newly-appearing risk in the production process are reflected in time, the continuous monitoring and the self-adaptive optimization of the production process are realized, the production flow optimization accuracy of lithium ion battery production is further improved, and the intelligent traceability management of lithium ion battery production is realized.
In a specific embodiment, the process of executing step S101 may specifically include the following steps:
(1) Constructing a cloud edge cooperative framework, wherein the cloud edge cooperative framework comprises a cloud computing platform and edge computing nodes;
(2) Acquiring an initial production flow of a lithium ion battery production line, and dividing production areas of the initial production flow by adopting a rough set theory to obtain a plurality of flow areas;
(3) Disposing edge computing nodes in a plurality of process areas of the lithium ion battery production line respectively, and collecting initial raw material data, initial equipment state data, initial production environment data and initial product quality data of each process area in the lithium ion battery production line based on the edge computing nodes;
(4) Respectively carrying out data cleaning on the initial raw material data, the initial equipment state data, the initial production environment data and the initial product quality data of each flow area to obtain standard raw material data, standard equipment state data, standard production environment data and standard product quality data of each flow area;
(5) And carrying out data standardization and aggregation conversion on the standard raw material data, the standard equipment state data, the standard production environment data and the standard product quality data of each process area to obtain a production association data set of an initial production process, and transmitting the production association data set to a cloud computing platform through an edge computing node.
Specifically, a cloud edge cooperative architecture is constructed, and the data processing capability of the cloud computing platform is combined with the real-time data processing advantage of the edge computing nodes. And carrying out large-scale data analysis, storage and long-term trend analysis through a cloud computing platform, and collecting real-time data by the edge computing nodes and carrying out preliminary processing. And acquiring an initial production flow of the lithium ion battery production line, and dividing the production flow into areas through a rough set theory. Rough set theory is an effective tool to deal with uncertainty and ambiguity problems, and the management of the whole production flow is optimized by identifying the internal mode of data to divide the production area. For example, in the battery assembling process, the production line is divided into a plurality of critical process areas, such as a cell preparation area, an assembly area, a test area and the like, according to the types of production equipment, the differences of production environments (such as temperature and humidity conditions), raw material supply chains and the like. And respectively deploying edge computing nodes for the divided flow areas. The nodes are proximate to the data source and are capable of collecting initial raw material data, equipment status data, production environment data, and product quality data for each region in real time. For example, in a cell preparation area, edge computing nodes can collect batch information of raw materials, chemical component analysis results, and the like; collecting equipment states, operating parameters and the like of an automatic assembly line in an assembly area; and in the test area, various index data of the battery performance test are collected. And cleaning the collected data to remove invalid, wrong or incomplete data records so as to ensure the quality of the data. For example, for raw material data, it may be necessary to reject records that are obscured by labels or that do not meet production standards in chemical composition; for device state data, it is necessary to filter out abnormal device operation records, such as a sudden shutdown or abnormal parameters. And after cleaning, obtaining standard data sets with higher quality in each process area, wherein the data sets more accurately reflect the actual conditions of the production process. And carrying out data standardization and integrated conversion on the standard raw material data, the standard equipment state data, the standard production environment data and the standard product quality data of each flow area, unifying data formats and improving the usability of the data. The production association data set is transmitted to the cloud computing platform by the edge computing node. And on the cloud platform, carrying out deep mining and trend prediction on the data by utilizing the data processing and analyzing capability. For example, by analyzing the relationship between raw material data and product quality data, key raw material indicators affecting battery performance are identified; by analyzing the equipment state data and the production environment data, equipment faults and maintenance requirements are predicted, so that adjustment or maintenance is performed in advance, and the risk of production interruption is reduced.
In a specific embodiment, the process of executing step S102 may specifically include the following steps:
(1) Carrying out abnormal characteristic identification on the production related data set through a cyclic variation self-encoder to obtain a plurality of first abnormal data characteristics;
(2) Performing feature screening on the first abnormal data features through a preset target value to obtain second abnormal data features;
(3) Carrying out risk factor identification on the plurality of second abnormal data features by adopting a network analytic hierarchy process to obtain a plurality of production risk factors;
(4) And (3) carrying out comprehensive weight calculation on the plurality of production risk factors by adopting a fuzzy comprehensive evaluation method to obtain the comprehensive weight value of each production risk factor.
Specifically, the production-related data set is subjected to abnormal feature identification through a cyclic variation self-encoder. The cyclic variation self-encoder generates new data samples by learning a high-dimensional distribution of input data, and is suitable for time-series data, so that the cyclic variation self-encoder can effectively identify abnormal modes in the production process. For example, during battery assembly, the cyclic variation may learn normal production patterns from sensor data collected on the production line, such as temperature, pressure, etc., and identify abnormal data features that deviate from those patterns, which may be indicative of equipment failure or raw material problems. And (3) screening the first abnormal data characteristic by setting a preset target value, such as a percentage threshold value that the abnormal data exceeds a normal range, so as to obtain a second abnormal data characteristic which is more representative, reduce the interference of accidental abnormality on an analysis result, and ensure that the identified abnormal characteristic is related to potential risks in the production process. For example, if the reject ratio of a certain battery pack suddenly increases, by analyzing the first abnormal data feature (such as a high temperature alarm record during a certain period of time) associated with the first abnormal data feature and setting a reasonable threshold value, the second abnormal data feature which truly affects the quality of the battery pack can be effectively identified. And carrying out risk factor identification on the screened second abnormal data characteristics by adopting a network analytic hierarchy process. The network analytic hierarchy process is a decision support tool, and takes the dependency relationship and feedback loop between decision elements into consideration. By constructing a multi-level decision model, network analysis helps identify and evaluate the relative importance of different production risk factors. For example, by network hierarchy analysis, high temperature alarms identified in battery testing are compared with abnormal data characteristics of other production links to determine which factors are critical production risk factors that lead to reduced battery performance. And carrying out comprehensive weight calculation on the identified multiple production risk factors by adopting a fuzzy comprehensive evaluation method. And quantifying the influence degree of each risk factor by establishing a factor set, an evaluation set and a weight set and utilizing the principle of fuzzy mathematics. The comprehensive weight value of each production risk factor is obtained through calculation, and the relative importance and urgency of each risk factor in the production process are intuitively reflected.
In a specific embodiment, the performing step uses a network hierarchical analysis method to identify risk factors for the plurality of second abnormal data features, and the process of obtaining the plurality of production risk factors may specifically include the following steps:
(1) Constructing a relationship network structure between a risk criterion and a risk factor by adopting a network analytic hierarchy process;
(2) Performing paired comparison on the relation between each layer in the relation network structure and the target layer to construct a paired comparison matrix;
(3) Carrying out feature vector solving on the plurality of second abnormal data features and the paired comparison matrixes to obtain feature vectors of each second abnormal data feature;
(4) Carrying out consistency test on the feature vector of each second abnormal data feature to obtain a consistency test result of each feature vector;
(5) And generating a plurality of production risk factors of a plurality of second abnormal data features according to the consistency test result of each feature vector.
Specifically, a relationship network structure between risk criteria and risk factors is constructed by a network analytic hierarchy process. The network analytic hierarchy process is used as a decision support tool, which allows complex decision problems to be analyzed on multiple levels, and the dependence and feedback among decision elements are considered. For example, dividing a production link into several critical stages, such as material preparation, assembly, testing, etc., each stage may be subject to different risk factors, such as raw material quality instability, equipment aging, operating irregularities, etc. Network hierarchies identify and evaluate these risk factors as a whole and build a network relationship structure that contains all of these elements. And (3) carrying out paired comparison on the relation between each layer in the network structure and the target layer, constructing a paired comparison matrix, and determining the importance of the factors relative to the production flow optimization target. For example, in assessing the importance of raw material quality control relative to improving product consistency, it may be desirable to compare the relative importance of other factors such as equipment maintenance. The comparison results between each pair of factors are quantized and populated into a pair-wise comparison matrix. And carrying out feature vector solving according to the pair comparison matrix to obtain feature vectors of each second abnormal data feature, namely importance weights of all factors relative to the target. The feature vector reflects the relative magnitude of the impact of each risk factor on the final decision goal given the evaluation criteria. And carrying out consistency check on the feature vector of each second abnormal data feature, and verifying consistency and reliability of the pair-wise comparison matrix. Consistency checking is accomplished by calculating a Consistency Ratio (CR), where a CR value within an acceptable range (e.g., less than 0.1) means that the results of the pair-wise comparisons are consistent and the decision process is reliable. And generating a plurality of production risk factors of a plurality of second abnormal data features according to the consistency test result of each feature vector. The importance and consistency assessment of all relevant risk factors are comprehensively considered, and the important risk factors are determined, which are important to pay attention to and manage in the current production flow.
In a specific embodiment, the performing step uses a fuzzy comprehensive evaluation method to perform comprehensive weight calculation on a plurality of production risk factors, and the process of obtaining the comprehensive weight value of each production risk factor may specifically include the following steps:
(1) Determining an evaluation set and a factor set of a plurality of production risk factors by adopting a fuzzy comprehensive evaluation method, wherein the evaluation set is used for evaluating risk grades, the risk grades comprise low risk, medium risk and high risk, and the factor set comprises a plurality of production risk factors;
(2) Generating an initial weight value of each production risk factor according to the evaluation set and the factor set;
(3) Constructing an initial fuzzy relation matrix of each production risk factor according to the initial weight value and the evaluation set of each production risk factor;
(4) Calculating a comprehensive fuzzy relation matrix according to the initial fuzzy relation matrix of each production risk factor, and performing fuzzy synthesis operation on the comprehensive fuzzy relation matrix to obtain a comprehensive evaluation result vector of each production risk factor;
(5) And carrying out comprehensive weight calculation on the initial weight value of each production risk factor according to the comprehensive evaluation result vector to obtain the comprehensive weight value of each production risk factor.
Specifically, an evaluation set and a factor set of a plurality of production risk factors are determined by a fuzzy comprehensive evaluation method. The evaluation set is used to describe different levels of risk, e.g., three levels of low risk, medium risk, and high risk, reflecting the severity of the risk. The factor set then contains various risk factors that affect the production process, such as raw material quality, equipment failure rate, operator skill level, etc., which are the subject of the evaluation process. An initial weight value is assigned to each production risk factor based on the set of evaluations and the set of factors. The initial weight value reflects the relative importance of each risk factor throughout the production process. And constructing an initial fuzzy relation matrix of each production risk factor by using the initial weight value and the evaluation set of each production risk factor. The matrix describes the membership of each risk factor at different risk levels, i.e., the likelihood that the risk factor is rated as each risk level. For example, membership with a high failure rate of the device may be higher at a high risk level, indicating that it has a more serious impact on the production process. And calculating a comprehensive fuzzy relation matrix according to the initial fuzzy relation matrix of each production risk factor. And obtaining a global view by integrating the fuzzy relation matrix of all risk factors, and reflecting the overall influence of all risk factors in the whole production process. And carrying out fuzzy synthesis operation on the comprehensive fuzzy relation matrix to obtain a comprehensive evaluation result vector of each production risk factor. The vector provides a quantified assessment of the risk level of each risk factor in the overall production process. And based on the comprehensive evaluation result vector, carrying out comprehensive weight calculation on the initial weight value of each production risk factor to obtain the comprehensive weight value of each production risk factor. And correcting and optimizing the original evaluation so that the weight of the risk factors can more accurately reflect the actual influence of the risk factors in the production process. For example, if some risk factors which are considered to be less influenced, such as operator skill levels, are found by fuzzy comprehensive evaluation, and the influence in actual production is greater, then the comprehensive weight values of these factors will correspondingly increase, indicating that the production manager needs to pay more attention to management and improvement in these aspects.
In a specific embodiment, the process of executing step S103 may specifically include the following steps:
(1) Defining a plurality of relation nodes of the initial production flow according to a plurality of production risk factors;
(2) Initializing a tracing graph of a plurality of relation nodes based on a graph theory algorithm to obtain an initial tracing graph;
(3) Carrying out graph node weighted analysis on the initial traceability graph according to the comprehensive weight value of each production risk factor to obtain a causal traceability graph of the initial production flow;
(4) And labeling the stock and the flow of the initial production flow based on the causal tracing map, and obtaining stock flow diagrams of a plurality of production risk factors.
Specifically, a plurality of relationship nodes of the initial production flow are defined based on the identified plurality of production risk factors. These nodes represent key risk points that may be encountered during the production process, such as varying raw material quality, aging of the production equipment, lack of operator skill, etc. And initializing the tracing graph of the relation node based on a graph theory algorithm to obtain an initial tracing graph. Graph theory algorithms, such as minimum spanning tree or network flow analysis, can connect all nodes with a minimum of connecting lines, revealing interactions between them. For example, in the battery assembly link, if an operation error is found to occur frequently, the "operator skill deficiency" may be used as a node, which is connected with other related nodes (such as "product quality defect") through a graph theory algorithm. And carrying out graph node weighted analysis on the initial traceability graph according to the comprehensive weight value of each production risk factor to obtain a causal traceability graph of the initial production flow. The comprehensive weight value reflects the importance of each risk factor on the influence of the production flow. By weighting the nodes, which risk factors are key factors, namely factors which have the greatest influence on the production flow, are obtained. And labeling the stock and the flow of the initial production flow based on the causal tracing map, and obtaining stock flow diagrams of a plurality of production risk factors. Stock refers to static resources in the production process, such as raw material storage, equipment quantity, etc., while flow refers to dynamic changes in the production process, such as raw material consumption rate, product yield rate, etc. By marking the stock and the flow of the causal traceability graph, the production risk factors are understood how to influence the configuration and the utilization of production resources so as to influence the whole production flow. For example, if the quality of raw materials is not uniform, resulting in frequent production interruptions, the flow of raw materials will be shown intermittent in the inventory flow graph, and the inventory of the production line may be increased by accumulated semi-finished products.
In a specific embodiment, the process of executing step S104 may specifically include the following steps:
(1) Carrying out production flow optimization point analysis on the causal tracing graph to obtain a first flow optimization point;
(2) Performing production flow optimization point analysis on the stock flow graph to obtain a second flow optimization point;
(3) Performing optimization point comprehensive analysis and screening on the first flow optimization point and the second flow optimization point to obtain a production flow optimization point;
(4) Initializing a process optimization strategy of the production process optimization point through a genetic algorithm to obtain a plurality of first process optimization strategies;
(5) Performing fitness calculation on the plurality of first process optimization strategies to obtain fitness values of each first process optimization strategy;
(6) Performing strategy screening on the plurality of first process optimization strategies according to the fitness value to obtain a plurality of second process optimization strategies;
(7) Carrying out optimization solution on a plurality of second process optimization strategies to obtain a target process optimization strategy;
(8) And carrying out flow optimization on the initial production flow according to the target flow optimization strategy to obtain the target production flow.
Specifically, the causal tracing graph is subjected to production flow optimization point analysis, the reasons causing low efficiency or quality problems are identified, and a first flow optimization point is obtained. For example, if the causal traceability graph shows that the quality of raw materials is different, resulting in an increase in product defect rate, then improving the raw material inspection process is a key optimization point. Likewise, analysis of the stock flow graph can reveal how production resources flow and translate between different production loops, determining a second set of flow optimization points. If the stock flow graph indicates that the resource usage of a certain production area is inefficient, such as that the backlog of semi-finished products indicates an imbalance in production cadence, it is indicated that a production plan needs to be adjusted or a workstation layout needs to be optimized. And carrying out comprehensive analysis and screening on the optimization points of the first flow optimization point and the second flow optimization point, and determining comprehensive flow optimization points capable of improving production efficiency and product quality to the greatest extent. Considering the potential influence and implementation difficulty of each optimization point, the finally selected optimization point is ensured to solve the key problem and has feasibility and cost effectiveness. And initializing a flow optimization strategy for the screened production flow optimization points through a genetic algorithm. The genetic algorithm is a search algorithm imitating a natural evolution process, and the optimal solution is searched in the candidate solution set in an iterative mode through operations such as selection, crossover, mutation and the like. In this process, each flow optimization strategy is considered an "individual" and the specifics of the strategy (e.g., adjusting raw material inspection criteria, redesigning production layout, etc.) constitute the individual's "genes". By initialization, a "population" is generated that contains a plurality of possible optimization strategies. Fitness calculations are performed on each individual in the population, evaluating the potential contribution of each process optimization strategy to the production process improvement. Fitness calculation takes into account factors such as cost, expected benefit and risk of policy implementation. And screening a batch of process optimization strategies with higher potential value, namely a second batch of process optimization strategies, based on the fitness value. And carrying out optimization solution on the second batch of flow optimization strategies to determine a final target flow optimization strategy. Through multiple generations of iteration, the overall fitness of the population is continuously improved, and finally an optimal strategy capable of maximizing the production efficiency and the product quality is obtained. And adjusting and optimizing the initial production flow according to the target flow optimization strategy to realize the target production flow. This may include specific measures to adjust raw material procurement strategies, improve quality control flows, optimize production layout and cadence, etc.
The method for tracing management of lithium ion battery production in the embodiment of the present application is described above, and the tracing management system for lithium ion battery production in the embodiment of the present application is described below, referring to fig. 2, an embodiment of the tracing management system for lithium ion battery production in the embodiment of the present application includes:
The acquisition module 201 is used for constructing a cloud edge cooperative framework and acquiring a production association data set of an initial production flow in a lithium ion battery production line;
the computing module 202 is configured to perform risk factor identification and comprehensive weight calculation on the production association data set to obtain a plurality of production risk factors and a comprehensive weight value of each production risk factor;
the creating module 203 is configured to create a causal trace map and a stock flow map of the initial production flow according to the multiple production risk factors and the comprehensive weight values;
and the optimizing module 204 is used for carrying out production flow optimizing point analysis on the causal tracing graph and the stock flow graph to obtain production flow optimizing points and carrying out flow optimization on the initial production flow to obtain the target production flow.
Through the cooperation of the components, a cloud edge cooperation framework is constructed, and a production association data set of each flow area is collected, so that comprehensive data support is provided for the whole lithium ion battery production flow. The method not only increases the transparency of the production flow, but also greatly improves the traceability of each production stage, and provides a data base for problem diagnosis and quality control in the production process. The production related data set is subjected to deep analysis through a cyclic variation self-encoder, a network analytic hierarchy process and a fuzzy comprehensive evaluation method, so that key risk factors in the production process can be accurately identified, comprehensive weight values of the key risk factors are calculated, and the optimization points of the production process can be subjected to deep analysis through creating a causal tracing graph and a stock flow graph of the initial production process, so that efficiency bottlenecks and quality risk points in the production process can be identified. By formulating and implementing specific flow optimization strategies aiming at the optimization points, the production efficiency can be effectively improved, the product quality is ensured, and the production cost is reduced. Based on the comprehensive weight calculation and the process optimization analysis result, the continuous dynamic adjustment and optimization of the production process can be performed. The causal tracing graph and the stock flow graph are updated regularly, so that the change and the newly-appearing risk in the production process are reflected in time, the continuous monitoring and the self-adaptive optimization of the production process are realized, the production flow optimization accuracy of lithium ion battery production is further improved, and the intelligent traceability management of lithium ion battery production is realized.
It will be clearly understood by those skilled in the art that, for convenience and brevity of description, the specific working processes of the above-described systems, systems and units may refer to the corresponding processes in the foregoing method embodiments, which are not repeated herein.
The integrated units, if implemented in the form of software functional units and sold or used as stand-alone products, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present application may be embodied essentially or in part or all of the technical solution or in part in the form of a software product stored in a storage medium, including instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method according to the embodiments of the present application. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a read-only memory (ROM), a random access memory (random acceS memory, RAM), a magnetic disk, or an optical disk, or other various media capable of storing program codes.
The above embodiments are only for illustrating the technical solution of the present application, and not for limiting the same; although the application has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims (6)

Carrying out risk factor identification and comprehensive weight calculation on the production associated data set to obtain a plurality of production risk factors and comprehensive weight values of each production risk factor; the method specifically comprises the following steps: carrying out abnormal characteristic identification on the production association data set through a cyclic variation self-encoder to obtain a plurality of first abnormal data characteristics; feature screening is carried out on the first abnormal data features through a preset target value to obtain second abnormal data features; carrying out risk factor identification on the plurality of second abnormal data features by adopting a network analytic hierarchy process to obtain a plurality of production risk factors; carrying out comprehensive weight calculation on the plurality of production risk factors by adopting a fuzzy comprehensive evaluation method to obtain a comprehensive weight value of each production risk factor;
Creating a causal traceability map and a stock flow map of the initial production flow according to the plurality of production risk factors and the comprehensive weight value; the method specifically comprises the following steps: defining a plurality of relationship nodes of the initial production flow according to the plurality of production risk factors; initializing a tracing graph of the plurality of relation nodes based on a graph theory algorithm to obtain an initial tracing graph; carrying out graph node weighted analysis on the initial traceability graph according to the comprehensive weight value of each production risk factor to obtain a causal traceability graph of the initial production flow; labeling the stock and the flow of the initial production flow based on the causal tracing graph to obtain stock flow graphs of the production risk factors, wherein the stock refers to static resources in the production flow, including raw material storage and equipment quantity, and the flow refers to dynamic changes in the production process, including raw material consumption rate and product output rate;
Performing production flow optimization point analysis on the causal tracing graph and the stock flow graph to obtain production flow optimization points and performing flow optimization on the initial production flow to obtain a target production flow; the method specifically comprises the following steps: carrying out production flow optimization point analysis on the causal tracing graph to obtain a first flow optimization point; performing production flow optimization point analysis on the stock flow graph to obtain a second flow optimization point; performing optimization point comprehensive analysis and screening on the first flow optimization point and the second flow optimization point to obtain a production flow optimization point; initializing a process optimization strategy of the production process optimization point through a genetic algorithm to obtain a plurality of first process optimization strategies; performing fitness calculation on the plurality of first process optimization strategies to obtain fitness values of each first process optimization strategy; performing strategy screening on the plurality of first process optimization strategies according to the fitness value to obtain a plurality of second process optimization strategies; carrying out optimization solution on the plurality of second process optimization strategies to obtain a target process optimization strategy; and carrying out flow optimization on the initial production flow according to the target flow optimization strategy to obtain a target production flow.
The calculation module is used for carrying out risk factor identification and comprehensive weight calculation on the production association data set to obtain a plurality of production risk factors and comprehensive weight values of each production risk factor; the computing module is specifically configured to: carrying out abnormal characteristic identification on the production association data set through a cyclic variation self-encoder to obtain a plurality of first abnormal data characteristics; feature screening is carried out on the first abnormal data features through a preset target value to obtain second abnormal data features; carrying out risk factor identification on the plurality of second abnormal data features by adopting a network analytic hierarchy process to obtain a plurality of production risk factors; carrying out comprehensive weight calculation on the plurality of production risk factors by adopting a fuzzy comprehensive evaluation method to obtain a comprehensive weight value of each production risk factor;
The creation module is used for creating a causal traceability map and a stock flow map of the initial production flow according to the plurality of production risk factors and the comprehensive weight value; the creation module is specifically configured to: defining a plurality of relationship nodes of the initial production flow according to the plurality of production risk factors; initializing a tracing graph of the plurality of relation nodes based on a graph theory algorithm to obtain an initial tracing graph; carrying out graph node weighted analysis on the initial traceability graph according to the comprehensive weight value of each production risk factor to obtain a causal traceability graph of the initial production flow; labeling the stock and the flow of the initial production flow based on the causal tracing graph to obtain stock flow graphs of the production risk factors, wherein the stock refers to static resources in the production flow, including raw material storage and equipment quantity, and the flow refers to dynamic changes in the production process, including raw material consumption rate and product output rate;
The optimizing module is used for carrying out production flow optimizing point analysis on the causal tracing graph and the stock flow graph to obtain production flow optimizing points and carrying out flow optimization on the initial production flow to obtain a target production flow; the method specifically comprises the following steps: carrying out production flow optimization point analysis on the causal tracing graph to obtain a first flow optimization point; performing production flow optimization point analysis on the stock flow graph to obtain a second flow optimization point; performing optimization point comprehensive analysis and screening on the first flow optimization point and the second flow optimization point to obtain a production flow optimization point; initializing a process optimization strategy of the production process optimization point through a genetic algorithm to obtain a plurality of first process optimization strategies; performing fitness calculation on the plurality of first process optimization strategies to obtain fitness values of each first process optimization strategy; performing strategy screening on the plurality of first process optimization strategies according to the fitness value to obtain a plurality of second process optimization strategies; carrying out optimization solution on the plurality of second process optimization strategies to obtain a target process optimization strategy; and carrying out flow optimization on the initial production flow according to the target flow optimization strategy to obtain a target production flow.
CN202410414673.5A2024-04-082024-04-08Traceability management method and traceability management system for lithium ion battery productionActiveCN118014165B (en)

Priority Applications (1)

Application NumberPriority DateFiling DateTitle
CN202410414673.5ACN118014165B (en)2024-04-082024-04-08Traceability management method and traceability management system for lithium ion battery production

Applications Claiming Priority (1)

Application NumberPriority DateFiling DateTitle
CN202410414673.5ACN118014165B (en)2024-04-082024-04-08Traceability management method and traceability management system for lithium ion battery production

Publications (2)

Publication NumberPublication Date
CN118014165A CN118014165A (en)2024-05-10
CN118014165Btrue CN118014165B (en)2024-06-25

Family

ID=90952339

Family Applications (1)

Application NumberTitlePriority DateFiling Date
CN202410414673.5AActiveCN118014165B (en)2024-04-082024-04-08Traceability management method and traceability management system for lithium ion battery production

Country Status (1)

CountryLink
CN (1)CN118014165B (en)

Families Citing this family (5)

* Cited by examiner, † Cited by third party
Publication numberPriority datePublication dateAssigneeTitle
CN118733599B (en)*2024-09-022024-11-15南通国轩新能源科技有限公司Full-flow data tracing method and system based on energy storage cloud platform
CN119338438B (en)*2024-10-112025-09-19祥符实验室Intelligent hardware for production safety decision
CN119338184B (en)*2024-10-162025-08-19广东嘉尚新能源科技有限公司Intelligent management method for production of soft package battery
CN119313361B (en)*2024-12-132025-07-11浙江工业大学 Method and system for constructing production process data set for edge computing of food quality traceability
CN120067601B (en)*2025-04-272025-06-27广州萃取生物科技有限公司NLP-based wheat germ production abnormal root cause tracing method and system

Citations (1)

* Cited by examiner, † Cited by third party
Publication numberPriority datePublication dateAssigneeTitle
CN103324165A (en)*2013-05-272013-09-25西北工业大学Process route optimization method considering production line stability

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication numberPriority datePublication dateAssigneeTitle
CN111737642A (en)*2020-05-132020-10-02海洋石油工程股份有限公司 Comprehensive evaluation method of failure risk of submarine pipeline based on fuzzy network analysis method
CN116976725A (en)*2023-07-052023-10-31中国标准化研究院Evaluation method for enterprise standardized implementation effect
CN117478358A (en)*2023-09-272024-01-30新华三信息安全技术有限公司Decision recommendation method and device

Patent Citations (1)

* Cited by examiner, † Cited by third party
Publication numberPriority datePublication dateAssigneeTitle
CN103324165A (en)*2013-05-272013-09-25西北工业大学Process route optimization method considering production line stability

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
基于Meta图的供应链形式化模型研究;陈弘健;生佳根;秦强;王志凯;;计算机与数字工程;20171120(11);第232-235+247页*

Also Published As

Publication numberPublication date
CN118014165A (en)2024-05-10

Similar Documents

PublicationPublication DateTitle
CN118014165B (en)Traceability management method and traceability management system for lithium ion battery production
CN117930815B (en)Wind turbine generator remote fault diagnosis method and system based on cloud platform
CN117670378B (en)Food safety monitoring method and system based on big data
CN118261439B (en)Internet intelligent manufacturing service platform
CN118011990B (en)Industrial data quality monitoring and improving system based on artificial intelligence
CN112200327B (en) A MES equipment maintenance early warning method and system
CN118194730B (en)Chip wafer test optimization method
CN119272580B (en) A performance simulation method for power modules
CN114252794B (en) A method and device for predicting the remaining life of a dismantled smart electric energy meter
CN119295037A (en) A method and system for managing the entire life cycle of mine electromechanical equipment
CN118780719A (en) A material management method and system for power grid engineering projects based on the Internet of Things
CN120178824A (en) Production control method and system based on digital factory
Friederich et al.A framework for validating data-driven discrete-event simulation models of cyber-physical production systems
Mateo Casalí et al.An industry maturity model for implementing Machine Learning operations in manufacturing
CN120217158A (en) Asset operation and maintenance decision-making management platform and management method based on data fusion
CN118797538B (en) Industrial energy storage system management method and system based on multi-source data
CN113780788B (en)Modeling method for health evaluation of embedded equipment
CN118982347A (en) IT operation and maintenance service management method based on big data
CN117574303A (en)Construction condition monitoring and early warning method, device, equipment and storage medium
CN119440964B (en) A device information processing method, device and computer device based on ontology modeling
CN119918430B (en) An intelligent simulation system for lyocell fiber production line based on digital twin
CN119203044B (en) Organizational digital transformation and upgrading system and method based on A-DFCT technology
TWI230349B (en)Method and apparatus for analyzing manufacturing data
CN119991097B (en)Method, equipment and medium for evaluating production capacity guarantee degree based on health degree score
Fatemi et al.Presenting a multi-objective intelligent dynamic model of preventive maintenance using data mining

Legal Events

DateCodeTitleDescription
PB01Publication
PB01Publication
SE01Entry into force of request for substantive examination
SE01Entry into force of request for substantive examination
GR01Patent grant
GR01Patent grant

[8]ページ先頭

©2009-2025 Movatter.jp