WO2023202098A1 - 锂电池化成阶段产品质量预测方法及系统 - Google Patents

锂电池化成阶段产品质量预测方法及系统 Download PDF

Info

Publication number
WO2023202098A1
WO2023202098A1 PCT/CN2022/138144 CN2022138144W WO2023202098A1 WO 2023202098 A1 WO2023202098 A1 WO 2023202098A1 CN 2022138144 W CN2022138144 W CN 2022138144W WO 2023202098 A1 WO2023202098 A1 WO 2023202098A1
Authority
WO
WIPO (PCT)
Prior art keywords
lithium battery
product quality
database
model
quality prediction
Prior art date
Application number
PCT/CN2022/138144
Other languages
English (en)
French (fr)
Inventor
杨之乐
周邦昱
姚文娇
郭媛君
刘凯龙
李慷
张艳辉
冯伟
王尧
Original Assignee
深圳先进技术研究院
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 深圳先进技术研究院 filed Critical 深圳先进技术研究院
Publication of WO2023202098A1 publication Critical patent/WO2023202098A1/zh

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063—Operations research, analysis or management
    • G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
    • G06Q10/06395—Quality analysis or management
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G06F18/243—Classification techniques relating to the number of classes
    • G06F18/24323—Tree-organised classifiers
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • H—ELECTRICITY
    • H01—ELECTRIC ELEMENTS
    • H01M—PROCESSES OR MEANS, e.g. BATTERIES, FOR THE DIRECT CONVERSION OF CHEMICAL ENERGY INTO ELECTRICAL ENERGY
    • H01M10/00—Secondary cells; Manufacture thereof
    • H01M10/05—Accumulators with non-aqueous electrolyte
    • H01M10/052—Li-accumulators
    • H01M10/0525—Rocking-chair batteries, i.e. batteries with lithium insertion or intercalation in both electrodes; Lithium-ion batteries
    • H—ELECTRICITY
    • H01—ELECTRIC ELEMENTS
    • H01M—PROCESSES OR MEANS, e.g. BATTERIES, FOR THE DIRECT CONVERSION OF CHEMICAL ENERGY INTO ELECTRICAL ENERGY
    • H01M10/00—Secondary cells; Manufacture thereof
    • H01M10/42—Methods or arrangements for servicing or maintenance of secondary cells or secondary half-cells
    • H01M10/44—Methods for charging or discharging
    • H01M10/446—Initial charging measures

Definitions

  • the invention belongs to the technical field of lithium-ion batteries, and particularly relates to a method and system for predicting product quality in the formation stage of lithium batteries.
  • Lithium-ion batteries are an important energy storage technology that will play an important role in the future transformation of the electronic mobility and energy industries.
  • lithium battery manufacturing still has high production costs and high environmental impact due to expensive materials, high process fluctuations with high scrap rates, and high energy requirements.
  • the lack of in-depth understanding of the lithium battery production process and its impact on lithium battery quality and performance makes it difficult to plan, control and execute production. Therefore, adopting a systematic approach to gain a deeper understanding of the interconnections between processes and product quality and performance is of great significance to improving enterprise economic efficiency.
  • the first charge curve and discharge curve of the graphite anode material of lithium battery do not completely overlap.
  • the difference between the charge capacity and the discharge capacity is called irreversible capacity.
  • the formation of irreversible capacity is mainly related to the formation of SEI film and other side reactions.
  • the SEI film is a solid electrolyte membrane that is ion-conductive but electron-non-conductive.
  • the main purpose of the formation is to form a complete SEI film on the surface of the negative electrode, so that the battery has stable cycle capability.
  • Random Forest is an ensemble learning method for classification, regression, and other tasks that operates by building multiple decision trees at training time and outputting the class as a mode (classification) or an average prediction (regression) of the class. Random forest can process very high-dimensional data without feature selection, and can detect the factors that influence the results and the influence between features. If a large portion of features are missing, accuracy can still be maintained using the random forest algorithm.
  • the material cost of lithium batteries accounts for about 70% of the total manufacturing cost, and process deviations and waste highly affect manufacturing costs.
  • the scrap rate may range from around 5% to 15%, and there may even be cases where up to 40% of produced lithium batteries are defective or need to be repaired in later production.
  • Embodiments of the present invention provide a method and device for predicting product quality in the formation stage of lithium batteries, so as to at least solve the technical problem of providing basis and convenience for technicians to improve and control battery production and enhance product quality.
  • a method for predicting product quality in the formation stage of lithium batteries includes the following steps:
  • Step S10 Establish a battery production database
  • Step S20 Extract and screen suitable features in the battery production database
  • Step S30 Establish a lithium battery product quality prediction model based on the characteristics
  • Step S40 Predict the quality of the lithium battery according to the lithium battery formation quality prediction model.
  • the step of preprocessing the data in the battery production database is also included.
  • the preprocessing includes subjecting the above data to missing value processing, outlier processing, dummy coding processing and normalization processing, and storing the preprocessed data in the initial database.
  • step S10 the step of establishing a battery production database specifically includes the following steps:
  • Step S11 Collect the initial raw materials and intermediate products of the chemical formation reaction process, perform product analysis and inspection on the initial raw materials and intermediate products, and extract the first characteristic data of the initial raw materials and intermediate products.
  • the first characteristic data includes but is not limited to Electrode materials and dosage, electrolyte composition and dosage, inorganic additives and dosage, binder dosage, formation current and voltage, formation temperature and time, battery moisture content, workshop humidity, pole piece and battery thickness;
  • Step S12 Extract second characteristic data reflecting performance characteristics of the final products that have completed the aging process in the same batch.
  • the second characteristic data includes, but is not limited to, self-discharge rate, maximum capacity, and battery health status after several cycles of charge and discharge;
  • Step S13 Merge the first characteristic data and the second characteristic data to establish a battery production database.
  • step S20 the step of extracting and screening suitable features in the battery production database specifically includes the following steps:
  • Step S21 Establish a copy of the database in the battery production database to form a feature screening database, divide the feature screening database into the original training set and the original test set according to the K ratio, and determine the parameters of the random forest regression model;
  • Step S22 Train the random forest regression model, predict the quality features that reflect the performance indicators of the final product, calculate and evaluate the indicators and feature importance of the regression model, and arrange them in descending order according to the feature importance;
  • Step S23 Delete the data corresponding to the last-ranked feature in the feature screening database to form a new feature screening database
  • Step S24 Randomly divide the new feature screening database into a training set and a test set according to the proportion of K, and determine the parameters of the random forest regression model;
  • Step S25 Determine whether the results of the random forest regression model in predicting the final product quality of the battery just meet the accuracy requirements or the number of remaining features is equal to the preset value and the error of the model also meets the requirements, stop filtering features, and the features at this time
  • the screening database was used as a database for training and testing random forest regression models.
  • the K ratio is 8:2
  • a random search method is used to determine the parameter value range of the random forest regression model
  • a grid search method is used to determine the parameters of the random forest regression model.
  • the indicators for evaluating the regression model include mean absolute error, mean square error, root mean square error and mean absolute percentage error.
  • step S30 the step of establishing a lithium battery product quality prediction model based on the characteristics specifically includes the following steps:
  • Step S31 Divide the database for training and testing the random forest regression model into a final training set and a final test set;
  • Step S32 Determine the parameter value range and random forest regression model parameters of the training and testing random forest regression model
  • Step S33 Use the final training set data to train the training and testing random forest regression model, obtain the prediction results of the final product quality indicators in the lithium battery formation stage, test its performance on the final test set, and calculate errors and characteristics Importance;
  • Step S34 Collect the battery production database generated subsequently in the formation stage, perform data preprocessing on the battery production database generated subsequently in the formation stage, and add it to the database for training and testing the random forest regression model;
  • Step S35 Train the training and testing random forest regression models according to the above steps S31 ⁇ 33, and compare the errors of the new and old models in predicting the final product quality, and determine whether the new model is better than the old model. If so, take the new model as Lithium battery quality prediction model, if not, maintain the old model, and obtain the lithium battery formed product quality prediction model.
  • step S40 the step of predicting the quality of the lithium battery according to the lithium battery formation quality prediction model specifically includes the following steps:
  • the second characteristic data is used as the target data of the lithium battery formed product quality prediction model, and the prediction value can be obtained by running the model.
  • a lithium battery formation stage product quality prediction system including:
  • Database building unit used to establish battery production database
  • a feature screening unit used to extract and screen suitable features in the battery production database
  • a lithium battery production quality prediction model used to establish a lithium battery product quality prediction model based on the characteristics
  • a prediction unit is used to predict the quality of the lithium battery according to the lithium battery formation quality prediction model.
  • a storage medium that stores program files capable of implementing any one of the above-mentioned methods for predicting product quality in the lithium battery formation stage.
  • a processor the processor is used to run a program, wherein when the program is running, it executes any one of the product quality prediction methods in the lithium battery formation stage.
  • the lithium battery formation stage product quality prediction method and system in the embodiment of the present invention establishes a battery production database, extracts and screens suitable features in the battery production database, and establishes a lithium battery formation product quality prediction model based on the features.
  • the lithium battery formation quality prediction model predicts the lithium battery quality.
  • the present invention uses the random forest regression model in machine learning to analyze the product data of the lithium battery formation stage to obtain the quality prediction model.
  • the advantage of the random forest model is that it can reveal intermediate products.
  • the impact and importance of characteristics on the performance of the final product provide convenience and basis for technicians to improve the production, design, control and operation of lithium batteries. Compared with traditional methods, it has the advantages of simple operation, high accuracy and fast speed.
  • Figure 1 is a step flow chart of the product quality prediction method in the lithium battery formation stage provided by Embodiment 1 of the present invention
  • Figure 2 is a flow chart of steps for establishing a battery production database provided by Embodiment 1 of the present invention
  • Figure 3 is a flow chart of steps for extracting and screening suitable features in the battery production database provided in Embodiment 1 of the present invention
  • Figure 4 is a flow chart of steps for establishing a quality prediction model for lithium battery formation products based on the characteristics provided in Embodiment 1 of the present invention.
  • Figure 5 is a schematic structural diagram of a random forest-based product quality prediction system in the formation stage of lithium batteries provided in Embodiment 2 of the present invention.
  • a method for predicting product quality in the formation stage of lithium batteries is provided, which includes the following steps S10 to S40. The implementation of each step is described in detail below.
  • Step S10 Establish a battery production database.
  • Figure 2 is a flow chart of steps for establishing a battery production database provided in this embodiment, including the following steps:
  • Step S11 Collect the initial raw materials and intermediate products of the chemical formation reaction process, perform product analysis and inspection on the initial raw materials and intermediate products, and extract the first characteristic data of the initial raw materials and intermediate products.
  • the first characteristic data includes but is not limited to Electrode materials and dosage, electrolyte composition and dosage, inorganic additives and dosage, binder dosage, formation current and voltage, formation temperature and time, battery moisture content, workshop humidity, pole piece and battery thickness.
  • Step S12 Extract second characteristic data reflecting performance characteristics of the final products that have completed the aging process in the same batch.
  • the second characteristic data includes, but is not limited to, self-discharge rate, maximum capacity, and battery health status after several cycles of charge and discharge.
  • Step S13 Merge the first characteristic data and the second characteristic data to establish a battery production database.
  • the step of preprocessing the data in the battery production database is also included.
  • the preprocessing includes performing missing value processing, outlier processing, dummy coding processing and normalization processing on the above data, and storing the preprocessed data in the initial database.
  • the above missing value processing includes deleting data containing missing values.
  • outlier processing includes: According to the definition of normal distribution, the probability of a distance beyond 3 ⁇ from the mean is P(
  • >3 ⁇ ) ⁇ 0.003. When the distance of a sample from the mean is greater than 3 ⁇ , the sample is deemed to be For outliers, delete data containing outliers.
  • the dummy coding process includes One-Hot coding, which is also called one-bit effective coding. It mainly uses an N-bit status register to code N states. Each state has its own independent register bit, and in Only one is valid at any time.
  • the normalization process is a linear transformation of the original data, so that the result falls into the [0,1] interval.
  • max is the maximum value of sample data
  • min is the minimum value of sample data. Note that each time new data is added to the database, it needs to be recalculated.
  • Step S20 Extract and screen suitable features in the battery production database.
  • Figure 3 is a flow chart of steps for extracting and screening suitable features in the battery production database in this embodiment, which specifically includes the following steps:
  • Step S21 Establish a copy of the database in the battery production database to form a feature screening database, divide the feature screening database into the original training set and the original test set according to the K ratio, and determine the parameters of the random forest regression model.
  • the K ratio is 8:2
  • a random search method is used to determine the parameter value range of the random forest regression model
  • a grid search method is used to determine the parameters of the random forest regression model.
  • Step S22 Train the random forest regression model, predict the quality features that reflect the performance indicators of the final product, calculate and evaluate the indicators and feature importance of the regression model, and arrange them in descending order according to the feature importance.
  • the indicators for evaluating the regression model include mean absolute error, mean square error, root mean square error and mean absolute percentage error.
  • Step S23 Delete the data corresponding to the last-ranked feature in the feature screening database to form a new feature screening database.
  • Step S24 Randomly divide the new feature screening database into a training set and a test set according to the proportion of K, and determine the parameters of the random forest regression model.
  • Step S25 Determine whether the result of the random forest regression model in predicting the final product quality of the battery just meets the accuracy requirements or the number of remaining features is equal to the preset value and the error of the model also meets the requirements. If yes, stop filtering features. If not, Repeat steps S22 ⁇ 24, and the feature screening database at this time is used as the database for training and testing the random forest regression model.
  • Step S30 Establish a lithium battery product quality prediction model based on the characteristics.
  • Figure 4 is a flow chart of steps for establishing a quality prediction model for lithium battery formation products based on the characteristics described in this embodiment, which specifically includes the following steps:
  • Step S31 Divide the database for training and testing the random forest regression model into a final training set and a final test set.
  • the database obtained above for training and testing the random forest regression model is divided into a training set and a test set in a ratio of 8:2.
  • Step S32 Determine the parameter value range and random forest regression model parameters of the training and testing random forest regression model.
  • the random search method is used to determine the parameter value range of the random forest regression model, and then the grid search method is used to determine the appropriate random forest regression model parameters.
  • Step S33 Use the final training set data to train the training and testing random forest regression model, obtain the prediction results of the final product quality indicators in the lithium battery formation stage, test its performance on the final test set, and calculate errors and characteristics Importance.
  • the quality indicators of the final product in the lithium battery formation stage include, but are not limited to, self-discharge rate, maximum capacity, and battery health status after several cycles of charge and discharge.
  • a quality prediction model that meets the requirements can be obtained by adjusting the parameters of the random forest regression model, such as increasing the depth and number of trees in the model, adding features, enriching and improving the data set, etc.
  • Step S34 Collect the battery production database generated subsequently in the formation stage, perform data preprocessing on the battery production database generated subsequently in the formation stage, and add it to the database for training and testing the random forest regression model.
  • Step S35 Train the training and testing random forest regression models according to the above steps S31 ⁇ 33, and compare the errors of the new and old models in predicting the final product quality, and determine whether the new model is better than the old model. If so, take the new model as Lithium battery quality prediction model, if not, maintain the old model, and obtain the lithium battery formed product quality prediction model.
  • Step S40 Predict the quality of the lithium battery according to the lithium battery formation quality prediction model.
  • the step of predicting the quality of the lithium battery according to the lithium battery formation quality prediction model specifically includes the following steps:
  • the first characteristic data is used as the input data of the lithium battery formed product quality prediction model; the second characteristic data is used as the target data of the lithium battery formed product quality prediction model, and the predicted value can be obtained by running the model.
  • the purpose of this invention is to add tags to each communication link and hardware resources on the basis of the classic von Neumann structure, speed up the process's access to resources, and more targeted correspondence between needs and resources to reduce the complexity of tasks. Run time, improve computer resource utilization, reduce mutual interference between applications that share resources, and reduce some concurrent lock problems that may otherwise occur, thereby reducing long-tail delays.
  • the invention aims to add tags to the process starting from the input device, and use the control center to control, judge and adjust the priority allocation of underlying hardware resources, thereby realizing top-down tag-based resource allocation and assigning different processes to the process according to the information provided by the tag. resources to solve problems, thereby improving the utilization of computer resources.
  • the embodiment of the present invention proposes a method for predicting product quality in the lithium battery formation stage.
  • the random forest regression model in machine learning is used to analyze the product data in the lithium battery formation stage to obtain a quality prediction model.
  • the advantage of the random forest model is that it can reveal intermediate products.
  • the impact and importance of characteristics on the performance of the final product provide convenience and basis for technicians to improve the production, design, control and operation of lithium batteries. Compared with traditional methods, it has the advantages of simple operation, high accuracy and fast speed.
  • a schematic structural diagram of a product quality prediction system in the formation stage of lithium batteries including:
  • Database building unit 110 used to establish a battery production database
  • Feature screening unit 120 used to extract and screen suitable features in the battery production database
  • Lithium battery production quality prediction model 130 used to establish a lithium battery product quality prediction model based on the characteristics
  • the prediction unit 140 is configured to predict the quality of the lithium battery according to the lithium battery formation quality prediction model.
  • the embodiment of the present invention proposes a product quality prediction system in the lithium battery formation stage.
  • the random forest regression model in machine learning is used to analyze the product data in the lithium battery formation stage to obtain a quality prediction model.
  • the advantage of the random forest model is that it can reveal intermediate products.
  • the impact and importance of characteristics on the performance of the final product provide convenience and basis for technicians to improve the production, design, control and operation of lithium batteries. Compared with traditional methods, it has the advantages of simple operation, high accuracy and fast speed.
  • a storage medium that stores program files capable of implementing any of the above product quality prediction methods in the formation stage of lithium batteries.
  • a processor is used to run a program, wherein when the program is running, it executes any one of the above-mentioned product quality prediction methods in the lithium battery formation stage.
  • the random forest regression model in machine learning is used to analyze the product data of the lithium battery formation stage to obtain a quality prediction model.
  • the advantage of the random forest model is that it can reveal the impact of intermediate product characteristics on the final product performance. The impact and importance provide convenience and basis for technicians to improve the production, design, control and operation of lithium batteries. Compared with traditional methods, it has the advantages of simple operation, high accuracy and fast speed.
  • the disclosed technical content can be implemented in other ways.
  • the system embodiments described above are only illustrative.
  • the division of units can be a logical function division. In actual implementation, there may be other division methods.
  • multiple units or components can be combined or integrated into Another system, or some features can be ignored, or not implemented.
  • the coupling or direct coupling or communication connection between each other shown or discussed may be through some interfaces, and the indirect coupling or communication connection of the units or modules may be in electrical or other forms.
  • Units described as separate components may or may not be physically separate, and components shown as units may or may not be physical units, that is, they may be located in one place, or they may be distributed over multiple units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.
  • each functional unit in various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit.
  • the above integrated units can be implemented in the form of hardware or software functional units.
  • Integrated units may be stored in a computer-readable storage medium if they are implemented in the form of software functional units and sold or used as independent products.
  • the technical solution of the present invention is essentially or contributes to the existing technology or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium , including several instructions to cause a computer device (which can be a personal computer, a server or a network device, etc.) to execute all or part of the steps of the methods of various embodiments of the present invention.
  • the aforementioned storage media include: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, magnetic disk or optical disk and other media that can store program code.

Landscapes

  • Engineering & Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Human Resources & Organizations (AREA)
  • Strategic Management (AREA)
  • Economics (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Chemical & Material Sciences (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Development Economics (AREA)
  • Physics & Mathematics (AREA)
  • Tourism & Hospitality (AREA)
  • Quality & Reliability (AREA)
  • General Chemical & Material Sciences (AREA)
  • Electrochemistry (AREA)
  • Chemical Kinetics & Catalysis (AREA)
  • Educational Administration (AREA)
  • Manufacturing & Machinery (AREA)
  • Data Mining & Analysis (AREA)
  • Game Theory and Decision Science (AREA)
  • General Business, Economics & Management (AREA)
  • Marketing (AREA)
  • Operations Research (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Artificial Intelligence (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Materials Engineering (AREA)
  • General Engineering & Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Evolutionary Biology (AREA)
  • Evolutionary Computation (AREA)
  • Battery Electrode And Active Subsutance (AREA)

Abstract

本发明提供的锂电池化成阶段产品质量预测方法及系统，建立电池生产数据库，提取并筛选所述电池生产数据库中合适的特征，根据所述特征建立锂电池化成产品质量预测模型，根据所述锂电池化成质量预测模型对锂电池质量进行预测，本发明利用机器学习中随机森林回归模型对锂电池化成阶段产品数据进行分析，得到质量预测模型，由于随机森林模型的优势在于能够揭示中间产品特征对最终产品性能的影响以及重要程度，为技术人员改进锂电池生产、设计、控制和操作提供便利和依据。比于传统的方法，具有操作简单、准确性高、速度快等优点。

Description

锂电池化成阶段产品质量预测方法及系统 技术领域
本发明属于锂离子电池技术领域，特别涉及一种锂电池化成阶段产品质量预测方法及系统。
 背景技术
锂离子电池一种重要的储能技术，在未来的电子移动和能源行业转型中发挥着重要作用。然而，由于昂贵的材料、高废品率的高工艺波动和高能源需求，锂电池制造仍然具有高生产成本和高环境影响。由于缺乏对锂电池生产过程及其对锂电池质量和性能的影响的深入了解，难以计划、控制和执行生产。因此，采用系统的方法来深入了解过程与产品质量和性能之间的相互联系对提高企业经济效益具有重要意义。
锂电池石墨负极材料的首次充电曲线和放电曲线并不完全重合，充电容量和放电容量的差值称不可逆容量，不可逆容量的形成主要与形成SEI膜和其他副反应有关。SEI膜是一种离子可导，电子不可导的固体电解质膜，化成的主要目的是使负极表面形成完整的SEI膜，从而使电池具有稳定的循环能力。
随机森林是用于分类，回归和其他任务的集成学习方法，其通过在训练时构建多个决策树并输出作为类的模式（分类）或平均预测（回归）的类来操作。随机森林能处理很高维度的数据，并且不用做特征选择，能够检测到特征对结果的影响因子及特征之间的影响。如果有很大一部分的特征遗失，用随机森林算法仍然可以维持准确度。
锂电池材料成本约占总制造成本的70%，工艺偏差和废料高度影响制造成本。报废率可能在5%左右到15%之间不等，甚至可能存在高达40%的已生产锂电池有缺陷，或需要在后期生产中修复的情况。
 技术问题
本发明实施例提供了一种锂电池化成阶段产品质量预测方法及装置，以至少解决为技术人员改进和控制电池生产，提升产品质量提供依据和便利的技术问题。
 技术解决方案
根据本发明的一实施例，提供了一种锂电池化成阶段产品质量预测方法，包括以下步骤：
步骤S10：建立电池生产数据库；
步骤S20：提取并筛选所述电池生产数据库中合适的特征；
步骤S30：根据所述特征建立锂电池化成产品质量预测模型；
步骤S40：根据所述锂电池化成质量预测模型对锂电池质量进行预测。
在其中一些实施例中，在建立电池生产数据库的步骤后，进行下一步步骤之前，还包括对所述电池生产数据库中的数据进行预处理的步骤。
在其中一些实施例中，所述预处理包括将上述的数据进行缺失值处理、异常值处理、哑编码处理及归一化处理，并将预处理后的数据存入初始数据库中。
在其中一些实施例中，在步骤S10，建立电池生产数据库的步骤中，具体包括下述步骤：
步骤S11：收集化成反应过程的初始原材料和中间产品，并对所述初始原材料和中间产品进行产品分析和检验，提取初始原材料和中间产品的第一特征数据，所述第一特征数据包括不限于电极材料及用量，电解液成分及用量，无机添加剂及用量、黏结剂用量、化成的电流和电压、化成时的温度和时间、电池的水分含量、车间湿度、极片和电池的厚度；
步骤S12：对同一批次完成老化工艺的最终产品提取反映性能特征的第二特征数据，所述第二特征数据包括不限于自放电率，最大容量，若干次循环充放电后的电池健康状态；
步骤S13：将所述第一特征数据与所述第二特征数据合并，建立电池生产数据库。
在其中一些实施例中，在步骤S20，提取并筛选所述电池生产数据库中合适的特征的步骤中，具体包括下述步骤：
步骤S21：建立所述电池生产数据库中的数据库副本，形成特征筛选数据库，将所述特征筛选数据库按照K比例划分原始训练集和原始测试集，并确定随机森林回归模型的参数；
步骤S22：训练所述随机森林回归模型，对反映最终产品性能指标的质量特征进行预测，并计算评估回归模型的指标和特征重要度，且按照特征重要度大小降序排列；
步骤S23：删除所述特征筛选数据库中排名末尾的特征对应的数据，形成用于新的特征筛选数据库；
步骤S24：将所述新的特征筛选数据库按所述K的比例随机地划分为训练集和测试集，并确定随机森林回归模型的参数；
步骤S25：判断所述随机森林回归模型预测电池最终产品质量的结果是否刚好满足精度需求或剩余的特征数量等于预先设定的值且模型的误差也满足要求，停止筛选特征，且此时的特征筛选数据库作为训练和测试随机森林回归模型的数据库。
在其中一些实施例中，所述K比例取8:2，用随机搜索法确定所述随机森林回归模型的参数值域，采用网格搜索法确定所述随机森林回归模型的参数。
在其中一些实施例中，所述评估回归模型的指标包括平均绝对误差、均方误差、均方根误差和平均绝对百分比误差。
在其中一些实施例中，在步骤S30，根据所述特征建立锂电池化成产品质量预测模型的步骤中，具体包括下述步骤：
步骤S31：将所述训练和测试随机森林回归模型的数据库划分为最终训练集和最终测试集；
步骤S32：确定所述训练和测试随机森林回归模型的参数值域及随机森林回归模型参数；
步骤S33：用所述最终训练集数据训练所述训练和测试随机森林回归模型，取得对锂电池化成阶段最终产品质量指标的预测结果并在所述最终测试集上测试其性能，计算误差和特征重要度；
步骤S34：收集化成阶段后续产生的电池生产数据库，并将所述化成阶段后续产生的电池生产数据库进行数据预处理，添加到所述训练和测试随机森林回归模型的数据库中；
步骤S35：根据上述步骤S31~33对所述训练和测试随机森林回归模型进行训练，并比较新旧模型对最终产品质量预测的误差，判断新模型是否优于旧模型，若是，则取新模型作为锂电池质量预测模型，若否，维持旧模型，即得到所述锂电池化成产品质量预测模型。
在其中一些实施例中，在步骤S40，根据所述锂电池化成质量预测模型对锂电池质量进行预测的步骤中，具体包括下述步骤：
将所述第一特征数据作为所述锂电池化成产品质量预测模型的输入数据；
将所述第二特征数据作为所述锂电池化成产品质量预测模型的目标数据，运行模型即可得到预测值。
根据本发明的另一实施例，提供了一种锂电池化成阶段产品质量预测系统，包括：
数据库构建单元，用于建立电池生产数据库；
特征筛选单元，用于提取并筛选所述电池生产数据库中合适的特征；
锂电池生产质量预测模型，用于根据所述特征建立锂电池化成产品质量预测模型；
预测单元，用于根据所述锂电池化成质量预测模型对锂电池质量进行预测。
一种存储介质，所述存储介质存储有能够实现上述任意一项所述锂电池化成阶段产品质量预测方法的程序文件。
一种处理器，所述处理器用于运行程序，其中，所述程序运行时执行任意一项所述的锂电池化成阶段产品质量预测方法。
 有益效果
本发明实施例中的锂电池化成阶段产品质量预测方法及系统，建立电池生产数据库，提取并筛选所述电池生产数据库中合适的特征，根据所述特征建立锂电池化成产品质量预测模型，根据所述锂电池化成质量预测模型对锂电池质量进行预测，本发明利用机器学习中随机森林回归模型对锂电池化成阶段产品数据进行分析，得到质量预测模型，由于随机森林模型的优势在于能够揭示中间产品特征对最终产品性能的影响以及重要程度，为技术人员改进锂电池生产、设计、控制和操作提供便利和依据。比于传统的方法，具有操作简单、准确性高、速度快等优点。
 附图说明
此处所说明的附图用来提供对本发明的进一步理解，构成本申请的一部分，本发明的示意性实施例及其说明用于解释本发明，并不构成对本发明的不当限定。在附图中：
图1为本发明实施例1提供的锂电池化成阶段产品质量预测方法的步骤流程图；
图2为本发明实施例1提供的建立电池生产数据库的步骤流程图；
图3为本发明实施例1提供的提取并筛选所述电池生产数据库中合适的特征的步骤流程图；
图4为本发明实施例1提供的根据所述特征建立锂电池化成产品质量预测模型的步骤流程图。
图5为本发明实施例2提供的一种基于随机森林的锂电池化成阶段产品质量预测系统结构示意图。
 本发明的实施方式
为了使本技术领域的人员更好地理解本发明方案，下面将结合本发明实施例中的附图，对本发明实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例仅仅是本发明一部分的实施例，而不是全部的实施例。基于本发明中的实施例，本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例，都应当属于本发明保护的范围。
需要说明的是，本发明的说明书和权利要求书及上述附图中的术语“第一”、“第二”等是用于区别类似的对象，而不必用于描述特定的顺序或先后次序。应该理解这样使用的数据在适当情况下可以互换，以便这里描述的本发明的实施例能够以除了在这里图示或描述的那些以外的顺序实施。此外，术语“包括”和“具有”以及他们的任何变形，意图在于覆盖不排他的包含，例如，包含了一系列步骤或单元的过程、方法、系统、产品或设备不必限于清楚地列出的那些步骤或单元，而是可包括没有清楚地列出的或对于这些过程、方法、产品或设备固有的其它步骤或单元。
实施例1
请参阅图1，根据本发明一实施例，提供了一种锂电池化成阶段产品质量预测方法，包括以下步骤S10~S40，以下详细说明每个步骤的实现方案。
步骤S10：建立电池生产数据库。
请参阅图2，为本实施例提供的建立电池生产数据库的步骤流程图，包括下述步骤：
步骤S11：收集化成反应过程的初始原材料和中间产品，并对所述初始原材料和中间产品进行产品分析和检验，提取初始原材料和中间产品的第一特征数据，所述第一特征数据包括不限于电极材料及用量，电解液成分及用量，无机添加剂及用量、黏结剂用量、化成的电流和电压、化成时的温度和时间、电池的水分含量、车间湿度、极片和电池的厚度。
步骤S12：对同一批次完成老化工艺的最终产品提取反映性能特征的第二特征数据，所述第二特征数据包括不限于自放电率，最大容量，若干次循环充放电后的电池健康状态。
步骤S13：将所述第一特征数据与所述第二特征数据合并，建立电池生产数据库。
在本实施例中，在建立电池生产数据库的步骤后，进行下一步步骤之前，还包括对所述电池生产数据库中的数据进行预处理的步骤。
在本实施例中，所述预处理包括将上述的数据进行缺失值处理、异常值处理、哑编码处理及归一化处理，并将预处理后的数据存入初始数据库中。
进一步地，上述缺失值处理包括删除含缺失值的数据。
进一步地，异常值处理包括根据正态分布的定义可知，距离平均值3δ之外的概率为 P(|x-μ|>3δ) <= 0.003，当样本距离平均值大于3δ，则认定该样本为异常值，删除含异常值的数据。
进一步地，哑编码处理包括用One-Hot编码，其又称为一位有效编码，主要是采用N位状态寄存器来对N个状态进行编码，每个状态都由他独立的寄存器位，并且在任意时候只有一位有效。
进一步地，归一化处理是对原始数据的线性变换，使结果落到[0,1]区间，转换函数如下：x*=(x-min)/(max-min)；
其中：max是样本数据最大值，min是样本数据最小值。注意每次新数据加入数据库需要重新计算。
步骤S20：提取并筛选所述电池生产数据库中合适的特征。
请参阅图3，为本实施例在提取并筛选所述电池生产数据库中合适的特征的步骤流程图，具体包括下述步骤：
步骤S21：建立所述电池生产数据库中的数据库副本，形成特征筛选数据库，将所述特征筛选数据库按照K比例划分原始训练集和原始测试集，并确定随机森林回归模型的参数。
具体地，所述K比例取8:2，用随机搜索法确定所述随机森林回归模型的参数值域，采用网格搜索法确定所述随机森林回归模型的参数。
可以理解，对于随机森林中的每一颗决策树,使用相应的OOB(袋外数据)数据来计算它的袋外数据误差，记为EO1。随机地对袋外数据OOB所有样本的特征X加入噪声干扰,再次计算它的袋外数据误差，记为EO2。假设随机森林中有N棵树,那么对于特征X的重要度=∑(EO2-EO1)/N。
步骤S22：训练所述随机森林回归模型，对反映最终产品性能指标的质量特征进行预测，并计算评估回归模型的指标和特征重要度，且按照特征重要度大小降序排列。
具体地，所述评估回归模型的指标包括平均绝对误差、均方误差、均方根误差和平均绝对百分比误差。
步骤S23：删除所述特征筛选数据库中排名末尾的特征对应的数据，形成用于新的特征筛选数据库。
步骤S24：将所述新的特征筛选数据库按所述K的比例随机地划分为训练集和测试集，并确定随机森林回归模型的参数。
可以理解，K的比例在整个筛选特征的过程中应保持一致，以提高预测准确性。
步骤S25：判断所述随机森林回归模型预测电池最终产品质量的结果是否刚好满足精度需求或剩余的特征数量等于预先设定的值且模型的误差也满足要求，若是，停止筛选特征，若否，重复步骤S22~24，且此时的特征筛选数据库作为训练和测试随机森林回归模型的数据库。
可以理解，本实施例中所述的结果刚好满足精度需求即表示下一次重复后精度不满足要求。
步骤S30：根据所述特征建立锂电池化成产品质量预测模型。
请参阅图4，为本实施例提供的根据所述特征建立锂电池化成产品质量预测模型的步骤流程图，具体包括下述步骤：
步骤S31：将所述训练和测试随机森林回归模型的数据库划分为最终训练集和最终测试集。
在实际中，将上述得到的训练和测试随机森林回归模型的数据库按8:2比例划分为训练集和测试集。
步骤S32：确定所述训练和测试随机森林回归模型的参数值域及随机森林回归模型参数。
具体地，用随机搜索法确定随机森林回归模型参数值域，然后采用网格搜索法确定适合的随机森林回归模型参数。
步骤S33：用所述最终训练集数据训练所述训练和测试随机森林回归模型，取得对锂电池化成阶段最终产品质量指标的预测结果并在所述最终测试集上测试其性能，计算误差和特征重要度。
本实施例中，锂电池化成阶段最终产品质量指标包括不限于自放电率，最大容量，若干次循环充放电后的电池健康状态。
可以理解，对于不符合性能要求的模型，通过调整随机森林回归模型参数，比如加大模型中树的深度和数量等、增加特征、丰富和完善数据集等手段以得到符合要求的质量预测模型。
步骤S34：收集化成阶段后续产生的电池生产数据库，并将所述化成阶段后续产生的电池生产数据库进行数据预处理，添加到所述训练和测试随机森林回归模型的数据库中。
步骤S35：根据上述步骤S31~33对所述训练和测试随机森林回归模型进行训练，并比较新旧模型对最终产品质量预测的误差，判断新模型是否优于旧模型，若是，则取新模型作为锂电池质量预测模型，若否，维持旧模型，即得到所述锂电池化成产品质量预测模型。
步骤S40：根据所述锂电池化成质量预测模型对锂电池质量进行预测。
在本实施例中，根据所述锂电池化成质量预测模型对锂电池质量进行预测的步骤中，具体包括下述步骤：
将所述第一特征数据作为所述锂电池化成产品质量预测模型的输入数据；将所述第二特征数据作为所述锂电池化成产品质量预测模型的目标数据，运行模型即可得到预测值。
本发明目的是在经典的冯诺依曼结构的基础上，在各通信环节以及硬件资源添加标签，加快进程访问资源的速度，以及将需求与资源进行更有针对性的对应，减小任务的运行时间，提高计算机资源的利用率，减小共享资源的应用程序的相互干扰，以及减少部分原本可能发生的并发锁问题，从而降低长尾延迟。本发明旨在从输入设备开始，将进程添加标签，利用控制中心来控制判断调整底层硬件资源的优先级分配，从而实现自顶向下基于标签的资源分配，按照标签提供的信息给进程分配不同的资源来解决问题，从而提升计算机资源的利用率。
本发明实施例提出一种锂电池化成阶段产品质量预测方法，利用机器学习中随机森林回归模型对锂电池化成阶段产品数据进行分析，得到质量预测模型，由于随机森林模型的优势在于能够揭示中间产品特征对最终产品性能的影响以及重要程度，为技术人员改进锂电池生产、设计、控制和操作提供便利和依据。比于传统的方法，具有操作简单、准确性高、速度快等优点。
实施例2
请参阅图5，根据本发明的另一实施例，提供了一种锂电池化成阶段产品质量预测系统的结构示意图，包括：
数据库构建单元110，用于建立电池生产数据库；
特征筛选单元120，用于提取并筛选所述电池生产数据库中合适的特征；
锂电池生产质量预测模型130，用于根据所述特征建立锂电池化成产品质量预测模型；
预测单元140，用于根据所述锂电池化成质量预测模型对锂电池质量进行预测。
本实施例提供的基于随机森林的锂电池化成阶段产品质量预测系统，其详细的实现方案在实施例1中已有详细说明，这里不再赘述。
本发明实施例提出一种锂电池化成阶段产品质量预测系统，利用机器学习中随机森林回归模型对锂电池化成阶段产品数据进行分析，得到质量预测模型，由于随机森林模型的优势在于能够揭示中间产品特征对最终产品性能的影响以及重要程度，为技术人员改进锂电池生产、设计、控制和操作提供便利和依据。比于传统的方法，具有操作简单、准确性高、速度快等优点。
实施例3
一种存储介质，存储介质存储有能够实现上述任意一项锂电池化成阶段产品质量预测方法的程序文件。
实施例4
一种处理器，处理器用于运行程序，其中，程序运行时执行上述任意一项的锂电池化成阶段产品质量预测方法。
本发明实施例的技术优点至少在于：利用机器学习中随机森林回归模型对锂电池化成阶段产品数据进行分析，得到质量预测模型，由于随机森林模型的优势在于能够揭示中间产品特征对最终产品性能的影响以及重要程度，为技术人员改进锂电池生产、设计、控制和操作提供便利和依据。比于传统的方法，具有操作简单、准确性高、速度快等优点。
上述本发明实施例序号仅仅为了描述，不代表实施例的优劣。
在本发明的上述实施例中，对各个实施例的描述都各有侧重，某个实施例中没有详述的部分，可以参见其他实施例的相关描述。
在本申请所提供的几个实施例中，应该理解到，所揭露的技术内容，可通过其它的方式实现。其中，以上所描述的系统实施例仅仅是示意性的，例如单元的划分，可以为一种逻辑功能划分，实际实现时可以有另外的划分方式，例如多个单元或组件可以结合或者可以集成到另一个系统，或一些特征可以忽略，或不执行。另一点，所显示或讨论的相互之间的耦合或直接耦合或通信连接可以是通过一些接口，单元或模块的间接耦合或通信连接，可以是电性或其它的形式。
作为分离部件说明的单元可以是或者也可以不是物理上分开的，作为单元显示的部件可以是或者也可以不是物理单元，即可以位于一个地方，或者也可以分布到多个单元上。可以根据实际的需要选择其中的部分或者全部单元来实现本实施例方案的目的。
另外，在本发明各个实施例中的各功能单元可以集成在一个处理单元中，也可以是各个单元单独物理存在，也可以两个或两个以上单元集成在一个单元中。上述集成的单元既可以采用硬件的形式实现，也可以采用软件功能单元的形式实现。
集成的单元如果以软件功能单元的形式实现并作为独立的产品销售或使用时，可以存储在一个计算机可读取存储介质中。基于这样的理解，本发明的技术方案本质上或者说对现有技术做出贡献的部分或者该技术方案的全部或部分可以以软件产品的形式体现出来，该计算机软件产品存储在一个存储介质中，包括若干指令用以使得一台计算机设备(可为个人计算机、服务器或者网络设备等)执行本发明各个实施例方法的全部或部分步骤。而前述的存储介质包括：U盘、只读存储器(ROM，Read-Only Memory)、随机存取存储器(RAM，Random Access Memory)、移动硬盘、磁碟或者光盘等各种可以存储程序代码的介质。
以上所述仅是本发明的优选实施方式，应当指出，对于本技术领域的普通技术人员来说，在不脱离本发明原理的前提下，还可以做出若干改进和润饰，这些改进和润饰也应视为本发明的保护范围。

Claims (12)

  1. 一种锂电池化成阶段产品质量预测方法，其特征在于，包括以下步骤：
    步骤S10：建立电池生产数据库；
    步骤S20：提取并筛选所述电池生产数据库中合适的特征；
    步骤S30：根据所述特征建立锂电池化成产品质量预测模型；
    步骤S40：根据所述锂电池化成质量预测模型对锂电池质量进行预测。
  2. 根据权利要求1所述的锂电池化成阶段产品质量预测方法，其特征在于，在建立电池生产数据库的步骤后，进行下一步步骤之前，还包括对所述电池生产数据库中的数据进行预处理的步骤。
  3. 根据权利要求2所述的锂电池化成阶段产品质量预测方法，其特征在于，所述预处理包括将上述的数据进行缺失值处理、异常值处理、哑编码处理及归一化处理，并将预处理后的数据存入初始数据库中。
  4. 根据权利要求1所述的锂电池化成阶段产品质量预测方法，其特征在于，在步骤S10，建立电池生产数据库的步骤中，具体包括下述步骤：
    步骤S11：收集化成反应过程的初始原材料和中间产品，并对所述初始原材料和中间产品进行产品分析和检验，提取初始原材料和中间产品的第一特征数据，所述第一特征数据包括不限于电极材料及用量，电解液成分及用量，无机添加剂及用量、黏结剂用量、化成的电流和电压、化成时的温度和时间、电池的水分含量、车间湿度、极片和电池的厚度；
    步骤S12：对同一批次完成老化工艺的最终产品提取反映性能特征的第二特征数据，所述第二特征数据包括不限于自放电率，最大容量，若干次循环充放电后的电池健康状态；
    步骤S13：将所述第一特征数据与所述第二特征数据合并，建立电池生产数据库。
  5. 根据权利要求4所述的锂电池化成阶段产品质量预测方法，其特征在于，在步骤S20，提取并筛选所述电池生产数据库中合适的特征的步骤中，具体包括下述步骤：
    步骤S21：建立所述电池生产数据库中的数据库副本，形成特征筛选数据库，将所述特征筛选数据库按照K比例划分原始训练集和原始测试集，并确定随机森林回归模型的参数；
    步骤S22：训练所述随机森林回归模型，对反映最终产品性能指标的质量特征进行预测，并计算评估回归模型的指标和特征重要度，且按照特征重要度大小降序排列；
    步骤S23：删除所述特征筛选数据库中排名末尾的特征对应的数据，形成用于新的特征筛选数据库；
    步骤S24：将所述新的特征筛选数据库按所述K的比例随机地划分为训练集和测试集，并确定随机森林回归模型的参数；
    步骤S25：判断所述随机森林回归模型预测电池最终产品质量的结果是否刚好满足精度需求或剩余的特征数量等于预先设定的值且模型的误差也满足要求，若是，停止筛选特征，若否，重复步骤S22~24，且此时的特征筛选数据库作为训练和测试随机森林回归模型的数据库。
  6. 根据权利要求5所述的锂电池化成阶段产品质量预测方法，其特征在于，所述K比例取8:2，用随机搜索法确定所述随机森林回归模型的参数值域，采用网格搜索法确定所述随机森林回归模型的参数。
  7. 根据权利要求5所述的锂电池化成阶段产品质量预测方法，其特征在于，所述评估回归模型的指标包括平均绝对误差、均方误差、均方根误差和平均绝对百分比误差。
  8. 根据权利要求5所述的锂电池化成阶段产品质量预测方法，其特征在于，在步骤S30，根据所述特征建立锂电池化成产品质量预测模型的步骤中，具体包括下述步骤：
    步骤S31：将所述训练和测试随机森林回归模型的数据库划分为最终训练集和最终测试集；
    步骤S32：确定所述训练和测试随机森林回归模型的参数值域及随机森林回归模型参数；
    步骤S33：用所述最终训练集数据训练所述训练和测试随机森林回归模型，取得对锂电池化成阶段最终产品质量指标的预测结果并在所述最终测试集上测试其性能，计算误差和特征重要度；
    步骤S34：收集化成阶段后续产生的电池生产数据库，并将所述化成阶段后续产生的电池生产数据库进行数据预处理，添加到所述训练和测试随机森林回归模型的数据库中；
    步骤S35：根据上述步骤S31~33对所述训练和测试随机森林回归模型进行训练，并比较新旧模型对最终产品质量预测的误差，判断新模型是否优于旧模型，若是，则取新模型作为锂电池质量预测模型，若否，维持旧模型，即得到所述锂电池化成产品质量预测模型。
  9. 根据权利要求8所述的锂电池化成阶段产品质量预测方法，其特征在于，在步骤S40，根据所述锂电池化成质量预测模型对锂电池质量进行预测的步骤中，具体包括下述步骤：
    将所述第一特征数据作为所述锂电池化成产品质量预测模型的输入数据；
    将所述第二特征数据作为所述锂电池化成产品质量预测模型的目标数据，运行模型即可得到预测值。
  10. 一种锂电池化成阶段产品质量预测系统，其特征在于，包括：
    数据库构建单元，用于建立电池生产数据库；
    特征筛选单元，用于提取并筛选所述电池生产数据库中合适的特征；
    锂电池生产质量预测模型，用于根据所述特征建立锂电池化成产品质量预测模型；
    预测单元，用于根据所述锂电池化成质量预测模型对锂电池质量进行预测。
  11. 一种存储介质，其特征在于，所述存储介质存储有能够实现权利要求1至9中任意一项所述锂电池化成阶段产品质量预测方法的程序文件。
  12. 一种处理器，其特征在于，所述处理器用于运行程序，其中，所述程序运行时执行权利要求1至9中任意一项所述的锂电池化成阶段产品质量预测方法。
PCT/CN2022/138144 2022-04-19 2022-12-09 锂电池化成阶段产品质量预测方法及系统 WO2023202098A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202210408892.3A CN114819583A (zh) 2022-04-19 2022-04-19 锂电池化成阶段产品质量预测方法及系统
CN202210408892.3 2022-04-19

Publications (1)

Publication Number Publication Date
WO2023202098A1 true WO2023202098A1 (zh) 2023-10-26

Family

ID=82506574

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2022/138144 WO2023202098A1 (zh) 2022-04-19 2022-12-09 锂电池化成阶段产品质量预测方法及系统

Country Status (2)

Country Link
CN (1) CN114819583A (zh)
WO (1) WO2023202098A1 (zh)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117783887A (zh) * 2024-02-28 2024-03-29 深圳市神通天下科技有限公司 一种锂离子电池电芯配组筛选方法
CN117908493A (zh) * 2024-01-26 2024-04-19 深圳市磐锋精密技术有限公司 一种基于大数据的锂电池生产监测系统及其方法
CN118503894A (zh) * 2024-07-22 2024-08-16 山东高质新能源检测有限公司 基于工序指标数据分析的锂电池质量检测系统

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114819583A (zh) * 2022-04-19 2022-07-29 深圳先进技术研究院 锂电池化成阶段产品质量预测方法及系统
CN115098704B (zh) * 2022-08-24 2023-01-06 深圳市信润富联数字科技有限公司 电池极片厚度预测方法、装置、设备及可读存储介质
CN116307405B (zh) * 2023-05-25 2023-08-04 日照鲁光电子科技有限公司 一种基于生产数据的二极管性能预测方法及系统
CN118095579B (zh) * 2024-04-26 2024-08-30 宁德时代新能源科技股份有限公司 制程参数的确定方法、装置及系统、电子设备和存储介质

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111044928A (zh) * 2019-12-31 2020-04-21 福州大学 一种锂电池健康状态估计方法
CN112560287A (zh) * 2020-12-28 2021-03-26 杭州师范大学 基于随机森林回归的锂离子电池寿命预测系统
CN113125960A (zh) * 2019-12-31 2021-07-16 河北工业大学 一种基于随机森林模型的车载锂离子电池荷电状态预测方法
CN113687250A (zh) * 2021-08-18 2021-11-23 蜂巢能源科技有限公司 电芯容量预测方法、装置、电子设备及介质
CN114819583A (zh) * 2022-04-19 2022-07-29 深圳先进技术研究院 锂电池化成阶段产品质量预测方法及系统

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111044928A (zh) * 2019-12-31 2020-04-21 福州大学 一种锂电池健康状态估计方法
CN113125960A (zh) * 2019-12-31 2021-07-16 河北工业大学 一种基于随机森林模型的车载锂离子电池荷电状态预测方法
CN112560287A (zh) * 2020-12-28 2021-03-26 杭州师范大学 基于随机森林回归的锂离子电池寿命预测系统
CN113687250A (zh) * 2021-08-18 2021-11-23 蜂巢能源科技有限公司 电芯容量预测方法、装置、电子设备及介质
CN114819583A (zh) * 2022-04-19 2022-07-29 深圳先进技术研究院 锂电池化成阶段产品质量预测方法及系统

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117908493A (zh) * 2024-01-26 2024-04-19 深圳市磐锋精密技术有限公司 一种基于大数据的锂电池生产监测系统及其方法
CN117783887A (zh) * 2024-02-28 2024-03-29 深圳市神通天下科技有限公司 一种锂离子电池电芯配组筛选方法
CN117783887B (zh) * 2024-02-28 2024-05-14 深圳市神通天下科技有限公司 一种锂离子电池电芯配组筛选方法
CN118503894A (zh) * 2024-07-22 2024-08-16 山东高质新能源检测有限公司 基于工序指标数据分析的锂电池质量检测系统

Also Published As

Publication number Publication date
CN114819583A (zh) 2022-07-29

Similar Documents

Publication Publication Date Title
WO2023202098A1 (zh) 锂电池化成阶段产品质量预测方法及系统
Turetskyy et al. Battery production design using multi-output machine learning models
WO2021088207A1 (zh) 云计算集群混部作业调度方法、装置、服务器及存储装置
CN107506865B (zh) 一种基于lssvm优化的负荷预测方法及系统
Liu et al. Machine learning boosting the development of advanced lithium batteries
CN114757468B (zh) 一种面向流程挖掘中流程执行异常的根源分析方法
CN106021985B (zh) 一种基因组数据压缩方法
CN114490316B (zh) 一种基于损失函数的单元测试用例自动生成方法
Garg et al. Aging model development based on multidisciplinary parameters for lithium‐ion batteries
CN118275903B (zh) 一种基于数据分析的电池性能测试方法
CN113657030A (zh) 一种基于高斯过程回归的锂电池剩余使用寿命预测方法
CN115301559A (zh) 用于退役电池回收利用的高效筛选分拣方法及系统
Galvez‐Aranda et al. Time‐Dependent Deep Learning Manufacturing Process Model for Battery Electrode Microstructure Prediction
Han et al. A self‐adaptive, data‐driven method to predict the cycling life of lithium‐ion batteries
Zhang et al. Remaining useful life prediction of lithium-ion batteries based on TCN-DCN fusion model combined with IRRS filtering
CN109921462A (zh) 一种基于lstm的新能源消纳能力评估方法及系统
CN117574774A (zh) 极片的烘干参数的确定方法、装置、电子设备及存储介质
Cui et al. Machine learning approach for solving inconsistency problems of Li‐ion batteries during the manufacturing stage
CN117171710A (zh) 一种电力系统的故障诊断方法及故障诊断装置
CN116204849A (zh) 一种面向数字孪生应用的数据与模型融合方法
KR20200065820A (ko) 스마트 팩토리 도입을 위한 디지털트윈 모델링 기반의 에너지 및 보안 효율성 분석 시스템
Gao et al. Software defect prediction based on adaboost algorithm under imbalance distribution
CN111062827A (zh) 一种基于人工智能模式的工程监理方法
CN114859233A (zh) 一种面向全生命利用的锂离子电池容量损失预测方法
CN113190544A (zh) 一种面向企业的mes数据抽取和清洗方法

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 22938316

Country of ref document: EP

Kind code of ref document: A1