Method, system, device and medium for predicting moisture of power battery

Abstract

The application discloses a kind of power battery moisture prediction method, system, equipment and medium, the method includes: at least one target feature data in target cell before entering vacuum baking process, in vacuum baking process and after completing vacuum baking process;Target feature data is input into moisture prediction model to generate the moisture prediction result of target cell, wherein the moisture prediction model is obtained by training sample data set constructed by multidimensional feature data in power battery process, multidimensional feature data at least includes process parameters, equipment parameters and environmental parameters.Such, by collecting multidimensional feature data of power battery process to train moisture prediction model, the moisture of cell can be accurately predicted in baking whole process, solve the problem that traditional detection is large, high and cannot be evaluated and traced back comprehensively.

Description

Technical Field

[0001] This application relates to the field of power battery manufacturing technology, specifically to a method, system, device, and medium for predicting the moisture content of power batteries. Background Technology

[0002] In the manufacturing process of power batteries, the moisture content of the cells is one of the important indicators affecting battery performance and safety. To remove residual moisture inside the cells, vacuum baking is usually required before assembly. For cells that have undergone vacuum baking, the industry generally uses sampling disassembly for moisture testing. This offline testing method requires destroying the cell structure, is complex and time-consuming, and is limited by the sampling ratio, making it difficult to comprehensively assess the moisture content of all cells on the production line. Furthermore, since the test results are only available after the baking process is completed, it is impossible to pre-judge the moisture state of the cells during or before baking, resulting in a significant time lag in production control. In addition, the final moisture content of the cells is affected by multiple processes and factors throughout the entire manufacturing chain, from raw material input to vacuum baking completion. A single offline sampling result cannot reflect individual differences within a batch, nor can it pinpoint the specific production stage causing moisture fluctuations. As power battery manufacturing moves towards large-scale and intelligent production, how to achieve efficient prediction and full-process control of the moisture state of all cells without disassembling them has become a continuous concern in the industry. Summary of the Invention

[0003] This application provides a method, system, device, and medium for predicting the moisture content of power batteries. The method for predicting the moisture content of power batteries in this application trains a moisture prediction model by collecting multi-dimensional feature data of the power battery manufacturing process. It can accurately predict the moisture content of the battery cells throughout the baking process, solving the problems of traditional detection methods being highly destructive, having high latency, and being unable to comprehensively assess and trace the source.

[0004] On one hand, embodiments of this application provide a method for predicting the moisture content of a power battery, including: Acquire target characteristic data of at least one of the following: before entering the vacuum baking process, during the vacuum baking process, and / or after completing the vacuum baking process; The target feature data is input into the moisture prediction model to generate the moisture prediction result of the target battery cell. The moisture prediction model is trained by a sample dataset constructed from multi-dimensional feature data in the power battery manufacturing process. The multi-dimensional feature data includes at least process parameters, equipment parameters, and environmental parameters.

[0005] In some embodiments, the moisture prediction method further includes: Based on the unique identifier of the target battery cell, the multi-dimensional feature data corresponding to the target battery cell in the manufacturing process are traced and obtained; Determine whether the multidimensional feature data for each dimension is within the range of the standard parameters corresponding to that dimension; If there is target multidimensional feature data that is not within the corresponding standard parameter range, the target multidimensional feature data is marked as an abnormal parameter, and the unique identifier of the corresponding target cell is recorded.

[0006] In some embodiments, the power battery manufacturing process includes at least one of the following steps: feeding, stirring, coating, rolling and slitting, winding or stacking, welding, helium inspection, and vacuum baking.

[0007] In some embodiments, the training process of the moisture prediction model includes: The sample dataset is divided into a training set and a test set according to a preset ratio; The prediction model is trained using the training set, and the trained prediction model is evaluated using the test set. Based on the evaluation results, the parameters of the prediction model are tuned, and the tuning method includes one of grid search, random search, Bayesian optimization, and cross-validation. The prediction model that meets the preset accuracy requirements is used as the moisture prediction model.

[0008] In some embodiments, the moisture prediction method further includes: Evaluate the performance metrics of the moisture prediction model, wherein the performance metrics include at least one of the following: mean square error, root mean square error, mean absolute error, maximum absolute error, coefficient of determination, and mean absolute percentage error. The performance metrics and their corresponding visualization charts are stored. The visualization charts include at least one of the following: feature importance chart, prediction comparison chart, residual chart, scatter plot, and error distribution chart.

[0009] In some embodiments, the moisture prediction model is built based on a machine learning algorithm, which includes at least one of a neural network model, a tree model, a Bayesian model, a clustering model, a support vector machine model, and a recurrent neural network model.

[0010] In some embodiments, the moisture prediction model is used to predict the moisture content of at least one type of battery cell: prismatic laminated cells, prismatic wound cells, and cylindrical cells.

[0011] In some embodiments, the moisture prediction model is used to predict at least one of the moisture content of the positive electrode of the battery cell, the moisture content of the negative electrode of the battery cell, or the moisture content of the separator.

[0012] In some embodiments, the moisture prediction result of the target battery cell is the moisture content of the target battery cell after the vacuum baking process is completed.

[0013] On the other hand, embodiments of this application provide a moisture prediction system for a power battery, the moisture prediction system comprising: The acquisition module is used to acquire at least one target feature data of the target cell before entering the vacuum baking process, during the vacuum baking process, and after completing the vacuum baking process; The prediction module is used to input the target feature data into the moisture prediction model to generate the moisture prediction result of the target battery cell. The multidimensional feature data includes at least process parameters, equipment parameters and environmental parameters.

[0014] On the other hand, embodiments of this application provide an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it performs the moisture prediction method for a power battery as described in any of the previous embodiments.

[0015] On the other hand, embodiments of this application provide a non-volatile computer-readable storage medium storing a computer program, which, when executed by a processor, performs the moisture prediction method for a power battery as described in the above embodiments.

[0016] The moisture prediction method, moisture prediction system, electronic device, and storage medium for power batteries provided in this application training a moisture prediction model by collecting multi-dimensional characteristic data of the power battery's production process, equipment, and environment, can predict the moisture content of the target cell at multiple stages before, during, and after vacuum baking, significantly improving the accuracy and reliability of moisture prediction. Furthermore, it can achieve a comprehensive assessment of the moisture content of all cells without disassembling them, avoiding the damage to the cells caused by traditional sampling methods and solving the problems of offline detection lag and inability to control the process. At the same time, it can correlate and analyze the influencing factors of multiple processes, accurately locate the source of moisture fluctuations, improve the level of production control refinement, and effectively adapt to the needs of power batteries. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a schematic diagram of a moisture prediction system provided in an embodiment of this application.

[0019] Figure 2-5 This is a flowchart illustrating the method for predicting the moisture content of a power battery provided in an embodiment of this application.

[0020] Figure 6 This is a schematic diagram of the module of the power battery moisture prediction system provided in the embodiments of this application.

[0021] Figure 7 A schematic diagram of the modules of an electronic device provided in an embodiment of this application. Detailed Implementation

[0022] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0023] This application provides a method, system, electronic device, and medium for predicting the moisture content of power batteries. It can accurately predict the moisture content of battery cells throughout the baking process, solving the problems of traditional detection methods being highly destructive, having high latency, and being unable to provide comprehensive assessment and traceability.

[0024] Specifically, the moisture prediction method for power batteries in this application embodiment can be executed by an electronic device, which can be a terminal or a server. The terminal can be a computer, laptop, or other similar device, and may also include a client, which can be a user client, a browser client, an instant messaging client, or a mini-program, etc.

[0025] A server can be a standalone physical server, a server cluster or a distributed system composed of multiple physical servers, and can also provide cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, and content delivery networks. ContentDeliveryNetwork , CDN Cloud servers that provide basic cloud computing services such as big data and artificial intelligence platforms.

[0026] It should be noted that, in the embodiments of this application, the executing entity of the power battery moisture prediction method can be a terminal device or a server. The terminal device can be a local terminal device or the aforementioned client device. This application does not limit the type of executing entity.

[0027] For example, in conjunction with the above description, Figure 1This application illustrates a moisture prediction system 1000 for implementing a moisture prediction method for a power battery. The moisture prediction system 1000 may include at least one terminal 1001, at least one server 1002, at least one database 1003, and a network. The user-held terminal 1001 can connect to different servers via the network. The terminal can be any device with computing hardware capable of supporting and executing queries for corresponding software application tools.

[0028] Furthermore, when the moisture prediction system 1000 includes multiple terminals, multiple servers, and multiple networks, different terminals can connect to each other through different networks and different servers. The network can be a wireless network or a wired network; for example, the wireless network could be a wireless local area network (WLAN). WLAN ),local area network( LAN ), cellular network, 2 G Network, 3 G Network, 4 G Network, 5G G Networks, etc. Additionally, different terminals can connect to other terminals or servers using their own Bluetooth networks or hotspot networks. Furthermore, the system 100 can include multiple databases, each coupled to different servers.

[0029] It should be noted that, Figure 1 The schematic diagram of the moisture prediction system shown is merely an example. The moisture prediction system 1000 described in this application embodiment is intended to more clearly illustrate the technical solution of this application embodiment and does not constitute a limitation on the technical solution provided in this application embodiment. As those skilled in the art will know, with the evolution of moisture prediction systems and the emergence of new business scenarios, the technical solution provided in this application embodiment is also applicable to similar technical problems.

[0030] The technical solution of this application will be described in detail below through specific embodiments. It should be noted that the following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments.

[0031] Please see Figure 2 , Figure 2 This is a flowchart illustrating a method for predicting the moisture content of a power battery according to an embodiment of this application. It should be noted that the steps shown may be executed in a different logical order than those shown in the flowchart. The method may include the following steps: Step 110: Obtain at least one target feature data of the target cell before entering the vacuum baking process, during the vacuum baking process, and after completing the vacuum baking process; The target feature data refers to the set of all available feature parameters corresponding to the target battery cell and actually generated at the prediction time node, extracted under different prediction timings. The prediction timing can be pre-baking, during-baking, or post-baking prediction. For example, it is possible to obtain only the target feature data of the target battery cell before entering the vacuum baking process, or only the target feature data of the target battery cell before entering the vacuum baking process and during the vacuum baking process, or target feature data before entering the vacuum baking process, during the vacuum baking process, and after completing the vacuum baking process. It is worth noting that although the specific parameter items covered by the target feature data under different prediction timings may differ somewhat, they all cover at least one or more of the following categories: process parameters, equipment parameters, and environmental parameters.

[0032] When the prediction time is before entering the vacuum baking process, the target characteristic data may include, but are not limited to: raw material moisture-related parameters (such as the moisture content of positive electrode powder, negative electrode powder, and separator), coating moisture-related parameters (such as the moisture value of the electrode after coating, the temperature of each section of the coating oven, and the coating speed), process parameters of the completed preceding processes (such as stirring revolution speed, stirring time, stirring vacuum degree, and roller pressure), equipment parameters of the completed preceding processes (such as the motor current of the stirring tank, the coating roller pressure, and the roller temperature), and environmental parameters corresponding to each completed process (such as the environmental dew point of each process), and other factors related to the moisture content after baking.

[0033] When the prediction timing is when the target battery cell is in the vacuum baking process, the target characteristic data includes two parts: first, the relevant factors affecting post-baking moisture generated in the preceding processes completed before the target battery cell enters the vacuum baking process; second, the parameters actually generated by the target battery cell in the current vacuum baking process and related to post-baking moisture. It should be noted that since the baking process is not yet complete, some parameters (such as the final holding time and total air exchange rate) are partial or cumulative values ​​at the current moment, rather than the final values ​​after the complete process is completed.

[0034] When the prediction timing is after the target cell has completed the vacuum baking process, the target characteristic data includes the relevant process parameters, equipment parameters, and environmental parameters that affect the moisture content of the target cell after baking in all the preceding process steps, as well as the complete process parameters and equipment parameters of the vacuum baking process itself.

[0035] In addition, during real-time data acquisition, the timestamps of each data acquisition are recorded synchronously and bound to the unique identifier of the target battery cell and the baking process stage to form a dynamic time-series feature dataset. If data acquisition is interrupted, a supplementary acquisition mechanism is automatically triggered. If supplementary acquisition fails, data is marked as missing and the reason for the anomaly is recorded.

[0036] In this embodiment, the target feature data can originate from multiple data sources on the power battery production line. These data sources may include, but are not limited to, Manufacturing Execution System (MES), Internet of Things (IoT) systems, Quality Management System (QMS), environmental monitoring systems, and Programmable Logic Controllers (PLCs) related to the target battery cell. The QMS is used for product quality data collection, quality inspection, anomaly detection, defect statistics, traceability analysis, and quality control, achieving end-to-end quality monitoring and management from raw materials and processes to finished products. The IoT system, through sensors, smart meters, industrial equipment, and other hardware terminals, enables data collection, transmission, and interconnection between devices and systems, providing real-time equipment data, environmental data, and operational status data for the production process. The Manufacturing Execution System is a shop floor-level production management and real-time execution system used to monitor, schedule, and record the production process, process parameters, equipment status, and production data from material input to finished product, achieving transparent and digital management of the production process.

[0037] As an optional implementation, target characteristic data of the target battery cells can be collected in real time or in batches from the aforementioned data source systems via data interfaces or direct database connections. For example, target characteristic data can be obtained from the aforementioned information systems through one or more of the following: database query tools, data interface packet capture tools, system log export tools, and production line data acquisition platforms.

[0038] Step 120: Input the target feature data into the moisture prediction model to generate the moisture prediction result of the target battery cell.

[0039] It is worth noting that before the target feature data is input into the moisture prediction model, it can be normalized to form standardized target feature data that can be directly input into the moisture prediction model. The standardized target feature data is then input into the prediction model to obtain the moisture prediction results.

[0040] Moisture prediction models can be deployed in the cloud, on local servers, or in edge computing gateways close to the production line to reduce network latency.

[0041] The moisture prediction model is used to predict the moisture content of a target battery cell; that is, the moisture prediction result can be the moisture content of the target battery cell. For example, the moisture prediction result can be at least one of the moisture content of the positive electrode, the negative electrode, or the separator in the target cell; the target battery cell can be a prismatic laminated cell, a prismatic wound cell, a cylindrical cell, or other battery types, such as pouch polymer batteries, energy storage batteries, sodium-ion batteries, etc.

[0042] In addition, in some other implementations, the moisture prediction model can be extended to predict other quality indicators of the battery cell, in addition to predicting the moisture content of the battery cell (such as electrode peeling force, wetting effect after liquid injection, etc.).

[0043] It is also worth noting that the moisture prediction model is trained on a sample dataset constructed from multidimensional feature data in the power battery manufacturing process. The multidimensional feature data includes at least process parameters, equipment parameters, and environmental parameters, and these parameters are relevant factors affecting the moisture content after baking.

[0044] Multidimensional feature data refers to historical sample data with labeled moisture values, used to train and build moisture prediction models. Multidimensional feature data and target feature data maintain consistency in parameter dimensions and data structure; the only difference lies in their intended use.

[0045] Multidimensional feature data can be all parameters of the power battery in all processes (from feeding to vacuum baking of the cell). The power battery process includes at least one of the following processes: feeding process, stirring process, coating process, rolling and slitting process, winding or stacking process, welding process, helium inspection process, and vacuum baking process.

[0046] The system can obtain multi-dimensional feature data from information systems such as the Manufacturing Execution System (MES) system, Internet of Things (IoT) system, Quality Management System (QMS) system, environmental monitoring system, and Programmable Logic Controller (PLC) system of power batteries or power battery production lines.

[0047] For example, database query tools, data interface packet capture tools, system log export tools, and production line data acquisition tools can be used to retrieve equipment parameters, process parameters, environmental parameters, and incoming material parameters for the entire process of the target battery cell from feeding, mixing, coating, rolling and slitting, stacking / rolling, Xray-1, pre-welding, final welding, pre-helium inspection, Xray-2 to vacuum baking from the above-mentioned information system.

[0048] In constructing the sample dataset, data can first be associated according to timestamps and unique cell identifiers, aggregating the feature data of the same cell from the start of material feeding to the end of vacuum baking into a single data sample. Each data sample is accompanied by a moisture label value, which is the actual moisture content of the cell after vacuum baking, measured by a high-precision moisture detection instrument (such as a Karl Fischer moisture analyzer). Next, feature processing is performed on the cleaned data, such as logarithmic transformation, normalization, or standardization of certain parameters. Then, feature selection is performed based on the correlation coefficient and contribution rate with the post-baking moisture content, combined with the judgment of on-site process personnel, to finally obtain the sample dataset, thereby reducing data dimensionality and improving model training efficiency.

[0049] Thus, the aforementioned moisture prediction method acquires at least one target feature data of the target battery cell before, during, and after the vacuum baking process. This target feature data is then input into a moisture prediction model trained with multi-dimensional feature data covering process parameters, equipment parameters, and environmental parameters to generate a moisture prediction result for the target battery cell. This achieves non-destructive, real-time, and accurate prediction of the moisture content of the battery cell before, during, or after vacuum baking. Compared to traditional disassembly and offline testing methods, this method does not require damage to the battery cell structure, significantly reducing testing costs and material waste. Furthermore, due to the flexible timing of prediction, moisture prediction and sorting can be performed before baking, dynamic process control can be implemented during baking, and 100% full inspection can be achieved after baking, effectively eliminating the time lag of traditional testing methods. In addition, the input features of the prediction model integrate multi-dimensional parameters from the entire power battery manufacturing chain, from raw materials to vacuum baking, overcoming the shortcomings of insufficient prediction accuracy caused by relying solely on parameters from a single baking process. This makes the prediction results closer to the actual moisture state of the battery cell, significantly improving the accuracy and reliability of moisture prediction.

[0050] In another alternative implementation, online moisture detection hardware such as near-infrared spectral sensors, microwave moisture sensors, or radio frequency moisture detection probes can be added to the production line to collect the original moisture signal of the battery cell or electrode in real time before or during the vacuum baking process. This real-time moisture signal is used as a supplementary dimension of the target feature data and is input into the moisture prediction model along with process parameters, equipment parameters, and environmental parameters to generate the final moisture prediction result.

[0051] Please see Figure 3 Optionally, the training process of the moisture prediction model can be implemented by steps 121-124.

[0052] Step 121: Divide the sample dataset into a training set and a test set according to a preset ratio.

[0053] The training set is used for model fitting and training, while the test set is used to evaluate the model's generalization ability. Preset ratios can be 9:1, 8:2, 7:3, or 6:4, etc.

[0054] Specifically, when the total number of samples is relatively sufficient, a ratio of 9:1 or 8:2 can be used to ensure that the training set contains enough samples for the model to learn fully; when the total number of samples is relatively limited, a ratio of 7:3 can be used to reserve more test samples while ensuring the number of training samples to more accurately evaluate the model's generalization ability.

[0055] The partitioning process can be performed using random sampling or stratified sampling to ensure that the distribution ratio of moisture content in the training and test sets is consistent with that of the original dataset, thus avoiding evaluation distortion caused by data distribution bias.

[0056] Step 122: Train the prediction model using the training set and evaluate the trained prediction model using the test set.

[0057] After partitioning the dataset, the pre-defined prediction model is iteratively trained using the feature data from the training set as input and the corresponding moisture label values ​​as output. During training, the loss value of the prediction model is recorded in real time, and the accuracy of the trained prediction model is evaluated using the test set. The purpose of the evaluation is to test the model's ability to predict unseen samples, i.e., its generalization performance.

[0058] As an optional implementation method, the evaluation process may use at least one of the following indicators: Mean Squared Error (MSE), Root Mean Squared Error (RMSE), Mean Absolute Error (MAE), Coefficient of Determination (R-Squared, R²), Maximum Absolute Error, Mean Absolute Percentage Error (MAPE), etc.

[0059] Among them, the mean square error is the average of the sum of squares of the differences between the predicted and actual values; the smaller the value, the higher the prediction accuracy. The root mean square error is the arithmetic square root of the mean square error, and its dimension is consistent with the original moisture value, making it easier to intuitively understand the magnitude of the error. The mean absolute error is the average of the absolute values ​​of the differences between the predicted and actual values; it is not sensitive to outliers and can reflect the typical level of prediction error. The coefficient of determination usually ranges from 0 to 1; the closer the value is to 1, the stronger the model's explanatory power for moisture changes. The maximum absolute error represents the extreme case of prediction deviation for a single sample and is used to evaluate the model's performance in the worst case. The mean absolute percentage error is expressed as a percentage of the deviation of the prediction error from the actual value, which is convenient for cross-dimensional comparison.

[0060] The preset prediction model can be an initial model built based on any one or more machine learning algorithms. These algorithms include, but are not limited to, at least one of the following: neural network models, random forest models, decision tree (DT) models, Naive Bayes (NB) models, clustering models, support vector machine (SVM) models, long short-term memory (LSTM) network models, and recurrent neural network (RNN) models. One or more machine learning algorithms can be selected for modeling based on business needs and data characteristics; there are no specific limitations.

[0061] It should be understood that the above embodiments are merely illustrative examples of this application and are not intended to limit the scope of protection of this application. Inspired by the technical concept of this application, those skilled in the art can conceive of various alternative implementation methods, all of which should fall within the scope of protection of this application. For example, machine learning algorithms are not limited to the types listed above; other supervised learning regression algorithms can also be used, such as Support Vector Regression (SVR), Lightweight Gradient Boosting Machine (LightGBM), CatBoost, etc.

[0062] Step 123: Optimize the parameters of the moisture prediction model based on the evaluation results. The optimization method includes one of the following: grid search, random search, Bayesian optimization, and cross-validation.

[0063] Specifically, the test evaluation results will be compared with preset accuracy requirements, which can be set according to actual production needs. For example, R² can be set to be no less than 0.85, MAPE no more than 5%, or MAE no more than a certain absolute moisture value (such as 20 ppm). If the evaluation results show that the performance of the prediction model does not meet the preset accuracy requirements, the parameters of the prediction model will be tuned.

[0064] Parameter tuning can employ at least one of the following methods: grid search, random search, Bayesian optimization, or manual tuning of specific parameters. Grid search predefines a set of candidate values ​​for each hyperparameter to be tuned, iterates through all combinations of candidate values ​​for training and evaluation, and selects the hyperparameter combination that optimizes the evaluation metrics. Random search randomly samples a set of hyperparameter combinations in the hyperparameter space according to a predefined probability distribution for training and evaluation, and selects the optimal combination within a given search budget. Bayesian optimization constructs a surrogate probability model between hyperparameters and model performance based on historical evaluation results, using a sampling function to guide the selection of the next set of hyperparameters, approximating the global optimum in fewer iterations, and is suitable for scenarios where single training sessions are time-consuming. Manual tuning of specific parameters involves adjusting specific hyperparameters based on business experience and model performance, such as increasing regularization coefficients for overfitting problems or increasing model complexity for underfitting problems.

[0065] During parameter tuning, the model must be re-evaluated after each adjustment. The tuning process terminates when the evaluation metrics meet the preset accuracy requirements. If the accuracy requirements cannot be met after multiple tunings, consider changing the prediction algorithm, expanding the sample data volume, or re-performing feature engineering.

[0066] Step 124: Select the moisture prediction model that meets the preset accuracy requirements as the moisture prediction model.

[0067] Once the performance of the prediction model meets the preset accuracy requirements, the trained model parameters, model structure, and necessary preprocessing logic are persistently saved as a model file. For example, the pickle or joblib libraries in Python can be used for serialization and saving, or common model exchange formats such as ONNX and PMML can be used for export to facilitate cross-platform deployment and use. The saved moisture prediction model can be deployed to servers in the production environment, edge computing devices, or cloud platforms for online prediction at any time.

[0068] If the performance of the prediction model fails to meet the preset accuracy requirements after evaluation, return to step 122 for retraining until the model accuracy meets the requirements. At the same time, the trained model is stored and synchronized to the production line execution terminal.

[0069] In this way, by dividing the sample dataset into training and test sets according to a preset ratio, the prediction model is trained using the training set and independently evaluated using the test set. This allows the model to test its generalization ability with unseen samples during the training phase, effectively avoiding overfitting or underfitting problems caused by improper data partitioning. Based on this, the parameters of the prediction model are fine-tuned according to the evaluation results until the preset accuracy requirements are met. This enables the final moisture prediction model to output stable and reliable high-precision prediction results of cell moisture content in the actual production environment, reducing the adverse impact of prediction deviations caused by insufficient model performance on subsequent cell sorting and process decisions.

[0070] It should be understood that the above description of the training process is merely illustrative and is not intended to limit the scope of this application. In practical applications, those skilled in the art can make adjustments according to specific needs.

[0071] In an alternative implementation, the moisture prediction model can also employ a hybrid modeling architecture. This involves not only relying on data-driven machine learning algorithms but also coupling the model with battery electrochemical mechanisms, thermodynamic principles, or moisture diffusion mechanism formulas. For example, the differential equation for moisture diffusion inside the battery cell under vacuum heating conditions can be embedded as a constraint in the loss function of the neural network model. Alternatively, a serial coupling approach can be used, where the mechanistic model first outputs the moisture prediction value, and then the machine learning model corrects the residuals of the mechanistic model. This approach has a more rigorous theoretical foundation, stronger model interpretability, and can compensate for the shortcomings of purely data-driven models in scenarios with small samples or sparse data.

[0072] In addition, the above-mentioned training model transmits the original production data to a unified central server for model training. When a power battery manufacturer has multiple production plants in multiple locations, if each plant does not wish to transmit the original production data to a unified central server for model training for reasons such as data security or privacy protection, federated learning or distributed training architecture can be used to build the moisture prediction model.

[0073] Please see Figure 4 Optionally, after the moisture prediction model has been trained and meets preset accuracy requirements, the method may further include a systematic evaluation of the model's performance and a visualization and storage step for the evaluation results. Specifically, this includes: Step 125: Evaluate the performance metrics of the moisture prediction model. The performance metrics include at least one of the following: mean square error, root mean square error, mean absolute error, maximum absolute error, coefficient of determination, and mean absolute percentage error.

[0074] Step 126: Storage performance metrics and corresponding visualization charts. The visualization charts include at least one of the following: feature importance chart, prediction comparison chart, residual chart, scatter plot, and error distribution chart.

[0075] The feature importance plot displays the contribution of each input feature to the model's prediction results in the form of a bar chart, arranged from highest to lowest importance score. The prediction comparison plot simultaneously plots the actual moisture value and the model's predicted moisture value for each sample in the form of a line chart or scatter plot, with the horizontal axis representing the sample number and the vertical axis representing the moisture value. The residual plot, in the form of a scatter plot, shows the relationship between the prediction residual (actual value minus predicted value) and the predicted value or sample number for each sample. The predicted vs. actual value scatter plot uses the model's predicted moisture value on the horizontal axis and the actual measured moisture value on the vertical axis, with one scatter point corresponding to each sample. An ideal diagonal line with a slope of 1 is also plotted; the closer the scatter points are to this diagonal line, the more accurate the prediction. The error distribution plot displays the statistical distribution characteristics of the prediction error in the form of a histogram or density curve.

[0076] As an optional implementation, performance metric values ​​can be stored in a model evaluation record table in a relational database (such as MySQL or PostgreSQL). Table fields include, but are not limited to: model version number, evaluation time, metric name and corresponding value, and the time range covered by the training data. Visualization charts can be stored in a local file system, a network-attached storage device, or a cloud object storage service. The file naming convention can include the model version number and evaluation timestamp for subsequent retrieval and version comparison. It can also generate a model evaluation report, integrating performance metrics and visualization charts into a single document (such as PDF) for use in model review, version release, and quality audit.

[0077] In scenarios involving continuous iterative modeling, multiple versions of the model can be evaluated separately, and their respective performance metrics and visualizations can be stored. When a newly trained model version is compared with an already deployed online model version, the performance metrics of the two can be compared to determine whether to perform a model replacement update.

[0078] In this way, by conducting multi-dimensional quantitative evaluation of the performance indicators after the moisture prediction model is trained, and storing the evaluated performance indicators and corresponding visualization charts, a comprehensive record of the model's prediction accuracy and stability is achieved. This not only provides a data foundation for model version management and continuous iterative optimization, but also enables process and quality management personnel to intuitively identify key characteristics affecting moisture through charts, grasp the distribution patterns and trends of prediction errors, and thus carry out targeted process parameter control and model tuning, ensuring that the moisture prediction model maintains high accuracy and high reliability in the actual production environment for a long time.

[0079] Please see Figure 5 Optionally, after generating the moisture prediction result for the target battery cell, the method of this application may further include the following application steps: Step 130: Based on the unique identifier of the target cell, trace and obtain the multi-dimensional feature data corresponding to the target cell in the manufacturing process; A unique identifier for a target battery cell refers to the identification code assigned to each cell during the production process, which can uniquely distinguish it from other cells. This unique identifier can be a cell barcode, QR code, RFID tag code, or a serial number automatically generated by a Manufacturing Execution System (MES). This unique identifier is used throughout the entire battery cell production process, and is bound to key data collection nodes at each stage, ensuring that the cell's process history data can be fully traced using this unique identifier.

[0080] For example, a join query can be executed in a relational database using SQL statements, or data can be retrieved from multiple systems and aggregated in memory using an application programming interface (API). The query results are then arranged sequentially by process order or categorized by parameter type to form a complete process parameter file for the target battery cell, which can be used for subsequent dimension-by-dimensional judgment.

[0081] Step 140: Determine whether the multidimensional feature data of each dimension is within the range of the standard parameters corresponding to that dimension; Specifically, after obtaining the complete process parameter file, the value of the feature data for each dimension can be determined to be within the range of the preset standard parameters corresponding to that dimension.

[0082] The standard parameter range can be preset based on at least one of the following methods: process specification documents, historical statistical patterns, equipment manufacturer recommended values, or experimental verification results.

[0083] Step 150: If there is target multidimensional feature data that is not within the corresponding standard parameter range, mark the target multidimensional feature data as abnormal parameters and record the unique identifier of the corresponding target cell.

[0084] The marking methods include, but are not limited to, data-level marking, interface display marking, and message warning marking. For example, adding an exception flag field to the corresponding data record in the database and assigning it the value "Exception" or "1"; After marking the abnormal parameters, record the unique identifier of the target cell corresponding to the target multidimensional feature data marked as abnormal parameters.

[0085] As an optional implementation, a list of abnormal battery cells can be generated and stored in a specified database table or exported as a spreadsheet file for subsequent quality traceability, batch locking, or product sorting.

[0086] Furthermore, the risk level of the target battery cell can be assessed based on the number and severity of abnormal parameters. For example, if one parameter slightly exceeds the standard range while all parameters strongly related to moisture are normal, the cell can be classified as low-risk and can be used normally but requires close monitoring. If multiple key parameters strongly related to moisture exceed the standard range simultaneously, the cell can be classified as high-risk, and a diversion command for the cell will be automatically triggered.

[0087] Thus, by generating the moisture prediction results of the target battery cell, and tracing back to the unique identifier of the target battery cell to obtain multi-dimensional characteristic data such as process parameters, equipment parameters, and environmental parameters corresponding to the battery cell in the entire process, and judging whether each characteristic data is within the preset standard parameter range, abnormal parameters that exceed the range are marked and the corresponding unique identifier of the battery cell are recorded, realizing automatic and precise location of the root cause of moisture anomalies at the battery cell level. Compared with the traditional offline sampling inspection method that can only detect excessive moisture but cannot quickly pinpoint the cause, this solution allows process and quality management personnel to directly locate the specific abnormal process, abnormal parameter item, and its deviation degree without manual inspection of each step, significantly shortening the response cycle of anomaly analysis. At the same time, by associating and recording abnormal parameters with the unique identifier of the battery cell, a complete abnormal battery cell file can be formed, providing accurate data support for subsequent batch identification, product sorting, process optimization, and supplier quality traceability, effectively reducing the risk of moisture consistency fluctuations caused by implicit deviations in process parameters, and ensuring the quality stability and yield level of power battery production.

[0088] In an alternative implementation, based on the aforementioned abnormal parameter tracing and marking, the function of AI automatically recommending process correction values ​​can be further added. That is, after identifying abnormal parameters and their degree of deviation, the system automatically calculates and recommends the process correction measures to be taken based on a preset process adjustment rule library or reinforcement learning algorithm (such as suggesting to extend the baking and heat preservation time by several minutes, increase the baking temperature by several degrees Celsius, or increase the number of air exchanges), and directly sends the correction instructions to the corresponding production equipment for adjustment through the equipment interface.

[0089] All of the above technical solutions can be combined in any way to form optional embodiments of this application, and will not be described in detail here.

[0090] To facilitate better implementation of the power battery moisture prediction method of this application embodiments, this application also provides a moisture prediction system. Please refer to... Figure 6 , Figure 6 This is a schematic diagram of the structure of a moisture prediction system provided in an embodiment of this application. The moisture prediction system 200 may include: The acquisition module 210 is used to acquire at least one target feature data of the target cell before entering the vacuum baking process, during the vacuum baking process, and after completing the vacuum baking process.

[0091] The prediction module 220 is used to input the target feature data into the moisture prediction model to generate the moisture prediction result of the target battery cell. The moisture prediction model is trained by a sample dataset constructed from multi-dimensional feature data in the power battery manufacturing process. The multi-dimensional feature data includes at least process parameters, equipment parameters and environmental parameters.

[0092] Thus, the aforementioned moisture prediction system 200 constructs a prediction model by collecting multi-dimensional characteristic data on the process, equipment, and environment of the power battery production process. It can predict the moisture content of the target battery cell at multiple stages before, during, and after vacuum baking. It can achieve a comprehensive assessment of the moisture content of the entire battery cell without disassembling the cell. This avoids the damage to the battery cell caused by traditional sampling inspection methods and solves the problems of offline detection lag and inability to control the process. At the same time, it can correlate and analyze the influencing factors of multiple processes, accurately locate the source of moisture fluctuations, improve the level of precision in production control, effectively adapt to the needs of power batteries, and ensure the performance and production safety of power batteries.

[0093] Each unit in the aforementioned moisture prediction system can be implemented entirely or partially through software, hardware, or a combination thereof. These units can be embedded in or independent of the processor within the electronic device, or stored in the memory of the electronic device in software form, so that the processor can call and execute the corresponding operations of each unit.

[0094] The moisture prediction system 200 can be integrated into a terminal or server that has storage and a processor and thus computing power, or the moisture prediction system 200 can be the terminal or server.

[0095] Optionally, this application also provides an electronic device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described method embodiments.

[0096] Figure 7 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. The electronic device may be a terminal or a server. Figure 7As shown, the electronic device 300 includes a processor 301 with one or more processing cores, a memory 302 with one or more computer-readable storage media, and a computer program stored in the memory 302 and executable on the processor. The processor 301 and the memory 302 are electrically connected. Those skilled in the art will understand that the electronic device structure shown in the figures does not constitute a limitation on the electronic device, and may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0097] The processor 301 is the control center of the electronic device 300. It connects various parts of the electronic device 300 through various interfaces and lines. By running or loading software programs and / or modules stored in the memory 302, and calling data stored in the memory 302, it executes various functions of the electronic device 300 and processes data, thereby performing overall processing of the electronic device 300.

[0098] In this embodiment, the processor 301 in the electronic device 300 loads the instructions corresponding to the processes of one or more computer programs into the memory 302 according to the following steps, and the processor 301 runs the computer programs stored in the memory 302 to realize various functions: Acquire target characteristic data of the target cell before entering the vacuum baking process, during the vacuum baking process, and / or after completing the vacuum baking process; The target feature data is input into the moisture prediction model to generate the moisture prediction result of the target battery cell. The moisture prediction model is trained by a sample dataset constructed from multi-dimensional feature data in the power battery manufacturing process. The multi-dimensional feature data includes at least process parameters, equipment parameters and environmental parameters.

[0099] For details on the implementation of each of the above operations, please refer to the previous examples, which will not be repeated here.

[0100] Optional, such as Figure 7 As shown, the electronic device 300 also includes: a display screen 303, a radio frequency circuit 304, an audio circuit 305, an input unit 306, and a power supply 307. The processor 301 is electrically connected to the display screen 303, the radio frequency circuit 304, the audio circuit 305, the input unit 306, and the power supply 307. Those skilled in the art will understand that... Figure 7 The electronic device structure shown does not constitute a limitation on the electronic device and may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0101] The display screen 303 can be used to display a graphical user interface (GUI) and receive operation commands generated by the user interacting with the GUI. The display screen 303 may include a display panel and a touch panel. The display panel can be used to display information input by the user or information provided to the user, as well as various graphical user interfaces of the electronic device. These graphical user interfaces can be composed of graphics, text, icons, video, and any combination thereof. The touch panel can be used to collect touch operations performed by the user on or near it (such as operations performed by the user using a finger, stylus, or any suitable object or accessory on or near the touch panel), generate corresponding operation commands, and execute the corresponding program. Optionally, the touch panel may include a touch detection system and a touch controller. The touch detection system detects the user's touch location and the signal generated by the touch operation, and transmits the signal to the touch controller. The touch controller receives touch information from the touch detection system, converts it into touch point coordinates, sends it to the processor 301, and can receive and execute commands from the processor 301. The touch panel can cover the display panel. When the touch panel detects a touch operation on or near it, it transmits the information to the processor 301 to determine the type of touch event. Subsequently, the processor 301 provides corresponding visual output on the display panel according to the type of touch event. In this embodiment, the touch panel and the display panel can be integrated into the display screen 303 to achieve input and output functions. However, in some embodiments, the touch panel and the display screen 303 can be implemented as two independent components to achieve input and output functions. That is, the display screen 303 can also be used as part of the input unit 306 to achieve input functions.

[0102] The radio frequency circuit 304 can be used to transmit and receive radio frequency signals to establish wireless communication with network devices or other electronic devices, and to transmit and receive signals with network devices or other electronic devices.

[0103] Audio circuit 305 can be used to provide an audio interface between a user and an electronic device via a speaker or microphone. Audio circuit 305 converts received audio data into electrical signals, transmits them to the speaker, and the speaker converts them into sound signals for output. Conversely, the microphone converts collected sound signals into electrical signals, which are then received by audio circuit 305, converted back into audio data, and then processed by processor 301 before being transmitted via radio frequency circuit 304 to, for example, another electronic device, or output to memory 302 for further processing. Audio circuit 305 may also include an earphone jack to facilitate communication between peripheral headphones and electronic devices.

[0104] The input unit 306 can be used to receive input numbers, characters, or object feature information (such as fingerprints, irises, facial information, etc.), and to generate keyboard, mouse, joystick, optical, or trackball signal inputs related to user settings and function control.

[0105] Power supply 307 is used to supply power to various components of electronic device 300. Optionally, power supply 307 can be logically connected to processor 301 through a power management system, thereby enabling functions such as charging, discharging, and power consumption management through the power management system. Power supply 307 may also include one or more DC or AC power supplies, recharging systems, power fault detection circuits, power converters or inverters, power status indicators, and other arbitrary components.

[0106] although Figure 7 As not shown in the diagram, the electronic device 300 may also include a camera, sensor, wireless fidelity module, Bluetooth module, etc., which will not be described in detail here.

[0107] This application also provides a computer-readable storage medium for storing a computer program. This computer-readable storage medium can be applied to an electronic device, and the computer program causes the electronic device to execute the corresponding processes in the power battery moisture prediction method of the embodiments of this application; for the sake of brevity, these will not be elaborated further here.

[0108] This application also provides a computer program comprising computer instructions stored in a computer-readable storage medium. The processor of an electronic device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the electronic device to perform the corresponding process in the power battery moisture prediction method of this application. For the sake of brevity, further details are omitted here.

[0109] It should be understood that the processor in this application may be an integrated circuit chip with signal processing capabilities. In implementation, the steps of the above method embodiments can be completed by integrated logic circuits in the processor's hardware or by instructions in software form. The processor described above can be a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the methods, steps, and logic block diagrams disclosed in the embodiments of this application. The general-purpose processor can be a microprocessor or any conventional processor, etc. The steps of the power battery moisture prediction method disclosed in the embodiments of this application can be directly manifested as execution by a hardware decoding processor, or execution by a combination of hardware and software modules in the decoding processor. The software modules can be located in random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, or other mature storage media in the art. This storage medium is located in memory, and the processor reads information from the memory and, in conjunction with its hardware, completes the steps of the above method.

[0110] It is understood that the memory in the embodiments of this application can be volatile memory or non-volatile memory, or may include both volatile and non-volatile memory. The non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. The volatile memory can be random access memory (RAM), which is used as an external cache. By way of example, but not limitation, many forms of RAM are available, such as Static Random Access Memory (SRAM), Dynamic Random Access Memory (DRAM), Synchronous DRAM (SDRAM), Double Data Rate SDRAM (DDR SDRAM), Enhanced Synchronous DRAM (ESDRAM), Synchlink DRAM (SLDRAM), and Direct Rambus RAM (DR RAM). It should be noted that the memory used in the systems and methods described herein is intended to include, but is not limited to, these and any other suitable types of memory.

[0111] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0112] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, units, and processes described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0113] In the embodiments of this application, the terms "module" or "unit" refer to a computer program or part of a computer program that has a predetermined function and works with other related parts to achieve a predetermined goal, and can be implemented wholly or partially using software, hardware (such as processing circuitry or memory), or a combination thereof. Similarly, a processor (or multiple processors or memory) can be used to implement one or more modules or units. Furthermore, each module or unit can be part of an overall module or unit that includes the functionality of that module or unit.

[0114] In the embodiments provided in this application, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the system embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the mutual coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between systems or units may be electrical, mechanical, or other forms.

[0115] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0116] In addition, the functional units in this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0117] If a function is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause an electronic device (which may be a personal computer or a server) to execute all or part of the steps of the methods in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, ROM, RAM, magnetic disks, or optical disks.

[0118] The above are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for predicting the moisture content of a power battery, characterized in that, The moisture prediction method includes: Acquire at least one target characteristic data of the target cell before entering the vacuum baking process, during the vacuum baking process, and after completing the vacuum baking process; The target feature data is input into the moisture prediction model to generate the moisture prediction result of the target battery cell. The moisture prediction model is trained by a sample dataset constructed from multi-dimensional feature data in the power battery manufacturing process. The multi-dimensional feature data includes at least process parameters, equipment parameters, and environmental parameters.

2. The moisture prediction method according to claim 1, characterized in that, The moisture prediction method also includes: Based on the unique identifier of the target battery cell, the multi-dimensional feature data corresponding to the target battery cell in the manufacturing process are traced and obtained; Determine whether the multidimensional feature data for each dimension is within the range of the standard parameters corresponding to that dimension; If there is target multidimensional feature data that is not within the corresponding standard parameter range, the target multidimensional feature data is marked as an abnormal parameter, and the unique identifier of the corresponding target cell is recorded.

3. The moisture prediction method according to claim 1, characterized in that, The power battery manufacturing process includes at least one of the following steps: feeding, mixing, coating, rolling and slitting, winding or stacking, welding, helium testing or vacuum baking.

4. The moisture prediction method according to claim 1, characterized in that, The training process of the moisture prediction model includes: The sample dataset is divided into a training set and a test set according to a preset ratio; The training set is used to train the preset prediction model, and the test set is used to evaluate the trained prediction model. Based on the evaluation results, the parameters of the prediction model are tuned. The tuning method includes one of the following: grid search, random search, Bayesian optimization, and cross-validation. The prediction model that meets the preset accuracy requirements is used as the moisture prediction model.

5. The moisture prediction method according to claim 4, characterized in that, The moisture prediction method also includes: Evaluate the performance metrics of the moisture prediction model, wherein the performance metrics include at least one of the following: mean square error, root mean square error, mean absolute error, maximum absolute error, coefficient of determination, and mean absolute percentage error. The performance metrics and their corresponding visualization charts are stored. The visualization charts include at least one of the following: feature importance chart, prediction comparison chart, residual chart, scatter plot, and error distribution chart.

6. The moisture prediction method according to claim 1, characterized in that, The moisture prediction model is built based on machine learning algorithms, which include at least one of the following: neural network model, tree model, Bayesian model, clustering model, support vector machine model, recurrent neural network model, and long short-term memory network model.

7. The moisture prediction method according to claim 1, characterized in that, The moisture prediction model is used to predict the moisture content of at least one of the following cell types: square laminated cells, square wound cells, and cylindrical cells.

8. The moisture prediction method according to claim 1, characterized in that, The moisture prediction model is used to predict at least one of the moisture content of the positive electrode of the battery cell, the moisture content of the negative electrode of the battery cell, or the moisture content of the separator.

9. The moisture prediction method according to any one of claims 1 to 8, characterized in that, The moisture prediction result of the target battery cell is the moisture content of the target battery cell after the vacuum baking process is completed.

10. A moisture prediction system for a power battery, characterized in that, include: The acquisition module is used to acquire at least one target feature data of the target cell before entering the vacuum baking process, during the vacuum baking process, and after completing the vacuum baking process; The prediction module is used to input the target feature data into the moisture prediction model to generate the moisture prediction result of the target battery cell. The moisture prediction model is trained by a sample dataset constructed from multi-dimensional feature data in the power battery manufacturing process. The multi-dimensional feature data includes at least process parameters, equipment parameters, and environmental parameters.

11. The moisture prediction system according to claim 10, characterized in that, It also includes an exception management module, which is used for: Based on the unique identifier of the target battery cell, the multi-dimensional feature data corresponding to the target battery cell in the manufacturing process is retrieved retrospectively. Determine whether the multidimensional feature data for each dimension is within the range of the standard parameters corresponding to that dimension; If there is target multidimensional feature data that is not within the corresponding standard parameter range, the target multidimensional feature data is marked as an abnormal parameter, and the unique identifier of the corresponding target cell is recorded.

12. An electronic device, characterized in that, It includes a processor, a memory, and a computer program, wherein the computer program is stored in the memory, and when the computer program is executed by the processor, the processor performs the moisture prediction method for the power battery according to any one of claims 1-9.

13. A non-volatile computer-readable storage medium containing a computer program, characterized in that, When the computer program is executed by the processor, the processor performs the moisture prediction method for the power battery according to any one of claims 1-9.

Images