Method and system for predicting air tightness of battery pack box body

Abstract

The invention relates to the technical field of batteries, and discloses a battery pack box air tightness prediction method and system. The battery pack box airtightness prediction method comprises the steps of performing airtightness detection on a to-be-detected battery pack box, recording process data of the whole airtightness detection, and extracting airtightness characteristic parameters from the process data; inputting the air tightness characteristic parameters into a pre-trained air tightness prediction model, wherein the air tightness prediction model outputs an air tightness prediction probability value; comparing the airtight prediction probability value with a preset prediction probability threshold value; when the airtight prediction probability value is greater than the prediction probability threshold value, judging that the airtight state of the to-be-detected battery pack box body is good; otherwise, sending out airtight state early warning. The method focuses on the pressure change data in the air inflation process of the airtight instrument, information related to the airtight state in the data is deeply mined through multi-aspect feature extraction, and compared with a traditional detection mode of only paying attention to pressure attenuation, the method can more comprehensively reflect the airtight characteristics of the box body.

Description

Technical Field

[0001] This invention relates to the field of battery technology, and in particular to a method and system for predicting the airtightness of a battery pack housing. Background Technology

[0002] In the manufacturing and daily operation and maintenance of energy storage battery packs, the airtightness testing of the enclosure is a crucial step in ensuring battery safety and reliability. Traditional airtightness testing relies on an airtightness tester to pressurize the battery pack enclosure and judge the airtightness by observing the degree of pressure drop. Figure 1 As shown.

[0003] However, existing methods are mostly post-event detections, only able to determine the current airtightness and unable to effectively predict future airtightness trends. With the large-scale application of energy storage systems, battery packs are subject to vibration, temperature changes, and material aging during long-term operation, leading to a gradual deterioration in airtightness. If predictive models could be built based on pressure change data during the airtightness testing process, identifying potential airtightness degradation risks in advance, the operational efficiency and safety of energy storage systems could be significantly improved. However, current technologies have not fully explored the potential value of airtightness testing data and lack efficient and accurate airtightness prediction methods. Summary of the Invention

[0004] Based on this, the purpose of this invention is to provide a method and system for predicting the airtightness of a battery pack enclosure. By analyzing the dynamic characteristics of pressure changes during the inflation process of the airtightness detector, the airtightness status of the battery pack enclosure can be predicted in advance, so that maintenance personnel can take timely measures to reduce safety hazards and improve the operational reliability of the energy storage system.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: This invention provides a method for predicting the airtightness of a battery pack enclosure, comprising the following steps: S1. Perform airtightness testing on the battery pack housing under test and record the entire airtightness testing process data, and extract airtightness characteristic parameters from the process data; S2. Input the airtightness feature parameters into the pre-trained airtightness prediction model, and the airtightness prediction model outputs the airtightness prediction probability value; S3. Compare the airtightness prediction probability value with a preset prediction probability threshold; when the airtightness prediction probability value is greater than the prediction probability threshold, determine that the airtightness of the battery pack under test is good; otherwise, issue an airtightness warning.

[0006] As a further improvement to the above-mentioned solution of the present invention, in step S1, the air tightness test is performed by using an air tightness tester to inflate the battery pack housing to a set pressure value, then holding the pressure for a set time before depressurizing; the process data includes pressure data and time data, and the pressure data of the battery pack housing during the air tightness test is collected and recorded by the air tightness tester at a predetermined frequency.

[0007] As a further improvement to the above-described solution of the present invention, step S1, extracting airtightness characteristic parameters from the process data, includes basic feature extraction: Extract the average pressure change rate k1 of the battery pack casing during the inflation phase: k1 = (P 充气结束 -P 充气开始 ) / (t 充气结束 -t 充气开始 ), where P 充气开始 P 充气结束 These represent the pressure values ​​at the start and end of the inflation phase, respectively, t. 充气开始 t 充气结束 These are the start and end times of the inflation phase, respectively. Extract the inflation time t required for the internal pressure of the battery pack under test to reach the predetermined pressure value. 充气 ; Extract the pressure fluctuation amplitude of the battery pack casing during the entire pressure holding phase: P 保压max -P 保压min ; Extract the average pressure change rate k2 of the battery pack casing during the depressurization phase: k2 = (P 泄压结束 -P 泄压开始 ) / (t 泄压结束 -t 泄压开始 ), where P 泄压开始 P 泄压结束 These represent the pressure values ​​at the start and end of the depressurization phase, respectively, t. 泄压开始 t 泄压结束 These are the start and end times of the pressure relief phase, respectively.

[0008] As a further improvement to the above-mentioned solution of the present invention, the extraction of airtightness characteristic parameters from the airtightness detection process data also includes high-order feature extraction: calculating the root mean square value, kurtosis and skewness of the pressure data during the pressure holding stage, and using fast Fourier transform to perform frequency domain analysis on the pressure fluctuation amplitude during the pressure holding stage to extract the main frequency components and corresponding amplitudes.

[0009] As a further improvement to the above-mentioned solution of the present invention, in step S2, the method for constructing the air tightness prediction model is as follows: the air tightness prediction model is constructed using the gradient boosting decision tree algorithm, and the air tightness prediction model is optimized by combining the XGBoost algorithm.

[0010] As a further improvement to the above-mentioned solution of the present invention, in step S2, the training method of the airtightness prediction model is as follows: Obtain historical airtightness test data of battery pack enclosures of the same type, and divide the historical airtightness test data into training set, verification set and test set according to a preset ratio; The airtightness prediction model is iteratively trained and its parameters are adjusted using the training set. The hyperparameters of the airtightness prediction model are then fine-tuned using the validation set. Finally, the performance of the airtightness prediction model is evaluated using the test set.

[0011] As a further improvement to the above-mentioned solution of the present invention, the preset ratio is 6-8:1-2:1-2; the model hyperparameters include the maximum depth of the gradient boosting decision tree, the learning rate, and the subsampling rate.

[0012] As a further improvement to the above-mentioned solution of the present invention, the battery pack housing airtightness prediction method further includes the following steps: S4. Obtain the predicted air tightness probability values ​​of the same battery pack housing at multiple detection time points within its life cycle and perform time series analysis to predict the future air tightness status change trend of the battery pack housing; when the predicted change trend shows that the air tightness status of the battery pack housing continues to decline, generate a leak location prediction and maintenance suggestions.

[0013] The present invention also provides a battery pack housing airtightness prediction system, comprising: The data acquisition and processing module performs airtightness testing on the battery pack housing under test and records the entire airtightness testing process data, and extracts airtightness characteristic parameters from the process data; The prediction model module is used to input the airtightness characteristic parameters and output the airtightness prediction probability value; The early warning module is used to compare the air tightness prediction probability value with a preset prediction probability threshold; when the air tightness prediction probability value is greater than the prediction probability threshold, it is determined that the air tightness of the battery pack under test is good; otherwise, an air tightness warning is issued.

[0014] The present invention also provides a computer-readable storage medium comprising a stored computer program, wherein, when the computer program is executed, it controls the device containing the computer-readable storage medium to perform the battery pack enclosure airtightness prediction method as described above.

[0015] Compared with the prior art, the present invention has the following beneficial effects: This invention focuses on pressure change data during the inflation process of an airtightness meter. Through multi-faceted feature extraction, it deeply mines information related to the airtightness status within the data. Compared to traditional detection methods that only focus on pressure decay, this invention provides a more comprehensive reflection of the airtightness characteristics of the enclosure. This invention not only predicts the current airtightness status but also performs trend analysis based on prediction results at different time points, proactively anticipating the risk of airtightness performance degradation. It transforms passive detection into proactive prevention, reducing the probability of energy storage system failures due to airtightness issues, thereby improving product quality and economic benefits. Attached Figure Description

[0016] Figure 1 The test steps and pressure trend change graph for testing the air tightness of battery pack housing using a traditional air tightness tester; Figure 2 A flowchart of a method for predicting the airtightness of a battery pack enclosure is provided in this embodiment of the invention; Figure 3 This is a pressure change curve of the battery pack housing during the airtightness test in an embodiment of the present invention; Figure 4 This is a flowchart illustrating the pressure data extraction process of the battery pack housing during airtightness testing in an embodiment of the present invention. Figure 5 This is a partial structural schematic diagram of a battery pack enclosure airtightness prediction system provided in an embodiment of the present invention. Detailed Implementation

[0017] To facilitate understanding of the present invention, a more comprehensive description will be given below with reference to specific embodiments. However, the present invention can be implemented in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of the disclosure of the present invention.

[0018] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention.

[0019] Reference Figure 2 This embodiment proposes a method for predicting the airtightness of a battery pack housing, which includes the following steps: S1. Perform airtightness testing on the battery pack casing and record the entire airtightness testing process data, extracting airtightness characteristic parameters from the process data.

[0020] like Figure 3As shown, an airtightness tester is used to inflate the battery pack casing to a set pressure value, hold the pressure for a set time, and then depressurize. When performing airtightness testing on the battery pack casing, ensure that the airtightness tester is sealed to the casing.

[0021] It is important to note that when using this method to test the battery pack enclosure, it is necessary to ensure that the same battery pack enclosure is in the same testing environment. Temperature and humidity sensors can be deployed simultaneously to collect temperature and humidity data of the testing environment in real time. Atmospheric pressure data of the environment can be obtained through a barometric pressure sensor. The environmental parameters and the pressure data of the airtightness tester can be time-stamped to construct a multi-dimensional data set.

[0022] The process data includes pressure data and time data. The airtightness tester collects and records the pressure data of the battery pack housing during the airtightness test at a predetermined frequency (e.g., 100Hz).

[0023] In this embodiment, the pressure value and holding time should be set reasonably according to the specific specifications of the battery pack housing.

[0024] Extracting airtightness characteristic parameters from process data includes basic feature extraction and higher-order feature extraction.

[0025] Basic feature extraction: Extract the average pressure change rate k1 of the battery pack casing during the inflation phase: k1 = (P 充气结束 -P 充气开始 ) / (t 充气结束 -t 充气开始 ), where P 充气开始 P 充气结束 These represent the pressure values ​​at the start and end of the inflation phase, respectively, t. 充气开始 t 充气结束 These are the start and end times of the inflation phase, respectively. Extract the inflation time t required for the internal pressure of the battery pack under test to reach the predetermined pressure value. 充气 ; Extract the pressure fluctuation range of the battery pack casing during the entire pressure holding phase: P 保压max -P 保压min ; Extract the average pressure change rate k2 of the battery pack casing during the depressurization phase: k2 = (P 泄压结束 -P 泄压开始 ) / (t 泄压结束 -t 泄压开始 ), where P 泄压开始 P 泄压结束 These represent the pressure values ​​at the start and end of the depressurization phase, respectively, t. 泄压开始 t 泄压结束 These are the start and end times of the pressure relief phase, respectively.

[0026] High-order feature extraction: The root mean square (RMS), kurtosis, and skewness of the pressure data during the pressure holding phase are calculated. A Fast Fourier Transform (FFT) is then used to perform frequency domain analysis on the pressure fluctuation amplitude during the pressure holding phase, extracting the main frequency components and their corresponding amplitudes. The RMS value characterizes the overall energy level of the pressure fluctuation during the pressure holding phase; kurtosis characterizes the steepness of the pressure distribution relative to a normal distribution during the pressure holding phase; and skewness characterizes the direction and degree of asymmetry in the pressure distribution during the pressure holding phase. Simultaneously, a Fast Fourier Transform is used to perform frequency domain analysis on the pressure fluctuation signal during the pressure holding phase, converting the time-domain pressure data to the frequency domain and extracting the top N main frequency components with the highest energy proportions and their corresponding amplitudes as frequency domain features, where N is a preset positive integer.

[0027] It should be noted that, as Figure 4 As shown, to ensure data accuracy and prevent random occurrences, the same battery pack casing needs to be tested following the above steps to obtain multiple sets of data. The average of these multiple tests will be used as the basis for subsequent steps. Selecting time points is to collect data on the pattern of pressure changes over time. The slope at different time intervals is also important: for example, the slope at 20s, 30s, and 60s of inflation.

[0028] S2. Input the airtightness feature parameters into the pre-trained airtightness prediction model, and the airtightness prediction model outputs the airtightness prediction probability value.

[0029] In this embodiment, a gradient boosting decision tree (GBDT) algorithm is used to construct an airtightness prediction model, which is then optimized using the XGBoost algorithm. GBDT can effectively handle nonlinear data, has a strong ability to rank feature importance, and is suitable for predicting airtightness based on multi-dimensional pressure features. Simultaneously, the XGBoost algorithm is used to optimize the model, improving training efficiency and prediction accuracy.

[0030] The training method for the airtightness prediction model is as follows: Acquire a large amount of historical airtightness test data for similar battery pack enclosures, and divide the historical airtightness test data into training set, validation set, and test set according to a preset ratio; The airtightness prediction model is iteratively trained and its parameters are adjusted using the training set. The hyperparameters of the airtightness prediction model are then fine-tuned using the validation set. Finally, the performance of the airtightness prediction model is evaluated using the test set.

[0031] It should be noted that the aforementioned historical airtightness test data covers the test results of energy storage battery pack enclosures under different service years and operating conditions. During the training process, historical pressure data is used as input, and historical actual airtightness test results (pass / fail) are used as labels. The model parameters are adjusted through iterative training; the hyperparameters are fine-tuned using the validation set; and finally, the model performance is evaluated using the test set.

[0032] In this embodiment, the preset ratio is 7:2:1; the hyperparameters include the maximum depth of the gradient boosting decision tree, the learning rate, and the subsampling rate.

[0033] S3. Compare the air tightness prediction probability value with the preset prediction probability threshold; when the air tightness prediction probability value is greater than the prediction probability threshold, determine that the air tightness of the battery pack under test is good; otherwise, issue an air tightness warning.

[0034] In this embodiment, the prediction probability threshold is a threshold based on historical data statistics. Using the historical airtightness test data of a large number of similar battery pack boxes as samples, the trained airtightness prediction model is used to predict the airtightness of these samples, obtaining the airtightness prediction probability value for each sample. Distribution maps of the airtightness prediction probability values ​​for qualified and unqualified samples are then plotted. The prediction probability threshold can be selected as the probability value that minimizes the total classification error rate (e.g., misclassifying qualified as unqualified, or misclassifying unqualified as qualified), or it can be determined according to actual quality control requirements (e.g., a greater emphasis on preventing missed detections).

[0035] S4. Obtain the air tightness prediction probability values ​​of the same battery pack housing at multiple detection time points (e.g., prediction results of the past 10 detections) within its life cycle and perform time series analysis. Use algorithms such as moving average and exponential smoothing to predict the future air tightness status change trend of the battery pack housing. When the predicted change trend shows that the air tightness status of the battery pack housing continues to decline, generate a leak location prediction and maintenance suggestions.

[0036] In this embodiment, the possible leak location is determined by combining a large amount of historical poor airtightness data. Maintenance recommendations are given based on the predicted leak location, such as local sealing treatment or replacement of seals, providing a basis for decision-making by maintenance personnel.

[0037] Reference Figure 5 This embodiment also provides a battery pack enclosure airtightness prediction system, which employs the battery pack enclosure airtightness prediction method as described above, including: The data acquisition and processing module performs airtightness testing on the battery pack enclosure under test and records the entire process data, extracting airtightness characteristic parameters from the process data. It mainly includes a data acquisition module and a data processing module. The data acquisition module includes pressure sensors, temperature sensors, and humidity sensors. The pressure sensor is the core data entry point for the airtightness prediction scheme: it directly collects the pressure value inside the enclosure, while the temperature and humidity sensors help eliminate environmental interference. The data processing module is the execution layer of the airtightness prediction scheme, mainly including a PLC controller, a digital-to-analog converter module, and relays. The PLC controller executes the inflation control and pressure holding timing in the scheme and automatically triggers the acquisition of pressure data. The digital-to-analog converter module converts the analog signals from the pressure sensors into digital signals, supporting the plotting of pressure change curves. The relays, in conjunction with the PLC controller, realize the start and stop of the inflation / pressure holding equipment. When the pressure value does not meet the standard ("less than the preset pressure value" in the scheme), the testing process can be automatically terminated.

[0038] The prediction model module is used to input airtightness characteristic parameters and output airtightness prediction probability values; The early warning module is used to compare the air tightness prediction probability value with the preset prediction probability threshold. When the air tightness prediction probability value is greater than the prediction probability threshold, it is determined that the air tightness of the battery pack under test is good; otherwise, an air tightness warning is issued.

[0039] The transmission layer transmits the processed pressure data to the computer, providing a platform for pressure feature extraction (such as calculating slope and drawing line graphs). In the application layer, the data storage unit retains the pressure data, supporting the traceability of airtightness test results; the visualization analysis platform presents time-pressure change curves, intuitively displaying airtightness performance (such as pressure drop rate), and finally outputs the airtightness test results in the solution.

[0040] This embodiment also provides a computer-readable storage medium, which includes a stored computer program, wherein, when the computer program is running, it controls the device where the computer-readable storage medium is located to execute the battery pack enclosure airtightness prediction method as described above.

[0041] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0042] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these all fall within the protection scope of the present invention. Therefore, the protection scope of this invention patent should be determined by the appended claims.

Claims

1. A method for predicting the airtightness of a battery pack casing, characterized in that, It includes the following steps: S1. Perform airtightness testing on the battery pack housing under test and record the entire airtightness testing process data, and extract airtightness characteristic parameters from the process data; S2. Input the airtightness feature parameters into the pre-trained airtightness prediction model, and the airtightness prediction model outputs the airtightness prediction probability value; S3. Compare the airtightness prediction probability value with a preset prediction probability threshold; when the airtightness prediction probability value is greater than the prediction probability threshold, determine that the airtightness of the battery pack under test is good; otherwise, issue an airtightness warning.

2. The method for predicting the airtightness of a battery pack casing according to claim 1, characterized in that, In step S1, the airtightness test involves using an airtightness tester to inflate the battery pack housing to a set pressure value, holding the pressure for a set time, and then depressurizing. The process data includes pressure data and time data. The airtightness tester collects and records the pressure data of the battery pack housing during the airtightness test at a predetermined frequency.

3. The method for predicting the airtightness of a battery pack casing according to claim 2, characterized in that, In step S1, extracting airtightness characteristic parameters from the process data includes basic feature extraction: extract The average pressure change rate k1 of the battery pack casing during the inflation phase is: k1=(P 充气结束 -P 充气开始 ) / (t 充气结束 -t 充气开始 ), where P 充气开始 P 充气结束 These represent the pressure values ​​at the start and end of the inflation phase, respectively, t. 充气开始 t 充气结束 These are the start and end times of the inflation phase, respectively. Extract the inflation time t required for the internal pressure of the battery pack under test to reach the predetermined pressure value. 充气 ; Extract the pressure fluctuation amplitude of the battery pack casing during the entire pressure holding phase: P 保压max -P 保压min ; Extract the average pressure change rate k2 of the battery pack casing during the depressurization phase: k2 = (P 泄压结束 -P 泄压开始 ) / (t 泄压结束 -t 泄压开始 ), where P 泄压开始 P 泄压结束 These represent the pressure values ​​at the start and end of the depressurization phase, respectively, t. 泄压开始 t 泄压结束 These are the start and end times of the pressure relief phase, respectively.

4. The method for predicting the airtightness of a battery pack casing according to claim 3, characterized in that, Extracting airtightness characteristic parameters from the airtightness testing process data also includes high-order feature extraction: calculating the root mean square value, kurtosis, and skewness of the pressure data during the pressure holding stage, and using fast Fourier transform to perform frequency domain analysis on the pressure fluctuation amplitude during the pressure holding stage to extract the main frequency components and their corresponding amplitudes.

5. The method for predicting the airtightness of a battery pack casing according to claim 1, characterized in that, In step S2, the method for constructing the airtightness prediction model is as follows: the airtightness prediction model is constructed using the gradient boosting decision tree algorithm, and the airtightness prediction model is optimized by combining the XGBoost algorithm.

6. The method for predicting the airtightness of a battery pack casing according to claim 1, characterized in that, In step S2, the training method for the airtightness prediction model is as follows: Obtain historical airtightness test data of battery pack enclosures of the same type, and divide the historical airtightness test data into training set, verification set and test set according to a preset ratio; The airtightness prediction model is iteratively trained and its parameters are adjusted using the training set. The hyperparameters of the airtightness prediction model are then fine-tuned using the validation set. Finally, the performance of the airtightness prediction model is evaluated using the test set.

7. The method for predicting the airtightness of a battery pack casing according to claim 6, characterized in that, The preset ratio is 6-8:1-2:1-2; the model hyperparameters include the maximum depth of the gradient boosting decision tree, the learning rate, and the subsampling rate.

8. The method for predicting the airtightness of a battery pack casing according to claim 1, characterized in that, The method for predicting the airtightness of the battery pack enclosure also includes the following steps: S4. Obtain the predicted air tightness probability values ​​of the same battery pack housing at multiple detection time points within its life cycle and perform time series analysis to predict the future air tightness status change trend of the battery pack housing; when the predicted change trend shows that the air tightness status of the battery pack housing continues to decline, generate a leak location prediction and maintenance suggestions.

9. A battery pack housing airtightness prediction system, characterized in that, It includes: The data acquisition and processing module performs airtightness testing on the battery pack housing under test and records the entire airtightness testing process data, and extracts airtightness characteristic parameters from the process data; The prediction model module is used to input the airtightness characteristic parameters and output the airtightness prediction probability value; The early warning module is used to compare the air tightness prediction probability value with a preset prediction probability threshold; when the air tightness prediction probability value is greater than the prediction probability threshold, it is determined that the air tightness of the battery pack under test is good; otherwise, an air tightness warning is issued.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored computer program, wherein, when the computer program is executed, it controls the device containing the computer-readable storage medium to perform the battery pack enclosure airtightness prediction method as described in any one of claims 1-8.

Images