A lithium iron phosphate energy storage cell group packaging method based on consistency monitoring
Patent Information
- Application Number
- CN202410769191.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-06-14
- Publication Date
- 2026-09-25
- Estimated Expiration
- 2044-06-14
AI Technical Summary
其不足之处在于：1、只考虑到了电芯出厂时的一致性，并未考虑长期使用下的电芯一致性问题，使得磷酸铁锂储能电池在使用初期效果好，随着使用时间加长，电池衰减速度变快，无法发挥出磷酸铁锂储能电池结构稳定可还原性强的优势；2、对于一致性相差较大的电芯，并未给出合理的处理方式
[0042]该实施例的优点在于，针对一致性较差的电芯，根据未来要使用的场景，提供了一种筛选标准，该筛选标准可选出能满足特定使用场景的一致性较差电芯，提供了电芯的使用率，可有效降低电芯的生产成本。
Smart Images
Figure CN118606762B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of lithium iron phosphate energy storage battery technology, and in particular to a method for packaging lithium iron phosphate energy storage cells based on consistency monitoring. Background Technology
[0002] When lithium iron phosphate (LFP) is used as the cathode material in batteries, ferrous iron (Fe2+) participates in charge transfer. Ferrous iron readily reacts with oxygen to form ferric iron (Fe3+), requiring reduction during production. Over-reduction can leave residual elemental iron in the cathode material, leading to short circuits. Insufficient reduction reduces the proportion of LFP in the cathode material, resulting in insufficient battery capacity. This is due to the chemical properties of the material, and the problem cannot be completely solved. Therefore, LFP batteries inherently suffer from poor cell consistency. In practical use, poor cell consistency can lead to overcharging and over-discharging of some cells, as well as low energy utilization.
[0003] Currently, the resistance-capacitance sorting method is generally used during battery pack assembly, which calculates the discharge capacity and assigns cells with similar static blockage consistency. Its shortcomings are: 1. It only considers the consistency of cells at the factory, neglecting the consistency issues under long-term use. This results in lithium iron phosphate batteries performing well initially, but with prolonged use, the battery degradation rate accelerates, failing to fully utilize the advantages of the stable structure and strong reversibility of lithium iron phosphate batteries; 2. It does not provide a reasonable handling method for cells with significant differences in consistency. Summary of the Invention
[0004] This invention discloses a method for packaging lithium iron phosphate energy storage cells based on consistency monitoring. The specific method is as follows:
[0005] Construct a model for predicting the degradation rate of lithium iron phosphate battery cells;
[0006] The same charge and discharge test was conducted on several lithium iron phosphate cells to be packaged, and the charging capacity evaluation index of each lithium iron phosphate cell was calculated, and the degradation evaluation index of each lithium iron phosphate cell was predicted by the degradation rate prediction model.
[0007] Clustering algorithms were used to classify the charging capacity evaluation index and degradation evaluation index of lithium iron phosphate cells.
[0008] Based on the distance between each lithium iron phosphate cell and the center of its category, lithium iron phosphate cells of the same category that are less than the distance threshold are divided into the same group of packable cells.
[0009] Select all lithium iron phosphate cells whose distance is greater than the distance threshold to form a group to be divided;
[0010] Based on the usage scenario, the lithium iron phosphate cells in the subgroups are screened, and those that meet the requirements are retained.
[0011] Among the qualified lithium iron phosphate cells, the product of the charging capacity evaluation index and the first weight, plus the product of the attenuation evaluation index and the second weight, is used as the group score.
[0012] Select lithium iron phosphate cells whose group scores are within the preset range for grouping and packaging.
[0013] The advantage of this embodiment is that it takes into account the current capacity and degradation rate during packing, thus ensuring that the packed lithium iron phosphate batteries will not exhibit inconsistencies over a longer period, improving battery efficiency and product quality. A packing strategy is also provided for cells with poor consistency, increasing the utilization rate of good cells.
[0014] Furthermore, a model for predicting the degradation rate of lithium iron phosphate battery cells is constructed, and the specific method is as follows:
[0015] Construct a neural network prediction model;
[0016] Select brand-new sample cells for charge and discharge tests. Perform more than a preset number of charge and discharge cycles on the sample cells and record the amount of electricity required to fully charge the sample cells each time.
[0017] The rate of degradation of the sample cell is calculated by subtracting the amount of electricity required to fully charge a brand-new sample cell from the amount required for the last charge, and then dividing the result by the number of charge-discharge cycles.
[0018] Training samples are constructed using the amount of electricity required to fully charge a sample cell a predetermined number of times and the degradation rate of the sample cell. A neural network prediction model is trained using several training samples, and the trained neural network prediction model is used as the degradation rate prediction model for lithium iron phosphate cells.
[0019] The advantage of this embodiment is that the future degradation rate of the battery cell can be predicted through a deep learning model, and the prediction accuracy will also improve as the number of data samples increases.
[0020] Furthermore, the degradation evaluation index of each lithium iron phosphate cell is predicted using a degradation rate prediction model. The specific method is as follows:
[0021] During the charge-discharge test, the amount of electricity required to fully charge each lithium iron phosphate cell for the first preset number of cycles was recorded.
[0022] Input the amount of electricity required to fully charge the lithium iron phosphate battery cell for a preset number of times into the lithium iron phosphate battery cell degradation rate prediction model, and output the degradation rate of the lithium iron phosphate battery cell.
[0023] The degradation evaluation index of the lithium iron phosphate battery cell is calculated by subtracting the degradation rate of the current lithium iron phosphate battery cell from the maximum degradation rate of historical lithium iron phosphate battery cells, and then dividing by the maximum degradation rate of historical lithium iron phosphate battery cells.
[0024] The advantage of this embodiment is that it constructs an evaluation index to quantify the future degradation rate of the battery cell, providing a data foundation for subsequent packaging.
[0025] Furthermore, the charging capacity evaluation index of each lithium iron phosphate cell is calculated using the following method:
[0026] During the discharge process of the charge and discharge test, the charging capacity of each lithium iron phosphate cell is calculated by multiplying the discharge current by the discharge time.
[0027] The charging capacity evaluation index of the lithium iron phosphate battery cell is calculated by dividing the charging capacity of the lithium iron phosphate battery cell by the maximum charging capacity of historical lithium iron phosphate battery cells.
[0028] Furthermore, the clustering algorithm is the k-means clustering algorithm, and the specific classification method is as follows:
[0029] The degradation evaluation index and charging capacity evaluation index of all lithium iron phosphate cells to be assembled are used as two-dimensional classification data;
[0030] Set the number of classification clusters and initialize the cluster centers;
[0031] Calculate the Euclidean distance from each category of data to the cluster center;
[0032] Based on Euclidean distance, assign each category of data to the nearest cluster;
[0033] Calculate the center of each cluster;
[0034] Repeat the iteration until convergence.
[0035] The advantage of this embodiment is that it uses a clustering algorithm to automatically cluster two-dimensional data, which can easily sort out cells that can be packaged, resulting in good sorting effect and low sorting cost.
[0036] Furthermore, the lithium iron phosphate cells in the subgroups were screened using the following method:
[0037] Determine the application scenarios for lithium iron phosphate batteries after packaging;
[0038] Calculate the average charge / discharge frequency, average single-use capacity, and average battery life of lithium iron phosphate batteries in this usage scenario;
[0039] The number of battery cycles, i.e. the number of cycles of a lithium iron phosphate cell, is calculated by multiplying the average battery life by the average charge and discharge frequency.
[0040] The minimum charging capacity of a lithium iron phosphate cell at the end of its service life is calculated based on the average single-use capacity of the battery.
[0041] The maximum charging capacity of each lithium iron phosphate cell to be assembled is subtracted from the product of the number of cycles and the decay rate of that lithium iron phosphate cell. The result is compared with the minimum charging capacity of the lithium iron phosphate cell at the end of its use. If the requirement is met, the lithium iron phosphate cell is retained as a pre-selected cell. If the requirement is not met, the lithium iron phosphate cell is eliminated.
[0042] The advantage of this embodiment is that it provides a screening criterion for battery cells with poor consistency, based on the future usage scenario. This screening criterion can select battery cells with poor consistency that can meet the specific usage scenario, thereby improving the utilization rate of battery cells and effectively reducing the production cost of battery cells.
[0043] Furthermore, the first weight and the second weight are determined based on the charging and discharging frequency of the lithium iron phosphate battery after assembly. The faster the charging and discharging frequency, the greater the second weight, and the slower the charging and discharging frequency, the greater the first weight.
[0044] The advantage of this embodiment is that the faster the charging and discharging frequency is in a specific usage scenario, the faster the degradation rate of all battery cells is. In this case, a battery cell with a slow degradation rate but a lower capacity can be selected to match a battery cell with a larger capacity.
[0045] Furthermore, the calculation methods for the first and second weights are as follows:
[0046] Set a reference charge / discharge frequency. At the reference charge / discharge frequency, both the first and second weights are 0.5.
[0047] The weighting correction parameter is calculated by dividing the difference between the actual charge / discharge frequency of the lithium iron phosphate battery and the reference charge / discharge frequency by the reference charge / discharge frequency.
[0048] The second weight value is obtained by multiplying the sum of the weight adjustment parameter plus 1 by 0.5, and the first weight value is obtained by subtracting the second weight value from 1.
[0049] Other advantages, objectives, and features of the invention will be set forth in part in the description which follows, and in part will be apparent to those skilled in the art from the following examination, or may be learned from practice of the invention. The objectives and other advantages of the invention can be realized and obtained through the following description. Attached Figure Description
[0050] The accompanying drawings of this invention are described below.
[0051] Figure 1 This is a schematic diagram of the process of the present invention.
[0052] Figure 2A schematic diagram of the process for obtaining battery cell degradation evaluation indicators.
[0053] Figure 3 This is a schematic diagram of the k-means clustering algorithm.
[0054] Figure 4 A schematic diagram for selecting a number of remaining battery cells that meet the requirements. Detailed Implementation
[0055] The present invention will be further described below with reference to the accompanying drawings and embodiments.
[0056] A method for packaging lithium iron phosphate energy storage cells based on consistency monitoring, such as Figure 1 As shown, the specific steps are as follows:
[0057] S1. Conduct three charge-discharge cycles on several lithium iron phosphate cells that have come off the production line, and record the amount of electricity required to fully charge each cell during each charge-discharge cycle (CE). i，1 CE i，2 CE i，3 ), discharge current and discharge time, where CE i，1 This represents the amount of electricity required to fully charge cell i for the first time.
[0058] S2, The amount of electricity required to fully charge the battery cell through three charge-discharge cycles (CE) i，1 CE i，2 CE i，3 Input the lithium iron phosphate battery cell degradation rate prediction model f bp (CE i，1 CE i，2 CE i，3 To obtain the degradation evaluation index ae of the battery cell. i .
[0059] Specifically, obtain the degradation evaluation index ae of the battery cell. i ,like Figure 2 As shown, the specific steps are as follows:
[0060] S21. Construct a BP neural network prediction model;
[0061] S22. Select brand-new sample cells for charge-discharge tests. Perform several or more charge-discharge cycles on the sample cells until the sample cells degrade to the point of being unusable. Record the amount of electricity required to fully charge the sample cells each time. in The amount of electricity required for the j-th charge of sample cell numbered i;
[0062] S23. The difference between the amount of energy required to fully charge a brand-new sample cell and the amount of energy required for the last charge, divided by the number of charge-discharge cycles, is used as the degradation rate of the sample cell. The specific formula is as follows:
[0063]
[0064] S24. Construct training samples using the amount of electricity required for the first three full charges of the sample cell and the decay rate of the sample cell. A neural network prediction model is trained using backpropagation (BP) with several training samples. The trained neural network prediction model is then used as the prediction model for the degradation rate of lithium iron phosphate battery cells. bp (X) = av i .
[0065] S25, The amount of electricity required to fully charge the lithium iron phosphate battery cell for the first three charges (CE) i,1 CE i,2 CE i,3 Input the lithium iron phosphate battery cell degradation rate prediction model f bp (X), outputting the degradation rate av of the lithium iron phosphate battery cell. i ;
[0066] Based on the historical maximum degradation rate of lithium iron phosphate cells (av) max Subtract the degradation rate (av) of the lithium iron phosphate battery cell i Then divide by the historical maximum degradation rate of lithium iron phosphate cells (av). max The degradation evaluation index ae of the lithium iron phosphate battery cell was calculated. i The calculation formula is as follows:
[0067]
[0068] S3. In the first discharge test, record the discharge time td of the battery cell. i and discharge current cd i The charging capacity c of the battery cell is calculated. i The charging capacity evaluation index ci is calculated based on the charging capacity of the battery cell. i The calculation formula is as follows:
[0069] c i =td i ×cd i
[0070]
[0071] S4. Construct a classification array g using the cell degradation evaluation index and the charging capacity evaluation index. i (ae i av iThe k-means clustering algorithm is used to classify several class arrays.
[0072] The clustering algorithm is the k-means clustering algorithm, such as... Figure 3 As shown, the specific classification method is as follows:
[0073] S41. Use the degradation evaluation index and charging capacity evaluation index of all lithium iron phosphate cells to be packaged as two-dimensional classification data.
[0074] S42. Set the number of classification clusters and initialize the cluster centers;
[0075] S43. Calculate the Euclidean distance from each category data point to the cluster center.
[0076] S44. Assign each category of data to the nearest cluster based on Euclidean distance;
[0077] S45. Calculate the center of each cluster;
[0078] S46. Repeat the iteration until convergence.
[0079] S5. Using the cluster centers calculated by the k-means clustering algorithm as a benchmark, and combining the distance between the cell and the center, select the cells that can be packaged, i.e., d n，i <th d , where d n，i Let th be the distance from the i-th cell to the n-th center. d This is the distance threshold.
[0080] S6. Analyze the future application scenarios of lithium iron phosphate batteries and select several remaining cells that meet the requirements from the remaining cells.
[0081] Select a number of remaining battery cells that meet the requirements, such as Figure 4 As shown, the specific steps are as follows:
[0082] S61. Calculate the average charge / discharge frequency f, average single-use capacity ct, and average battery life l of the lithium iron phosphate battery under this usage scenario.
[0083] S62. Multiply the average battery life by the average charge / discharge frequency l×f to calculate the battery cycle number n, which is the cycle number of the lithium iron phosphate cell.
[0084] S63. Calculate the minimum charging capacity c of the lithium iron phosphate cell at the end of its service life based on the average single-charge capacity of the battery. min .
[0085] S64. Subtract the product of the number of cycles and the rate of degradation of each lithium iron phosphate cell from the maximum charging capacity of each lithium iron phosphate cell to be assembled, i.e., c max-n.ae, the results are compared with the minimum charging capacity of lithium iron phosphate cells at the end of their service life. min If the requirements are met, the lithium iron phosphate cell will be retained as a pre-selected cell; otherwise, it will be rejected.
[0086] S7. Calculate the first weight value and the second weight value according to the future use scenarios of lithium iron phosphate batteries. The first weight and the second weight are determined according to the charging and discharging frequency of lithium iron phosphate batteries after packaging. The faster the charging and discharging frequency, the larger the second weight, and the slower the charging and discharging frequency, the larger the first weight.
[0087] Specifically, the calculation methods for the first and second weights are as follows:
[0088] Set a reference charge / discharge frequency. At the reference charge / discharge frequency, both the first and second weights are 0.5.
[0089] The weighting correction parameter is calculated by dividing the difference between the actual charge / discharge frequency of the lithium iron phosphate battery and the reference charge / discharge frequency by the reference charge / discharge frequency.
[0090] The second weight value is obtained by multiplying the sum of the weight adjustment parameter plus 1 by 0.5, and the first weight value is obtained by subtracting the second weight value from 1.
[0091] S8. Calculate the group score for each cell based on the first weight value, the second weight data, the cell charging capacity evaluation index, and the cell degradation evaluation index.
[0092] S9. Based on the group scoring, complete the packaging of the remaining battery cells.
[0093] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.
Claims
1. A method for assembling lithium iron phosphate energy storage cells based on consistency monitoring, characterized in that, The specific method is as follows: Construct a model for predicting the degradation rate of lithium iron phosphate battery cells; The same charge and discharge test was conducted on several lithium iron phosphate cells to be packaged, and the charging capacity evaluation index of each lithium iron phosphate cell was calculated, and the degradation evaluation index of each lithium iron phosphate cell was predicted by the degradation rate prediction model. Clustering algorithms were used to classify the charging capacity evaluation index and degradation evaluation index of lithium iron phosphate cells. Based on the distance between each lithium iron phosphate cell and the center of its category, lithium iron phosphate cells of the same category that are less than the distance threshold are divided into the same group of packable cells. Select all lithium iron phosphate cells whose distance is greater than the distance threshold to form a group to be divided; Based on the usage scenario, the lithium iron phosphate cells in the subgroups are screened, and those that meet the requirements are retained. Among the qualified lithium iron phosphate cells, the product of the charging capacity evaluation index and the first weight, plus the product of the attenuation evaluation index and the second weight, is used as the group score. Select lithium iron phosphate cells whose group scores are within a preset range for grouping and packaging; A model for predicting the degradation rate of lithium iron phosphate battery cells is constructed, and the specific method is as follows: Construct a neural network prediction model; Select brand-new sample cells for charge and discharge tests. Perform more than a preset number of charge and discharge cycles on the sample cells and record the amount of electricity required to fully charge the sample cells each time. The rate of degradation of the sample cell is calculated by subtracting the amount of electricity required to fully charge a brand-new sample cell from the amount required for the last charge, and then dividing the result by the number of charge-discharge cycles. Training samples are constructed using the amount of electricity required to fully charge a sample cell a predetermined number of times and the degradation rate of the sample cell. A neural network prediction model is trained using several training samples, and the trained neural network prediction model is used as the degradation rate prediction model for lithium iron phosphate cells. The specific method for screening lithium iron phosphate cells in the subgroup is as follows: Determine the application scenarios for lithium iron phosphate batteries after packaging; Calculate the average charge / discharge frequency, average single-use capacity, and average battery life of lithium iron phosphate batteries in this usage scenario; The number of battery cycles, i.e. the number of cycles of a lithium iron phosphate cell, is calculated by multiplying the average battery life by the average charge and discharge frequency. The minimum charging capacity of a lithium iron phosphate cell at the end of its service life is calculated based on the average single-use capacity of the battery. The maximum charging capacity of each lithium iron phosphate cell to be assembled is subtracted from the product of the number of cycles and the decay rate of that lithium iron phosphate cell. The result is compared with the minimum charging capacity of the lithium iron phosphate cell at the end of its use. If the requirement is met, the lithium iron phosphate cell is retained as a pre-selected cell. If the requirement is not met, the lithium iron phosphate cell is eliminated.
2. The lithium iron phosphate energy storage cell packaging method based on consistency monitoring as described in claim 1, characterized in that, The degradation evaluation index of each lithium iron phosphate cell is predicted using a degradation rate prediction model. The specific method is as follows: During the charge-discharge test, the amount of electricity required to fully charge each lithium iron phosphate cell for the first preset number of cycles was recorded; Input the amount of electricity required to fully charge the lithium iron phosphate battery cell for a preset number of times into the lithium iron phosphate battery cell degradation rate prediction model, and output the degradation rate of the lithium iron phosphate battery cell. The degradation evaluation index of the lithium iron phosphate battery cell is calculated by subtracting the degradation rate of the current lithium iron phosphate battery cell from the maximum degradation rate of historical lithium iron phosphate battery cells, and then dividing by the maximum degradation rate of historical lithium iron phosphate battery cells.
3. The lithium iron phosphate energy storage cell packaging method based on consistency monitoring as described in claim 1, characterized in that, The specific method for calculating the charging capacity evaluation index of each lithium iron phosphate cell is as follows: During the discharge process of the charge and discharge test, the charging capacity of each lithium iron phosphate cell is calculated by multiplying the discharge current by the discharge time. The charging capacity evaluation index of the lithium iron phosphate battery cell is calculated by dividing the charging capacity of the lithium iron phosphate battery cell by the maximum charging capacity of historical lithium iron phosphate battery cells.
4. The lithium iron phosphate energy storage cell packaging method based on consistency monitoring as described in claim 1, characterized in that, The clustering algorithm used is the k-means clustering algorithm, and the specific classification method is as follows: The degradation evaluation index and charging capacity evaluation index of all lithium iron phosphate cells to be assembled are used as two-dimensional classification data; Set the number of classification clusters and initialize the cluster centers; Calculate the Euclidean distance from each category of data to the cluster center; Based on Euclidean distance, assign each category of data to the nearest cluster; Calculate the center of each cluster; Repeat the iteration until convergence.
5. The lithium iron phosphate energy storage cell packaging method based on consistency monitoring as described in claim 1, characterized in that, The first weight and the second weight are determined based on the charging and discharging frequency of the lithium iron phosphate battery after it is packaged. The faster the charging and discharging frequency, the greater the second weight, and the slower the charging and discharging frequency, the greater the first weight.
6. The lithium iron phosphate energy storage cell packaging method based on consistency monitoring as described in claim 5, characterized in that, The calculation methods for the first and second weights are as follows: Set a reference charge / discharge frequency. At the reference charge / discharge frequency, both the first and second weights are 0.
5. The weighting correction parameter is calculated by dividing the difference between the actual charge / discharge frequency of the lithium iron phosphate battery and the reference charge / discharge frequency by the reference charge / discharge frequency. The second weight value is obtained by multiplying the sum of the weight adjustment parameter plus 1 by 0.5, and the first weight value is obtained by subtracting the second weight value from 1.
Citation Information
Patent Citations
Battery cell screening method, device and equipment and storage medium
CN116713211A