A lithium battery grouping method and device based on multiple clustering

Abstract

This application provides a lithium battery grouping method and apparatus based on multiple clustering. The method includes performing cluster analysis on each battery cell according to relevant parameters to obtain multiple first clusters; testing first preset parameters of all cells in each first cluster; performing cluster analysis on all cells in each first cluster according to the first preset parameters to obtain multiple second clusters; testing second preset parameters of all cells in each second cluster; performing cluster analysis on all cells in each second cluster according to the second preset parameters to obtain multiple third clusters; and selecting cells that meet the requirements from the third clusters for grouping according to preset grouping requirements. This application improves the cycle life of the battery system obtained by grouping the battery cells by incorporating parameters such as the battery cell's production process parameters, raw material parameters, and production date into the battery cell grouping and using multiple clustering to select the most similar cells for grouping.

Description

Technical Field

[0001] This application relates to the field of battery manufacturing technology, and in particular to a lithium battery grouping method and apparatus, computer equipment and storage medium based on multiple clustering. Background Technology

[0002] With the development of technology, lithium batteries have become one of the important energy sources for mankind, providing continuous and stable power for equipment in various fields, and gradually becoming an indispensable product in human technological development. Among them, lithium batteries are widely used due to their high energy density and long cycle life.

[0003] To enable lithium batteries to provide higher performance, such as increased current, multiple cells are typically connected in series or parallel to form a battery pack. However, when multiple power batteries are connected in series or parallel, the cycle life of the battery system is lower than that of a single cell due to differences in voltage, internal resistance, and rated capacity among the cells. Battery parameters such as voltage, internal resistance, and rated capacity are affected by manufacturing process parameters and raw materials, but existing battery packing methods do not take into account these factors.

[0004] Therefore, there is an urgent need to propose a new method for lithium battery packing to solve the above problems. Summary of the Invention

[0005] In order to solve one or more of the technical problems existing in the prior art, the present application provides a lithium battery grouping method and apparatus, computer equipment and storage medium based on multiple clustering, so as to solve the problem that the cycle life of the battery system is lower than that of the single cell due to the differences in voltage, internal resistance, rated capacity and other aspects of each cell in the group.

[0006] To achieve the above objectives, the technical solution adopted by this application to solve its technical problem is as follows:

[0007] In a first aspect, this application provides a lithium battery grouping method based on multiple clustering, the method comprising:

[0008] Based on the relevant parameters of each battery cell, a cluster analysis is performed on each battery cell to obtain multiple first-class clusters. The relevant parameters include production process parameters, raw material parameters, and production date.

[0009] Test the first preset parameters of all cells in each first cluster, and perform cluster analysis on all cells in each first cluster based on the first preset parameters to obtain multiple second clusters;

[0010] Test the second preset parameters of all cells in each second cluster, and perform cluster analysis on all cells in each second cluster according to the second preset parameters to obtain multiple third clusters;

[0011] According to the preset grouping requirements, select cells that meet the requirements from the third cluster for grouping.

[0012] Preferably, the production process parameters include at least one of the following: electrolyte injection volume of the battery cell, baking temperature, and rolling pressure of the electrode sheet; the raw material parameters include at least one of the following: positive electrode material, electrolyte material, negative electrode material, and separator material of the battery cell; and the date parameters include at least the formation end time of the battery cell.

[0013] In one specific embodiment, the step of performing cluster analysis on the various battery cells based on their relevant parameters to obtain multiple first clusters includes:

[0014] The relevant parameters are encoded, and the encoded relevant parameters are normalized to obtain a first normalization result;

[0015] Based on the first normalization result, the Mean-Shift clustering algorithm is used to perform cluster analysis on each of the battery cells to obtain multiple first-class clusters.

[0016] In a specific embodiment, the step of performing cluster analysis on each battery cell using the Mean-Shift clustering algorithm based on the first normalization result to obtain multiple first clusters includes:

[0017] Step 1: Randomly select the first normalized result of the relevant parameters of any cell x as the initial center point, and generate a sphere S with a preset window size h as the radius. h Calculate S h All battery cells x i The mean M of the vector generated to cell x h (x), the mean M h The formula for calculating (x) is as follows:

[0018]

[0019] Where n is S h The total number of all cells in the system, where K(·) is the Gaussian kernel function;

[0020] Step 2: Move the window based on battery x and S. h All batteries inside x i The mean M of the vector generated by battery x h (x) Calculate the next center point C, and the formula for calculating the next center point C is:

[0021] C:=M h (x)+x

[0022] Iterative calculation of the mean M h (x) and the next center point C, up to the mean M h If (x) is less than 0.01, then the center at this point is set as the center of the first type of cluster;

[0023] Step 3: Repeat steps 1 to 2 until all cells are classified into the cluster with the highest access frequency, resulting in multiple first clusters.

[0024] In one specific embodiment, the test involves setting a first preset parameter for all cells in each first cluster, and then performing cluster analysis on all cells in each first cluster based on the first preset parameter to obtain multiple second clusters, including:

[0025] The test obtains the first preset parameters of all cells in each of the first clusters, and the first preset parameters include at least the first preset discharge voltage V of the cell.

[0026] The first preset discharge voltage V is feature extracted using refined composite multi-scale spread entropy (RCMDE) to obtain the RCMDE features of the first preset discharge voltage V.

[0027] Normalize the parameters in the first preset parameters other than the first preset discharge voltage V to obtain a second normalization result;

[0028] Based on the second normalization result corresponding to all cells in each first cluster and the RCMDE feature of the first preset discharge voltage V, the Mean-Shift clustering algorithm is used to perform cluster analysis on all cells in each first cluster to obtain multiple second clusters.

[0029] In a specific embodiment, the step of using refined composite multi-scale spread entropy (RCMDE) to extract features from the first preset discharge voltage V and obtaining the RCMDE features of the first preset discharge voltage V includes:

[0030] Obtain the length L of the first preset discharge voltage V, and continuously divide the first preset discharge voltage V into multiple small segments in the interval [1, τ] to obtain V = {v1, v2, ..., v...} L}, calculate the average value of each small segment, and arrange the average values ​​of each small segment in order to obtain a coarse-grained sequence at τ scales, where the k-th coarse-grained sequence at the τ-th scale is... The formula for calculating the j-th element is:

[0031]

[0032] Calculate the probability of the scattering pattern of each coarse-grained sequence at each scale, and calculate the average probability of the scattering pattern of the coarse-grained sequence at each scale based on the probability of the scattering pattern.

[0033] The RCMDE characteristic of the preset discharge voltage V is calculated based on the average value, and the calculation formula is as follows:

[0034]

[0035] in, coarse-grained sequence Distribution pattern The average probability, where m is the embedding dimension, c is the class, d is the time delay parameter, and τ is the scaling factor.

[0036] In one specific embodiment, the second preset parameter is tested for all cells in each second cluster, and cluster analysis is performed on all cells in each second cluster based on the second preset parameter to obtain multiple third clusters, including:

[0037] The test obtains a second preset parameter for all cells in each of the second clusters, the second preset parameter including at least the second preset discharge voltage V of the cell. a ;

[0038] The second preset discharge voltage V is analyzed using the refined composite multiscale distribution entropy (RCMDE). a Feature extraction is performed to obtain the second preset discharge voltage V. a RCMDE features;

[0039] Of the second preset parameters, excluding the second preset discharge voltage V a Other parameters are normalized to obtain a third normalization result;

[0040] Based on the third normalization result corresponding to all cells in each of the second clusters and the second preset discharge voltage V a The RCMDE features are analyzed by using the Mean-Shift clustering algorithm to cluster all cells in each second cluster, resulting in multiple third clusters.

[0041] In a specific embodiment, the step of selecting cells that meet the requirements from each of the third clusters for pairing according to preset pairing requirements includes:

[0042] Select any cell i from any of the third clusters, located near the cluster center. Using cell i as the center and a preset threshold h as the radius, form a sphere S. h1 :

[0043] Determine the ball S h1 Is the number of cells included lower than the required number Num for the group?

[0044] If the ball S h1 If the number of battery cells included is not less than the required number Num, then from the ball S h1 Num cells are randomly selected from the pool for grouping; otherwise, the center value cs of the sphere is updated. The formula for calculating the center value cs is:

[0045]

[0046] Where M is the sphere S h1 The number of cells in the battery, M < Num, R(·) is a function that takes the parameters of the nearest cell;

[0047] The sphere S is updated with the updated cs as the center and the preset threshold h as the radius. h1 If the updated ball S h1 If the number of battery cells included is not less than the required number Num, then from the updated sphere S h1 Num cells are randomly selected for grouping; otherwise, the preset threshold h is adjusted until the number of cells remaining in the corresponding third cluster is insufficient for grouping.

[0048] Secondly, corresponding to the above-mentioned lithium battery grouping method based on multiple clustering, this application also provides a lithium battery grouping device based on multiple clustering, the device comprising:

[0049] The first clustering module is used to perform clustering analysis on each battery cell according to the relevant parameters of each battery cell to obtain multiple first clusters. The relevant parameters include production process parameters, raw material parameters and production date.

[0050] The second clustering module is used to test the first preset parameters of all cells in each first cluster, and to perform clustering analysis on all cells in each first cluster according to the first preset parameters to obtain multiple second clusters;

[0051] The third clustering module is used to test the second preset parameters of all cells in each second cluster, and to perform clustering analysis on all cells in each second cluster based on the second preset parameters to obtain multiple third clusters:

[0052] The cell selection module is used to select cells that meet the requirements from the third cluster according to the preset grouping requirements for grouping.

[0053] Thirdly, a computer device is also provided, including a memory and a processor, wherein the memory stores a computer program that can run on the processor, and when the computer program is executed by the processor, the lithium battery grouping method based on multiple clustering is implemented.

[0054] Fourthly, a computer-readable storage medium is also provided, wherein a computer program is stored therein, and when the computer program is executed, it implements the lithium battery grouping method based on multiple clustering.

[0055] The beneficial effects of the technical solutions provided in this application are:

[0056] This application provides a lithium battery grouping method, apparatus, computer equipment, and storage medium based on multiple clustering. The method includes performing cluster analysis on each battery cell according to relevant parameters to obtain multiple first clusters. The relevant parameters include production process parameters, raw material parameters, and production date. First preset parameters of all cells in each first cluster are tested. Cluster analysis is then performed on all cells in each first cluster according to the first preset parameters to obtain multiple second clusters. Second preset parameters of all cells in each second cluster are tested. Cluster analysis is then performed on all cells in each second cluster according to the second preset parameters to obtain multiple third clusters. Cells that meet the preset grouping requirements are selected from the third clusters for grouping. This application incorporates parameters such as the battery cell's production process parameters, raw material parameters, and production date into the cell grouping, and uses multiple clustering to select the most similar cells for grouping. This ensures that the voltage, internal resistance, rated capacity, and other parameters of each cell in the resulting battery system are as close as possible, thereby improving the cycle life of the resulting battery system. Attached Figure Description

[0057] 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.

[0058] Figure 1 This is a flowchart of a lithium battery grouping method based on multiple clustering provided in an embodiment of this application;

[0059] Figure 2 This is a schematic diagram of the structure of the lithium battery matching device based on multiple clustering provided in the embodiments of this application;

[0060] Figure 3 This is a structural example diagram of the computer device provided in the embodiments of this application. Detailed Implementation

[0061] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions in 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 in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0062] As described in the background art, the existing battery grouping method does not take into account the influence of process parameters and raw materials during the manufacturing process. Process parameters and raw materials during the manufacturing process can affect parameters such as voltage, internal resistance, and rated capacity of the battery cells. Furthermore, the differences in voltage, internal resistance, and rated capacity among the cells of a battery group can lead to a lower cycle life of the battery system than that of a single cell.

[0063] To address one or more of the aforementioned problems, this application creatively proposes a novel lithium battery grouping method based on multiple clustering. In this method, during the lithium battery grouping process, parameters such as the cell's production process parameters, raw material parameters, and production date are incorporated into the cell grouping. Multiple clustering is then used to select the most similar cells for grouping, ensuring that the voltage, internal resistance, rated capacity, and other parameters of each cell in the resulting battery system are as close as possible, thereby improving the cycle life of the resulting battery system.

[0064] The solution of this application will now be described in detail with reference to the accompanying drawings and various embodiments.

[0065] Example 1

[0066] To implement the solution of this application, embodiments of this application provide a lithium battery grouping method based on multiple clustering, referring to... Figure 1 As shown, the method includes the following steps:

[0067] S100: Cluster analysis is performed on each battery cell based on its relevant parameters to obtain multiple first-class clusters. The relevant parameters include production process parameters, raw material parameters, and production date.

[0068] Specifically, since parameters such as voltage, internal resistance, and rated capacity of battery cells are affected by process parameters and raw materials during the manufacturing process, and the differences in voltage, internal resistance, and rated capacity among the cells of a group of batteries can lead to a lower cycle life of the battery system than that of a single cell, the embodiments of this application take into account parameters such as the production process parameters, raw material parameters, and production date of the battery cells in order to improve the cycle life and other performance of the battery system obtained by grouping.

[0069] Preferably, the production process parameters include at least one of the following: electrolyte injection volume of the battery cell, baking temperature, and rolling pressure of the electrode sheet; the raw material parameters include at least one of the following: positive electrode material, electrolyte material, negative electrode material, and separator material of the battery cell; and the date parameters include at least the formation end time of the battery cell.

[0070] S200: Test the first preset parameters of all cells in each first cluster, and perform cluster analysis on all cells in each first cluster according to the first preset parameters to obtain multiple second clusters.

[0071] In this embodiment, the first preset parameter is not specifically limited; users can set it according to actual needs. As an example rather than a restrictive description, the first preset parameter includes, but is not limited to, parameters such as the cell's discharge capacity, DC internal resistance, discharge voltage, and temperature rise. The clustering algorithm used for cluster analysis includes, but is not limited to, the Mean-Shift clustering algorithm.

[0072] S300: Test the second preset parameters of all cells in each second cluster, and perform cluster analysis on all cells in each second cluster according to the second preset parameters to obtain multiple third clusters.

[0073] In this embodiment, the second preset parameter is not specifically limited; users can set it according to actual needs. This is an illustrative rather than restrictive explanation. The second preset parameter includes, but is not limited to, parameters such as the cell's open-circuit voltage, discharge capacity, DC internal resistance, discharge voltage, and temperature rise. The clustering algorithm used in the clustering analysis includes, but is not limited to, the Mean-Shift clustering algorithm.

[0074] S400: Select cells that meet the requirements from the third cluster according to the preset grouping requirements and group them.

[0075] As a preferred implementation, in this embodiment of the application, the step of performing cluster analysis on the various battery cells based on their relevant parameters to obtain multiple first clusters includes:

[0076] S110: Encode the relevant parameters and normalize the encoded relevant parameters to obtain a first normalization result.

[0077] S120: Based on the first normalization result, the Mean-Shift clustering algorithm is used to perform cluster analysis on each of the battery cells to obtain multiple first-class clusters.

[0078] In practice, the batch number M_pos of the positive electrode raw materials is recorded before the production of the i-th cell. i Batch of electrolyte raw materials M_ele i Batch M_neg of negative electrode raw materials i , Batch M_sep of diaphragm raw materials i Record the electrolyte injection volume NUM_liquid during the production of the i-th battery. i Baking temperature Bak_T i The rolling pressure P of the j-th electrode i，j Record the battery formation end time. i The i-th cell is any one of the cells to be matched, and the j-th electrode is any one of the electrodes of the i-th cell.

[0079] After obtaining the relevant parameters of each cell in the group to be matched, each parameter is encoded. Specifically, the LabelIEncode method can be used for encoding to obtain the batch ME_pos of the positive electrode raw material. i Batch of electrolyte raw materials ME_ele i Batch ME_neg of negative electrode raw materials i , Batch number of membrane raw materials ME_sep i Then convert the battery into an end time. i Split into Month i and Days i .

[0080] Then, the encoded parameters are normalized. In practice, the Min-Max method can be used to normalize the above parameters. The normalization formula is as follows:

[0081]

[0082] Where, x i Let x be any relevant parameter of the i-th cell (i.e., any one of the production process parameters, raw material parameters, and production date), min(x) is the minimum or preset value (such as an empirical value) of the corresponding relevant parameter among all cells; max(x) is the maximum or preset value (such as an empirical value) of the corresponding relevant parameter among all cells.

[0083] After Min-Max normalization, the production process parameters, raw material parameters, and production date parameters of the i-th battery cell are obtained as follows: Based on normalized parameters The Mean-Shift clustering method is used to cluster all battery cells. Mean-Shift clustering attempts to find the densest region of data points by continuously updating the cluster centers until the final cluster centers meet the termination condition. Moreover, this clustering method does not require manual specification of the number of clusters.

[0084] As a preferred implementation, in this embodiment of the application, the step of performing cluster analysis on each battery cell using the Mean-Shift clustering algorithm based on the first normalization result to obtain multiple first clusters includes:

[0085] Step 1: Randomly select the first normalized result of the relevant parameters of any cell x as the initial center point, and generate a sphere S with a preset window size h as the radius. h Calculate S h All battery cells x i The mean M of the vector generated to cell x h (x), the mean M h The formula for calculating (x) is as follows:

[0086]

[0087] Where n is S h The total number of all cells in the system, where K(·) is the Gaussian kernel function;

[0088] Specifically, for a batch of battery cells to be paired, a battery x is randomly selected, and a sphere S is generated with the first normalized result of its relevant parameters as the initial center point (i.e., the center of the sphere) and a preset window size h as the radius. h .

[0089] Step 2: Move the window based on battery x and S. h All batteries inside x i The mean M of the vector generated by battery x h (x) Calculate the next center point C, and the formula for calculating the next center point C is:

[0090] C:=M h (x)+x

[0091] Iterative calculation of the mean M h (x) and the next center point C, up to the mean M h If (x) is less than 0.01, then the center at this point is set as the center of the first type of cluster;

[0092] Specifically, the mean M is calculated iteratively. h (x) and the next center point C can cause the center point C to gradually move towards a higher density region until the mean (M) is reached. h If (x) is less than 0.01, then the center is a cluster center.

[0093] Step 3: Repeat steps 1 to 2 until all cells are classified into the cluster with the highest access frequency, resulting in multiple first clusters.

[0094] Therefore, based on the batch M_pos of the cathode raw material i Batch of electrolyte raw materials M_ele i Batch M_neg of negative electrode raw materials i , Batch M_sep of diaphragm raw materials i The electrolyte injection volume NUM_liquid recorded during the production of the i-th battery. i Baking temperature Bak_T i The rolling pressure P of the j-th electrode i，j Battery formation end time i Multiple first-class clusters were obtained through clustering algorithms.

[0095] As a preferred implementation, in this embodiment of the application, the step of testing all cells in each first cluster with a first preset parameter, and performing cluster analysis on all cells in each first cluster based on the first preset parameter to obtain multiple second clusters includes:

[0096] S210: Test and obtain the first preset parameters of all cells in each of the first clusters, wherein the first preset parameters include at least the first preset discharge voltage V of the cell.

[0097] Specifically, the first preset parameter can be set according to actual needs. As an example rather than a limiting description, the i-th cell in one of the multiple first clusters can be placed in the battery performance testing cabinet to test its discharge capacity C under preset charge and discharge conditions (such as 0.33C). N The battery's DC internal resistance R is tested for 10 seconds using an HPPC test with a 5C current under a preset state of charge (e.g., 50%). 10 Perform three charge-discharge cycles on the battery using a high-rate current, and record the discharge voltage V (i.e., the first preset discharge voltage) and its temperature rise T during the last cycle. R The high-rate current is a current that will not cause battery malfunctions and is permitted by the design specifications. Preferably, the discharge capacity C can be determined after the test. N 10s DC internal resistance R 10 Cells that do not meet the requirements for grouping are rejected.

[0098] S220: Use refined composite multi-scale spread entropy (RCMDE) to extract features from the first preset discharge voltage V and obtain the RCMDE features of the first preset discharge voltage V.

[0099] As a preferred implementation method, the calculation steps are as follows:

[0100] S221: Obtain the length L of the first preset discharge voltage V, where the length L is related to the data acquisition frequency. In the fine composite multi-scale scattering entropy algorithm, the first preset discharge voltage V is continuously divided into multiple small segments in [1, τ] to obtain V = {v1, v2, ..., v...}. L First, calculate the average value of each small segment. Then, arrange the average values ​​of each small segment in order to form a coarse-grained sequence, resulting in coarse-grained sequences at τ scales. The k-th coarse-grained sequence at the τ-th scale is... The formula for calculating the j-th element is:

[0101]

[0102] S222: Calculate the probability of the scattering pattern of each coarse-grained sequence at each scale, and calculate the average probability of the scattering pattern of the coarse-grained sequence at each scale based on the probability of the scattering pattern.

[0103] The specific calculation process is as follows:

[0104] 1. Map each coarse-grained sequence using the normal distribution function. For example, map the k-th coarse-grained sequence... Mapped to y The mapping formula is as follows:

[0105]

[0106] Where μ and σ represent the mean and standard deviation, respectively.

[0107] 2. Use linear transformation to transform y j Mapping to the range [1, 2, ..., c], the mapping formula is as follows:

[0108]

[0109] Where R is the floor function and c is the number of categories. For y j The number of sequence classes.

[0110] 3. Calculate the embedding vector sequence composed of embedding dimension m and time delay d. The calculation formula is as follows:

[0111]

[0112] Where i = 1, 2, ..., N-(m-1)d

[0113] 4. Calculate the distribution pattern

[0114] like but The corresponding distribution pattern is Each pattern consists of c data points, and each value has m possible values, thus there are c corresponding scatter patterns. m kind.

[0115] 5. Calculate each distribution pattern probability The calculation formula is as follows:

[0116]

[0117] in Represents the embedding vector Mapping to scattering pattern The number of items.

[0118] S223: Calculate the RCMDE characteristic of the preset discharge voltage V based on the average value, using the following formula:

[0119]

[0120] in, coarse-grained sequence Distribution pattern The average probability, where m is the embedding dimension, c is the class, d is the time delay parameter, and τ is the scaling factor.

[0121] Preferably, the embedding dimension m can be 2 or 3, the category c is selected as an integer from 4 to 8, the time delay parameter d is selected as 1, and the scaling factor τ can be selected based on the actual effect. More preferably, the scaling factor τ is set to 20.

[0122] S230: Normalize the parameters other than the first preset discharge voltage V in the first preset parameters to obtain a second normalization result.

[0123] Specifically, the Min-Max method can also be used for normalization here, and the normalization formula is:

[0124]

[0125] Where, x i The discharge capacity C of the i-th cell in one of the multiple first-class clusters mentioned above. N 10s DC internal resistance R10 Or temperature rise T R w represents the weight. The weight for each parameter type can be determined based on experience and the impact of different parameter types on battery consistency. For example, to increase the number of cycles, the 10s DC internal resistance R can be increased. 10 The weight w is set to a large value, such as 5. These are the normalized parameters (i.e., the second normalization result). min(x) is the minimum value of a certain parameter among all cells in one of the multiple first-class clusters mentioned above; max(x) is the maximum value of a certain parameter among all cells in one of the multiple first-class clusters mentioned above. The normalized battery discharge capacity is obtained after the Min-Max method. 10s DC internal resistance and temperature rise

[0126] S240: Based on the second normalization result corresponding to all cells in each first cluster and the RCMDE feature of the first preset discharge voltage V, the Mean-Shift clustering algorithm is used to perform cluster analysis on all cells in each first cluster to obtain multiple second clusters.

[0127] Specifically, the clustering process can be referred to in step S100, which will not be repeated here.

[0128] As a preferred implementation, in this embodiment of the application, the step of testing all cells in each second cluster with second preset parameters, and performing cluster analysis on all cells in each second cluster based on the second preset parameters to obtain multiple third clusters includes:

[0129] S310: Test and obtain the second preset parameters of all cells in each of the second clusters, the second preset parameters including at least the second preset discharge voltage V of the cell. a .

[0130] Specifically, the second preset parameter can be set according to actual needs. As an example rather than a limiting illustration, the i-th cell in one of the multiple second clusters can be placed in a high-temperature aging chamber for a period of time (e.g., 72 hours). The high-temperature aging temperature can be the maximum cell storage temperature in the design specifications minus 10°C. After aging, the open-circuit voltage OCV of the cell is measured. After standing at room temperature for a period of time (e.g., 1 hour), the discharge capacity C of the cell after aging under preset discharge conditions (e.g., 0.33°C) is measured and tested. a The battery's DC internal resistance R after 10 seconds of aging was tested using an HPPC test with a 5C current under a preset state of charge (e.g., 50%). a10The aged battery was charged and discharged for 3 cycles using a high-rate current, and the discharge voltage V of the last cycle was recorded. a (i.e., the second preset discharge voltage) and its temperature rise T aR .

[0131] S320: Use refined composite multi-scale distribution entropy RCMDE to measure the second preset discharge voltage V. a Feature extraction is performed to obtain the second preset discharge voltage V. a The RCMDE features.

[0132] Specifically, here, refined composite multi-scale spread entropy (RCMDE) is used to evaluate the second preset discharge voltage V. a The process of feature extraction can be referred to in step S220, and will not be repeated here.

[0133] S330: For the second preset parameters excluding the second preset discharge voltage V a Other parameters are normalized to obtain a third normalization result.

[0134] Specifically, the Min-Max method can also be used for normalization here, and the normalization formula is:

[0135]

[0136] Where, x i After the aging process is completed for the i-th cell in one of the multiple second-category cells mentioned above, the open-circuit voltage OCV and the discharge capacity C after aging are measured. a The DC internal resistance R after 10 seconds of aging a10 Or the temperature rise T after aging aR w represents the weight. The weight for each parameter type can be determined based on experience and the impact of different parameter types on battery consistency. For example, to increase the number of cycles, the 10-second DC internal resistance R after aging can be increased. a10 The weight w is set to a large value, such as 5. These are the normalized parameters (i.e., the third normalization result). min(x) is the minimum value of a parameter among all cells in one of the multiple second-class clusters mentioned above; max(x) is the maximum value of a parameter among all cells in one of the multiple second-class clusters mentioned above. The normalized parameters are obtained using the Min-Max method. The open-circuit voltage of the battery is measured after aging. Discharge capacity after aging 10s DC internal resistance after aging Or the temperature rise after aging

[0137] S340: Based on the third normalization result corresponding to all cells in each of the second clusters and the second preset discharge voltage V a The RCMDE features are analyzed by using the Mean-Shift clustering algorithm to cluster all cells in each second cluster, resulting in multiple third clusters.

[0138] Specifically, the clustering process can also refer to the relevant content in step S100, which will not be repeated here.

[0139] If the consistency of cells in the battery system is highly required, as a preferred implementation method, in this embodiment, the step of selecting cells that meet the requirements from each of the third clusters for grouping according to preset grouping requirements includes:

[0140] Select any cell i from any of the third clusters, located near the cluster center. Using cell i as the center and a preset threshold h as the radius, form a sphere S. h1 The preset threshold h can be adjusted as needed, and there is no particular limitation on it in this application.

[0141] Determine the ball S h1 Is the number of cells included lower than the required number Num for the group?

[0142] If the ball S h1 If the number of battery cells included is not less than the required number Num, then from the ball S h1 Num cells are randomly selected from the pool for grouping; otherwise, the center value cs of the sphere is updated. The formula for calculating the center value cs is:

[0143]

[0144] Where M is the sphere S h1 The number of cells in the battery, M < Num, R(·) is a function that takes the parameters of the nearest cell;

[0145] The sphere S is updated with the updated cs as the center and the preset threshold h as the radius. h1 If the updated ball S h1 If the number of battery cells included is not less than the required number Num, then from the updated sphere S h1 Num cells are randomly selected for grouping; otherwise, the preset threshold h is adjusted until the number of cells remaining in the corresponding third cluster is insufficient for grouping.

[0146] If the consistency requirement of the battery cells in the battery system is not high, the required number of Num cells can be randomly selected from multiple third-category clusters for pairing, until the number of cells remaining in the third-category clusters cannot be paired. The remaining unpaired cells and the cells removed during the clustering process can be used in the next lithium battery pairing process based on multiple clustering.

[0147] Furthermore, if the consistency requirements of the battery cells in the battery system are extremely high, other parameters for evaluating battery performance can be added to the first and second preset parameters, such as SOC-OCV parameters, 10s AC internal resistance, etc., which will not be listed here.

[0148] Example 2

[0149] Corresponding to Embodiment 1 above, this application also provides a training device for a cell expansion force prediction model. In this embodiment, content that is the same as or similar to that in Embodiment 1 above can be referred to the above description and will not be repeated hereafter. (Refer to...) Figure 2 The device includes:

[0150] The first clustering module is used to perform clustering analysis on each battery cell according to the relevant parameters of each battery cell to obtain multiple first clusters. The relevant parameters include production process parameters, raw material parameters and production date.

[0151] The second clustering module is used to test the first preset parameters of all cells in each first cluster, and to perform clustering analysis on all cells in each first cluster according to the first preset parameters to obtain multiple second clusters;

[0152] The third clustering module is used to test the second preset parameters of all cells in each second cluster, and to perform clustering analysis on all cells in each second cluster according to the second preset parameters to obtain multiple third clusters;

[0153] The cell selection module is used to select cells that meet the requirements from the third cluster according to the preset grouping requirements for grouping.

[0154] Example 3

[0155] Corresponding to Embodiment 1 or 2 above, this application also provides a computer device, including: a processor and a memory, wherein the memory stores a computer program that can run on the processor, and when the computer program is executed by the processor, it executes the lithium battery grouping method based on multiple clustering provided in any of the above embodiments.

[0156] in, Figure 3An exemplary computer device 1500 is shown, which may specifically include a processor 1510, a video display adapter 1511, a disk drive 1512, an input / output interface 1513, a network interface 1514, and a memory 1520. The processor 1510, video display adapter 1511, disk drive 1512, input / output interface 1513, network interface 1514, and memory 1520 can communicate with each other via a communication bus 1530.

[0157] The processor 1510 can be implemented using a general-purpose CPU (Central Processing Unit), microprocessor, application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solution provided by the present invention.

[0158] The memory 1520 can be implemented in the form of ROM (Read Only Memory), RAM (Random Access Memory), static storage device, dynamic storage device, etc. The memory 1520 can store the operating system 1521 for controlling the operation of the electronic device, and the basic input / output system (BIOS) for controlling the low-level operations of the electronic device. Additionally, it can store a web browser 1523, a data storage management system 1524, and a device identification information processing system 1525, etc. The aforementioned device identification information processing system 1525 can be the application program that specifically implements the aforementioned steps in this embodiment of the invention. In summary, when implementing the technical solution provided by this invention through software or firmware, the relevant program code is stored in the memory 1520 and is called and executed by the processor 1510.

[0159] Input / output interface 1513 is used to connect input / output modules to realize information input and output. Input / output modules can be configured as components in the device (not shown in the figure) or externally connected to the device to provide corresponding functions. Input devices may include keyboards, mice, touch screens, microphones, various sensors, etc., and output devices may include displays, speakers, vibrators, indicator lights, etc.

[0160] Network interface 1514 is used to connect a communication module (not shown in the figure) to enable communication between this device and other devices. The communication module can communicate via wired means (such as USB, Ethernet cable, etc.) or wireless means (such as mobile network, WIFI, Bluetooth, etc.).

[0161] The bus includes a pathway for transmitting information between various components of the device (e.g., processor 1510, video display adapter 1511, disk drive 1512, input / output interface 1513, network interface 1514, and memory 1520).

[0162] In addition, the electronic device can also obtain information on specific claim conditions from the virtual resource object claim condition information database for condition judgment, and so on.

[0163] It should be noted that although the above-described device only shows the processor 1510, video display adapter 1511, disk drive 1512, input / output interface 1513, network interface 1514, memory 1520, bus, etc., in specific implementations, the device may also include other components necessary for normal operation. Furthermore, those skilled in the art will understand that the above-described device may only include the components necessary for implementing the present invention, and not necessarily all the components shown in the figures.

[0164] Example 4

[0165] Corresponding to embodiments one to three above, this application also provides a computer-readable storage medium. In this embodiment, the content that is the same as or similar to embodiments one to three above can be referred to the above description and will not be repeated hereafter.

[0166] The computer-readable storage medium stores a computer program thereon, which, when executed by a processor, implements the lithium battery grouping method based on multiple clustering as described above.

[0167] In some implementations of this application, when the computer program is executed by a processor, it can also implement the steps corresponding to the method described in Embodiment 1. Please refer to the detailed description in Embodiment 1, which will not be repeated here.

[0168] As can be seen from the above description of the embodiments, those skilled in the art can clearly understand that the present invention can be implemented by means of software plus necessary general-purpose hardware platforms. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in various embodiments or some parts of the embodiments of the present invention.

[0169] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, for system or system embodiments, since they are basically similar to method embodiments, the description is relatively simple, and relevant parts can be referred to the descriptions in the method embodiments. The systems and system embodiments described above are merely illustrative. 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 modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without creative effort.

[0170] The technical solution provided by this invention has been described in detail above. Specific examples have been used to illustrate the principles and implementation methods of this invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of this invention. Furthermore, those skilled in the art will recognize that, based on the ideas of this invention, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this invention.

Claims

1. A lithium battery grouping method based on multiple clustering, characterized in that, The method includes: Based on the relevant parameters of each battery cell, a cluster analysis is performed on each battery cell to obtain multiple first-class clusters. The relevant parameters include production process parameters, raw material parameters, and production date. Test the first preset parameters of all cells in each first cluster, and perform cluster analysis on all cells in each first cluster based on the first preset parameters to obtain multiple second clusters. This includes: testing and obtaining the first preset parameters of all cells in each first cluster, where the first preset parameters include at least the first preset discharge voltage V of the cell; and using refined composite multi-scale distribution entropy. Feature extraction is performed on the first preset discharge voltage V to obtain the first preset discharge voltage V. Features; normalize the parameters other than the first preset discharge voltage V in the first preset parameters to obtain a second normalization result; based on the second normalization result corresponding to all cells in each first cluster and the first preset discharge voltage V... The feature uses the Mean-Shift clustering algorithm to perform cluster analysis on all cells in each first cluster to obtain multiple second clusters; Test the second preset parameters of all cells in each second cluster, and perform cluster analysis on all cells in each second cluster according to the second preset parameters to obtain multiple third clusters; According to the preset grouping requirements, select cells that meet the requirements from the third cluster for grouping.

2. The lithium battery grouping method based on multiple clustering according to claim 1, characterized in that, The step of performing cluster analysis on each battery cell based on relevant parameters to obtain multiple first-class clusters includes: The relevant parameters are encoded, and the encoded relevant parameters are normalized to obtain a first normalization result; Based on the first normalization result, the Mean-Shift clustering algorithm is used to perform cluster analysis on each of the battery cells to obtain multiple first-class clusters.

3. The lithium battery grouping method based on multiple clustering according to claim 2, characterized in that, The Mean-Shift clustering algorithm is used to perform cluster analysis on each battery cell based on the first normalization result, resulting in multiple first clusters, including: Step 1: Randomly select any battery cell The first normalized result of the relevant parameters is used as the initial center point, with a preset window size. Generate a sphere with radius . ,calculate All battery cells To the battery cell The mean of the generated vectors The mean The calculation formula is as follows: in, for The number of all battery cells in the system. The Gaussian kernel function; Step 2: Move the window according to the battery. as well as All batteries inside To battery The mean of the generated vectors Calculate the next center point The next center point The calculation formula is: Iterative calculation of the mean and the next center point until the mean If the value is less than 0.01, then the center at this point is set as the center of the first type of cluster; Step 3: Repeat steps 1 to 2 until all cells are classified into the cluster with the highest access frequency, resulting in multiple first clusters.

4. The lithium battery grouping method based on multiple clustering according to claim 1, characterized in that, The use of refined composite multi-scale distribution entropy Feature extraction is performed on the first preset discharge voltage V to obtain the first preset discharge voltage V. Features include: Obtain the length of the first preset discharge voltage V The first preset discharge voltage by Divide into multiple small segments continuously to obtain Calculate the average value of each segment, and then arrange the average values ​​of each segment in order. The coarse-grained sequence at the nth scale, where the nth The first of the scales coarse-grained sequences No. The formula for calculating each element is: Calculate the probability of the scattering pattern of each coarse-grained sequence at each scale, and calculate the average probability of the scattering pattern of the coarse-grained sequence at each scale based on the probability of the scattering pattern. The first preset discharge voltage V is calculated based on the average value. Features, calculated using the following formula: in, coarse-grained sequence Distribution pattern The average probability, where m is the embedding dimension, c is the class, and d is the time delay parameter. is the scale factor.

5. The lithium battery grouping method based on multiple clustering according to any one of claims 1 to 3, characterized in that, The test involves setting a second preset parameter for all cells in each of the second clusters, and then performing cluster analysis on all cells in each of the second clusters based on the second preset parameter to obtain multiple third clusters, including: The test obtains a second preset parameter for all cells in each of the second type of cluster, the second preset parameter including at least the second preset discharge voltage of the cell. ; Using refined composite multi-scale distribution entropy For the second preset discharge voltage Feature extraction is performed to obtain the second preset discharge voltage. of feature; Of the second preset parameters, excluding the second preset discharge voltage Other parameters are normalized to obtain a third normalization result; Based on the third normalization result corresponding to all cells in each of the second clusters and the second preset discharge voltage of The feature uses the Mean-Shift clustering algorithm to perform cluster analysis on all cells in each second cluster to obtain multiple third clusters.

6. The lithium battery grouping method based on multiple clustering according to any one of claims 1 to 3, characterized in that, The step of selecting and grouping battery cells that meet the requirements from the third cluster according to the preset grouping requirements includes: Select any cell from any of the third clusters, located near the cluster center. With the battery cell Centered on the sphere, with a preset threshold Form a sphere with radius . ; Determine the ball Is the number of cells included lower than the required number Num for the group? If the ball If the number of battery cells included is not less than the required number Num, then from the ball... Select Num cells from the sample for pairing; otherwise, update the core of the sphere. , center of the ball The calculation formula is: in, For the ball The number of battery cells in China , A function to retrieve the parameters of the most recent battery cell; After the update Center of the ball, preset threshold Update the sphere for radius If the updated ball If the number of battery cells included is not less than the required number Num, then from the updated sphere Num cells are randomly selected for grouping; otherwise, the preset threshold is adjusted. This continues until the number of cells remaining in the corresponding third cluster is insufficient for pairing.

7. A lithium battery grouping device based on multiple clustering, characterized in that, The device includes: The first clustering module is used to perform clustering analysis on each battery cell according to the relevant parameters of each battery cell to obtain multiple first clusters. The relevant parameters include production process parameters, raw material parameters and production date. The second clustering module is used to test the first preset parameters of all cells in each first cluster, and to perform clustering analysis on all cells in each first cluster based on the first preset parameters to obtain multiple second clusters. This includes: testing and obtaining the first preset parameters of all cells in each first cluster, wherein the first preset parameters include at least the first preset discharge voltage V of the cell; and using refined composite multi-scale distribution entropy. Feature extraction is performed on the first preset discharge voltage V to obtain the first preset discharge voltage V. Features; normalize the parameters other than the first preset discharge voltage V in the first preset parameters to obtain a second normalization result; based on the second normalization result corresponding to all cells in each first cluster and the first preset discharge voltage V... The feature uses the Mean-Shift clustering algorithm to perform cluster analysis on all cells in each first cluster to obtain multiple second clusters; The third clustering module is used to test the second preset parameters of all cells in each second cluster, and to perform clustering analysis on all cells in each second cluster according to the second preset parameters to obtain multiple third clusters; The cell selection module is used to select cells that meet the requirements from the third cluster according to the preset grouping requirements for grouping.

8. A computer device, characterized in that, It includes a memory and a processor, wherein the memory stores a computer program that can run on the processor, and when the computer program is executed by the processor, it implements the lithium battery grouping method based on multiple clustering as described in any one of claims 1 to 6.

9. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed, it implements the lithium battery grouping method based on multiple clustering as described in any one of claims 1 to 6.

Images