Battery classification method, battery module manufacturing method, and battery classification system

Abstract

The invention provides a battery classification method, a manufacturing method of a battery module, and a battery classification system. The method for classifying batteries includes causing a computer to execute the following steps. The method includes: acquiring manufacturing conditions of a plurality of batteries; predicting the durability of the plurality of batteries on the basis of the acquired manufacturing conditions on the basis of the relationship between the manufacturing conditions of the batteries and the durability of the batteries; and classifying the plurality of batteries according to the similarity of the predicted durability quality. The method of the present disclosure of manufacturing a battery module includes connecting batteries having similar durability properties classified by the method of the present disclosure of classifying the batteries in parallel or in series with each other.

Description

Technical Field

[0001] The present disclosure relates to a battery classification method, a battery module manufacturing method, and a battery classification system. Background Art

[0002] In order to obtain a desired battery capacity, a technique of using a plurality of batteries in combination is known.

[0003] For example, Japanese Patent Application Laid-Open No. 2009-240154 discloses a method for controlling charge and discharge of multiple batteries. When multiple batteries are used alternately, by matching the usage conditions, the degradation of the individual batteries can be averaged and the battery life can be extended. Summary of the Invention

[0004] The present inventors have discovered that when a plurality of batteries are used in combination, for example, during factory inspection, even if the input / output characteristics and other performance of the batteries are the same, differences may occur in the durability of the batteries.

[0005] An object of the present disclosure is to provide a battery classification method based on differences in durability performance when a plurality of batteries are used in combination, a battery module manufacturing method including such a method, and a battery classification system.

[0006] The inventors of the present application have discovered that the above-mentioned problems can be solved by the following means.

[0007] Method 1

[0008] A battery classification method includes causing a computer to execute the following steps:

[0009] Obtaining manufacturing conditions for multiple batteries;

[0010] Based on the relationship between the manufacturing conditions of the batteries and the durability of the batteries, predicting the durability of the plurality of batteries according to the obtained manufacturing conditions; and

[0011] The plurality of batteries are classified according to the predicted similarities in the durability performance.

[0012] Method 2: In the method described in Method 1,

[0013] The manufacturing conditions are the dew point of the atmosphere when the positive electrode mixture material and / or the negative electrode mixture material is laminated on the substrate or the ratio of the weight per unit area of ​​the positive electrode mixture material to the weight per unit area of ​​the negative electrode mixture material.

[0014] Method 3: In the method described in Method 1,

[0015] The relationship is a relationship predetermined by machine learning.

[0016] Method 4

[0017] A method for manufacturing a battery module, comprising:

[0018] Batteries having similar durability performance classified by the method described in any one of Embodiments 1 to 3 are connected in parallel or in series.

[0019] Method 5

[0020] A battery classification system having:

[0021] an acquiring unit, which acquires manufacturing conditions of a plurality of batteries;

[0022] a prediction unit that predicts the durability performance of the plurality of batteries based on the manufacturing conditions acquired by the acquisition unit, based on a relationship between the manufacturing conditions of the batteries and the durability performance of the batteries; and

[0023] The classification unit classifies the plurality of batteries based on the similarity of the durability performance predicted by the prediction unit.

[0024] According to the present disclosure, it is possible to provide a battery classification method based on differences in durability performance when a plurality of batteries are used in combination, a battery module manufacturing method including such a method, and a battery classification system. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Features, advantages, and technical and industrial significance of exemplary embodiments of the present invention will be described below with reference to the accompanying drawings, in which like reference numerals represent like elements, and in which:

[0026] Figure 1 is a flow chart illustrating a method of the present disclosure for classifying batteries.

[0027] Figure 2 This is a graph showing an example of the relationship between the dew point of the atmosphere when the positive electrode layer material is laminated on the substrate (positive electrode lamination environment dew point) and the SHAP value.

[0028] Figure 3 This is a graph showing an example of the relationship between the dew point of the atmosphere when the negative electrode composite material is laminated on the substrate (negative electrode lamination environment dew point) and the SHAP value.

[0029] Figure 4 This is a graph showing an example of the relationship between the ratio of the weight per unit area of ​​the positive electrode composite material to the weight per unit area of ​​the negative electrode composite material (positive electrode weight per unit area / negative electrode weight per unit area) and the SHAP value.

[0030] Figure 5A This is a schematic diagram illustrating battery modules connected in parallel by the method of manufacturing the battery module disclosed herein.

[0031] Figure 5B This is a schematic diagram illustrating battery modules connected in series by the method of manufacturing the battery module disclosed herein.

[0032] Figure 6 This is a schematic diagram showing an example of a battery classification system of the present disclosure. DETAILED DESCRIPTION

[0033] Hereinafter, embodiments of the present disclosure will be described in detail. However, the present disclosure is not limited to the following embodiments, and can be implemented with various modifications within the scope of the disclosed gist.

[0034] Battery classification method

[0035] The method of classifying batteries disclosed herein includes causing a computer to execute the following steps.

[0036] Obtaining manufacturing conditions for multiple batteries;

[0037] Based on the relationship between the manufacturing conditions of the batteries and the durability performance of the batteries, predicting the durability performance of a plurality of batteries based on the obtained manufacturing conditions; and

[0038] The plurality of batteries are classified according to the similarity of the predicted durability performance.

[0039] The present disclosure author and others believe that when multiple batteries are used in combination, one of the reasons for the difference in the durability of each battery is sometimes due to the manufacturing conditions of the battery. Therefore, the present disclosure author and others studied the correlation between manufacturing conditions and durability through machine learning. More specifically, by constructing LightGBM, one of the data analysis methods known as "teacher learning", the manufacturing conditions of the explanatory variables with high SHAP values ​​were extracted for the durability as the target variable, and the correlation was studied. In addition, the SHAP value refers to the contribution of the explanatory variable to the target variable.

[0040] The present inventors have discovered that the durability performance of multiple batteries can be predicted based on the relationship between battery manufacturing conditions and battery durability performance determined in this manner, using the manufacturing conditions obtained as actual measured values. Furthermore, the present inventors have discovered that by classifying multiple batteries based on the similarity of the predicted durability performance, batteries with similar durability performance can be combined and used.

[0041] In addition, the battery used in the present disclosure may be, for example, a lithium-ion secondary battery. As an application of the battery, for example, a power source for a hybrid electric vehicle (HEV), a plug-in hybrid electric vehicle (PHEV), a battery electric vehicle (BEV), a gasoline vehicle, a diesel vehicle, or the like can be cited. In particular, it is preferably used as a driving power source for a hybrid electric vehicle (HEV), a plug-in hybrid electric vehicle (PHEV), or a battery electric vehicle (BEV). In addition, the battery in the present disclosure can be used as a power source for mobile objects other than vehicles (such as railways, ships, and airplanes), and can also be used as a power source for electrical products such as information processing devices.

[0042] Figure 1 is a flow chart showing the method of the present disclosure for classifying batteries. Figure 1 , describing each step of the method of the present disclosure for classifying batteries.

[0043] Manufacturing condition acquisition process

[0044] like Figure 1 As shown, the method of the present disclosure includes causing a computer to execute a process of acquiring manufacturing conditions of a plurality of batteries ( S101 ).

[0045] The manufacturing conditions may be the dew point of the atmosphere when the positive electrode composite material and / or the negative electrode composite material is laminated to the substrate. Figure 2 This is a graph showing an example of the relationship between the dew point of the atmosphere and the SHAP value when the positive electrode composite material is laminated on the substrate. Figure 3 This is a graph showing an example of the relationship between the dew point of the atmosphere when the negative electrode composite material is laminated on the substrate and the SHAP value.

[0046] According to the research of the present application discloser, Figure 2 The example in the paper shows a correlation between the dew point of the atmosphere when the positive electrode composite material is laminated onto the substrate and the SHAP value, i.e., the contribution of the dew point to durability performance. More specifically, when using a predetermined material and conducting studies under predetermined conditions other than the dew point, with -80.5°C as the threshold, the durability performance of multiple batteries differed. Specifically, when the dew point was below -80.5°C, the battery's durability was high, while when the dew point was above -80.5°C, the battery's durability was low.

[0047] In addition, as in Figure 3 For example, a correlation exists between the dew point of the atmosphere when the negative electrode composite material is laminated onto the substrate and the SHAP value, i.e., the contribution of the dew point to durability performance. More specifically, when a study was conducted using a predetermined material under predetermined conditions other than the dew point, with a threshold of -75°C, differences in the durability performance of multiple batteries were observed. Specifically, when the dew point was below -75°C, the battery's durability was high, while when the dew point was above -75°C, the battery's durability was low.

[0048] The manufacturing condition may also be the ratio of the weight per unit area of ​​the positive electrode composite material to the weight per unit area of ​​the negative electrode composite material. Figure 4 This is a graph showing an example of the relationship between the ratio of the weight per unit area of ​​the positive electrode layer material to the weight per unit area of ​​the negative electrode layer material and the SHAP value. In the present disclosure, weight per unit area means the mass of the electrode layer material per unit area.

[0049] According to the research of the present application discloser, Figure 4 The example in Figure 2 shows a correlation between the ratio of the weight per unit area of ​​the positive electrode composite material to the weight per unit area of ​​the negative electrode composite material and the SHAP value, i.e., the contribution of this ratio to durability performance. More specifically, when using predetermined materials and studying under predetermined conditions in addition to the aforementioned ratio, using thresholds of 1.123 and 1.167 resulted in differences in durability performance. Specifically, when the ratio was between 1.123 and 1.167, durability performance was high, while when the ratio was less than 1.123 or greater than 1.167, durability performance was low.

[0050] As a method for laminating the positive electrode composite material and the negative electrode composite material on the substrate, a method of preparing a composite material slurry by adding a dispersant to each composite material, applying the slurry to the substrate, and drying the slurry is exemplified.

[0051] For the purposes of this disclosure, a "multi-layer material" refers to a composition capable of constituting a positive electrode active material layer or a negative electrode active material layer, either as such or by further comprising other components. Furthermore, for the purposes of this disclosure, a "multi-layer material slurry" refers to a slurry that, in addition to the "multi-layer material," further comprises a dispersant, and can be applied and dried to form a positive electrode active material layer or a negative electrode active material layer.

[0052] Durability performance prediction process

[0053] like Figure 1As shown, the method of the present disclosure includes causing a computer to execute a process ( S102 ) of predicting the durability performance of a plurality of batteries based on the acquired manufacturing conditions according to the relationship between the manufacturing conditions of the batteries and the durability performance of the batteries.

[0054] In the method disclosed herein, the relationship between the manufacturing conditions of the battery and the durability performance of the battery can be a relationship predetermined by machine learning. Specifically, it can be based on the above Figures 2 to 4 The relationship of the diagram shown in the example.

[0055] For example, according to Figure 2 The graph shown in the example can predict the durability performance of multiple batteries based on the dew point of the atmosphere when the positive electrode composite material is laminated on the substrate obtained in the manufacturing condition acquisition process. Figure 2 As shown in , when the dew point is -80.5°C or lower, which is a threshold value, it can be predicted that the durability performance is high. When the dew point is higher than -80.5°C, it can be predicted that the durability performance is low.

[0056] According to Figure 3 The graph shown in the example can predict the durability performance of multiple batteries based on the dew point of the atmosphere when the negative electrode composite material is laminated on the substrate obtained in the manufacturing condition acquisition process. Figure 3 As shown in , when the dew point is -75°C or lower, which is a threshold value, it can be predicted that the durability performance is high. When the dew point is higher than -75°C, it can be predicted that the durability performance is low.

[0057] According to Figure 4 The chart shown in the example can predict the durability performance of multiple batteries based on the ratio of the unit area weight of the positive electrode composite material to the unit area weight of the negative electrode composite material obtained in the manufacturing condition acquisition step. Figure 4 As shown in , when the ratio is 1.123 or more and 1.167 or less, which is a threshold value, it can be predicted that the durability performance is high. When the ratio is less than 1.12 or greater than 1.167, it can be predicted that the durability performance is low.

[0058] Battery sorting process

[0059] As in Figure 1 As shown in , the method of the present disclosure includes classifying a plurality of batteries according to similarities of predicted durability performance ( S103 ).

[0060] In the method disclosed herein, it is possible to Figures 2 to 4 The graph shown in the example determines the similarity of durability performance.

[0061] For example, according to Figure 2In the graph illustrated in Figure 2 when the dew point of the atmosphere at the time of laminating the positive electrode composite material onto the base material is -80.5°C or lower as a threshold value, the SHAP value is set to -0.004. Further, according to the graph illustrated in

[0062] when the dew point of the atmosphere at the time of laminating the positive electrode composite material onto the base material is greater than -80.5°C, the SHAP value is set to 0.004. In this case, for batteries with equal SHAP values, it can be determined that their durability performances are similar to each other. Thus, multiple batteries can be classified according to the similarity of the durability performance. Figure 3 In the graph illustrated in Figure 3 when the dew point of the atmosphere at the time of laminating the negative electrode composite material onto the base material is -75°C or lower as a threshold value, the SHAP value is set to -0.005. Further, according to the graph illustrated in

[0063] when the dew point of the atmosphere at the time of laminating the negative electrode composite material onto the base material is greater than -75°C, the SHAP value is set to 0.004. In this case, for multiple batteries with equal SHAP values, it can be determined that their durability performances are similar to each other. Thus, multiple batteries can be classified according to the similarity of the durability performance.

[0064] ･SHAP value = -0.009… Grade A

[0065] ･ -0.009 < SHAP value < 0.008… Grade B

[0066] ･SHAP value = 0.008… Grade C.

[0067] In addition, in this case, the SHAP value is a value obtained by adding the dew point of the atmosphere at the time of laminating the positive electrode composite material onto the base material and the dew point of the atmosphere at the time of laminating the negative electrode composite material onto the base material. Further, in the order of durability performance from high to low, it is set as Grade A, Grade B, and Grade C.

[0068] According to the graph illustrated in Figure 4 when the ratio of the weight per unit area of the positive electrode composite material to the weight per unit area of the negative electrode composite material is 1.123 or more and 1.167 or less as a threshold value, the SHAP value is set to -0.02. Further, according to the graph illustrated in Figure 4 when this ratio is less than 1.123, the SHAP value is set to 0.07. According to the graph illustrated in Figure 4In the example graph shown in Figure 2, the SHAP value can be set to 0.06 when the ratio is greater than 1.167. In this case, batteries with equal SHAP values ​​can be determined to have similar durability performance. This allows multiple batteries to be categorized based on the similarity of their durability performance.

[0069] Furthermore, a plurality of batteries can be classified as follows based on the SHAP values ​​set as described above.

[0070] ･SHAP value = -0.02…Rank A

[0071] ･SHAP value ≤ 0.06…Rank B.

[0072] In this case, the durability performance is ranked from high to low, and the grades are A and B.

[0073] Method for manufacturing battery module

[0074] As in Figure 5A 、 5B As illustrated in FIG, the method of manufacturing a battery module of the present disclosure includes connecting batteries having similar durability performance, which are classified by the method of classifying batteries of the present disclosure, in parallel or in series. Figure 5A 1 is a schematic diagram illustrating battery modules connected in parallel by the method of manufacturing the battery module disclosed herein. Figure 5B This is a schematic diagram illustrating battery modules connected in series by the method of manufacturing the battery module disclosed herein.

[0075] The authors of this application have discovered that combining batteries with different durability performance can cause the following problems. Specifically, when batteries with different durability performance are connected in parallel, current loads vary significantly, potentially leading to rapid battery degradation. Furthermore, from the perspective of battery stability, connecting batteries with different durability performance in series is not preferred.

[0076] In this regard, the present inventors have discovered that in a battery module manufactured by the method disclosed herein, since batteries having similar durability performance are connected in parallel or in series, the above-mentioned problems are less likely to occur.

[0077] Regarding the method of classifying batteries disclosed herein, reference can be made to the above description related to the method of classifying batteries disclosed herein.

[0078] There is no particular limitation on the method of connecting batteries having similar durability in parallel or in series. Figure 5A 、 5BAs an example, a method of connecting batteries of rank A, rank B, or rank C in parallel or in series is described. The method of connecting batteries in parallel or in series is not particularly limited, and a method commonly used in this field can be employed.

[0079] Battery classification system

[0080] As in Figure 6 As shown in the example, the battery classification system 1 of the present disclosure includes an acquisition unit 11, a prediction unit 12, and a classification unit 13. The acquisition unit 11 acquires the manufacturing conditions of multiple batteries. The prediction unit 12 predicts the durability performance of the multiple batteries based on the manufacturing conditions acquired by the acquisition unit, based on the relationship between the battery manufacturing conditions and the battery durability performance. The classification unit 13 classifies the multiple batteries based on the similarity of the durability performance predicted by the prediction unit. The battery classification system of the present disclosure can execute the method of the present disclosure for classifying batteries.

[0081] As in Figure 6 As shown in the example, the battery classification system of the present disclosure may include a CPU (Central Processing Unit) 10, a memory 20, an input port 30, and a sensor 40. In this case, the CPU can function as an acquisition unit, a prediction unit, and a classification unit. Signals from sensors and the like can be input to the CPU via the input port.

[0082] Acquisition Department

[0083] The acquisition unit 11 acquires the manufacturing conditions of the plurality of batteries. More specifically, the acquisition unit 11 can acquire the manufacturing conditions of the plurality of batteries based on signals from the sensor 40 that measures parameters related to the manufacturing conditions of the batteries. The acquisition unit 11 can be one or more CPUs 10 and their peripheral circuits.

[0084] When the manufacturing condition is the dew point of the atmosphere when the positive electrode composite material and / or the negative electrode composite material is laminated on the substrate, the acquisition unit 11 can acquire the above-mentioned dew points as the manufacturing condition based on a signal from a dew point sensor, for example.

[0085] When the manufacturing condition is the ratio of the weight per unit area of ​​the positive electrode mixture material to the weight per unit area of ​​the negative electrode mixture material, the acquisition unit 11 can acquire the ratio as the manufacturing condition based on a signal from a mass sensor, for example.

[0086] Forecasting Department

[0087] The prediction unit 12 predicts the durability of multiple batteries based on the manufacturing conditions acquired by the acquisition unit 11, based on the relationship between the manufacturing conditions of the batteries and the durability of the batteries. Regarding the relationship between the manufacturing conditions of the batteries and the durability of the batteries, reference can be made to the above-mentioned description related to the method of classifying batteries disclosed herein. This relationship can be stored in the memory 20, such as a volatile semiconductor memory (such as RAM), a non-volatile semiconductor memory (such as ROM), a hard disk drive (HDD), a solid-state drive (SSD), or an optical recording medium. The prediction unit 12 can predict the durability of multiple batteries based on information related to the manufacturing conditions of the multiple batteries acquired by the acquisition unit 11 and the above-mentioned relationship read from the memory 20. The prediction unit 12 can be one or more CPUs (Central Processing Units) and their peripheral circuits.

[0088] Classification Department

[0089] The classification unit 13 classifies the plurality of batteries according to the similarity of the predicted durability performance. The classification unit 13 can classify the plurality of batteries according to the information related to the similarity of the durability performance predicted by the prediction unit 12. For example, the plurality of batteries can be classified by labeling batteries with similar durability performance, grouping batteries with similar durability performance together, etc. Regarding the method of determining the similarity of durability performance, reference can be made to the above-mentioned description related to the method of classifying batteries disclosed in the present invention. The classification unit 13 can be one or more CPUs (Central Processing Units) and their peripheral circuits. In addition, the classification results can be stored in the memory 20, such as a volatile semiconductor memory (such as RAM), a non-volatile semiconductor memory (such as ROM), a hard disk drive (HDD), a solid-state drive (SSD), or an optical recording medium.

Claims

1. A battery classification method, comprising causing a computer to execute the following steps: Obtaining manufacturing conditions for multiple batteries; Based on the relationship between the manufacturing conditions of the batteries and the durability of the batteries, predicting the durability of the plurality of batteries according to the obtained manufacturing conditions; and The plurality of batteries are classified according to the predicted similarities in the durability performance.

2. The method according to claim 1, wherein The manufacturing conditions are the dew point of the atmosphere when the positive electrode mixture material and / or the negative electrode mixture material is laminated on the substrate or the ratio of the weight per unit area of ​​the positive electrode mixture material to the weight per unit area of ​​the negative electrode mixture material.

3. The method according to claim 1, wherein The relationship is a relationship predetermined by machine learning.

4. A method for manufacturing a battery module, comprising: Batteries having similar durability performance classified by the method according to any one of claims 1 to 3 are connected in parallel or in series.

5. A battery classification system comprising: an acquiring unit, which acquires manufacturing conditions of a plurality of batteries; a prediction unit that predicts the durability performance of the plurality of batteries based on the manufacturing conditions acquired by the acquisition unit, based on a relationship between the manufacturing conditions of the batteries and the durability performance of the batteries; and The classification unit classifies the plurality of batteries based on the similarity of the durability performance predicted by the prediction unit.

Images