Secondary battery life prediction method

Abstract

To provide a secondary battery life prediction method which can improve accuracy of secondary battery life prediction when recycling a secondary battery.SOLUTION: There is a control part that has a function to calculate usage history information and SOH of a secondary battery as a life prediction index. The secondary battery life prediction method includes: a first index calculation process (steps ST8 to ST11) of calculating a life prediction index of the secondary battery collected after primary use; and a second index calculation process (step ST13) of calculating a life prediction index for the secondary battery collected after secondary use, wherein the remaining life of the secondary battery is predicted by identifying parameters for life prediction through machine learning of calculation results of the first index calculation process and calculation results of the second index calculation process (step ST15).SELECTED DRAWING: Figure 3

Description

[Technical field]

[0001] The present invention relates to a method for predicting the life span of a secondary battery, and more particularly to a method for improving the accuracy of predicting the life span of a secondary battery when recycling the secondary battery. [Background technology]

[0002] There is a growing need to recycle (secondarily reuse) secondary batteries installed in electric vehicles (hereinafter sometimes referred to as vehicles) for purposes other than vehicles after the vehicles are discarded. For example, there is a need to recycle them into small electric carts, electric wheelchairs, automatic transport robots, household batteries, disaster batteries, etc.

[0003] When recycling secondary batteries, it is useful to predict the lifespan (hereinafter sometimes referred to as remaining lifespan) of secondary batteries with high accuracy, as this leads to an accurate understanding of the utility value of secondary batteries. For this reason, there has been a demand for technology to predict the remaining lifespan of secondary batteries with high accuracy.

[0004] Patent Document 1 discloses a technology for evaluating the deterioration of secondary batteries when recycling them. This patent document 1 discloses that the usage history of secondary batteries is provided as added value to provide a more accurate state to the recycler of the secondary battery and to promote the use and protection of the secondary battery. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] JP 2022-147162 A Summary of the Invention [Problem to be solved by the invention]

[0006] However, Patent Document 1 does not disclose any technology for predicting the remaining life of a secondary battery with high accuracy.

[0007] In the future, in order to promote the Sustainable Development Goals (SDGz), technology that can predict the remaining lifespan of secondary batteries with high accuracy will become important in order to revitalize the secondary battery recycling market.

[0008] The present invention has been made in consideration of the above points, and its object is to provide a method for predicting the life span of a secondary battery that can improve the accuracy of predicting the life span (prediction of remaining life span) of the secondary battery when recycling the secondary battery. [Means for solving the problem]

[0009] The solution of the present invention for achieving the above object is directed to a method for predicting the remaining life of a secondary battery, which uses a secondary battery database that accumulates the usage history of the secondary battery during primary use and the usage history of the secondary battery during secondary use, and a control unit having a function of calculating at least one of the usage history information, remaining life information, SOC (State Of Charge), and SOH (State Of Health) of the secondary battery as a life prediction index, and includes a first index calculation step of calculating the life prediction index of the secondary battery collected after the primary use, and a second index calculation step of calculating the life prediction index of the secondary battery collected after the secondary use, and is characterized in that the remaining life of the secondary battery is predicted by identifying parameters for life prediction through machine learning of at least the calculation results in the first index calculation step and the calculation results in the second index calculation step.

[0010] With this specification, it has been difficult to obtain sufficient accuracy in predicting the life span of secondary batteries using only the measurement results after primary use. However, by using machine learning to identify parameters for life span prediction using the calculation results of the life span prediction index of secondary batteries collected after primary use and the calculation results of the life span prediction index of secondary batteries collected after secondary use (the calculation results of the life span prediction index in a state that is further deteriorated than after primary use), it is possible to improve the accuracy of life span prediction for secondary batteries. Effect of the Invention

[0011] In the present invention, the remaining life of the secondary battery is predicted by identifying life prediction parameters through machine learning of the calculation results of the life prediction index of the secondary battery collected after the primary use and the calculation results of the life prediction index of the secondary battery collected after the secondary use, thereby improving the accuracy of the life prediction of the secondary battery. [Brief description of the drawings]

[0012] [Figure 1] 1 is a diagram showing a schematic configuration of a secondary battery life prediction system according to an embodiment. [Diagram 2] 1 is a diagram for explaining factors that deteriorate the accuracy of SOH measurement of a secondary battery. [Diagram 3] 4 is a flowchart showing a procedure of a secondary battery life prediction process according to the embodiment. FIG. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0013] Hereinafter, an embodiment of the present invention will be described with reference to the drawings. In this embodiment, a case where the present invention is applied as a method for predicting the life of a secondary battery (also called a driving battery) mounted on an electric vehicle will be described.

[0014] -Outline of secondary battery life prediction system- Fig. 1 is a diagram showing a schematic configuration of a secondary battery life prediction system 1 according to this embodiment. As shown in Fig. 1, the secondary battery life prediction system 1 includes a primary usage history database 2, a primary usage SOH measurement unit 3, a secondary usage history database 4, a secondary usage SOH measurement unit 5, and a life prediction device 6, which are connected to each other so that they can transmit and receive information.

[0015] The primary usage history database 2 accumulates information on the usage history (primary usage history) of the secondary battery BA mounted in an electric vehicle (hereinafter sometimes simply referred to as a vehicle) VE. Specifically, the primary usage history database 2 stores individual information (secondary battery ID) of the secondary battery BA mounted in the vehicle VE and history information on the usage state related to charging and discharging of the secondary battery BA, which are linked together. The usage state related to charging and discharging of the secondary battery BA includes a number of parameters that affect the life of the secondary battery BA, such as the number of times it has been overcharged during charging, the number of times it has been charged in an extremely low temperature environment, and the number of times it has been charged in a high temperature environment. The primary usage history database 2 may be mounted in the vehicle VE, or may be a device (such as a data server) capable of communicating information with the vehicle VE.

[0016] The primary use SOH measurement unit 3 measures the SOH of a secondary battery BA that has been collected (after primary use has ended) in a storage facility VA after the vehicle VE has been scrapped. A well-known method can be used to measure this SOH. This SOH measurement result 1MR is obtained as a predicted value of the remaining life at the time when the primary use ends (hereinafter referred to as the primary measurement result 1MR). Information on this primary measurement result 1MR can be output to the life prediction device 6. Therefore, the primary use SOH measurement unit 3 corresponds to the control unit (a control unit having a function of calculating a life prediction index for a secondary battery) in the present invention, and the primary measurement result 1MR corresponds to the life prediction index (a life prediction index for a secondary battery collected after primary use) in the present invention.

[0017] The secondary usage history database 4 accumulates information on the usage history (secondary usage history) of the secondary battery BA when the secondary battery BA is used secondarily by a secondary usage device (e.g., a small electric cart or an electric wheelchair, etc.) SU. Specifically, the secondary usage history database 4 also stores, in association with individual information (secondary battery ID) of the secondary battery BA and history information on the usage state relating to charging and discharging of the secondary battery BA. The usage state relating to charging and discharging of the secondary battery BA also includes the above-mentioned multiple parameters that affect the lifespan of the secondary battery BA.

[0018] The secondary use SOH measurement unit 5 measures the SOH of a secondary battery BA that has been recovered (secondary use completed) at a recovery facility CF after secondary use in a secondary use device SU has ended. This SOH measurement result 2MR is obtained as a predicted value of the remaining life at the time of the end of secondary use (hereinafter referred to as the secondary measurement result 2MR). Information on this secondary measurement result 2MR can be output to a life prediction device 6. Therefore, the secondary use SOH measurement unit 5 also corresponds to a control unit in the present invention, and the secondary measurement result 2MR also corresponds to a life prediction index in the present invention (a life prediction index for a secondary battery recovered after secondary use).

[0019] The life prediction device 6 utilizes the primary usage history information, the secondary usage history information, the primary measurement result 1MR, and the secondary measurement result 2MR to identify parameters of a predefined prediction formula (a formula for predicting the remaining life of the secondary battery BA). That is, the device 6 receives the primary usage history information from the primary usage history database 2, the secondary usage history information from the secondary usage history database 4, the primary measurement result 1MR from the primary usage SOH measurement unit 3, and the secondary measurement result 2MR from the secondary usage SOH measurement unit 5, and predicts the life (remaining life) of the secondary battery BA by identifying parameters for life prediction through machine learning using these pieces of information. That is, the device 6 predicts the remaining life of the secondary battery BA by identifying parameters for life prediction using both the primary measurement result 1MR in a state in which the secondary battery BA has deteriorated (has had its life shortened) by being used for a predetermined period of time since it was new, and the secondary measurement result 2MR in a state in which the secondary battery BA has further deteriorated (has had its life shortened) by secondary use after the primary use. The life prediction parameters identified by this life prediction device 6 are output to the primary usage SOH measurement unit 3, and are reflected in the measurement of the SOH of the secondary battery BA in this primary usage SOH measurement unit 3 (the next measurement of the SOH of the secondary battery BA).

[0020] -Factors behind the deterioration of SOH prediction accuracy- In order to predict the remaining life of the secondary battery BA with high accuracy, it is necessary to measure the SOH of the secondary battery BA with high accuracy, but it is expected that the accuracy of the SOH measurement of the secondary battery BA will deteriorate due to the influence of various conditions. The factors that deteriorate the accuracy of the SOH measurement of the secondary battery BA will be explained below. Figure 2 is a diagram for explaining these factors.

[0021] Factors that cause the secondary battery BA to deteriorate include Li deposition, reduction in Li ions, rust, and reduced reaction amount of active materials due to high-speed charging, charging in an extremely low temperature environment, charging in a high temperature environment, and high charge state (overcharging), etc. In addition, the degree of deterioration due to each factor differs, so in order to accurately predict the remaining life of a secondary battery BA, it is necessary to accurately understand the deterioration state.

[0022] Consider the secondary battery α (see FIG. 2(a)) and secondary battery β (see FIG. 2(b)) in a usage state in which the degree of factors that cause the secondary battery BA to deteriorate is different, as shown in FIG. 2(a). As shown in FIG. 2(a), factors A, B, and C are dominant in the secondary battery α, and the deterioration progresses. On the other hand, as shown in FIG. 2(b), factor A is dominant in the secondary battery β, and the deterioration progresses. In this case, since the factors of deterioration of the secondary batteries α and β are different from each other, the amount of Li deposition and the amount of Li ion reduction are different, and the deterioration state of the secondary batteries is also different. The solid line in FIG. 2(c) shows an example of the deterioration progress (change in SOH) of the secondary battery α, and the dashed line in FIG. 2(c) shows an example of the deterioration progress of the secondary battery β. In other words, even if the SOH (corresponding to the remaining battery life) is the same when the secondary batteries α and β are used only for the period T1, the change in SOH in the subsequent period ΔT may be significantly different. For this reason, in order to accurately predict the remaining life of the secondary battery BA, it is necessary to accurately grasp the degradation state (the dominant factor of degradation).

[0023] In this embodiment, the remaining life of the secondary battery BA is predicted by identifying parameters for life prediction through machine learning using information on the primary usage history, information on the secondary usage history, the primary measurement result 1MR, and the secondary measurement result 2MR, making it possible to predict the remaining life of the secondary battery BA with high accuracy.

[0024] - Secondary battery life prediction processing - Next, the procedure of the secondary battery life prediction process according to this embodiment will be described with reference to the flowchart of FIG.

[0025] The procedure for this secondary battery life prediction process involves primary use data accumulation process (steps ST1 to ST7), life prediction process after primary use ends (steps ST8 to ST11), secondary use data accumulation process (step ST12), life prediction process after secondary use ends (step ST13), and data utilization process (steps ST14, ST15), in that order.

[0026] First, in step ST1, the ID (secondary battery ID) of the secondary battery to be charged is confirmed. After that, the database of the secondary battery BA confirmed (whose ID has been confirmed) in step ST1 is collated with the database of the secondary battery BA stored in the primary usage history database 2 (step ST2).

[0027] When charging of the secondary battery BA is started in step ST3, a charging log is output in step ST4, and this charging log is stored. While charging of the secondary battery continues in this manner, in step ST5, the information on the secondary battery ID, the charger, and the charging log is hashed (data is converted into an irregular character string using a hash function algorithm), and this information is sent to the primary usage history database 2 and registered in the primary usage history database 2 (step ST6). The information registered here may also include other information such as charging and discharging patterns, cell temperature, and environmental temperature.

[0028] After charging of the secondary battery BA is completed in step ST7, in a situation where the vehicle VE continues to be used, the processes of steps ST1 to ST7 are repeated every time the secondary battery BA is charged or discharged. This causes primary usage history information to be accumulated in the primary usage history database 2. That is, the aforementioned information on the secondary battery ID, charger, and charging log is hashed, and this information is transmitted to the primary usage history database 2 and registered therein.

[0029] When the primary use of the secondary battery BA ends due to, for example, the disposal of the vehicle VE, and the secondary battery BA is collected in the storage facility VA, a process of predicting the remaining life of the secondary battery BA is started in step ST8. In this prediction process, information stored in the primary use history database 2 is acquired to confirm the charging log (step ST9), and machine learning is performed by AI based on this charging log (step ST10), thereby calculating the remaining life at the present time (step ST11). The primary measurement result 1MR calculated here is output to the life prediction device 6. Therefore, the process of step ST11 corresponds to the first index calculation step in the present invention.

[0030] When the secondary battery BA is used secondarily by the secondary usage device SU, in step ST12, charging and discharging processing during this secondary usage (secondary usage data accumulation processing) is performed. This processing is the same as the above-mentioned primary usage data accumulation processing (steps ST1 to ST7). That is, each piece of information on the secondary battery ID, the charger, and the charging log during secondary usage is hashed, and this information is transmitted to the secondary usage history database 4 and registered in the secondary usage history database 4.

[0031] In step ST13, machine learning is performed by AI based on the registered charging log, and the final SOH (at the time when the secondary use ends) is measured. The SOH measured here is output to the life prediction device 6 as the secondary measurement result 2MR. Therefore, the process of step ST13 corresponds to the second index calculation step in the present invention.

[0032] In step ST14, the life prediction device 6 compares the primary measurement result 1MR with the secondary measurement result 2MR, thereby grasping the deterioration state of the secondary battery BA during primary use and the deterioration state of the secondary battery BA during secondary use. Then, in step ST15, the primary measurement result 1MR and the secondary measurement result 2MR and other information (information on the primary use history and information on the secondary use history, etc.) are utilized to identify parameters for life prediction by machine learning, thereby predicting the remaining life of the secondary battery BA. Then, the identified parameters are used for machine learning in step ST10, and are reflected in the measurement of the SOH of the secondary battery BA by the primary use SOH measurement unit 3, thereby contributing to improving the accuracy of life prediction after the end of the primary use of the secondary battery BA.

[0033] -Effects of the embodiment- As described above, according to this embodiment, the remaining life of the secondary battery BA is predicted by identifying parameters for life prediction through machine learning using the calculation result (first measurement result 1MR) of the life prediction index of the secondary battery BA collected after the primary use and the calculation result (second measurement result 2MR) of the life prediction index of the secondary battery BA collected after the secondary use, so that the accuracy of the life prediction of the secondary battery BA can be improved. For example, the remaining life after the primary use of other secondary batteries BA produced in the same lot (with the same physical properties) as the secondary battery BA whose remaining life has been predicted can be predicted with high accuracy.

[0034] -Other embodiments- The present invention is not limited to the above-described embodiment, and all modifications and applications within the scope of the claims and equivalents thereto are possible.

[0035] For example, in the above embodiment, the present invention has been described as being applied to a method for predicting the life span of a secondary battery BA mounted on an electric vehicle, but the present invention is not limited to this, and can also be applied to a method for predicting the life span of other secondary batteries BA.

[0036] In the above embodiment, the secondary battery BA is used for a second time after the primary use. The present invention is not limited to this, and can be applied to a tertiary use of the secondary battery BA or a case where the secondary battery BA is recycled more than once. In this case, the life prediction parameters are identified by machine learning of the calculation result of the life prediction index in a further deteriorated state, so that the accuracy of the life prediction of the secondary battery BA can be further improved.

[0037] In the above embodiment, the case where the measurement result of SOH is used as the life prediction index has been described. However, the present invention is not limited to this, and other values ​​(usage history information, remaining life information, SOC, etc.) correlated with deterioration of the secondary battery BA may be used as the life prediction index to predict the remaining life of the secondary battery BA, or multiple values ​​correlated with deterioration of the secondary battery BA may be used as the life prediction index to predict the remaining life of the secondary battery BA.

[0038] In addition, in the above embodiment, the secondary battery database is provided with a primary usage history database 2 and a secondary usage history database 4 separately, but a single database may be configured to function as both the primary usage history database 2 and the secondary usage history database 4. [Industrial Applicability]

[0039] INDUSTRIAL APPLICABILITY The present invention is applicable to a life prediction method for predicting the life of a secondary battery when recycling the secondary battery. [Explanation of symbols]

[0040] 2...Primary usage history database (secondary battery database) 3... Primary use SOH measurement unit (control unit) 4. Secondary usage history database (secondary battery database) 5...Secondary use SOH measurement unit (control unit) 6...Life prediction device BA...Secondary battery 1MR: 1st measurement result (life expectancy prediction index) 2MR: 2nd measurement result (life expectancy prediction index)

Claims

[Claim 1] A method for predicting a remaining life of a secondary battery, comprising: a secondary battery database that accumulates a usage history of the secondary battery during primary use and a usage history of the secondary battery during secondary use; and a control unit having a function of calculating at least one of usage history information, remaining life information, SOC, and SOH of the secondary battery as a life prediction index, a first index calculation step of calculating the life prediction index of the secondary battery collected after the primary use; a second index calculation step of calculating the life prediction index of the secondary battery collected after the secondary reuse, A method for predicting the remaining life of a secondary battery, comprising: identifying life prediction parameters through machine learning of at least the calculation results in the first index calculation process and the calculation results in the second index calculation process, thereby predicting the remaining life of the secondary battery.

Images