Deterioration state prediction method, deterioration state prediction device, and deterioration state prediction program

Abstract

To accurately predict the state of degradation of a secondary battery.SOLUTION: The present invention relates to a method for predicting the state of degradation which calculates a cause-by-cause degradation state 25 and predicts a degradation state 26 of a secondary battery 1 from a plurality of degradation states 25 for various different causes. The cause-by-cause degradation states 25 is calculated on the basis of a previous cause-by-cause degradation state 21 as a cause-by-cause degradation state 25 an arbitrary first time before the current cause-by-cause degradation state and a unit cause-by-cause degradation state as the degradation for one hour in consideration of a time dependency which is different according to cause of degradation and is according to a power method for the passage of time of the cause-by-cause degradation state 25 and the rate of degradation which is different by causes of degradation and is determined by the usage condition 22 at the time of prediction.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a degradation state prediction method, a degradation state prediction device, and a degradation state prediction program for predicting the degradation state of a chargeable and dischargeable secondary battery. [Background technology]

[0002] A secondary battery can supply stored power to various devices by discharging the power it has been charged with. Repeated charging and discharging or long-term storage of a secondary battery can cause deterioration, such as a decrease in full charge capacity. To ensure healthy and continuous use of the secondary battery, the deterioration state of the secondary battery may be predicted.

[0003] For example, Patent Document 1 discloses a method for calculating the degradation state of a secondary battery from a power law of time with the degradation rate as a coefficient. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] International Publication No. 2019 / 171688 Summary of the Invention [Problem to be solved by the invention]

[0005] However, the degradation process varies greatly depending on the conditions of use, and these conditions change constantly while the secondary battery is in use. In particular, when a single energy storage device is used for multiple applications, such as connecting an electric vehicle to the power grid for power leveling, the charging and discharging methods and conditions for the electric vehicle and for power leveling are significantly different. This makes it difficult to accurately predict the degradation state.

[0006] An object of the present invention is to accurately predict the degradation state of a secondary battery. [Means for solving the problem]

[0007] In order to achieve the above object, a characteristic configuration of a degradation state prediction method according to one embodiment of the present invention is a degradation state prediction method that calculates a cause-specific degradation state for each degradation cause and predicts a degradation state of a secondary battery from a plurality of the cause-specific degradation states, and each of the cause-specific degradation states is calculated based on a prior cause-specific degradation state, which is the cause-specific degradation state an arbitrary first time ago, and a unit cause-specific degradation state that deteriorates during the first time, taking into consideration time dependency of the cause-specific degradation state that differs depending on the degradation cause and that follows a power law with respect to elapsed time, and a degradation rate that differs depending on the degradation cause and that is determined by the usage conditions at the time of prediction.

[0008] A characteristic configuration of a degradation state prediction device according to one embodiment of the present invention is that the degradation state prediction device predicts the degradation state of a secondary battery, and comprises a degradation state acquisition unit for each degradation cause that acquires a degradation state for each degradation cause by a prior cause, which is a cause-specific degradation state acquired an arbitrary first time before for each degradation cause of the secondary battery; a usage condition acquisition unit that acquires the usage conditions at the time of prediction for predicting the degradation state of the secondary battery; a degradation rate coefficient calculation unit that determines a degradation rate coefficient corresponding to the degradation rate of the secondary battery for each degradation cause using the usage conditions; a cause-specific degradation state calculation unit that calculates the cause-specific degradation state for each degradation cause based on a time coefficient related to a predetermined elapsed time for each degradation cause, the degradation rate coefficient, and the degradation state for each prior cause; and a degradation state calculation unit that predicts the degradation state of the secondary battery from the multiple cause-specific degradation states calculated for each degradation cause.

[0009] A characteristic configuration of a degradation state prediction program according to one embodiment of the present invention is a degradation state prediction program for predicting the degradation state of a secondary battery, and has a computer execute the following functions: a function for acquiring a pre-cause-specific degradation state for each degradation cause of the secondary battery, which is a cause-specific degradation state obtained an arbitrary first time in advance for each of the degradation causes of the secondary battery; a function for acquiring usage conditions at the time of prediction for the degradation state of the secondary battery; a function for determining a degradation rate coefficient corresponding to the degradation rate of the secondary battery for each degradation cause using the usage conditions; a function for calculating the cause-specific degradation state for each degradation cause based on a time coefficient relating to a predetermined elapsed time for each degradation cause, the degradation rate coefficient, and the pre-cause-specific degradation state; and a function for predicting the degradation state of the secondary battery from the multiple cause-specific degradation states calculated for each of the degradation causes.

[0010] The deterioration of a secondary battery progresses due to the influence of multiple causes of deterioration. Furthermore, the rate of deterioration (degradation state) varies depending on the cause of deterioration, and the rate of deterioration for each cause changes in a complex manner depending on the conditions of use. While it is difficult to comprehensively calculate the degradation state of a secondary battery by taking into account multiple causes of deterioration, it is relatively easy to calculate the degradation state for each cause individually. Therefore, by calculating the degradation state for each cause of deterioration individually and predicting the degradation state of the secondary battery from the degradation state for each cause, it is possible to easily and accurately predict the degradation state of a secondary battery.

[0011] Furthermore, as a result of intensive research by the inventors, they have found that the deterioration state for each cause of deterioration has time dependency according to a power law and depends on the deterioration rate, which differs depending on the cause of deterioration. In other words, they have found that the deterioration state for each cause of deterioration can be calculated based on the deterioration rate and time dependency.

[0012] Here, even if the cause of deterioration is the same, the rate of deterioration changes from time to time depending on various usage conditions such as temperature conditions, charge / discharge current, battery voltage, etc. Therefore, the rate of deterioration is not uniform from the start of use to the time the deterioration state is predicted, and if a fixed rate of deterioration is used for each cause of deterioration, it is difficult to accurately calculate the deterioration state (cause-specific deterioration state).

[0013] Therefore, the cause-specific degradation state is calculated based on the prior cause-specific degradation state, which is the degradation state up to a certain time immediately before (the cause-specific degradation state), and the unit cause-specific degradation state, which is the degradation state from that time until a first hour has elapsed (the cause-specific degradation state). This allows only the degradation rate during the first hour to be considered, thereby minimizing the impact of changes in usage conditions on the degradation rate. As a result, it becomes possible to accurately predict the degradation state of a secondary battery.

[0014] Furthermore, instead of using a fixed degradation rate for each degradation cause, the degradation state (cause-specific degradation state) is calculated taking into account the degradation rate calculated according to the usage conditions at the time of prediction (current time). This allows the degradation state of the secondary battery to be predicted with high accuracy.

[0015] The deterioration state by cause is expressed as follows: the time of prediction is t, the first time is δ, and the deterioration state by cause is μ t , the deterioration state by each prior cause is μ (t-δ) , where α is a degradation rate coefficient corresponding to the degradation rate, and β is a time coefficient indicating the time dependency, μ t ={μ (t-δ) 1 / β -α 1 / β ·δ} β It may be found as:

[0016] With this configuration, the deterioration state of the secondary battery can be predicted easily and accurately.

[0017] The deterioration state may be determined by adding up all the calculated deterioration states for each cause.

[0018] With this configuration, the deterioration state can be predicted by easily and accurately taking into account the deterioration states for a plurality of causes, and the deterioration state of the secondary battery can be predicted easily and accurately.

[0019] The deterioration state for each prior cause may be an actual measurement value obtained by any method.

[0020] With this configuration, it is possible to predict the deterioration state by cause from the start of use to the prediction time by simply predicting the deterioration state by cause during the first time period using the actual deterioration state by cause prior to use, thereby enabling the deterioration state of the secondary battery to be predicted with high accuracy.

[0021] Furthermore, the deterioration state by prior cause may be a deterioration state by cause calculated based on the deterioration state by prior cause a second time before the first time and the deterioration state by unit cause between the second time and the first time.

[0022] With this configuration, by repeatedly predicting the deterioration state by cause over a short period of time during which changes in usage conditions are expected to be limited, it is possible to predict the deterioration state by cause from the start of use to the time of prediction, thereby enabling accurate prediction of the deterioration state of the secondary battery.

[0023] The deterioration rate may be determined using temperature information, current information, and voltage information as the usage conditions, where the temperature information includes at least one of cell temperature, ambient temperature, and air temperature, the current information includes at least one of charge current, discharge current, charge rate, discharge rate, total discharged amount of electricity, and total charged amount of electricity, and the voltage information may include at least one of average voltage, average state of charge, upper limit charge voltage, lower limit discharge voltage, time spent within a specified range of state of charge, and overvoltage.

[0024] With this configuration, it is possible to select appropriate usage conditions depending on the cause of deterioration and determine the deterioration rate using the selected usage conditions, thereby making it possible to accurately predict the deterioration state of the secondary battery.

[0025] In addition, the overvoltage may be determined by applying current to the secondary battery for a predetermined period of time and then stopping the application of current, and subtracting the voltage value of the secondary battery before the application of current to the secondary battery is started from the voltage value of the secondary battery when a predetermined third hour has elapsed since the application of current to the secondary battery is started.

[0026] With this configuration, the overvoltage can be easily and accurately determined, and the deterioration rate can be determined accurately, which in turn allows the deterioration state of the secondary battery to be predicted accurately.

[0027] In addition, the overvoltage may be determined by turning on and off the secondary battery for a predetermined period of time, and subtracting the voltage value of the secondary battery when a predetermined fourth hour has elapsed since the power supply was stopped from the voltage value of the secondary battery while it was powered on.

[0028] With this configuration, the overvoltage can be easily and accurately determined, and the deterioration rate can be determined accurately, which in turn allows the deterioration state of the secondary battery to be predicted accurately.

[0029] The overvoltage may be calculated as the product of the charging current and the internal resistance of the secondary battery.

[0030] With this configuration, the overvoltage can be easily and accurately determined, and the deterioration rate can be determined accurately, which in turn allows the deterioration state of the secondary battery to be predicted accurately.

[0031] The deterioration rate may also be determined using usage history data including the temperature information history, the current information history, and the voltage information history, and cell state history data including at least one of battery capacity history, resistance history, battery voltage history, output history, external dimensions, and confinement pressure.

[0032] With this configuration, the deterioration rate can be determined more accurately, and the deterioration state of the secondary battery can be predicted more accurately.

[0033] The deterioration rate may also be determined by machine learning the usage history data and the cell state history data.

[0034] With this configuration, the deterioration rate can be determined more accurately, and the deterioration state of the secondary battery can be predicted more accurately. [Brief explanation of the drawings]

[0035] [Figure 1] FIG. 2 is a diagram illustrating a functional configuration of a degradation state prediction device. [Figure 2] FIG. 10 is a diagram illustrating a flow for predicting a deterioration state. [Figure 3] FIG. 10 is a diagram comparing a predicted deterioration state taking one deterioration cause into consideration with an actual measurement value. [Figure 4] FIG. 10 is a diagram comparing a predicted deterioration state taking into account two causes of deterioration with an actual measurement value. [Figure 5] FIG. 10 is a diagram comparing a degradation state prediction using a degradation rate coefficient determined under one use condition with an actual measurement value. [Figure 6] FIG. 10 is a diagram comparing the predicted degradation state using the degradation rate coefficient determined from three usage conditions with the actual measured value. DETAILED DESCRIPTION OF THE INVENTION

[0036] Deterioration of secondary batteries occurs due to various reaction mechanisms (causes of deterioration), such as an increase in the resistive film (SEI) at the electrode-electrolyte interface, deposition of metallic Li (electrode metal) on the electrode surface, cracks in the electrode active material, crystalline phase transition, and gas generation due to electrolyte decomposition. Therefore, when predicting the state of deterioration 26 of a secondary battery (see FIG. 1), a state of deterioration 26 (cause-specific state of deterioration 25) is determined for each cause of deterioration, and the state of deterioration 26 of the secondary battery is determined from one or more state of deterioration by cause 25. Note that the state of deterioration 26 is a numerical value that decreases from 1 as deterioration progresses, where the chargeable capacity of an unused secondary battery is set to 1, and is also referred to as the SOH (State of Health).

[0037] The deterioration of a secondary battery can be approximated by a power law with respect to the time (elapsed time) since the start of use. In other words, the deterioration of a secondary battery has a time dependency that is proportional to the power law of time and depends on the deterioration rate. This proportional relationship can be expressed as a constant using the deterioration rate coefficient 23 (see Figure 1) described below, and a power exponent using the time coefficient 24 (see Figure 1) described below.

[0038] Here, the deterioration rate coefficient 23 depends on various use conditions 22 (see FIG. 1 ), such as the use environment, the use method, and the use mode. Furthermore, the use conditions 22 often change while the secondary battery is in use. Therefore, if the deterioration state 26 (cause-specific deterioration state 25) of the secondary battery over a long period (long elapsed time) is calculated (predicted) using a constant deterioration rate coefficient 23, errors occur according to variations in the use conditions 22, making it difficult to accurately predict the deterioration state 26 (cause-specific deterioration state 25) of the secondary battery.

[0039] Therefore, the secondary battery's deterioration state by cause 25 is calculated based on the deterioration state by prior cause 21 (see Figure 1) obtained in advance and the deterioration state by unit cause from the time the deterioration state by prior cause 21 is calculated (measured) until an arbitrary first time has elapsed.

[0040] The prior cause-specific degradation state 21 is the degradation state 26 (cause-specific degradation state 25) of the secondary battery at a time point one hour before the present time (the time point at which the degradation state 26 of the secondary battery is predicted). The prior cause-specific degradation state 21 may be a predicted value (cause-specific degradation state 25) calculated one hour before the present time. In this case, the cause-specific degradation state 25 calculated one hour before the present time and a second hour before the first time are acquired as the prior cause-specific degradation state 21, and the cause-specific degradation state 25 is calculated, and such calculation of the cause-specific degradation state 25 is repeated. The prior cause-specific degradation state 21 may also be an actual measurement value measured (measured) one hour before the present time by a predetermined means. The prior cause-specific degradation state 21 may also be a predicted corrected value obtained by correcting the predicted value (cause-specific degradation state 25) calculated one hour before the present time by taking into account the actual measurement value measured (measured) one hour before the present time by a predetermined means. In addition, the deterioration state by unit cause is the deterioration state 26 (deterioration state by cause 25) of the secondary battery during the period from when the deterioration state by prior cause 21 was calculated / measured (acquired) to the present time, i.e., during the first hour.

[0041] The current (prediction time) degradation state 26 (cause-specific degradation state 25) has a proportional relationship with the time coefficient 24 as a power exponent, and therefore the degradation state 26 (cause-specific degradation state 25) is calculated based on the prior cause-specific degradation state 21 and the unit cause-specific degradation state, taking into account time dependency in accordance with a power law with respect to elapsed time. In other words, the degradation state 26 (cause-specific degradation state 25) is calculated based on the prior cause-specific degradation state 21 and the unit cause-specific degradation state that deteriorates during a first time period, taking into account the time dependency of the cause-specific degradation state 25 in accordance with a power law with respect to elapsed time, which differs depending on the cause of deterioration, and the degradation rate, which differs depending on the cause of deterioration and is determined by the usage conditions 22 at the time of prediction. For example, the current (prediction time) degradation state 26 (cause-specific degradation state 25) is calculated using Equation 1, which will be described later.

[0042] By calculating the deterioration state 26 of the secondary battery as described above, the deterioration state 26 can be calculated using the cause-specific deterioration state 25 calculated for each cause of deterioration, and the deterioration state 26 of the secondary battery can be calculated with high accuracy.

[0043] Furthermore, by calculating the cause-specific degradation state 25 using the unit cause-specific degradation state for the first time period, the period for calculating the degradation state 26 (unit cause-specific degradation state) can be shortened, thereby suppressing the influence of variations in the use conditions 22. As a result, the degradation rate coefficient 23 can be determined with high accuracy, and the unit cause-specific degradation state (cause-specific degradation state 25) can be calculated with high accuracy.

[0044] Furthermore, since the degradation state 26 (cause-specific degradation state 25) can be predicted using the prior cause-specific degradation state 21, the amount of calculation (prediction) required is reduced, and the degradation state 26 (cause-specific degradation state 25) of the secondary battery can be calculated more accurately. In particular, when the prior cause-specific degradation state 21 is an actual measurement value or a predicted corrected value corrected by taking the actual measurement value into account, the actual prediction is reduced to only the first time period, and the degradation state 26 (cause-specific degradation state 25) of the secondary battery can be calculated more accurately. Note that even if the prior cause-specific degradation state 21 is a calculated value, the usage conditions 22 can be determined for a short period, and it is only necessary to calculate the degradation state 26 (cause-specific degradation state 25) for each short period, so the degradation state 26 (cause-specific degradation state 25) of the secondary battery can be calculated more accurately.

[0045] [Deterioration state prediction] 1 and 2, the functional configuration of a degradation state prediction device and a degradation state prediction method for predicting a degradation state 26 of a secondary battery will be described. In the following explanation, a configuration for predicting a degradation state 26 of a lithium ion battery 1 will be described as an example of a secondary battery.

[0046] Lithium-ion battery 1 is capable of charging and discharging, and supplies power to various devices that consume power by discharging the charged power. An ammeter 2 is connected in series to lithium-ion battery 1. Ammeter 2 measures the current value of lithium-ion battery 1. In addition, a voltmeter 3 is connected in parallel to lithium-ion battery 1. Voltmeter 3 measures the voltage value of lithium-ion battery 1.

[0047] A battery capacity measurement unit 5 is connected to the lithium ion battery 1, and the battery capacity measurement unit 5 measures the battery capacity stored in the lithium ion battery 1. In addition, a usage condition measurement unit 6 is provided in the vicinity of the lithium ion battery 1. The usage condition measurement unit 6 measures various usage conditions 22 of the lithium ion battery 1, such as the usage environment, including temperature (cell temperature), ambient temperature, and air temperature, as well as the usage method and usage mode.

[0048] The degradation state prediction device for predicting the degradation state 26 of a lithium ion battery 1 includes a control unit 10, a prior cause-specific degradation state acquisition unit 12, a usage condition acquisition unit 13, a degradation rate coefficient calculation unit 15, a cause-specific degradation state calculation unit 16, a degradation state calculation unit 18, and a memory unit 19.

[0049] The control unit 10 includes a processor such as a CPU, and controls the operation of each functional unit of the deterioration state predicting device.

[0050] The usage condition acquisition unit 13 acquires various usage conditions 22 for the lithium ion battery 1, such as the usage environment, usage method, and usage mode, acquired by the ammeter 2, voltmeter 3, usage condition measurement unit 6, or by other means (step #1 in FIG. 2). The acquired usage conditions 22 are the usage conditions 22 at the present time (when the degradation state 26 is predicted (prediction time)), that is, at the point in time when a first hour has elapsed since the prior cause-specific degradation state 21 (cause-specific degradation state 25) acquired by the prior cause-specific degradation state acquisition unit 12, described below, was calculated and acquired. The usage condition acquisition unit 13 stores the acquired usage conditions 22 in the memory unit 19.

[0051] The prior cause-specific degradation state acquisition unit 12 acquires the cause-specific degradation state 25 an arbitrary first time before the current time (time of prediction) as the prior cause-specific degradation state 21 for each degradation cause (step #2 in FIG. 2 ). The prior cause-specific degradation state 21 is the degradation state 26 (cause-specific degradation state 25) of the lithium ion battery 1 that has deteriorated up to the time that is the first time before the prediction. For example, when the degradation state 26 of the lithium ion battery 1 is predicted at a predetermined or arbitrary time interval, the prior cause-specific degradation state acquisition unit 12 acquires the cause-specific degradation state 25 calculated immediately before or in the past for each degradation cause. Alternatively, the prior cause-specific degradation state acquisition unit 12 acquires the cause-specific degradation state 25 acquired by any means an arbitrary first time before the current time for each degradation cause. The prior cause-specific degradation state acquisition unit 12 associates the acquired prior cause-specific degradation state 21 with the cause of degradation and stores it in the storage unit 19.

[0052] The deterioration rate coefficient calculation unit 15 determines a deterioration rate coefficient 23 for each deterioration cause based on the usage conditions 22 acquired by the usage condition acquisition unit 13 (step #3 in FIG. 2). As described above, the deterioration state 26 (cause-specific deterioration state 25) is proportional to the power of time. The deterioration rate coefficient 23 is a coefficient corresponding to the deterioration rate and corresponds to the constant of this proportional relationship (proportional equation).

[0053] The use conditions 22 used to determine the deterioration rate coefficient 23 can be set arbitrarily depending on the cause of deterioration. For example, the use conditions 22 may include temperature information, current information, and voltage information. These use conditions 22 are measured by the ammeter 2, voltmeter 3, use condition measurement unit 6, etc., and further calculated as necessary.

[0054] The temperature information includes at least one of the cell temperature, which is the temperature of the lithium-ion battery 1, the ambient temperature, which is the temperature around the lithium-ion battery 1, and the air temperature, and is selected depending on the cause of deterioration. The current information includes at least one of the charge current, discharge current, charge rate (C rate), discharge rate (C rate), total discharged amount of electricity, and total charged amount of electricity during charging and discharging of the lithium-ion battery 1, and is selected depending on the cause of deterioration. The voltage information includes at least one of the average voltage, average state of charge, upper limit voltage for charging, lower limit voltage for discharging, time spent within a specified state of charge range, and overvoltage of the lithium-ion battery 1, and is selected depending on the cause of deterioration.

[0055] Deterioration of the lithium ion battery 1 is affected by the usage state of the lithium ion battery 1, for example, how often the lithium ion battery 1 is used, whether it is continuously charged and discharged, whether there are periods when it is not in use, etc. Therefore, it is preferable that the usage conditions 22 include, in particular, the total discharged amount of electricity, the total charged amount of electricity, the time spent within a predetermined range of charge state, etc.

[0056] The deterioration rate coefficient 23 can be determined by any method using the selected use conditions 22, but may also be calculated by a predetermined calculation method that is determined in advance for each cause of deterioration.

[0057] For example, the deterioration rate coefficient 23 can be determined by substituting the acquired usage conditions 22 into a predetermined formula for calculating the deterioration rate coefficient 23. The formula for calculating the deterioration rate coefficient 23 can be determined by collecting multiple sets of selected usage conditions 22 and battery deterioration data for the lithium-ion battery 1 in advance and performing regression analysis on the collected data. Note that the regression analysis model can use, as a physicochemical model, the Arrhenius equation (Arrhenius model: a physicochemical empirical formula that expresses the temperature dependence of reaction rate) or the Butler-Volmer equation (a theoretical formula that expresses the relationship between reaction rate, overvoltage, and temperature), and as a statistical model (machine learning model), a linear model (linear multiple regression model), a generalized linear model, Gaussian process regression, random forest regression, neural network, support vector regression, etc. For example, a linear multiple regression model and an Arrhenius model can be combined.

[0058] Here, a time coefficient 24 for each degradation cause used to calculate the cause-specific degradation state 25 is calculated in advance and stored in the storage unit 19. Then, the stored time coefficient 24 is acquired for each degradation cause (step #4 in FIG. 2). As described above, the degradation state 26 is proportional to the power of time. The time coefficient 24 is a coefficient indicating the time dependency of the degradation state 26 and corresponds to a power exponent. Like the degradation rate coefficient 23, the time coefficient 24 can be calculated by regression analysis using a statistical model (machine learning model). The time coefficient 24 may be calculated by any of the functional blocks of the degradation state prediction device, that is, the control unit 10, the degradation rate coefficient calculation unit 15, or the cause-specific degradation state calculation unit 16, or may be calculated by a device other than the degradation state prediction device.

[0059] The cause-specific degradation state calculation unit 16 calculates the cause-specific degradation state 25 for each degradation cause based on the prior cause-specific degradation state 21, the degradation rate coefficient 23, and the time coefficient 24 (step #5 in FIG. 2 ). The cause-specific degradation state calculation unit 16 stores the calculated cause-specific degradation state 25 in the storage unit 19 in association with the degradation cause. The cause-specific degradation state calculation unit 16 may calculate the cause-specific degradation state 25 using any method, but as described above, the cause-specific degradation state calculation unit 16 calculates the cause-specific degradation state 25 by taking the prior cause-specific degradation state 21 into consideration using a proportional relationship with time in which the degradation rate coefficient 23 is a constant and the time coefficient 24 is a power exponent. In other words, the cause-specific degradation state calculation unit 16 calculates the cause-specific degradation state 25 based on the prior cause-specific degradation state 21 and the degradation state 26 (unit cause-specific degradation state) of the lithium-ion battery 1 at the elapsed time (first time) after the prior cause-specific degradation state 21 was acquired.

[0060] For example, the cause-specific degradation state calculation unit 16 may calculate the time of prediction as t, the first time as δ, and the cause-specific degradation state 25 at time t as μ t , the deterioration state by prior cause 21 is μ (t-δ) , where the deterioration rate coefficient 23 is α and the time coefficient 24 is β, μ t ={μ (t-δ) 1 / β -α 1 / β ·δ} β (Formula 1) The deterioration state 25 by cause at the time of prediction is calculated using the function shown below.

[0061] The degradation state calculation unit 18 calculates the degradation state 26 of the lithium ion battery 1 at the time of prediction from the degradation states 25 by cause stored for each degradation cause in the storage unit 19 (step #6 in FIG. 2). For example, the degradation state calculation unit 18 calculates the predicted value of the degradation state 26 of the lithium ion battery 1 by adding up all the calculated degradation states 25 by cause.

[0062] The state of health 26 calculated based on the above (Equation 1) is expressed as SOH t , the first time is δ, and the deterioration state 25 by each cause at time t is μ nt , the prior cause-specific deterioration state 21 for each cause-specific deterioration state 25 is μn(t-δ) , the deterioration rate coefficient 23 for each cause-specific deterioration state 25 is α n , the time coefficient 24 for each cause-specific deterioration state 25 is β n Then (n=1,2,3,...), SOH t =f(μ 1t ,μ 2t ,μ 3t ,···) μ 1t ={μ 1(t-δ) 1 / β1 -α1 1 / β1 ·δ} β1 μ 2t ={μ 2(t-δ) 1 / β2 -α2 1 / β2 ·δ} β2 μ 3t ={μ 3(t-δ) 1 / β3 -α3 1 / β3 ·δ} β3 ... (Formula 2) It is calculated as:

[0063] By predicting the degradation state 26 of the lithium-ion battery 1 using the degradation state prediction device configured as described above, the degradation rate coefficient 23 is determined for each cause of degradation using one or more appropriate usage conditions 22 selected according to the cause of degradation, making it possible to easily determine an appropriate degradation rate coefficient 23. As a result, the degradation state 26 of the lithium-ion battery 1 at the current time (at the time of prediction) can be predicted with high accuracy.

[0064] Furthermore, there are various causes of deterioration in the lithium-ion battery 1, and the state of progression of the deterioration is also diverse. For example, in the case of deterioration of the lithium-ion battery 1 due to the growth of a film on the surface of the negative electrode of the lithium-ion battery 1 (SEI film growth), the deterioration progresses rapidly in the initial stage of use of the lithium-ion battery 1, but the progression of the deterioration slows down once the reaction has progressed to a certain extent. In addition, in the case of deterioration of the lithium-ion battery 1 due to gas generation in the positive electrode, cracks in the positive electrode active material, deposition of metallic Li on the negative electrode, etc., the deterioration hardly progresses in the initial stage of use of the lithium-ion battery 1, but the rate of deterioration tends to increase as the deterioration progresses.

[0065] In response to these various causes of deterioration, the degradation state prediction device according to this embodiment determines more appropriate degradation rate coefficients 23 and time coefficients 24 according to usage conditions 22. That is, the degradation state prediction device determines a calculation method for the degradation rate coefficients 23 and time coefficients 24 for each cause of deterioration that reflects deterioration caused by SEI film growth, Li precipitation, gas generation, etc., from data previously acquired using an actual sample lithium ion battery 1. Then, using a model based on these degradation rate coefficients 23 and time coefficients 24, it is possible to accurately predict the degradation state 26 (cause-specific degradation state 25) of the lithium ion battery 1.

[0066] Furthermore, the degradation rate coefficient 23 for each degradation cause is affected by the usage conditions 22, which change constantly during use of the lithium-ion battery 1. Therefore, if the degradation rate coefficient 23 is set to a constant value for each degradation cause, it would be difficult to accurately calculate the degradation state 26 (cause-specific degradation state 25) of the lithium-ion battery 1. According to this embodiment, the prior cause-specific degradation state 21 for each degradation cause at a time prior to the current time (time of prediction) is acquired, and the cause-specific degradation state 25 can be calculated using a model based on the amount of change in the degradation state (unit cause-specific degradation state) by cause from the time of calculation or measurement of the prior cause-specific degradation state 21 to the time of prediction and the prior cause-specific degradation state 21. As a result, the influence of changes in the usage conditions 22 can be suppressed, and the current (time of prediction) degradation state 26 of the lithium-ion battery 1 can be accurately predicted.

[0067] [Comparison results] Next, with reference to FIG. 1, the results of an experiment on predicting the degradation state of the lithium ion battery 1 (secondary battery) according to this embodiment will be described using FIGS.

[0068] First, Figure 3 shows the experimental results when only Li consumption due to SEI film growth was considered as the cause of degradation, that is, when degradation state prediction was performed taking only one cause of degradation into consideration. In Figure 3, line 30 shows the results of degradation state prediction when only Li consumption due to SEI film growth was considered as the cause of degradation, and in the above (Equation 1), the first time was set to δ = 240 hours, the degradation rate coefficient 23 was set to α = 0.002, and the time coefficient 24 was set to β = 0.5. Figure 4 shows the results of degradation state prediction when Li consumption due to SEI film growth and high-temperature degradation (degradation due to gas generation) were considered as the causes of degradation, and in the above (Equation 2), the first time was set to δ = 240 hours, the degradation rate coefficient 23 for Li consumption was set to α = 0.002, the time coefficient 24 was set to β = 0.5, and the degradation rate coefficient 23 for high-temperature degradation was set to α = 0.5 × 10. ‐6 4 shows the experimental results when degradation state prediction was performed with time coefficient 24 set to β2 = 1.5. In FIG. 4, line 30 shows the result of degradation state prediction considering only Li consumption due to SEI film growth as the cause of degradation, line 31 shows the result of degradation state prediction considering only high-temperature degradation as the cause of degradation, and line 32 shows the result of degradation state prediction considering both Li consumption due to SEI film growth and high-temperature degradation as the causes of degradation.

[0069] As shown in Figures 3 and 4, the predicted results shown in Diagrams 30 and 31, which predict the state of deterioration by considering only one cause of deterioration, do not match the actual measured values ​​from experiments. On the other hand, as shown in Figure 4, Diagram 32, which predicts the state of deterioration by considering multiple causes of deterioration, closely matches the actual measured values ​​from experiments.

[0070] From the above, it can be seen that by calculating the degradation state 25 for each of the multiple degradation causes and calculating the degradation state 26 of the secondary battery from the multiple degradation states 25 for each of the multiple degradation causes, the degradation state 26 can be predicted with high accuracy.

[0071] 5 shows the relationship between the predicted results and the measured values ​​when the degradation state 26 is predicted by determining the degradation rate coefficient 23 for Li consumption and the degradation rate coefficient 23 for high-temperature degradation as shown below, taking into account Li consumption due to SEI film growth and high-temperature degradation (degradation due to gas generation) as causes of degradation and only considering temperature (cell temperature) as usage condition 22, and setting the first time period to δ = 240 hours, the time coefficient 24 for Li consumption to β1 = 0.5, and the time coefficient 24 for high-temperature degradation to β2 = 1.5 in (Equation 2) above. Also, FIG. 6 shows the relationship between the predicted results and the measured values ​​when the degradation state 26 is predicted by determining the degradation rate coefficient 23 for Li consumption and the degradation rate coefficient 23 for high-temperature degradation as shown in FIG. 5, taking into account temperature (cell temperature), average voltage, and charge rate as usage condition 22, and setting the first time period to δ = 240 hours, the time coefficient 24 for Li consumption to β1 = 0.5, and the time coefficient 24 for high-temperature degradation to β2 = 1.5 in (Equation 2) above.

[0072] The deterioration rate coefficient 23, which is used in the prediction shown in FIG. 5 and takes only the cell temperature into consideration, can be calculated by the following equation (3), where α is the deterioration rate coefficient 23, α0 is a predetermined reference coefficient, Ea is a constant that does not depend on the cell temperature, R is the gas constant (approximately 8.314), and T is the cell temperature (absolute temperature [K]). α=α0×exp(Ea / RT) (Equation 3)

[0073] Furthermore, the deterioration rate coefficient 23 used in the prediction shown in FIG. 6, which takes into account the cell temperature, average voltage, and charge rate, can be calculated by the following equation (4), where α is the deterioration rate coefficient 23, WV is a predetermined reference voltage, V is the average voltage at the time of prediction, WC is a predetermined reference charge rate, C is the charge rate of the charge performed at or immediately before the prediction, r is a predetermined constant, Ea is a constant independent of the cell temperature, R is a gas constant (approximately 8.314), and T is the cell temperature (absolute temperature [K]). α=(WV×V+r)×exp(Ea / RT)+WC×C (Formula 4)

[0074] In this case, the deterioration rate coefficient 23 can also be calculated by the following (Equation 5). α=(WV×V+WC×C+r)×exp(Ea / RT) (Formula 5)

[0075] Here, the experiment was conducted under different usage conditions 22 (usage environment), and the relationship between the predicted results and the actual measured values ​​is shown for the cases where the voltage of the lithium-ion battery 1 (secondary battery) was fixed (stored) at 4.15 V (4.15 V storage test), where the voltage was fixed (stored) at 3.67 V (3.67 V storage test), and where charge and discharge were repeated a predetermined number of times at a charge rate of 0.7 C so that the average voltage was 3.7 V (cycle test). In the figure, the closer the relationship between the predicted results and the actual measured values ​​is to a proportional relationship with a slope of 1 (solid line in the figure), the closer the match between the predicted results and the actual measured values.

[0076] As shown in Figure 5, when the voltage was fixed (stored) at 4.15 V (4.15 V storage test), the predicted results closely matched the measured values, but when the voltage was fixed (stored) at 3.67 V (3.67 V storage test) or when charge / discharge was repeated a predetermined number of times at a charge rate of 0.7 C so that the average voltage was 3.7 V (cycle test), the predicted results did not match the measured values. This shows that if the deterioration rate coefficient 23 is determined without using appropriate usage conditions 22 (usage environment), the deterioration cause and the deterioration rate coefficient 23 may not match, making it impossible to accurately predict the deterioration state 26.

[0077] On the other hand, as shown in FIG. 6, when a deterioration rate coefficient 23 is determined taking into consideration temperature, average voltage, and charge rate as usage conditions 22, and this deterioration rate coefficient 23 is used to predict a deterioration state 26, the predicted results and the actual measured values ​​are almost identical in all cases.

[0078] From the above, it can be seen that by determining the deterioration rate coefficient 23 using a plurality of usage conditions 22 that are in accordance with the cause of deterioration, it is possible to determine the deterioration rate coefficient 23 that is consistent with the cause of deterioration with high accuracy, and to predict the deterioration state 26 with high accuracy.

[0079] [Another embodiment] (1) In the above embodiment, the secondary battery to be predicted for the degradation state 26 is not limited to the lithium ion battery 1, but various secondary batteries that can be charged and discharged, such as non-aqueous secondary batteries, can also be the secondary battery to be predicted.

[0080] (2) In each of the above embodiments, the overvoltage, which is the usage condition 22 for determining the deterioration rate coefficient 23, can be determined by various methods as follows.

[0081] For example, the overvoltage can be determined by applying current to a secondary battery (lithium ion battery 1) for a predetermined period of time, then stopping the application of current, and subtracting the voltage value of the secondary battery before the application of current to the secondary battery from the voltage value of the secondary battery after a predetermined third time has elapsed since the application of current. The third time can be, for example, from 0.5 seconds to 60 seconds. By setting the third time in this manner, the overvoltage caused by the charge / discharge reaction on the electrode surface can be determined with high accuracy, and the deterioration rate coefficient 23 can be calculated with high accuracy.

[0082] Alternatively, the overvoltage may be determined by turning on and off the secondary battery (lithium ion battery 1) for a predetermined period of time, and subtracting the voltage value of the secondary battery a predetermined fourth time after the current is stopped from the voltage value of the secondary battery while it is powered on. The fourth time may be, for example, 0.5 seconds or more and 60 seconds or less. By setting the fourth time in this manner, the overvoltage can be determined with high accuracy, and the deterioration rate coefficient 23 can be calculated with high accuracy.

[0083] Moreover, the overvoltage can be calculated as the product of the charging current and the internal resistance of the secondary battery.

[0084] (3) The deterioration rate coefficient 23 may be determined using usage history data including a history of temperature information (past trends), a history of current information, and a history of voltage information, and cell state history data including at least one of a battery capacity history, a resistance history, a battery voltage history, an output history, an external dimension, and a restraining pressure.

[0085] For example, it can be determined by machine learning the usage history data and the cell state history data.

[0086] This makes it possible to more accurately determine the deterioration rate coefficient 23 by taking into account the history of the usage conditions 22. As a result, it is possible to accurately predict the deterioration state 26 (cause-specific deterioration state 25) of the secondary battery (lithium ion battery 1).

[0087] (4) In each of the above embodiments, the degradation state prediction device is not limited to being configured with functional blocks as shown in FIG. 1, but may be configured with any functional blocks. For example, each functional block of the degradation state prediction device may be further subdivided, or conversely, some or all of the functional blocks may be combined. Furthermore, the above degradation state prediction method may be executed by a degradation state prediction device of any configuration, not limited to the degradation state prediction device shown in FIG. 1. Furthermore, some or all of the functions of the degradation state prediction device may be configured by software. A program related to the software is stored in any storage device such as the storage unit 19, and executed by a processor (computer) such as a CPU included in the control unit 10, or a separately provided processor (computer).

[0088] The configurations disclosed in the above embodiments (including other embodiments, the same applies below) can be applied in combination with configurations disclosed in other embodiments, as long as no contradiction arises. Furthermore, the embodiments disclosed in this specification are examples, and the embodiments of the present invention are not limited to these, and can be modified as appropriate within the scope that does not deviate from the purpose of the present invention. [Industrial Applicability]

[0089] The present invention can be applied to predicting the state of deterioration of various secondary batteries, including lithium ion batteries. [Explanation of symbols]

[0090] 1. Lithium-ion battery (secondary battery) 12. Deterioration status acquisition unit for each prior cause 13 Usage conditions acquisition section 15 Deterioration rate coefficient calculation section 16. Cause-specific deterioration state calculation section 18 Deterioration state calculation unit 19 Memory section 21 Deterioration status by prior cause 22 Terms of Use 23 Degradation rate coefficient 24-hour factor 25 Deterioration status by cause 26 Deterioration state

Claims

1. A degradation state prediction method for calculating a degradation state for each cause of degradation and predicting a degradation state of a secondary battery from a plurality of the degradation states for each cause, comprising: A degradation state prediction method in which each of the cause-specific degradation states is calculated based on a prior cause-specific degradation state, which is the cause-specific degradation state an arbitrary first time ago, and a unit cause-specific degradation state that deteriorates during the first time, taking into account a time dependency that varies depending on the cause of deterioration and follows a power law with respect to the elapsed time of the cause-specific degradation state, and a degradation rate that varies depending on the cause of deterioration and is determined by the usage conditions at the time of prediction.

2. The deterioration state by cause is expressed as follows: the time of prediction is t, the first time is δ, and the deterioration state by cause is μ ｔ , the deterioration state by prior cause is μ （ｔ－δ） , where α is a degradation rate coefficient corresponding to the degradation rate, and β is a time coefficient indicating the time dependency, m ｔ ={μ （ｔ－δ） １／β -a １／β ・d} β The degradation state prediction method according to claim 1, wherein the degradation state prediction method is performed as follows:

3. 3. The degradation state prediction method according to claim 1, wherein the degradation state is determined by adding up all of the calculated degradation states for each cause.

4. The degradation state prediction method according to claim 1 , wherein the degradation state for each prior cause is an actual measurement value obtained by any method.

5. 4. A degradation state prediction method according to claim 1, wherein the deterioration state by pre-existing cause is calculated based on the deterioration state by pre-existing cause a second time before the first time and the deterioration state by unit cause between the second time and the first time.

6. the deterioration rate is determined using temperature information, current information, and voltage information as the usage conditions; the temperature information includes at least one of a cell temperature, an environmental temperature, and an air temperature; The current information includes at least one of a charging current, a discharging current, a charging rate, a discharging rate, a total discharging amount of electricity, and a total charging amount of electricity; 6. The degradation state prediction method according to claim 1, wherein the voltage information includes at least one of an average voltage, an average state of charge, an upper limit charge voltage, a lower limit discharge voltage, a residence time within a predetermined range of state of charge, and an overvoltage.

7. 7. The degradation state prediction method according to claim 6, wherein the overvoltage is determined by turning on and off the secondary battery for a predetermined period of time, and subtracting the voltage value of the secondary battery before turning on the secondary battery from the voltage value of the secondary battery when a predetermined third hour has elapsed since turning on the secondary battery.

8. 7. The degradation state prediction method according to claim 6, wherein the overvoltage is calculated by turning on and off the secondary battery for a predetermined period of time, and subtracting the voltage value of the secondary battery when a predetermined fourth time has elapsed since the power supply was stopped from the voltage value of the secondary battery while it was powered on.

9. 7. The method for predicting a state of deterioration according to claim 6, wherein the overvoltage is calculated as a product of the charging current and the internal resistance of the secondary battery.

10. The deterioration rate is usage history data including a history of the temperature information, a history of the current information, and a history of the voltage information; The degradation state prediction method according to any one of claims 6 to 9, wherein the degradation state is determined using cell state history data including at least one of battery capacity history, resistance history, battery voltage history, output history, external dimensions, and confinement pressure.

11. The degradation state prediction method according to claim 10 , wherein the degradation rate is determined by machine learning of the usage history data and the cell state history data.

12. A degradation state prediction device that predicts a degradation state of a secondary battery, a deterioration state acquisition unit for acquiring, for each cause of deterioration of the secondary battery, a deterioration state for each cause of deterioration that is a deterioration state for each cause acquired an arbitrary first time period ago; a usage condition acquisition unit that acquires usage conditions at a time when the degradation state of the secondary battery is predicted; a deterioration rate coefficient calculation unit that determines a deterioration rate coefficient corresponding to a deterioration rate of the secondary battery for each of the causes of deterioration using the use conditions; a cause-specific degradation state calculation unit that calculates the cause-specific degradation state for each of the deterioration causes based on a time coefficient relating to an elapsed time predetermined for each of the deterioration causes, the degradation rate coefficient, and the prior cause-specific degradation state; a degradation state prediction device comprising: a degradation state calculation unit that predicts the degradation state of the secondary battery from the plurality of cause-specific degradation states calculated for each of the degradation causes.

13. A degradation state prediction program for predicting a degradation state of a secondary battery, a function of acquiring, for each cause of deterioration of the secondary battery, a prior cause-specific deterioration state, which is a cause-specific deterioration state obtained an arbitrary first time before, for each cause of deterioration of the secondary battery; a function of acquiring a usage condition at the time of predicting the deterioration state of the secondary battery; a function of determining a deterioration rate coefficient corresponding to a deterioration rate of the secondary battery for each of the causes of deterioration using the use conditions; a function of calculating the cause-specific deterioration state for each of the deterioration causes based on a time coefficient relating to an elapsed time predetermined for each of the deterioration causes, the deterioration rate coefficient, and the deterioration state for each of the prior causes; and a function of predicting the degradation state of the secondary battery from the plurality of degradation states calculated for each of the causes of degradation.

Images