JP7710363B2 - Secondary battery state determination device, charge/discharge control device, and secondary battery system - Google Patents

Description

The present invention relates to a secondary battery state determination device, a charge/discharge control device, and a secondary battery system.

As background art in this technical field, the abstract of the following Patent Document 1 states, "The deterioration state of a secondary battery for a practical current value is accurately determined. A state determination device for a secondary battery having a positive electrode and a negative electrode, which determines a group of capacity reduction parameters A based on charge/discharge characteristics per reference amount of the positive electrode and the negative electrode and a current value A, and determines a group of resistance increase parameters B based on the charge/discharge characteristics per reference amount of the positive electrode and the negative electrode, the group of capacity reduction parameters A, and a current value B that is larger than the current value A, a method for determining a state of a secondary battery, a secondary battery system, and a charge/discharge control device having the state determination device."
It is possible to use the secondary battery and the method of determining the state of the secondary battery that are known from Patent Document 1, and in this respect the contents of Patent Document 1 are included in this specification.

International Publication No. 2015/045015

However, in the above-mentioned technology, there is a demand for obtaining the state of the secondary battery more appropriately.
The present invention has been made in consideration of the above-mentioned circumstances, and aims to provide a secondary battery state determination device, a charge/discharge control device, and a secondary battery system that can appropriately acquire the state of a secondary battery.

In order to solve the above problem, the device for determining the state of a secondary battery of the present invention includes a candidate value extraction unit that determines a plurality of candidate values for a parameter of the secondary battery at each of a plurality of times based on a state quantity that is a result of measuring the state of the secondary battery at the plurality of times, and a calculation unit that determines an estimate of the parameter and a change in the estimate over time based on the plurality of the candidate values and a predetermined function, wherein the state quantity includes measurement results of a voltage and a current in the secondary battery, and the parameters include an effective weight of a positive electrode active material in the secondary battery, an effective weight of a negative electrode active material, or a capacity difference between the positive electrode and the negative electrode, the candidate value extraction unit calculates the plurality of candidate values by an optimization method that uses pre-stored discharge curves of the positive electrode alone and the negative electrode alone and the state quantity, the function describes the deterioration behavior of the parameter, and the estimate value is the candidate value belonging to the candidate value set that is a combination of the plurality of candidate values at each time and has a minimum difference from the function .

According to the present invention, the state of the secondary battery can be appropriately obtained.

5 is a diagram showing a procedure for acquiring estimated values of degradation parameters in the first embodiment and a comparative example. FIG. 1 is a block diagram of a state determination device according to a first embodiment. FIG. 1 is a block diagram of a computer. FIG. 1 is a diagram illustrating an overview of parameter optimization. FIG. 11 is a diagram illustrating an example of time-dependent changes in a plurality of candidate values. FIG. 13 is a diagram showing a predicted curve and measurement results of the capacity retention rate of a secondary battery. FIG. 13 is a diagram showing the results of predicting the change over time in effective mass of a positive electrode active material in a comparative example. FIG. 13 is a diagram showing the results of predicting the change over time in effective mass of a positive electrode active material in an example. FIG. 1 is a diagram showing the error between the actual measured values and the predicted curves in Examples and Comparative Examples. FIG. 11 is a block diagram of a secondary battery system according to a second embodiment.

[Overview of the embodiment]
In recent years, efforts have been made to use secondary batteries such as lithium-ion batteries as onboard power sources for vehicles and as power sources for storing electricity in smart houses, in order to utilize energy efficiently. However, it is known that secondary batteries deteriorate in characteristics due to charging and discharging and storage. In particular, since the power sources for the above-mentioned applications are expected to be used for a long period of time, it is important to appropriately determine the deterioration state of the secondary battery in order to continue to use the product equipped with the secondary battery stably.

The deterioration of secondary battery characteristics is due to the deterioration of the active material contained in the electrodes of the secondary battery. An active material is one of the materials that make up the electrodes of a secondary battery, and plays a role in promoting the charging and discharging reactions of the secondary battery by storing and releasing ions. For example, in the case of a lithium-ion secondary battery, when the secondary battery is charged, lithium ions are released from the positive electrode active material and are stored in the negative electrode active material, promoting the charging reaction. The discharging reaction follows this reverse process.

To determine the degradation state of a secondary battery, it is desirable to accurately estimate the degradation state of the active material described above. In order to grasp the degradation state of the active material, a method of measuring the charge/discharge capacity of the secondary battery as well as the current/voltage response such as the resistance value is used. For example, by applying the technology of Patent Document 1, it is believed that it is possible to detect the degradation state of the active material of the positive and negative electrodes of the secondary battery by utilizing the charge/discharge curve of the secondary battery.

In addition, by applying the technology of Patent Document 1, it is believed that it is possible to reproduce the charge/discharge curve of the secondary battery by calculation based on the pre-stored discharge curves of the positive and negative electrodes alone, and in the process obtain the values of parameters corresponding to the effective weight of the positive electrode active material, the effective weight of the negative electrode active material, and the capacity difference between the positive and negative electrodes. All of these parameters are strongly correlated with the deterioration state of the active material, and hereinafter, parameters that are correlated with the deterioration state of the active material are collectively referred to as "deterioration parameters."

Furthermore, by applying the technology of Patent Document 1, it is believed that the value of the degradation parameter can be estimated by comparing the discharge curve calculated by the above procedure with the measured value of the discharge curve of the secondary battery. However, this method is based on the premise that there is no measurement error in the measured value of the discharge curve of the secondary battery. In general, the measured value contains a certain amount of measurement error due to the influence of the device and the measurement environment. Furthermore, even if measurements are taken with the same type of secondary battery, the measured value contains a small error due to differences in manufacturing conditions such as lot numbers (hereinafter referred to as sample dependency).

In this way, the measured value of the discharge curve of a secondary battery contains a certain measurement error and sample dependency. These can cause a decrease in the estimation accuracy when estimating the degradation parameters of a secondary battery, so it is difficult to obtain the estimated values with high accuracy by simply applying the technology of Patent Document 1. This means that the degradation state of a secondary battery cannot be accurately predicted using the estimated values of the degradation parameters.
The embodiment described below is configured in consideration of the above points, and makes it possible to accurately determine the degradation state of a secondary battery.

More specifically, in the embodiment described below, the state of the secondary battery is measured at various times, and two or more candidate values that serve as estimates of a specified degradation parameter are obtained from the measurement data at some of these times, and the change in each candidate value over time is obtained based on the obtained candidate values and a specified function. Next, of these candidate values, the "most likely candidate value" that is the most probable (the value estimated to be closest to the true value) is selected. Next, this most likely candidate value is adopted as an estimate of the degradation parameter, and this estimate is used to predict the degradation state of the secondary battery.

1 is a diagram showing a procedure for acquiring an estimated value of a degradation parameter A in a comparative example and a candidate value of the degradation parameter A in the first embodiment. Here, the comparative example is a technique to which the above-mentioned Patent Document 1 is applied.
In the method of the comparative example, when the state of the secondary battery is measured at time t1, an estimate AE(t1) of the deterioration parameter A at time t1 is obtained based on the measurement result. Thereafter, estimates AE(t2) and AE(t3) of the deterioration parameter A at times t2 and t3 are similarly obtained. On the other hand, in the first embodiment, when the state of the secondary battery is measured, a plurality of candidate values AC1(t), AC2(t), ... that are candidates for the estimate value are obtained for one measurement time t.

In the example shown in FIG. 1, in each embodiment, multiple candidate values AC1(t1), AC2(t1), ... for degradation parameter A are obtained at time t1. Thereafter, candidate values AC1(t2), AC2(t2), ... and candidate values AC1(t3), AC2(t3), ... are similarly obtained at times t2 and t3. Then, in the first embodiment, the most likely candidate value is selected from among these multiple candidate values to predict the degradation state of the secondary battery. As a result, according to the first embodiment, the degradation state of the secondary battery for a practical current value can be accurately determined.

[First embodiment]
Various embodiments will be described in detail below with reference to the drawings. In the following description, components having the same functions in all drawings will be given the same reference numerals, and repeated description thereof may be omitted. In the following description, a secondary battery (storage battery) may be simply referred to as a "battery."

Overall Configuration of First Embodiment
FIG. 2 is a block diagram of a state determining device 100 according to the first embodiment.
In FIG. 2, a secondary battery module 200 includes a plurality of secondary batteries 210 connected in series or in parallel.
The state determination device 100 is a device that determines the state of the secondary battery module 200 or each of the secondary batteries 210. In the following description, an example in which a lithium ion secondary battery is applied as the secondary battery 210 will be described, but the secondary battery is not limited to a lithium ion secondary battery.

In the following description, the operation of the state determining device 100 will be described as an operation for determining the state of each secondary battery 210, but the operation of the secondary battery module 200 can also be determined in a similar manner.
In FIG. 2 , the state determination device 100 includes a measurement unit 110, a deterioration parameter extraction unit 120 (candidate value extraction unit, candidate value extraction means), a memory 130, a calculation unit 140 (calculation means), a deterioration prediction unit 150, and an output unit 160.

FIG. 3 is a block diagram of a computer 900 .
The state determining device 100 shown in FIG. 2 includes one or more computers 900 shown in FIG.
3 , the computer 900 includes a CPU 901, a RAM 902, a ROM 903, a HDD 904, a communication I/F 905, an input/output I/F 906, and a media I/F 907. The communication I/F 905 is connected to a communication circuit 915. The input/output I/F 906 is connected to an input/output device 916. The media I/F 907 reads and writes data from a recording medium 917.

The ROM 903 stores control programs executed by the CPU, various data, etc. The CPU 901 executes application programs loaded into the RAM 902 to realize various functions. The inside of the state determination device 100 shown in Figure 2 above shows the functions realized by application programs, etc., as blocks.

Returning to FIG. 2, the measurement unit 110 measures the state quantities of the secondary battery 210 and stores them in the memory 130. Here, the state quantities of the secondary battery 210 refer to quantities related to the charge and discharge characteristics of the secondary battery 210, such as the capacity, resistance value, temperature, voltage, and current of the secondary battery 210 at the time of measurement.

The degradation parameter extraction unit 120 calculates a candidate value of the degradation parameter A (parameter) using the state quantity of the secondary battery acquired by the measurement unit 110 at each measurement time t. At this time, the degradation parameter extraction unit 120 calculates two or more candidate values AC1(t), AC2(t), ... (see FIG. 1) of the degradation parameter based on the state quantity at least at one of the times.

The degradation parameter extraction unit 120 may also determine candidate values for multiple types of degradation parameters (e.g., degradation parameters A, B, and C) based on the state quantities at each time. The degradation parameter extraction unit 120 may determine two or more candidate values for at least one type of degradation parameter among the degradation parameters A, B, and C. The degradation parameter extraction unit 120 associates the candidate values extracted as described above with the state quantities of the secondary battery and the parameter names, and stores them in the memory 130. The memory 130 also stores at least one type of function used by the calculation unit 140.

The calculation unit 140 compares the state quantities of the secondary battery stored in the memory 130 with multiple candidate values of the degradation parameters and a predetermined function. As a result, the calculation unit 140 identifies the most likely candidate value that is estimated to be closest to the true value from among the multiple candidate values. The details of the operation behavior of the calculation unit 140 will be described later.

The deterioration prediction unit 150 uses the most likely candidate value identified by the calculation unit 140 as an estimated value of the deterioration parameter to calculate the deterioration state (internal resistance and capacity reduction rate of the secondary battery 210, etc.) and remaining life of the secondary battery 210. The life of the secondary battery 210 is, for example, when the capacity of the secondary battery 210 falls below a predetermined lower limit value, or when the internal resistance of the secondary battery 210 exceeds a predetermined upper limit value, and it is preferable to consider the life of the secondary battery to be over when these conditions are reached.

When calculating the degradation state and remaining life, a predetermined function such as an exponential function or a power function described below is used to predict the change over time of the estimated value of the degradation parameters, and the degradation state of the secondary battery 210 at a future time is predicted based on the result. Note that the method of predicting the life, etc. in the degradation prediction unit 150 is not limited to the above-mentioned procedure. For example, machine learning or the like may be applied to the estimated value obtained by the calculation unit 140 to predict the change over time of the estimated value, and the degradation state of the secondary battery 210 may be predicted.

The output unit 160 outputs the predicted values of the deterioration state and remaining life of the secondary battery 210 determined by the deterioration prediction unit 150 to a display device for the user, a battery life management device (not shown), etc. Also, if it is determined that the secondary battery has already reached the end of its life at the time the predicted value is output, a warning is displayed.

<Details of the degradation parameter extraction unit 120>
Next, the detailed operation of the degradation parameter extraction section 120 will be described. First, the reason why the degradation parameter extraction section 120 obtains the estimated value (or candidate value) of the degradation parameter will be described.
It is believed that the technology described in Patent Document 1 can be applied to reproduce the charge/discharge curve of a secondary battery by calculation using the discharge curves of the positive and negative electrodes alone and the values of the deterioration parameters stored in advance. According to this method, the values of the deterioration parameters can be optimized so that the charge/discharge curve of the secondary battery reproduced by calculation matches the charge/discharge curve of the secondary battery actually measured. The optimized values of the deterioration parameters can be used as the estimated values (or candidate values) of the deterioration parameters.

This type of procedure is called parameter optimization or numerical optimization. If the measured values do not contain measurement errors, it is possible to obtain estimates close to the true values of the degradation parameters using the optimization method described above. However, in general, due to the influence of measurement errors, multiple solutions are obtained as candidates for the estimated value. The details are explained with reference to Figure 4.

FIG. 4 is a diagram showing an overview of parameter optimization.
The horizontal axis of Fig. 4 indicates the range of values that can be assumed as the parameter value Ax of the deterioration parameter A. A calculated value of a certain state quantity (e.g., battery voltage) of the secondary battery 210 can be obtained for all parameter values Ax on the horizontal axis of Fig. 4. In addition, the measured value of the state quantity (e.g., battery voltage) of the secondary battery 210 at a certain measurement time t is stored in the memory 130 as described above. Then, the difference between the calculated value and the measured value can be obtained for all parameter values Ax of the deterioration parameter A. This difference is designated as δ.

The vertical axis in Figure 4 is the absolute difference |δ|, which is the absolute value of the difference δ. In the example shown, the absolute difference |δ| has three minimum values at parameter values Ax = α, β, and γ. These minimum values are δ(α), δ(β), and δ(γ), respectively. The values α, β, and γ at which these minimum values appear are candidate values for the estimated value AE(t).

Here, assume that the measured value of the state quantity contains a measurement error of magnitude δn. In the example shown, the measurement error δn is larger than all of the minimum values δ(α), δ(β), and δ(γ), and all of the minimum values δ(α), δ(β), and δ(γ) are included within the range of the measurement error δn. Therefore, with only the information shown in Figure 4, it is difficult to determine which of the candidate values α, β, and γ should be adopted as the estimated value AE(t).

In this way, when the measured value of the state quantity contains a measurement error, multiple candidate values for the estimated value AE(t) of the degradation parameter A are obtained, and it may be difficult to uniquely determine the estimated value AE(t) of the degradation parameter A. This embodiment determines which of these candidate values α, β, γ, ... to adopt as the estimated value AE(t). Therefore, this embodiment is not limited to the secondary battery system described in Patent Document 1, but can also be widely applied to parameter estimation for various other physical systems.

<Details of Calculation Unit 140>
Next, the calculation unit 140 will be described in detail.
Assume that the degradation parameter extraction unit 120 calculates a plurality of candidate values AC1(t), AC2(t), AC3(t), ... as candidates of the estimated value AE(t) for the degradation parameter A at a certain measurement time t. Here, the candidate values AC1(t), AC2(t), AC3(t), ... may be considered to have the same meaning as the candidate values α, β, γ shown in FIG. 4. Therefore, the degradation parameter extraction unit 120 similarly calculates the candidate values AC1(tk), AC2(tk), AC3(tk), ... at a plurality of measurement times tk (where k=0, 1, 2, 3, ...). The estimated value AE(t) is one of these candidate values AC1(tk), AC2(tk), AC3(tk), .... If these candidate values are arranged in the order of the measurement time t, it is as shown in the following formula (1).

(AC1(t0), AC2(t0), AC3(t0),...) at t=t0
(AC1(t1), AC2(t1), AC3(t1),...) at t=t1
(AC1(t2), AC2(t2), AC3(t2),...) at t=t2
: …Formula (1)

Now, rearranging the candidate values in formula (1) gives the following formula (2).

(AC1(t0), AC1(t1), AC1(t2),...)
(AC2(t0), AC2(t1), AC2(t2),...)
(AC3(t0), AC3(t1), AC3(t2),...)
: …Equation (2)

Each row of equation (2) contains one candidate value of the degradation parameter A measured at each time. That is, each row of equation (1) indicates the change over time of the degradation parameter. Furthermore, equation (2) comprehensively lists all possible combinations for the candidate values of the degradation parameter A measured at each time. Therefore, the change over time of the most likely candidate value of the degradation parameter, i.e., the estimated value AE(t), is included in one of the rows shown in equation (2). Therefore, each row in equation (2) is referred to as a "candidate value set."

Next, we will describe a method for identifying a candidate value set that represents a change over time of the most likely candidate value from among the candidate value sets shown in equation (2). Since the deterioration parameter of a secondary battery is a quantity that indicates the degree of deterioration of the secondary battery, its value is considered to change monotonically over time. Therefore, sets that show non-monotonic time dependency can be excluded from the candidates for the most likely candidate value as non-physical candidate value sets.

FIG. 5 is a diagram showing an example of how a plurality of candidate values change over time.
Graph G1 in FIG. 5 shows the change over time of candidate value AC1(t), and graph G2 shows the change over time of candidate value AC2(t).
In the graph G1, the candidate value AC1(t) monotonically decreases with time. If the degradation parameter A monotonically decreases with time, the candidate value AC1(t) is valid in that respect. On the other hand, in the graph G2, the candidate value AC2(t) shows a non-monotonic time dependency. Therefore, the candidate value AC2(t) can be removed from the candidates of the most likely candidate value as an inappropriate candidate value of the degradation parameter A. This process is called the "inappropriate candidate value elimination process."

In addition, it is generally known that the deterioration of the characteristics of a secondary battery progresses qualitatively according to a specific functional form. For example, when graphite is used as the negative electrode material of a secondary battery, it is known that the deterioration of the characteristics (deterioration of capacity and resistance) progresses in proportion to the square root of the period of use of the secondary battery (root rule). Therefore, the calculation unit 140 of this embodiment narrows down the set of deterioration parameters shown in formula (2) by utilizing such knowledge and empirical rules regarding battery deterioration. The specific procedure is shown below. The memory 130 stores functions that describe the deterioration behavior of the secondary battery for each type of deterioration parameter based on empirical rules, etc. These functions are called behavior describing functions. As an example, the behavior describing function F(t) shown in the following formula (3) can be applied. In the following formula (3), G, H, and J are constants. The behavior describing function F(t) in formula (3) assumes that the deterioration characteristics of the secondary battery progress exponentially with respect to the measurement time t.

F(t)=GH×exp(-J×t)…Formula (3)

As another example, the power function shown in the following formula (4) can be applied as the behavior describing function F(t). This also includes the root rule mentioned above. In the following formula (4), G, H, and J are constants.

F(t)=G+H×t ^J ...Formula (4)

The behavior description function F(t) stored in memory 130 is not limited to the above, and various functions may be applied depending on the type of degradation parameter.

The calculation unit 140 then selects one of the candidate value sets shown in formula (2) and optimizes the constants G, H, and J in the behavior describing function F(t) so that the difference δf (not shown) between the selected candidate value set and the behavior describing function F(t) of formulas (3), (4), etc., is minimized. Next, the calculation unit 140 determines the difference δg (not shown) between the behavior describing function F(t) calculated using the optimized constants G, H, and J and the selected candidate value set.

The calculation unit 140 performs the above-mentioned operation on all candidate value sets shown in formula (2) or on candidate value sets that have not been excluded by the above-mentioned inappropriate candidate value exclusion process. As a result, the optimal values of the constants G, H, and J are found for each candidate value set, and the difference δg (not shown) between the function F(t) calculated using the optimal values of the constants G, H, and J and the selected candidate value set is found.

Here, the behavior describing function F(t) shown in equations (3) and (4) is a function that describes (empirically) the deterioration behavior of the secondary battery 210, so it is considered that the true values of the deterioration parameters also change over time in accordance with the behavior describing function F(t). Therefore, the calculation unit 140 can determine that the candidate value set for which the difference δg between the candidate value set and the behavior describing function F(t) is the smallest is the most likely candidate value set. The contents of this process will be explained again with reference to FIG. 5.

Graph G3 in FIG. 5 shows the change over time of candidate value AC3(t), and graph G4 shows the change over time of candidate value AC4(t).
In addition, graphs G3 and G4 also show examples of the above-mentioned behavior describing function F(t). In the illustrated example, candidate value AC3(t) is almost identical to behavior describing function F(t), so the difference δg between the candidate value set and behavior describing function F(t) is a small value. On the other hand, candidate value AC4(t) deviates from behavior describing function F(t), so the difference δg (not shown) between the two is large. Therefore, if the only candidates for the most likely candidate value are candidate value AC3(t) and candidate value AC4(t), calculation unit 140 selects candidate value AC3(t) as the most likely candidate value.

Next, the processing procedure in the calculation unit 140 will be described when multiple candidate value sets with similar differences δg between the candidate value set and the behavior describing function F(t) are generated. For the sake of simplicity, it is assumed that there are two most likely candidate values, candidate values AC5(t) and AC6(t) (not shown), and that the state quantities of the secondary battery 210 have been measured n+1 times at measurement times t0, t1, t2, t3, ..., tn.

First, the calculation unit 140 calculates candidate values AC5(t) and AC6(t) at n+1 measurement times t0, t1, t2, t3, ..., tn. Then, the calculation unit 140 calculates the difference δg between each candidate value set and the function F(t) for the candidate values AC5(t) and AC6(t), and assumes that the two are substantially identical.

Here, the constants Gx, Hx, Jx are the results of optimizing the constants G, H, J in the behavior describing function F(t) of equations (3), (4), etc., based on the candidate value AC5(t). Similarly, the constants Gy, Hy, Jy are the results of optimizing the constants G, H, J in the behavior describing function F(t) based on the candidate value AC6(t).

The calculation unit 140 calculates a predicted value ACE5(tm) of the candidate value AC5(t) at a future time tm (where m>n) based on the constants Gx, Hx, and Jx. Similarly, the calculation unit 140 calculates a predicted value ACE6(tm) of the candidate value AC6(t) at a future time tm based on the constants Gy, Hy, and Jy.

Next, when the current time actually reaches time tm, the measurement unit 110 measures the state quantity of the secondary battery 210 and stores it in the memory 130. Next, the deterioration parameter extraction unit 120 calculates the candidate values AC5(tm) and AC6(tm) at time tm based on the state quantity at time tm. Next, the calculation unit 140 obtains the difference δ5 = |AC5(tm) - ACE5(tm) | and the difference δ6 = |AC6(tm) - ACE6(tm) |. Then, if the relationship "δ5 ≦ δ6" is established for the differences δ5 and δ6, the calculation unit 140 selects the candidate value AC5(t) as the most likely candidate value, and otherwise selects the candidate value AC6(t) as the most likely candidate value. Through the above procedure, the calculation unit 140 identifies a set of the most likely candidate values from the set of candidate values in formula (2) and outputs the identified set of the most likely candidate values as the estimated value AE(t) of the deterioration parameter A.

Example of life prediction
Next, examples and comparative examples in which life prediction was performed will be described.
First, in the examples and comparative examples, a secondary battery 210 was applied in which the negative electrode active material was graphite and the positive electrode active material was _LiNiCoMnO2 . In the test, the charge rate of the secondary battery 210 was adjusted to 80%, and it was left standing in a thermostatic chamber controlled at a constant temperature. The left standing secondary battery was taken out of the thermostatic chamber every 30 days, and the charge rate dependency (SOC-OCV curve) of the capacity and open circuit voltage of the secondary battery was measured in an environment of 25°C.

After measuring the capacity, the secondary battery's charge rate was adjusted to 80% and it was again placed in the thermostatic chamber, and after another 30 days it was removed and the capacity was measured. This process was repeated for approximately 900 days (storage test). In this study, two thermostatic chamber temperatures were considered during the storage test: 60°C and 25°C.

FIG. 6 is a diagram showing a predicted curve and measurement results of the capacity maintenance rate of the secondary battery 210. In FIG.
Graph G10 in FIG. 6 is a graph in the case where the thermostatic chamber temperature during the storage test was 60° C., and graph G20 is a graph in the case where the thermostatic chamber temperature during the storage test was 25° C.
In graphs G10 and G20, actual measurement values 14 and 24 indicated by open circles and actual measurement values 16 and 26 indicated by filled circles are actual measurement values of capacity maintenance ratios.

Of these, actual measured values 14 and 24 represent data for approximately 200 days out of 900 days of test data, and were used as training data for prediction. The remaining actual measured values 16 and 26 were used as data for verifying the prediction results. The solid predicted curves L11 and L21 are predicted curves based on the actual measured values 14 and 24 and on an example according to the first embodiment. The dashed predicted curves L12 and L22 are predicted curves based on the actual measured values 14 and 24 and on a comparative example. Here, the comparative example was obtained using the method shown in Patent Document 1.

The procedure for creating the predicted curves L12 and L22 in the comparative example will be described below. In the method of the comparative example, the estimated values of the deterioration parameters of the secondary battery can be obtained by fitting the potential curves of the positive and negative electrodes to the charge/discharge curve of the secondary battery using the least squares method or the like. In order to obtain the predicted curves L12 and L22 in the comparative example, first, the fitting work using the least squares method was repeated for each charge/discharge curve obtained every 30 days to obtain the change over time of the deterioration parameters. Furthermore, the change over time of the deterioration parameters was extrapolated to obtain the change in the deterioration parameters into the future, and the battery capacity at a future point in time was calculated based on the result, thereby obtaining the predicted curves L12 and L22.

FIG. 7 is a diagram showing the results of predicting the change over time in the effective mass of the positive electrode active material in the comparative example.
That is, the table in the upper part of FIG. 7 shows the change over time of the estimated value of the effective mass of the positive electrode active material extracted from the test results at 60° C. That is, in the comparative example, the deterioration parameter is the “effective mass of the positive electrode active material”, and the estimated value of the deterioration parameter is obtained approximately every 30 days. The graph in the lower part of FIG. 7 shows the estimated value and its extrapolated value. The curve used for extrapolation (extrapolation curve) was applied with the exponential behavior describing function F(t) shown in Equation (3).

Next, a procedure for creating the predicted curves L11 and L21 according to this embodiment will be described.
In this embodiment, when determining the deterioration parameters by the least squares method, multiple candidate values were calculated. Specifically, multiple initial values of the parameters used for fitting were prepared when fitting the charge/discharge curve of the secondary battery to the potential curves of the positive and negative electrodes. The reason for this is that it is known that the candidate values of the deterioration parameters obtained as a result of fitting depend on the initial values. In fact, in fitting using the charge/discharge curve 180 days after the start of the test, as a result of examining about 100 types of initial values of the parameters used for fitting, multiple different values were obtained as candidate values of the deterioration parameters. The results are shown in FIG. 8.

Figure 8 shows the results of predicting the change over time in the effective mass of the positive electrode active material in this example. That is, the table in the upper part of Figure 8 shows multiple candidate values for the effective mass of the positive electrode active material extracted from the test results at 60°C, and the change over time in the estimated value selected from these candidate values. That is, in this example as well, the degradation parameter is the "effective mass of the positive electrode active material." The graph in the lower part of Figure 8 shows the estimated value and its extrapolated value. As shown in the table in the upper part of Figure 8, the candidate value of the degradation parameter is not always uniquely determined, and it can be seen that there are multiple candidate values due to the influence of measurement errors, etc.

In order to determine the change over time of the most likely candidate value of the degradation parameter, it is necessary to determine which candidate value is physically reasonable. Therefore, it is assumed that the change over time of the degradation parameter follows the exponential behavior description function F(t) shown in Equation (3), and a combination showing an exponential change over time behavior was selected as an estimated value from the combinations of candidate values shown in FIG. 8. The selected estimated value is shown at the right end of the table in the upper part of FIG. 8. In this embodiment, the estimated value obtained in this manner is considered to correctly describe the change over time of the degradation parameter. Thereafter, the extrapolated value of the capacity reduction parameter was obtained in the same manner as in the comparative example or the method of Patent Document 1, and the predicted curves L11 and L21 were obtained by calculating the battery capacity at a future point in time using the extrapolated value.

FIG. 9 is a diagram showing the error between the actual measured values and the predicted curves in the above examples and comparative examples.
That is, FIG. 9 shows the results of calculating the errors between the actual measured values 16 and 26 (see FIG. 26) that are the verification data and the predicted curves L11, L12, L21, and L22.
The root mean square error (RMSE) was used to define the error. The smaller the RMSE, the more accurate the life prediction. As is clear from Fig. 9, the prediction curves L11 and L21 according to this embodiment have smaller RMSE values than the prediction curves L12 and L22 according to the comparative example, and it can be seen that the life prediction was made with high accuracy.

[Second embodiment]
Next, a second embodiment will be described.
FIG. 10 is a block diagram of a secondary battery system 1 according to the second embodiment.
The secondary battery system 1 includes a secondary battery module 200 and a charge/discharge control device 300. The configuration of the secondary battery module 200 is the same as that of the first embodiment (see FIG. 2). The charge/discharge control device 300 includes a state determination device 100 and a charge/discharge control unit 310.

The charge/discharge control unit 310 controls the charge and discharge currents for the secondary battery 210 so that they fall within a specified current range. The state determination device 100 in this embodiment also includes an operation condition determination unit 170 in addition to the same configuration as that in the first embodiment (see FIG. 2). As described above, the deterioration prediction unit 150 predicts the current and future deterioration state (internal resistance, capacity reduction rate, etc.) of the secondary battery 210 based on the estimated value AE(t) of the deterioration parameter A.

Based on this degradation state, the operation condition determination unit 170 sets the voltage tolerance range (upper and lower limits) for the secondary battery 210 during operation, calculates the current range (upper and lower limits) of the charge/discharge current in which the battery voltage of the secondary battery 210 falls within this voltage tolerance range, and commands the charge/discharge control unit 310 to set the current range. This enables the charge/discharge control unit 310 to control the charge/discharge current so that it falls within a range corresponding to the estimated value AE(t) of the degradation parameter A, thereby enabling the secondary battery 210 to have a longer life.

[Effects of the embodiment]
As described above, according to the above-mentioned embodiment, the secondary battery state determination device 100 includes a candidate value extraction unit (120) that obtains a plurality of candidate values (AC1(tk), AC2(tk), AC3(tk), ...) for the parameter (A) of the secondary battery 210 at each of a plurality of times (tk; k = 0, 1, 2, 3, ...) based on a state quantity that is a result of measuring the state of the secondary battery 210 at a plurality of times (tk; k = 0, 1, 2, 3, ...), and a calculation unit 140 that obtains a time-dependent change (AE(t)) of an estimated value of the parameter (A) based on the plurality of candidate values (AC1(tk), AC2(tk), AC3(tk), ...) and a predetermined function (F(t)). As a result, the state of the secondary battery, such as the time-dependent change (AE(t)) of the characteristics of the deterioration parameters, can be appropriately obtained based on the plurality of candidate values (AC1(tk), AC2(tk), AC3(tk), ...).

Moreover, the secondary battery 210 has an electrode, the electrode contains an active material, and it is more preferable that the parameter (A) includes any one of the deactivation rate of the active material, the thickness of the coating film on the surface of the active material, and the resistance value of the coating film on the surface of the active material. This makes it possible to appropriately estimate any one of the deactivation rate of the active material, the thickness of the coating film on the surface of the active material, and the resistance value of the coating film on the surface of the active material.

Moreover, it is more preferable that the state quantity includes the battery voltage of the secondary battery 210, the function (F(t)) is a function including an exponential function or a power function, the parameter (A) is a quantity representing the degradation state of the secondary battery 210, and the calculation unit 140 outputs the estimated value (AE(t)) based on the function (F(t)) and the change in the battery voltage over time. This makes it possible to output a more appropriate estimated value (AE(t)) when the parameter (A) changes approximately as an exponential function or a power function.

Moreover, it is even more preferable that the secondary battery state determination device 100 further includes an operation condition determination unit 170 that sets an allowable voltage range during operation of the secondary battery 210 based on the estimated value (AE(t)) and determines the operation conditions of the secondary battery 210 so that the battery voltage of the secondary battery 210 falls within the allowable voltage range. This makes it possible to determine appropriate operation conditions for the secondary battery 210 based on the estimated value (AE(t)).

[Modification]
The present invention is not limited to the above-mentioned embodiment, and various modifications are possible. The above-mentioned embodiment is exemplified to explain the present invention in an easy-to-understand manner, and is not necessarily limited to those having all the configurations described. In addition, it is possible to replace a part of the configuration of one embodiment with the configuration of another embodiment, and it is also possible to add the configuration of another embodiment to the configuration of one embodiment. In addition, it is possible to delete a part of the configuration of each embodiment, or to add or replace other configurations. In addition, the control lines and information lines shown in the figure show those that are considered necessary for explanation, and do not necessarily show all the control lines and information lines necessary on the product. In reality, it may be considered that almost all the configurations are connected to each other. Possible modifications of the above-mentioned embodiment are, for example, as follows.

(1) In the above embodiment, when p candidate values AC1(t) to ACp(t) of the degradation parameter exist at each of n measurement times t0, t2, t3, ..., t(n-1), a set of the most likely candidate values is selected from the p candidate value sets. However, it is also possible to set all combinations of candidate values that can be traced in time series as candidate value sets, and select the most likely candidate value from among them. In the above example, the number of candidate value sets is ^pn , which becomes very large especially when the number of measurements n is large. However, if the most likely candidate value set is searched for using, for example, the Viterbi algorithm, the most likely candidate value set can be obtained with a relatively small amount of calculation.

(2) The hardware of the state determination device 100 in the above embodiment can be realized by a general computer, so programs for executing the various processes shown in Figures 2 and 10 may be stored on a storage medium or distributed via a transmission path.

(3) In the above embodiment, the processes shown in Figures 2 and 10 and the other processes described above are described as software processes using a program, but some or all of them may be replaced with hardware processes using an ASIC (Application Specific Integrated Circuit) or an FPGA (Field Programmable Gate Array), etc.

(4) The various processes performed in the above embodiment may be executed by a server computer via a network (not shown), and the various data stored in the above embodiment may also be stored in the server computer.

1 Secondary battery system 100 State determination device 120 Deterioration parameter extraction unit (candidate value extraction unit, candidate value extraction means)
140 Calculation section (calculation means)
170 Operation condition determination unit 200 Secondary battery module 210 Secondary battery 300 Charge/discharge control device 310 Charge/discharge control unit A Deterioration parameters (parameters)

Claims (7)

a candidate value extraction unit that determines a plurality of candidate values for a parameter of the secondary battery at each of a plurality of times based on a state quantity that is a result of measuring a state of the secondary battery at the plurality of times;
a calculation unit that calculates an estimated value of the parameter and a change in the estimated value over time based on a plurality of the candidate values and a predetermined function ;
the state quantity includes a measurement result of a voltage and a current in the secondary battery,
the parameter includes an effective weight of a positive electrode active material in the secondary battery, an effective weight of a negative electrode active material, or a capacity difference between a positive electrode and a negative electrode,
the candidate value extraction unit calculates a plurality of the candidate values by an optimization method using pre-stored discharge curves of the positive electrode alone and the negative electrode alone and the state quantities;
the function describes the degradation behavior of the parameter,
The estimated value is a candidate value that belongs to a candidate value set that is a combination of a plurality of the candidate values at each time and has a minimum difference from the function.
A secondary battery state determination device comprising: The device for determining a state of a secondary battery according to claim 1 , wherein the calculation unit predicts a deterioration state of the secondary battery at a future point in time using the estimated value and a change in the estimated value over time. The secondary battery includes an electrode, the electrode including an active material,
2. The device for determining a state of a secondary battery according to claim 1, wherein the parameters include any one of a deactivation rate of the active material, a film thickness on the surface of the active material, and a resistance value of the film on the surface of the active material. the state quantity includes a battery voltage of the secondary battery,
The function is a function including an exponential function or a power function,
the parameter is an amount representing a deterioration state of the secondary battery,
The device for determining a state of a secondary battery according to claim 1 , wherein the calculation unit outputs the estimated value based on the function and a change in the battery voltage over time. A secondary battery state determination device as described in any one of claims 1 to 4, further comprising an operational condition determination unit that sets an acceptable voltage range during operation of the secondary battery based on the estimated value, and determines the operational conditions of the secondary battery so that the battery voltage of the secondary battery falls within the acceptable voltage range. a candidate value extraction unit that determines a plurality of candidate values for a parameter of the secondary battery at each of a plurality of times based on a state quantity that is a result of measuring a state of the secondary battery at the plurality of times;
a calculation unit that calculates an estimated value of the parameter and a change in the estimated value over time based on a plurality of the candidate values and a predetermined function;
a charge/discharge control unit that controls charging/discharging of the secondary battery ,
the state quantity includes a measurement result of a voltage and a current in the secondary battery,
the parameter includes an effective weight of a positive electrode active material in the secondary battery, an effective weight of a negative electrode active material, or a capacity difference between a positive electrode and a negative electrode,
the candidate value extraction unit calculates a plurality of the candidate values by an optimization method using pre-stored discharge curves of the positive electrode alone and the negative electrode alone and the state quantities;
the function describes the degradation behavior of the parameter,
The estimated value is a candidate value that belongs to a candidate value set that is a combination of a plurality of the candidate values at each time and has a minimum difference from the function.
A charge/discharge control device comprising: a secondary battery module having a plurality of secondary batteries;
a candidate value extracting unit that obtains a plurality of candidate values for a parameter of the secondary battery at each of a plurality of times based on a state quantity that is a result of measuring a state of the secondary battery module at the plurality of times;
a calculation unit that calculates an estimated value of the parameter and a change in the estimated value over time based on a plurality of the candidate values and a predetermined function ;
the state quantity includes a measurement result of a voltage and a current in the secondary battery,
the parameter includes an effective weight of a positive electrode active material in the secondary battery, an effective weight of a negative electrode active material, or a capacity difference between a positive electrode and a negative electrode,
the candidate value extraction unit calculates a plurality of the candidate values by an optimization method using pre-stored discharge curves of the positive electrode alone and the negative electrode alone and the state quantities;
the function describes the degradation behavior of the parameter,
The estimated value is a candidate value that belongs to a candidate value set that is a combination of a plurality of the candidate values at each time and has a minimum difference from the function.
A secondary battery system comprising: