JP2018141665A - Battery management method, battery management device, and computer program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a new method for obtaining full charge capacity FCC on the assumption of a change in the full charge capacity resulting from current dependency, temperature dependency, manufacturing variation dependency, deterioration dependency, and so on.SOLUTION: Using a SOC and discharge current integrated quantity Q, an upper limit value FCCu of FCC is calculated from an equation of FCCu=Q/SOC. Due to the flow of discharge current I, a terminal voltage value V decreases lower than OCV by an amount of voltage drop I×R due to internal impedance R. Therefore, actual FCC becomes smaller than the upper limit value FCCu. Therefore, the FCC is found from an equation of FCC(I,T)=FCCu-I×R(I,T)×dQ/dV by subtracting an amount of loss Qloss caused by the flow of the discharge current I from the upper limit value FCCu. In the equation, R(I,T) represents the value of the internal impedance R corresponding to the current I and a temperature T.SELECTED DRAWING: Figure 6

Description

  The present invention relates to battery management.

  Patent Document 1 discloses a method for estimating a state of charge (SOC) using a Kalman filter. Patent document 2 discloses identifying the internal impedance of the battery and identifying the open circuit voltage (OCV) of the battery from the internal impedance.

Japanese Patent Laid-Open No. 2006-99123 Japanese Patent Laid-Open No. 2006-155671

  In order to obtain the SOC or the like, the total charge capacity (FCC) of the battery may be used. Therefore, accurately grasping the FCC is important for battery management. In general, when FCC of a primary battery is required, for example, FCC published as a specification value from a battery manufacturer is used. Moreover, FCC of a secondary battery is calculated | required by fully charging a secondary battery.

  However, since the primary battery cannot be charged, the method of obtaining the FCC by fully charging like a secondary battery cannot be used. If the manufacturer's specification value is unknown, it is difficult to obtain the FCC.

  In addition, the FCC may fluctuate due to various factors, and the FCC as a manufacturer specification value or the FCC obtained by full charging may not be reliable. The present inventor has found that FCC has current dependency, temperature dependency, manufacturing variation dependency, deterioration dependency, and the like. Because of these dependencies, the FCC as a specification value and the FCC obtained before use are not necessarily reliable.

  Therefore, a new method for obtaining the FCC on the assumption of the fluctuation of the FCC is required.

  In the embodiment, the total charge capacity (FCC) is calculated using an upper limit value of the total charge capacity and a loss that occurs according to the discharge current value of the battery. The upper limit value of the total charge capacity can be calculated using the state of charge of the battery and the accumulated discharge current.

It is a block diagram of a management apparatus and a sensor module. It is a flowchart of a management process. It is a figure which shows the example of an OCV-SOC curve. It is a flowchart of the process using a sequential least square method. It is an equivalent circuit of a battery. It is a figure which shows the change of OCV and a terminal voltage with respect to an electric current integrated value. It is a figure which shows the change of OCV with respect to SOC, and an internal impedance.

[1. Battery management method, management device, computer program]

(1) The battery management method according to the embodiment calculates the upper limit value FCC _u of the total charge capacity using the state of charge SOC of the battery and the accumulated discharge current Q, and the upper limit value FCC _u Calculating the total charge capacity FCC using a _loss Q _loss generated according to the discharge current value I of the battery. Even if the total charge capacity FCC varies, the total charge capacity FCC can be obtained by these calculations.

(2) The loss can be calculated using a voltage drop generated in the internal impedance R of the battery by the discharge current I and a voltage decrease characteristic of the battery as the discharge current integration amount increases. The voltage decrease characteristic of the battery as the discharge current integration amount increases may be a decrease characteristic of the open circuit voltage OCV as the discharge current integration amount Q increases, or a terminal associated with the increase of the discharge current integration amount Q. The decreasing characteristic of the voltage V may be sufficient.

(3) The state of charge SOC is preferably calculated from the open circuit voltage OCV of the battery. The state of charge SOC is calculated from the open circuit voltage OCV using, for example, the OCV-SOC relationship. The open circuit voltage OCV is preferably calculated by a sequential least square method. For the calculation of the open circuit voltage OCV by the successive least squares method, for example, the method described in Patent Document 2 can be used.

(4) The remaining usable time t of the battery can be calculated from the total charge capacity FCC.

(5) The deterioration state of the battery can be calculated from the total charge capacity FCC. Since the calculated total charge capacity FCC is a value reflecting any change due to deterioration, the deterioration state can be calculated from the calculated total charge capacity.

(6) The management method includes repeating the calculation of the state of charge, and repeating the calculation of the state of charge includes switching the calculation method of the state of charge from the first method to the second method. preferable. By switching the charging state calculation method during repetition, the charging state can be calculated by an appropriate method according to the situation.

(7) The first method is preferably a method of calculating the state of charge by a current integration method. The second method is preferably a method for calculating the state of charge from a closed circuit voltage of the battery.

(8) The closed circuit voltage is preferably calculated by a sequential least square method.

(9) It is preferable that the management method includes determining whether to switch from the first method to the second method based on the switching index. The switching index is preferably one or a plurality of indices selected from the group consisting of a battery state of charge, a closed circuit voltage, a change in open circuit voltage, an internal impedance, and a change in internal impedance.

(10) The battery is preferably a primary battery, but may be a secondary battery.

(11) A battery management device according to an embodiment includes a processor that executes processing for managing a battery. The processing uses the state of charge of the battery and the integrated amount of discharge current to calculate the upper limit value of the total charge capacity, the upper limit value, and a loss that occurs according to the discharge current value of the battery, And calculating the total charge capacity.

  The battery management device may be a device that is powered by the battery to be managed, or performs a process for managing the battery based on information obtained by communication from a device that is powered by the battery to be managed. It may be a device to execute.

(12) A computer program according to an embodiment causes a computer to execute processing for managing a battery. The processing uses the state of charge of the battery and the integrated amount of discharge current to calculate the upper limit value of the total charge capacity, the upper limit value, and a loss that occurs according to the discharge current value of the battery, And calculating the total charge capacity. The computer program is stored in a computer readable storage medium. A computer performs the process which manages a battery by running a computer program.

[2. Example of battery management]

[2.1 Management device and sensor module]

  FIG. 1 shows a management apparatus 100. The management apparatus 100 in FIG. 1 manages the sensor module 200. The management apparatus 100 includes a computer having a processor 110 and a memory 120. The memory 120 here may be a primary storage device or a secondary storage device.

  The processor 110 is connected to an input device 130 and an output device 140 via an interface (not shown). The input device 130 is, for example, a keyboard or a mouse. The output device 140 is, for example, a display. A radio device 150 is also connected to the processor 110 via an interface (not shown). The wireless device 150 performs wireless communication with the sensor module 200. The wireless device 150 is, for example, a wireless device for short-range wireless communication. Note that the communication between the management apparatus 100 and the sensor module 200 may be wired communication.

  The management apparatus 100 manages the battery 220 provided in the sensor module 200 in addition to managing the sensing information acquired by the sensor module 200. The memory 120 stores a computer program 121 that causes the processor 110 to execute these management processes. The memory 120 stores various information 122 and 123 used for management.

  The sensor module 200 includes a sensing device 230. The sensing device 230 senses a sensing target and outputs an electrical signal. The sensing target is, for example, speed, acceleration, atmospheric pressure, or temperature. The electrical signal output from the sensing device 230 is given to the processor 210.

  The sensor module 200 includes a battery 220. The battery 220 supplies electric power to the sensing device 230 that is a load. In FIG. 1, the battery 220 is illustrated to supply power only to the sensing device 230, but in reality, the battery 220 supplies power to all elements that require power in the processor 210 and the sensor module 200.

  The battery 220 is a primary battery, for example. The primary battery has a different total charge capacity (FCC) depending on an elapsed time since manufacture, a storage state, and manufacturing variations, and calculation of a state of charge (SOC) is not always easy. In the following description, the SOC indicates the remaining battery level as a state of charge. The remaining battery level is indicated by, for example, a value from 100% to 0%, and the SOC becomes 0% when the remaining battery level is exhausted.

  The sensor module 200 of FIG. 1 includes an ammeter 240, a voltmeter 250, and a thermometer 260. The ammeter 240 measures a discharge current value I from the battery 220 to the load. The voltmeter 250 measures the terminal voltage value V of the battery 220. Thermometer 260 measures temperature T of battery 220. Each electric signal indicating the discharge current value I, the terminal voltage value V, and the temperature T is given to the processor 210.

  The processor 210 transmits the acquired current value I, voltage value V, temperature T, and sensing information to the management apparatus 100 via the wireless device 270.

  The management apparatus 100 receives the current value I, the voltage value V, the temperature T, and sensing information via the wireless device 150. The received information is stored in the memory 120. The processor 110 manages the sensing information stored in the memory 120 and manages the battery 220 based on the current value I, the voltage value V, and the temperature T stored in the memory 120.

[2.2 Management processing]

  The processor 110 executes the process shown in FIG. 2 by executing the computer program 121. The process shown in FIG. 2 includes calculating the remaining capacity (SOC) of the battery 220 and calculating the total charge capacity (FCC) of the battery 220.

  In step S11, the processor 110 executes an initial setting process. In the initial setting, an initial SOC and an initial FCC are set. The initial SOC and the initial FCC are input from the input device 130 and stored in the memory 120, for example. The initial SOC is, for example, 100%, and the initial FCC is, for example, an FCC specification value published by a battery manufacturer.

  The process of FIG. 2 includes two methods for calculating the SOC. In the first method, the SOC is calculated by a current integration method. The first method corresponds to step S13. The second method is a method of calculating the open circuit voltage (OCV) of the battery 220 and calculating the SOC from the OCV using the relationship between the OCV and the SOC. The second method corresponds to step S14. In the second method, for example, the OCV is calculated by a sequential least square method. In the second method, as described later, the FCC and the remaining time t are also calculated. The calculated SOC, OCV, FCC, remaining time t, and the like are output from, for example, the output device 140 in step S15.

  The second method is useful when the battery 220 has the characteristic that the OCV decreases as the SOC decreases. If the battery 220 has such characteristics, the OCV-SOC relationship can be obtained in advance, and the SOC can be calculated from the OCV using the OSC-SOC relationship.

  However, some secondary batteries using a primary battery or iron olivicate have an OCV-SOC relationship as shown in FIG. In FIG. 3, in the first half region where the remaining amount of the battery is large and the SOC is relatively high (for example, the region where the SOC is 100% to 50%), the OCV becomes a substantially constant flat region even if the SOC decreases. In the flat first half region, even if the OCV is known, if the SOC is obtained from the OCV-SOC relationship, the error becomes large.

  On the other hand, in the OCV-SOC relationship of FIG. 3, the latter half region where the SOC is low is an inclined region where the OCV decreases as the SOC decreases. In the inclined second half region, the SOC can be calculated from the OCV with relatively high accuracy using the OCV-SOC relationship.

  In the process of FIG. 2, the processor 110 executes the state determination process in step S12 in order to calculate the SOC by the first method in the flat first half region and to calculate the SOC by the second method in the inclined second half region. If it is determined by the state determination that the region is a flat region, the first method of step S13 is executed. If it is determined that the region is an inclined region, the second method is executed.

  As shown in FIG. 2, steps S12 to S15 are repeated loop processing. Therefore, the calculation of SOC or the like is repeatedly calculated while the battery 220 is in use. During the repetition of the calculation of the SOC, the SOC is calculated by the first method because the remaining battery level is initially in the flat first half region. When the remaining battery level decreases, the second half region is inclined, so that the SOC calculation method is switched from the first method to the second method. A determination method for switching will be described later.

[2.3 First Method: Current Integration Method]

  In the current integration method used in the first method, the SOC is estimated based on the current value entering and exiting the battery. Here, since the battery 220 is a primary battery, only the discharge current output from the battery needs to be considered.

In the current integration method, assuming that the current flowing from time t _k to time t _{k + 1} is I, SOC (k + 1) at time k + 1 is expressed as the following equation (1).

  In the equation (1), the initial SOC set in the initial setting in step S11 is used as the SOC (0) when the time k = 0. In addition, the initial FCC set in the initial setting is used as the FCC in Expression (1).

  Since the first method does not use the OCV-SOC relationship as in the second method, the SOC can be calculated more accurately than in the second method in the first half region where the remaining amount is large and flat. However, since the initial FCC used in the current integration method is not necessarily reliable, the absolute value of the SOC is relatively low in the current integration method. On the other hand, in the current integration method, the amount of decrease in SOC is represented with relatively high accuracy. In the process of FIG. 2, the current integration method of the first method is used when the remaining amount is relatively large. When the remaining amount is large, the exact absolute value of the SOC is not so important, and the approximate decrease in the SOC is achieved. If you know the amount, it is often enough. For this reason, when the first method is used, the low reliability of the initial FCC is not a problem.

[2.4 Second Method: Sequential Least Squares Method]

  In the second method, unlike the first method, the SOC is calculated without using the initial FCC. As a result, the SOC can be calculated with relatively high accuracy based on the OCV-SOC relationship without being affected by the reliability of the initial FCC. In the processing of FIG. 2, the second method is used when the remaining amount is relatively low, and when the remaining amount is low, it is more important to accurately grasp how long the battery 220 can be used. It becomes. Therefore, as in the second method, it is advantageous that the SOC can be calculated without being affected by the reliability of the initial FCC.

  FIG. 4 shows details of step S14 corresponding to the second method. In the process of FIG. 4, in addition to the SOC, the remaining time of the FCC and the battery is calculated. The remaining time is the remaining time that the battery 220 can be used.

  As shown in FIG. 4, in step S <b> 21, the processor 110 obtains an accumulated current amount Q from the start of use of the battery 220 to the present from the discharge current value I of the battery 220. The obtained integrated current amount Q is stored in the memory 120.

In step S22, the processor 110 calculates the OCV. The OCV can be calculated from the terminal voltage value V of the battery. As is clear from the equivalent circuit model of the battery 220 shown in FIG. 5, the terminal voltage value V is expressed by the following equation (2). In the formula (2), V _IR is a voltage drop in the internal impedance R.

  In FIG. 5, the internal impedance of the battery 220 is a simple equivalent circuit model indicated by only the resistor R connected in series with the OCV. The equivalent circuit model of the battery 220 may be more complicated including a capacitor and the like.

From the above equation (2), the OCV can be obtained by removing the voltage drop _VIR in the internal impedance R from the terminal voltage value V. The voltage drop V _IR in the internal impedance R is calculated by, for example, the internal impedance R and the discharge current value I. Although the internal impedance R varies depending on the discharge current value I and the temperature T, the current value can be obtained by obtaining the dependency characteristic R (I, T) of the internal impedance R on the current value I and the temperature T in advance. From I and temperature T, the internal impedance can be uniquely determined. Note that R (I, T) obtained in advance is stored in the storage area 123 of the memory 120 and used for OCV calculation.

  In order to obtain the OCV with higher accuracy, it is preferable to perform OCV identification (system identification) by sequential transient response analysis using the sequential least square method, rather than directly applying Equation (2). An example of the OCV identification method is disclosed in Patent Document 2. Patent Document 2 discloses that an internal impedance of a battery is identified by a sequential least square method based on an equivalent circuit model of the battery, and an OCV is identified from the internal impedance. The equivalent circuit model of Patent Document 2 is a higher-order model including a capacitor. However, for the sake of easy understanding, an OCV identification method based on the successive least squares method will be described below based on the simple equivalent circuit model of FIG. explain.

First, the above equation (2) is differentiated by the forward method, and the following equation (3), (4), (5) is obtained by adding the observation noise term w (k).

  Here, y (k) in Expression (3) indicates the actual terminal voltage of the battery 220 at time k. Equation (4) consists of data available at time k and is called a regression vector. Equation (5) consists of unknown parameters.

In the successive least squares method, the sum of the squares of errors e (k) between the actual terminal voltage y (k) and the terminal voltage of the battery equivalent circuit model for the same V _IR (k) is minimized. Then, the parameter θ (k) of the equivalent circuit model is optimally obtained.

In this successive least square method, a forgetting element is used. In the successive least squares method using forgetting elements, first, an estimation parameter θ (k), initial values θ (0) and P (0) of a covariance matrix P (N), and a forgetting factor λ are set. The covariance matrix P (N) is a matrix related to the regression vector as shown in Expression (6).

In the successive least squares method, first, a gain G (k) is calculated from the covariance matrix P (N) based on Expression (7).

Next, an output prediction error e (k) shown in Expression (8) is obtained.

Then, the OCV included in the parameter θ (k) is calculated from the values G (k) and e (k) obtained by the equations (7) and (8) based on the equation (9).

Further, the covariance matrix P (k) is updated based on Expression (10).

  When the OCV is calculated, the processor 110 calculates the SOC in step S23. The processor 110 uses the OCV-SOC relationship 122 stored in the memory 120 for the calculation of the SOC. The OCV calculated by the successive least square method is converted into SOC based on the OCV-SOC relationship 112.

[2.5 FCC calculation]

When the SOC is calculated, the processor 110 calculates the FCC in step S24. The FCC is calculated using the upper limit FCC _u of the FCC and the loss Q _loss that occurs according to the discharge current value I of the battery 220. The upper limit FCC _u of FCC is calculated using the SOC calculated in step S23 and the discharge current integrated amount Q calculated in step S21.

The FCC is calculated based on the following formula, for example. First, the upper limit FCC _u of FCC is calculated by the following equation (11).

The upper limit value FCC _u calculated by the equation (11) is a theoretical FCC when the discharge current value I = 0. That is, the upper limit value FCC _u is an FCC when the terminal voltage value V of the battery 220 is equal to the OCV because the current value I = 0. However, as shown in FIG. 6, when the discharge current I flows, the terminal voltage value V becomes lower than the OCV by the voltage drop I × R due to the internal impedance R. For this reason, the actual FCC is smaller than the upper limit value FCC _u . Therefore, the FCC can be obtained by subtracting the loss Q _loss caused by the discharge current I from the upper limit FCC _u .

The FCC is calculated by, for example, the following formula (12). In the formula (12), FCC is expressed as FCC (I, T) because it has current dependency and temperature dependency.

In equation (12), the term I × R (I, T) × dQ / dV represents the loss Q _loss . In Formula (12), I is a discharge current value. R (I, T) is a value of the internal impedance R according to the current I and the temperature T. Since the internal impedance R changes depending on the current I and the temperature T, the loss Q _loss also has current dependency and temperature dependency. The internal impedance R is calculated based on the current I and the temperature T, and is stored in the memory area 123. dQ / dV indicates the slope of an approximate straight line in the slope region in which the OCV decreases as Q increases in the OCV change curve corresponding to the current accumulated amount Q as shown in FIG. dQ / dV may be an inclination of an approximate straight line of an inclined region in which the terminal voltage V decreases as Q increases in the terminal voltage V change curve corresponding to the integrated current amount as shown in FIG. The slope dQ / dV may be obtained in advance through experiments and stored in the memory 120, or may be calculated based on the OCV or terminal voltage V and the current integration amount Q. May be.

[2.6 Calculation of remaining time using FCC]

When FCC (I, T) is calculated, the processor 110 calculates the remaining usable time t of the battery 220 in step S25. The remaining time t is calculated by the following equation (13) using, for example, FCC (I, T) and Q.

  If it is necessary to calculate the remaining time t until FCC (I, T) is calculated in step S24, the processor 110 uses the initial FCC as FCC (I, T) in Expression (13). Since the initial FCC is not always reliable, the reliability of the remaining time t calculated from the initial FCC is low, but the remaining time t is used as an alternative until the FCC (I, T) is calculated in step S24. Can be calculated.

  Further, the processor may obtain data indicating the deterioration of the battery 220 in accordance with the magnitude or fluctuation of the FCC calculated in step S24. The degree of deterioration is calculated from, for example, the current FCC decrease with respect to the reference FCC, with the initial FCC or the FCC when the remaining battery level is high as the reference FCC. It can also be determined that the battery 220 has deteriorated when the current FCC becomes smaller than a predetermined threshold.

[2.7 Switching judgment]

  Hereinafter, an example of the determination method in step S12 of FIG. 2 will be described. The processor 110 monitors a predetermined switching index value, compares the switching index value with a threshold value, and changes the SOC calculation method from the first method (step S13) to the second method (step S13). Switch to S14).

  The first example of the switching index is a value calculated from the SOC itself or the SOC. For example, the processor 110 uses the SOC itself calculated by the first method as a switching index. When the SOC falls below a predetermined threshold (for example, 30%), the processor 110 considers that the SOC has entered the inclined region in FIG. The calculation method can be switched from the first method to the second method.

  The second example of the switching index is a value calculated from the OCV itself or the OCV. The processor 110 calculates the OCV by the same method as in step S22 before the SOC calculation method is switched to the second method, and the SOC calculation method is changed from the first method based on the calculated OCV. It is possible to switch to the second method. For example, when the OCV itself is used as a switching index, and the OCV itself falls below a predetermined threshold voltage (for example, 2.7 [V] in FIG. 3), it is considered that the OCV itself has entered the inclined region in FIG. Can be switched from the first method to the second method.

  The switching index may be dOCV / dSOC indicating a change in OCV with respect to a change in SOC. For example, dOCV / dSOC is calculated based on the SOC calculated by the first method in step S13 and the OCV calculated by the same method as in step S22. Since dOCV / dSOC increases in the inclined region of FIG. 3, when dOCV / dSOC is used as a switching index, the SOC calculation method is switched from the first method to the second method when the degree of change in OCV increases. be able to. If the degree of change in OCV is used as a switching index instead of the OCV itself, switching can be determined regardless of the type of battery. The index indicating the degree of change in OCV may be dOCV / dQ or dOCV / dt.

  A third example of the switching index is a value calculated from the internal impedance R itself or the internal impedance R of the battery 220. The processor 110 calculates the internal impedance by, for example, the same method as the internal impedance identification method described in Patent Document 2 before the SOC calculation method is switched to the second method, and based on the calculated internal impedance. Thus, the SOC calculation method can be switched from the first method to the second method. The internal impedance R increases as the SOC decreases, as shown in FIG. Therefore, for example, when the internal impedance R itself is used as a switching index, and the internal impedance R exceeds a threshold value, the SOC calculation method can be switched from the first method to the second method.

  The switching index may be dR / dSOC indicating a change in internal impedance R with respect to a change in SOC. If the degree of change of the internal impedance R is used as a switching index instead of the internal impedance itself, switching can be determined regardless of the type of battery. The index indicating the degree of change in internal impedance may be dR / dQ or dR / dt.

  As shown in FIG. 7, the increase in internal impedance R occurs prior to the decrease in OCV. Therefore, the processor 110 initially monitors the change in the internal impedance R, and when an increase in the internal impedance R is detected, starts monitoring the change in the OCV, and calculates the SOC based on the change in the OCV. The method can be switched from the first method to the second method. By doing in this way, the change of OCV can be detected in advance based on the change of the internal impedance R, and the change of the OCV can be detected with high accuracy.

  The processor 110 initially monitors only the value of the SOC itself. When the internal impedance R falls below a predetermined threshold (for example, 50%), the processor 110 monitors the change in the internal impedance R, and the increase in the internal impedance R Once detected, the SOC change may be monitored. Further, the processor 110 may monitor a change in the SOC when the internal impedance R falls below a predetermined threshold. In this way, monitoring for switching determination can be performed without increasing the processing load so much. That is, since the SOC is calculated in step S13, if only the SOC itself is monitored, it is not necessary to separately calculate an index for the determination in step S12, and an increase in processing load can be suppressed.

  Thus, it is preferable to perform the determination in step S12 by combining a plurality of switching indexes.

[3. Addendum]

  The present invention is not limited to the above embodiment, and various modifications are possible.

  In FIG. 2, the FCC calculation is performed only in step S14, and is not performed when the SOC calculation is performed in the first method of step S13. However, by omitting step S12 and step S13, the FCC is calculated. May be always calculated. In many secondary batteries, even when the SOC is high, the OCV decreases as the SOC decreases. Therefore, the FCC can be calculated by the same method as in step S14.

  This specification also naturally discloses the following battery management method. That is, the battery management method disclosed in the present specification includes repeating the calculation of the state of charge, and repeating the calculation of the state of charge includes changing the method of calculating the state of charge from the first method to the second method. Including switching to a method.

100 Management Device 110 Processor 120 Memory 130 Input Device 140 Output Device 150 Radio 200 Sensor Module 210 Processor 220 Battery 230 Sensing Device 240 Ammeter 250 Voltmeter 260 Thermometer 270 Radio

Claims (12)

A battery management method,
Using the state of charge of the battery and the accumulated discharge current, calculate the upper limit of the total charge capacity,
A battery management method, comprising: calculating the total charge capacity using the upper limit value and a loss generated according to a discharge current value of the battery. The method according to claim 1, wherein the loss is calculated using a voltage drop that occurs in an internal impedance of the battery due to the discharge current and a voltage decrease characteristic of the battery that accompanies an increase in the accumulated amount of discharge current. The state of charge is calculated from the open circuit voltage of the battery,
The method according to claim 1, wherein the open circuit voltage is calculated by a sequential least square method. The method according to claim 1, further comprising calculating a remaining usable time of the battery from the total charge capacity. The method according to claim 1, further comprising calculating a deterioration state of the battery from the total charge capacity. Further comprising repeating the calculation of the state of charge,
The method according to any one of claims 1 to 5, wherein repeating the calculation of the state of charge includes switching the method of calculating the state of charge from a first method to a second method. The first method is a method of calculating the state of charge by a current integration method,
The method according to claim 6, wherein the second method is a method of calculating the state of charge from a closed circuit voltage of the battery. The method according to claim 7, wherein the closed circuit voltage is calculated by a sequential least squares method. Determining whether to switch from the first method to the second method based on the switching index;
The switching index is one or a plurality of indices selected from the group consisting of a battery charge state, closed circuit voltage, change in open circuit voltage, internal impedance, and change in internal impedance. 2. The method according to item 1. The method according to claim 1, wherein the battery is a primary battery. A processor for performing processing for managing the battery;
The process is
Using the state of charge of the battery and the accumulated discharge current, a process for calculating an upper limit value of the total charge capacity;
A process of calculating the total charge capacity using the upper limit value and a loss generated according to the discharge current value of the battery;
Battery management device including A computer program for causing a computer to execute a process for managing a battery,
The process is
Using the state of charge of the battery and the accumulated discharge current, a process for calculating an upper limit value of the total charge capacity;
A process of calculating the total charge capacity using the upper limit value and a loss generated according to the discharge current value of the battery;
A computer program containing