JP2024031219A - Full charge capacity estimation method - Google Patents

Abstract

The present invention provides a technology that enables accurate estimation of the full charge capacity of a battery. A full charge capacity estimation method includes the step of generating a plurality of full charge capacity curves indicating changes in full charge capacity with respect to elapsed time of the battery, according to each of a plurality of stages of deterioration state of the battery. The method includes the step of generating a plurality of SOC-OCV characteristic curves indicating the relationship between the charging rate and the open-circuit voltage of the battery in accordance with each of the plurality of stages of deterioration states of the battery. The method includes the step of generating an optimal full charge capacity curve close to a current state of deterioration using a plurality of full charge capacity curves based on the temperature history and SOC history of the battery. The method includes generating an optimal SOC-OCV characteristic curve close to a current state of deterioration using a plurality of SOC-OCV characteristic curves based on temperature history, SOC history, and elapsed time of the battery. The method includes the step of estimating the full charge capacity of the battery based on the optimum full charge capacity curve and the optimum SOC-OCV characteristic curve. [Selection diagram] Figure 2

Description

The technology disclosed in this specification relates to a method for estimating the full charge capacity of a battery.

In the technique of Patent Document 1, a deterioration line (a curve indicating a change in full charge capacity with respect to battery usage) is corrected according to the ratio of the number of times of timer charging to the number of times of charging. The purpose of this is to improve the accuracy of estimating the full charge capacity of the battery.

JP2020-128970A

In the technique of Patent Document 1, the input information used for correcting the degraded line is the number of times of charging, and is limited. Therefore, it is difficult to ensure sufficient accuracy in estimating the full charge capacity. This specification provides a technique that enables accurate estimation of the full charge capacity of a battery.

In the configuration according to claim 1, the elapsed time of the battery is a time corresponding to an integrated value of the time during which the battery is used. According to this configuration, an optimal full charge capacity curve representing the current state of deterioration of the battery can be generated based on the temperature history and SOC history of the battery. Furthermore, an optimal SOC-OCV characteristic curve representing the current state of deterioration of the battery can be generated based on the temperature history, SOC history, and battery usage. Since the full charge capacity can be estimated based on the optimum full charge capacity curve and the optimum SOC-OCV characteristic curve, it is possible to ensure sufficient estimation accuracy.

1 is a block diagram of a full charge capacity estimation system 1. FIG. 7 is a flowchart illustrating an operation for estimating full charge capacity according to the first embodiment. 7 is a flowchart illustrating the operation of estimating the full charge capacity according to the second embodiment.

(Configuration of full charge capacity estimation system 1)
A full charge capacity estimation system 1 shown in FIG. 1 mainly includes an electric vehicle 10 and a data center 100. Note that the dotted lines in FIG. 1 represent signal lines. The electric vehicle 10 includes a battery 11, a power converter 12, a driving motor 13, a controller 14, a voltage sensor 20, a current sensor 21, a temperature sensor 22, a charger 24, a charging socket 23, and a data communication device 26. The battery 11 is a rechargeable secondary battery, specifically a lithium ion battery. The power converter 12 converts the DC power of the battery 11 into drive power (AC power) for the motor 13 . A power converter 12 supplies drive power to a motor 13. When the driver steps on the brake, the motor 13 is reversely driven to generate electricity. The power (regenerated power) obtained by power generation is converted into DC power by the power converter 12 and supplied to the battery 11 . That is, the battery 11 is charged with regenerated power. Electric vehicle 10 can receive power supply from an external power source (not shown) via charging socket 23 and charger 24 . This allows plug-in charging of the battery 11.

The current sensor 21 measures the charging/discharging current of the battery 11 and transmits the measured data to the controller 14. The temperature sensor 22 measures the temperature of the battery 11 and transmits the measured data to the controller 14. Voltage sensor 20 measures the voltage of battery 11 and transmits the measurement data to controller 14 . The open circuit voltage (OCV) of the battery 11 has a positive correlation with the remaining power amount (State of Charge: SOC). Therefore, the controller 14 can determine the SOC of the battery 11 from the OCV of the battery 11 using the correlation.

The controller 14 is a computer that centrally controls various elements that make up the electric vehicle 10. The controller 14 includes a CPU 15, a memory 16, and a timer 17. The memory 16 stores various control programs and the like. The CPU 15 performs control described below in accordance with a control program. The timer 17 measures the elapsed time T of the battery 11. The elapsed time T is a time corresponding to the cumulative value of the time during which the battery 11 is used. The elapsed time T of the battery 11 can be considered to be approximately equal to the elapsed time T of the electric vehicle 10.

The memory 16 also stores three full charge capacity curves FC1 to FC3 and three SOC-OCV characteristic curves SO1 to SO3. The full charge capacity curve is a curve that shows changes in the full charge capacity of the battery 11 with respect to elapsed time. The SOC-OCV characteristic curve is a curve showing the relationship between SOC (charging rate) and OCV (open circuit voltage) of a battery. The full charge capacity curves FC1 to FC3 and the SOC-OCV characteristic curves SO1 to SO3 can be created in advance at the vehicle design stage and stored in the memory 16. Note that the number of full charge capacity curves and SOC-OCV characteristic curves in this embodiment is merely an example, and can be determined as appropriate depending on the capacity of the memory 16 and the required estimation accuracy of the full charge capacity.

The data communication device 26 is a device that communicates various data with the data center 100 via wireless communication 27. The data center 100 includes a CPU 101 and a memory 102. The memory 102 stores various data received from the electric vehicle 10. CPU 101 executes various calculation processes based on the received data and transmits the calculation results to electric vehicle 10.

(Summary of full charge capacity curve FC and SOC-OCV characteristic curve SO)
Graph 3 in FIG. 2 shows an example of full charge capacity curves FC1 to FC3. The horizontal axis of graph 3 is the elapsed time of the battery 11, and the vertical axis is the full charge capacity.

An example of a method for creating full charge capacity curves FC1 to FC3 will be described. The full charge capacity curves FC1 to FC3 are curves whose parameters are the probability frequency F of battery temperature and the probability frequency G of SOC. The probability frequency F of battery temperature and the probability frequency G of SOC will be explained using graphs 1 and 2 of FIG. Graph 1 is a histogram, where the horizontal axis is the battery temperature and the vertical axis (degrees) is its probability frequency F. Graph 2 is also a histogram, the horizontal axis is the SOC, and the vertical axis (frequency) is its probability frequency G. The probability frequencies F and G are statistical probabilities, and can be obtained as approximate values by cumulatively increasing the number of times the battery temperature and SOC are measured. At the vehicle design stage, full charge capacity curves FC1 to FC3 are created for various combinations of the battery temperature probability frequency F and the SOC probability frequency G. That is, the full charge capacity curves FC1 to FC3 are full charge capacity curves corresponding to each of the plurality of stages of deterioration states of the battery 11.

Further, graph 4 in FIG. 2 shows an example of SOC-OCV characteristic curves SO1 to SO3. The horizontal axis of the graph 4 is the SOC of the battery 11, and the vertical axis is the OCV. An example of a method for creating the SOC-OCV characteristic curves SO1 to SO3 will be described. The SOC-OCV characteristic curves SO1 to SO3 are curves whose parameters are the probability frequency F of the battery temperature, the probability frequency G of the SOC, and the elapsed time T of the battery 11. At the vehicle design stage, SOC-OCV characteristic curves SO1 to SO3 are created for various combinations of the probability frequency F of the battery temperature, the probability frequency G of the SOC, and the elapsed time T of the battery 11. That is, the SOC-OCV characteristic curves SO1 to SO3 are SOC-OCV characteristic curves corresponding to each of the plurality of stages of deterioration states of the battery 11.

Note that there may be various methods for creating the full charge capacity curves FC1 to FC3 and the SOC-OCV characteristic curves SO1 to SO3. For example, it may be created by actually measuring the deterioration of a plurality of batteries 11 under different conditions. For example, as a user's usage, a plurality of curves may be created by interpolating between two curves actually measured under conditions where deterioration is least likely to occur and conditions where deterioration is most likely to occur.

(Full charge capacity estimation operation)
The operation for estimating the full charge capacity will be explained using the flowchart in FIG. 2 . This estimation operation is executed onboard the electric vehicle 10. Hereinafter, "Step 10" will be abbreviated as "S10". The flow in FIG. 2 may be executed at a predetermined period (eg, one day).

In S10, the CPU 15 updates information indicating how the battery 11 is used. Specifically, as explained in Graph 1, the probability frequency F of the battery temperature is cumulatively calculated. Further, as explained in connection with graph 2, the probability frequency G of the SOC is cumulatively calculated. Further, the timer 17 is used to cumulatively obtain the elapsed time T of the battery 11.

In S20, the CPU 15 generates an optimal full charge capacity curve close to the current deterioration state based on the battery temperature probability frequency F and the SOC probability frequency G updated in S10. There may be various generation methods. For example, the full charge capacity curve with the closest deterioration condition may be selected from among the full charge capacity curves FC1 to FC3 (see graph 3). Alternatively, for example, the optimum full charge capacity curve may be generated by selecting two full charge capacity curves with similar deterioration conditions and interpolating the two full charge capacity curves. In this embodiment, a case will be described in which the full charge capacity curve FC2 is selected as the optimum full charge capacity curve.

In S30, the CPU 15 generates an optimal SOC-OCV characteristic curve close to the current state of deterioration based on the probability frequency F of the battery temperature, the probability frequency G of the SOC, and the elapsed time T that were updated in S10. There may be various generation methods. For example, the full charge capacity curve with the closest deterioration condition may be selected from among the SOC-OCV characteristic curves SO1 to SO3 (see graph 4), or the optimum full charge capacity curve can be selected by interpolating multiple SOC-OCV characteristic curves. A capacity curve may also be generated. In this embodiment, a case will be described in which the SOC-OCV characteristic curve SO2 is selected as the optimum SOC-OCV characteristic curve.

In S40, the CPU 15 estimates the full charge capacity of the battery 11. Specifically, as shown in graph 5, the estimated full charge capacity value FE1 corresponding to the elapsed time T is determined in the optimal full charge capacity curve FC2.

In S50, the CPU 15 determines whether or not the battery 11 has been plug-in charged. If plug-in charging is not being performed, S60 is skipped, and if plug-in charging is being performed, the process advances to S60.

In S60, the CPU 15 estimates the full charge capacity of the battery 11. This will be explained in detail using graph 6. The CPU 15 calculates the OCV before charging and the OCV after charging. Next, ΔSOC, which is the difference between the SOC corresponding to the pre-charge OCV and the SOC corresponding to the post-charge OCV, is determined based on the optimal SOC-OCV characteristic curve SO2. That is, ΔSOC, which is the SOC increased by plug-in charging, is determined. Then, the estimated full charge capacity value FE2 can be determined using the following equation (1).
Estimated full charge capacity FE2 = integrated current value during plug-in charging ÷ ΔSOC x 100...Equation (1)

In S70, the CPU 15 arbitrates between the estimated full charge capacity value FE1 obtained in S40 and the estimated full charge capacity value FE2 obtained in S60. Reconciliation uses two estimates to produce one more accurate estimate. Mediation methods may vary. For example, a simple average or a weighted average of the two estimated values may be calculated.

(effect)
Based on the probability frequency F indicating the battery temperature history and the probability frequency G indicating the SOC history, an optimal full charge capacity curve representing the current state of deterioration of the battery can be generated (S20). Furthermore, an optimal SOC-OCV characteristic curve representing the current state of deterioration of the battery can be generated based on the temperature history, SOC history, and battery usage (S30). Therefore, it is possible to flexibly consider changes in the full charge capacity curve and the SOC-OCV characteristic curve over time depending on how the user uses the battery 11. This makes it possible to ensure sufficient accuracy in estimating the full charge capacity.

In the first embodiment, a mode has been described in which the generation of the optimum full charge capacity curve (S20) and the generation of the optimum SOC-OCV characteristic curve (S30) are performed on-board. In the second embodiment, a mode in which these calculation processes are performed in the data center 100 will be described.

The operation for estimating the full charge capacity in the second embodiment will be explained using the flowchart of FIG. 3. The memory 102 of the data center 100 stores in advance a plurality of full charge capacity curves generated under various conditions of battery temperature and SOC. Note that there may be various methods of generating the plurality of full charge capacity curves. For example, a plurality of full charge capacity curves may be generated using various parameters derived based on big data obtained from other vehicles equipped with the same battery.

In S110, the CPU 15 cumulatively calculates the probability frequency F of the battery temperature and the probability frequency G of the SOC. The calculation method is the same as the method explained in S10. Further, the CPU 15 generates time-series information on the voltage, current, temperature, and SOC of the battery 11. In S115, the CPU 15 transmits the various information generated in S110 to the data center 100 via the data communication device 26. In S125, the CPU 15 waits for the calculation processing in the data center 100 to be completed.

The CPU 101 of the data center 100 generates an optimal full charge capacity curve close to the current state of deterioration based on the received battery temperature probability frequency F and SOC probability frequency G. The generated optimal full charge capacity curve is then transmitted to the electric vehicle 10. Note that the generation method may be various as described in S20.

Furthermore, the CPU 101 generates an optimal SOC-OCV characteristic curve that approximates the current state of deterioration based on the received time-series information. The generated optimal SOC-OCV characteristic curve is then transmitted to the electric vehicle 10. The generation method may vary. An example of the generation method will be specifically explained using graph 7. The horizontal axis of graph 7 is SOC, and the vertical axis is battery voltage. The CPU 101 generates a graph 7 by shotgun plotting the received voltage and SOC time series information. Then, by image pattern recognition of the graph 7, an optimal full charge capacity curve is generated. Note that conditions for drawing shotgun plots may be set. For example, the range of [20° C.≦battery temperature≦30° C.] and [−50A≦battery current≦50A] may be plotted with respect to battery temperature and current.

In S135, the CPU 15 determines whether the optimal full charge capacity curve and the optimal SOC-OCV characteristic curve have been received from the data center 100 via the data communication device 26. If it has not been received (S135: NO), the process returns to S125 and waits. On the other hand, if received (S135: YES), the process advances to S140. The contents of each step from S140 to S170 are the same as each step from S40 to S70 of the first embodiment, so the description thereof will be omitted.

(effect)
Arithmetic processing for generating the optimal full charge capacity curve and the optimal SOC-OCV characteristic curve can be performed in the data center 100, which has a higher processing capacity than onboard the electric vehicle 10. Since high-load calculations can be performed, the optimal full charge capacity curve and the optimal SOC-OCV characteristic curve can be followed with higher precision. Furthermore, since various parameters derived based on big data obtained from other vehicles equipped with the same battery can be used, the number of samples can be dramatically increased. It becomes possible to further improve the estimation accuracy of the full charge capacity.

(Modified example)
The method of calculating the optimal SOC-OCV characteristic curve (S30) is not limited to the method based on the probability frequency F of the battery temperature, the probability frequency G of the SOC, and the elapsed time T. For example, the full charge capacity of the battery 11 may be estimated by a method other than the estimation method in S60 (eg, the estimation method in S40). Then, a curve whose deterioration condition is close to the estimated full charge capacity may be selected from among the SOC-OCV characteristic curves SO1 to SO3 (see graph 4). Note that the SOC-OCV characteristic curves SO1 to SO3 at this time are curves created for each full charge capacity. For example, the SOC-OCV characteristic curve is when SO1 is 100Ah, SO2 is 80Ah, and SO3 is 60Ah.

1: Full charge capacity estimation system 10: Electric vehicle 11: Battery 14: Controller 15: CPU 16: Memory 100: Data center

Claims (1)

A full charge capacity estimation method for estimating the full charge capacity of a battery, the method comprising:
generating a plurality of full charge capacity curves indicating changes in full charge capacity with respect to elapsed time of the battery according to each of the plurality of stages of deterioration state of the battery;
generating a plurality of SOC-OCV characteristic curves indicating the relationship between the charging rate and the open-circuit voltage of the battery according to each of the plurality of stages of deterioration states of the battery;
Generating an optimal full charge capacity curve close to the current state of deterioration using the plurality of full charge capacity curves based on the temperature history and SOC history of the battery;
Generating an optimal SOC-OCV characteristic curve close to a current state of deterioration using the plurality of SOC-OCV characteristic curves based on the temperature history, the SOC history, and the elapsed time of the battery;
estimating the full charge capacity of the battery based on the optimum full charge capacity curve and the optimum SOC-OCV characteristic curve;
A method for estimating full charge capacity.

Images