Battery analysis system, battery analysis method, battery analysis program, and recording medium

Abstract

An SOH calculating unit calculates an SOH on the basis of: the voltage, included in battery data, of a secondary battery in a first resting state; the voltage of the secondary battery in a second resting state; a state of charge (SOC) difference based on an SOC - open circuit voltage (OCV) curve of the secondary battery; the current full charge capacity of the secondary battery obtained on the basis of an integrated current value for a period between the first resting state and the second resting state; and an initial full charge capacity of the secondary battery. A regression curve generating unit performs curve regression using calculated sample data of a plurality of SOHs to generate a deterioration regression curve of the secondary battery. The regression curve generating unit assigns a weight in a range of 0 to 1 to each SOH.

Description

Battery analysis system, battery analysis method, battery analysis program, and recording medium

[0001] The present disclosure relates to a battery analysis system, a battery analysis method, a battery analysis program, and a recording medium for analyzing a battery state.

[0002] In recent years, remote battery monitoring services have been put into practical use, which aggregate battery data (such as voltage, current, and temperature) from battery packs installed in electric vehicles (EVs), power-assisted bicycles, and PCs on a cloud server. The cloud server that collects the battery data estimates the battery status (e.g., SOH (State of Health) estimation) and detects signs of abnormalities (e.g., signs of micro-short circuits and cell failure) based on the battery usage history, and provides the analysis results as feedback to the customer. In addition, battery analysis companies that provide remote battery monitoring services provide consultations regarding safe operation, improved convenience, and extended lifespan of battery packs based on the analysis results of the battery packs.

[0003] To determine the SOH, which is defined as the ratio of the current full charge capacity (FCC) to the initial full charge capacity of a battery, it is necessary to determine the current FCC. The current FCC can be determined by measuring the discharge capacity from a fully charged state to a fully discharged state, but since few users use batteries until they are fully discharged, it is practically difficult to determine the FCC using this method.

[0004] Therefore, the FCC is often calculated using a two-point OCV (Open Circuit Voltage) method (see, for example, Patent Document 1). In the two-point OCV method, the current FCC is estimated based on ΔSOC between the SOC (State of Charge) corresponding to a first OCV and the SOC corresponding to a second OCV, and the integrated current value for the period between the detection timing of the first OCV and the detection timing of the second OCV. In other words, the two-point OCV method is a method for estimating the integrated current value (FCC) corresponding to a 100% change in SOC from the integrated current value corresponding to ΔSOC.

[0005] Japanese Patent Application Laid-Open No. 2008-241358

[0006] In the two-point OCV method, when ΔSOC is small, the calculation error of SOH is likely to be large. When ΔSOC is small, the measurement error of the current sensor or voltage sensor has a large effect on the SOH calculation. Furthermore, when ΔSOC is small, the OCVs at the two points are likely to fall within the flat region of the SOC-OCV curve, and when the OCVs at the two points fall within the flat region, errors are likely to occur in the conversion from the OCVs at the two points to ΔSOC.

[0007] In order to accurately calculate ΔSOC, the basic OCV value needs to be accurate, but the relaxation to the OCV after charging or discharging depends on the rest time, temperature, SOC, deterioration state, etc., and depending on these conditions, the accuracy of obtaining the OCV may be low.

[0008] From the perspective of obtaining a highly accurate current integration value, it is desirable to set the two points in the two-point OCV method as the start and end points of constant current charging. For commercial vehicles, charging once a day is common, but for electrically assisted bicycles and consumer devices with short usage times, many users charge once every three to four days or once a week. When regressing a deterioration curve from multiple SOH sample data, it is desirable to increase the number of SOH calculation opportunities, but for devices that are charged infrequently, the number of SOH samples that can be calculated from the charging interval is reduced.

[0009] Therefore, it is possible to calculate the SOH from the discharge interval, but in the case of consumer devices with short usage times, the discharge interval is often divided into narrow SOC intervals. In this case, ΔSOC in each discharge interval becomes small, and as described above, the SOH calculation error is likely to become large.

[0010] The present disclosure has been made in consideration of these circumstances, and its purpose is to provide a technique for improving the accuracy of analysis related to SOH.

[0011] In order to solve the above problem, a battery analysis system according to one aspect of the present disclosure includes: a data acquisition unit that acquires battery data including a voltage and a current of a secondary battery; an SOH calculation unit that calculates an SOH based on the voltage of the secondary battery in a first resting state and the voltage of the secondary battery in a second resting state included in the battery data, the SOC difference based on an SOC-OCV curve of the secondary battery, a current full charge capacity of the secondary battery calculated based on an integrated current value for a period between the first resting state and the second resting state, and an initial full charge capacity of the secondary battery; and a regression curve generation unit that performs curve regression using sample data of the calculated SOHs to generate a degradation regression curve of the secondary battery. The regression curve generation unit assigns a weight to each SOH in a range of 0 to 1.

[0012] Any combination of the above components, and conversion of the expression of the present disclosure between devices, systems, methods, programs, recording media, etc. are also valid aspects of the present disclosure.

[0013] According to the present disclosure, the accuracy of analysis regarding SOH can be improved.

[0014] 7 is a diagram for explaining a battery analysis system according to an embodiment. FIG. 8 is a diagram illustrating an example of the configuration of a battery pack. FIG. 9 is a diagram illustrating an example of the configuration of a battery analysis system according to an embodiment. FIG. 10 is a diagram illustrating a specific image of an FCC estimation method. FIG. 11 is a diagram illustrating a graph of a deterioration curve of a secondary battery. FIG. 12 is a diagram illustrating an example of a state transition of a battery pack over a predetermined period. FIG. 13 is a diagram illustrating an example of a state transition of a battery pack over one charge / discharge cycle. FIG. 14 is a diagram illustrating a graph summarizing SOH reliability evaluations calculated from battery data for three discharge intervals set in the state transition period of the battery pack shown in FIG. 7. FIG. 15 is a diagram illustrating a graph summarizing SOH reliability evaluations calculated from battery data for six discharge intervals set in the state transition period of the battery pack shown in FIG. 7. FIG. 16 is a flowchart illustrating a flow of a battery analysis process using SOH by a battery analysis system according to an embodiment.

[0015] FIG. 1 is a diagram illustrating a battery analysis system 10 according to an embodiment. The battery analysis system 10 according to the embodiment is a system for analyzing a battery pack 30 mounted on a battery-equipped device (assumed to be an electric bicycle 1 in this embodiment). The battery analysis system 10 may be constructed, for example, on an in-house server installed in the in-house facility or data center of a business providing an analysis service for the battery pack 30. The battery analysis system 10 may also be constructed on a cloud server used based on a cloud service. The battery analysis system 10 may also be constructed on multiple servers distributed across multiple locations (data centers, in-house facilities). The multiple servers may be a combination of multiple in-house servers, a combination of multiple cloud servers, or a combination of an in-house server and a cloud server.

[0016] The electric bicycle 1 is equipped with a detachable, portable, and replaceable battery pack 30. The battery pack 30 is charged while attached to a charging slot of a charger (not shown). The charged battery pack 30 is removed from the charging slot by the user and attached to the attachment slot of the electric bicycle 1.

[0017] The battery pack 30 and the mobile communication terminal 20 (hereinafter, assumed to be a smartphone) carried by the user are connected via short-range wireless communication. Bluetooth (registered trademark), Wi-Fi (registered trademark), infrared communication, etc. can be used as short-range wireless communication. Hereinafter, in this embodiment, it is assumed that BLE (Bluetooth Low Energy) is used as short-range wireless communication. BLE is an extended standard of Bluetooth and is a low-power short-range wireless communication standard using the 2.4 GHz band.

[0018] The mobile communication terminal 20 can access the network 2 to which the battery analysis system 10 is connected. The mobile communication terminal 20 can access the network 2 via a mobile phone network (4G / 5G) or a wireless LAN.

[0019] The network 2 is a general term for communication paths such as the Internet, a dedicated line, and a virtual private network (VPN), and the communication medium and protocol are not important. Examples of communication media that can be used include a wired LAN, a wireless LAN, a mobile phone network, an optical fiber network, an ADSL network, and a CATV network. Examples of communication protocols that can be used include TCP (Transmission Control Protocol) / IP (Internet Protocol), UDP (User Datagram Protocol) / IP, and Ethernet (registered trademark).

[0020] 2 is a diagram showing an example of the configuration of the battery pack 30. The battery pack 30 includes a battery pack 31 and a battery management device 32. The battery pack 31 includes multiple cells E1-En connected in series. The number of cells connected in series is determined by the load specifications. The main loads of the electric bicycle 1 are the motor and inverter.

[0021] The cells may be lithium-ion battery cells, nickel-metal hydride battery cells, lead-acid battery cells, etc. In the following description, we will assume an example in which lithium-ion battery cells (nominal voltage: 3.6-3.7V) are used. Note that in each series stage of cells, multiple cells may be connected in parallel to increase capacity.

[0022] A switch SW1 for switching between electrical continuity with the load or charger is inserted in the power line connecting the battery pack 31 and the load or charger. A semiconductor switch or a relay can be used as the switch SW1.

[0023] The battery management device 32 includes a measurement unit 33, a control unit 34, a wireless communication unit 35, and an antenna 35a. The measurement unit 33 is configured as an AFE (Analog Front End) IC or an ASIC (Application Specific Integrated Circuit). The control unit 34 is configured as a microcontroller.

[0024] The measurement unit 33 is connected to each node of the multiple cells E1-En connected in series by multiple voltage measurement lines, and measures the voltage of each cell E1-En by measuring the voltage between each two adjacent voltage measurement lines.

[0025] The measurement unit 33 includes a multiplexer and an A / D converter. The multiplexer outputs the voltages of the multiple cells E1-En to the A / D converter in a predetermined order. The A / D converter converts the analog voltages input from the multiplexer into digital values. The measurement unit 33 transmits the converted digital voltage values ​​of the cells E1-En to the control unit 34 via the serial communication interface.

[0026] The measurement unit 33 measures the current flowing through the battery pack 31. A shunt resistor Rs is connected to a power line connecting the battery pack 31 to a load or a charger. A differential amplifier (not shown) amplifies the voltage across the shunt resistor Rs and outputs it to an A / D converter in the measurement unit 33. The A / D converter converts the analog voltage indicating the current flowing through the battery pack 31, which is input from the differential amplifier, into a digital value. The measurement unit 33 transmits the current value converted into a digital value to the control unit 34 via a serial communication interface.

[0027] A temperature sensor T1 (e.g., a thermistor) is installed on the surface of the battery pack 31. A divided voltage between the temperature sensor T1 and a voltage dividing resistor (not shown) is input to a measurement unit 33. An A / D converter in the measurement unit 33 converts the input analog voltage indicating the temperature into a digital value. The measurement unit 33 transmits the converted digital temperature value to the control unit 34 via a serial communication interface.

[0028] The control unit 34 manages the states of the cells E1-En based on the voltage values ​​of the cells E1-En, the current values ​​flowing through the battery pack 31, and the temperature values ​​of the battery pack 31 received from the measurement unit 33. When the control unit 34 detects overcharge, overdischarge, overcurrent, abnormally high temperature, or abnormally low temperature, it sends a shutoff signal for the switch SW1 to the measurement unit 33 to turn off the switch SW1.

[0029] The control unit 34 estimates the SOC by combining the OCV method and the current integration method. The OCV method is a method for estimating the SOC based on the OCV, which is based on the measured cell voltage, and the SOC-OCV curve of the cell. The SOC-OCV curve of the cell is created in advance by the battery manufacturer based on characteristic tests and is registered in the control unit 34 at the time of shipment.

[0030] The current integration method is a method for estimating the SOC based on the OCV at the start of cell charging and discharging and the integrated value of the measured current. With the current integration method, current measurement errors accumulate as the charging and discharging time increases. Therefore, it is preferable to use a weighted average of the SOC estimated by the current integration method and the SOC estimated by the OCV method.

[0031] The control unit 34 transmits battery data including the voltage, current, temperature, and SOC of each cell of the battery pack 30 to the wireless communication unit 35. The wireless communication unit 35 performs signal processing for short-range wireless communication. The wireless communication unit 35 pairs with the mobile communication terminal 20 and transmits the battery data received from the control unit 34 to the mobile communication terminal 20 via short-range wireless communication. The wireless communication unit 35 transmits the battery data to the mobile communication terminal 20 at a predetermined transmission cycle (for example, every one minute).

[0032] The portable communication terminal 20 accesses the network 2 and transfers the battery data received from the battery pack 30 to the battery analysis system 10 in real time. Alternatively, the portable communication terminal 20 temporarily stores the battery data received from the battery pack 30 in an internal buffer and transmits the stored battery data to the battery analysis system 10 in batches.

[0033] In addition, if the wireless communication unit 35 of the battery pack 30 is equipped with a module for connecting to a mobile phone network (4G / 5G), the wireless communication unit 35 can directly access the network 2 without going through the mobile communication terminal 20 and transmit battery data to the battery analysis system 10.

[0034] 3 is a diagram showing an example of the configuration of a battery analysis system 10 according to an embodiment. The battery analysis system 10 includes a control unit 11, a storage unit 12, and a communication unit 13. The communication unit 13 is an external communication interface (e.g., a network interface card (NIC)) for connecting to the network 2 via a wired or wireless connection.

[0035] The control unit 11 includes a data acquisition unit 111, an SOH calculation interval setting unit 112, an SOH calculation unit 113, an SOH calculation value evaluation unit 114, a regression curve generation unit 115, and a life prediction unit 116. The functions of the control unit 11 can be realized by a combination of hardware and software resources, or by hardware resources alone. Examples of hardware resources that can be used include a CPU, ROM, RAM, GPU, ASIC, FPGA, and other LSIs. Examples of software resources that can be used include programs such as an operating system and applications. The programs may also be recorded on a recording medium. Using this recording medium, the programs can be installed on the computer, for example. The recording medium on which the programs are recorded may be a non-transitory recording medium. The non-transitory recording medium is not particularly limited, and may be, for example, a CD-ROM or other recording medium.

[0036] The storage unit 12 includes a non-volatile recording medium such as an HDD or SSD, and stores various data. The storage unit 12 includes a battery data holding unit 121. The data acquisition unit 111 acquires battery data of the battery pack 30 via the network 2. The data acquisition unit 111 stores the acquired battery data in the battery data holding unit 121.

[0037] The SOH calculation unit 113 uses a two-point OCV method to calculate the SOH of each cell or the entire battery pack 31. When calculating the SOH of the entire battery pack 31, the SOH calculation unit 113 calculates the voltage of the battery pack 31 by adding up the cell voltages of the multiple series-connected cells E1-En included in the battery data.

[0038] The SOH calculation unit 113 calculates the SOC difference (ΔSOC) between the SOC in the first rest state and the SOC in the second rest state based on the voltage of the cell or battery pack 31 in the first rest state, the voltage of the cell or battery pack 31 in the second rest state, and the SOC-OCV curve of the cell or battery pack 31. The SOH calculation unit 113 calculates the current integrated value Q for the period between the first rest state and the second rest state based on the current value included in the battery data. The SOH calculation unit 113 calculates the current FCC of the cell or battery pack 31 based on the current integrated value Q and ΔSOC. The SOH calculation unit 113 calculates the SOH based on the current FCC and initial FCC of the cell or battery pack 31.

[0039] FIG. 4 is a diagram showing a specific image of the FCC estimation method. The SOH calculation unit 113 identifies two voltage points, the first rest state and the second rest state, and sets these as the OCV points. The SOH calculation unit 113 references the SOC-OCV curve to identify two SOC points corresponding to the two OCV points, and calculates a ΔSOC between the two SOC points. In the example shown in FIG. 4, the SOC points are 20% and 75%, and the ΔSOC is 55%.

[0040] The SOH calculation unit 113 calculates the integrated current amount (=charge / discharge capacity) Q between the two times when the OCVs at the two points were acquired. The SOH calculation unit 113 calculates the following (Equation 1) to estimate the FCC.

[0041] FCC=Q / ΔSOC (Equation 1) SOH is defined as the ratio of the current FCC to the initial FCC, and the lower the value (closer to 0%), the more advanced the deterioration. The SOH calculation unit 113 calculates the following (Equation 2) to estimate the SOH.

[0042] SOH=current FCC / initial FCC×100 (Equation 2) The regression curve generator 115 performs curve regression using a plurality of sample data of SOH calculated in time series for the cell or assembled battery 31, to generate a deterioration regression curve for the cell or assembled battery 31. For example, the least squares method can be used for the curve regression.

[0043] 5 is a graph showing the deterioration curve of a secondary battery. It is known that the deterioration of a secondary battery progresses in proportion to the square root of time (0.5 power law), as shown in the following (Equation 3).

[0044] SOH=w0+w1√t (Equation 3) where w0 is the initial value and w1 is the deterioration coefficient.

[0045] The regression curve generator 115 calculates the deterioration coefficient w1 in the above equation (3) by exponential curve regression of 0.5, with time t as the independent variable and SOH as the dependent variable. w0 is a common parameter and is usually set in the range of 1.0 to 1.1. If the actual initial capacity matches the nominal value, w0 is set to 1.0. If the nominal value is set to the minimum guaranteed amount and is set lower than the actual initial capacity, a value greater than 1.0 is set.

[0046] The life prediction unit 116 can predict the period (remaining life) until the SOH at which the battery pack 30 should be terminated is reached (70% in the example of FIG. 5 ) by substituting the SOH at which the battery pack 30 should be terminated into the generated degradation regression curve. Note that it is also possible to predict the SOH value at any future date and time by substituting any future date and time into the degradation regression curve.

[0047] FIG. 6 is a diagram showing an example of the state transition of the battery pack 30 over a predetermined period. The state transition of the battery pack 30 is classified into a charging section, a discharging section, and a resting section. In the charging section, the voltage and SOC of the battery pack 30 increase, in the discharging section, the voltage and SOC decrease, and in principle, the voltage and SOC are maintained constant in the resting section. The state of the battery pack 30 can be determined by whether the current value included in the battery data is "positive," "negative," or "0."

[0048] Empirical data shows that many users of the battery pack 30 mounted on the electric bicycle 1 use approximately 10-15% of the DOD capacity per day and charge the battery all at once on the weekend. In such cases, discharge intervals occur in a stepped pattern, and the length of each discharge interval is short. Charging intervals occur once a week, and the length of each charging interval is long. Similar trends are observed in other consumer devices, and basically, discharging intervals tend to occur frequently and be short, while charging intervals tend to occur infrequently and be long.

[0049] FIG. 7 is a diagram showing an example of the state transition of the battery pack 30 during one charge / discharge cycle. While it is generally desirable to calculate the SOH based on battery data from the charging interval where the current value is stable, opportunities to calculate the SOH are reduced. To generate a deterioration regression curve for a secondary battery as described above, typically, at least 25 to 40 SOH sample data points are required. When calculating the SOH from only the charging interval, it takes time from the start of operation of the battery pack 30 until the initial generation of the deterioration regression curve. Furthermore, it takes time for the deterioration phenomena that actually occur in the secondary battery during operation of the battery pack 30 to be reflected in the deterioration regression curve.

[0050] Therefore, it is possible to calculate the SOH from battery data during the discharging section. However, the current value during the discharging section changes irregularly, unlike the current value during the charging section, which is based on CC-CV charging. Furthermore, because battery data is discrete data, the current value at a given sampling timing is generally treated as continuing at the same value until the next sampling timing. If the current value at a sampling timing happens to be an outlier, the error in the integrated current value will increase. The longer the sampling period, the larger the error in the integrated current value during discharging.

[0051] As such, the accuracy of SOH calculation based on battery data in the discharging section tends to be lower than the accuracy of SOH calculation based on battery data in the charging section, and attention must be paid to the accuracy of SOH calculation based on battery data in the discharging section.

[0052] Generally, factors that have a large effect on the accuracy of SOH calculation include (1) ΔSOC, (2) the length of the downtime, (3) the temperature during downtime, and (4) the slope of the SOC-OCV curve.

[0053] Generally, the larger the ΔSOC, the higher the accuracy of SOH calculation. When ΔSOC is small, the numerator and denominator values ​​of the above (Equation 1) become smaller. The weight of the current value and voltage value of one sample increases, and the impact of measurement errors of the current sensor or voltage sensor on the numerator or denominator value of the above (Equation 1) becomes greater. In reality, it is difficult to reduce the measurement error of the current sensor or voltage sensor to zero. Furthermore, when ΔSOC is small, the OCVs at the two points are more likely to fall within the flat region of the SOC-OCV curve. If the OCVs at the two points fall within the flat region, errors are more likely to occur in the conversion from the OCVs at the two points to ΔSOC.

[0054] Conversely, when ΔSOC is large, variations in current values ​​due to measurement errors of the current sensor, etc., are smoothed by long-term integration. In this way, it can be said that the reliability of the SOH calculated from battery data in an interval where ΔSOC is small is low, and the reliability of the SOH calculated from battery data in an interval where ΔSOC is large is high.

[0055] The voltage values ​​included in the battery data are measured values ​​of the voltage across each cell E1-En. A secondary battery is an electrochemical product, and the measured voltage increases nonlinearly when a charging current flows through the secondary battery, and decreases nonlinearly when a discharging current flows through the secondary battery. The voltage measured when a current flows through the secondary battery is called the closed circuit voltage (CCV) or operating voltage. After charging and discharging are completed, the secondary battery gradually converges (relaxes) to an OCV that does not include an overvoltage component over time.

[0056] The time it takes to reach the OCV depends on factors such as the type of cell, temperature, and SOH. In recent years, lithium-ion battery cells have been developed that use anode materials containing silicon. These lithium-ion battery cells take longer to resolve polarization than conventional cells, requiring 1 to 10 hours or more for the cell to reach the OCV after charging and discharging. Furthermore, in low-temperature environments, the rate of chemical reactions within the cell slows, lengthening the time it takes to reach the OCV. Degraded cells with a reduced SOH also take longer to reach the OCV.

[0057] From the above findings, it is highly likely that the voltage measured when the resting time is long matches the OCV, while the voltage measured when the resting time is short is likely to deviate from the OCV. Furthermore, among the voltages measured when the temperature is high, it is highly likely that the voltage measured when the temperature is low is likely to deviate from the OCV. Furthermore, among the voltages measured when the resting time is long, it is highly likely that the voltage of a cell with a high SOH matches the OCV, while the voltage of a cell with a low SOH is likely to deviate from the OCV.

[0058] The SOC-OCV curve of a cell depends on the cell type. Some cells have a flat region of the SOC-OCV curve. The flat region of the SOC-OCV curve is a region where the SOC changes very little in response to changes in OCV, and errors are likely to occur when converting from two OCV points to ΔSOC, which corresponds to the amount of SOC change. Even a slight error in the measured voltage can lead to a major misreading of the SOC.

[0059] Fig. 8 is a graph showing a summary of SOH reliability evaluations calculated from battery data for three discharge intervals set during the state transition period of battery pack 30 shown in Fig. 7. SOH calculation interval setting unit 112 sets a first discharge interval between first pause interval (1) and second pause interval (2), a second discharge interval between second pause interval (2) and third pause interval (3), and a third discharge interval between third pause interval (3) and fourth pause interval (4).

[0060] The SOH calculation unit 113 calculates the SOH based on the battery data for each discharging interval. The SOH calculation value evaluation unit 114 evaluates the reliability of each SOH. In the example shown in Fig. 8, five SOH evaluation items are evaluated: (1) ΔSOC, (2) the length of the pause time at the start point, (3) the length of the pause time at the end point, (4) the temperature at the start of discharge, and (5) the temperature at the end of discharge.

[0061] The SOH calculated value evaluation unit 114 increases the evaluation score as the ΔSOC value increases. In the example shown in FIG. 8 , the ΔSOC value is classified into three categories: "small," "medium," and "large." The SOH calculated value evaluation unit 114 evaluates the evaluation score as "low" when the ΔSOC value is "small," "medium" when the ΔSOC value is "medium," and "high" when the ΔSOC value is "large." Note that the SOH calculated value evaluation unit 114 does not have to use three levels for evaluation, and may use two levels or four or more levels for evaluation. Note that the normalized numerical value itself may be used for the overall evaluation described below without classifying the evaluation score.

[0062] The longer the pause time at the start point, the higher the evaluation score is assigned by the SOH calculation value evaluation unit 114. In the example shown in Fig. 8, the length of the pause time at the start point is classified into three categories: "short," "medium," and "long," and the SOH calculation value evaluation unit 114 assigns the evaluation score as "low" when the pause time at the start point is "short," "medium" when it is "medium," and "high" when it is "long."

[0063] The longer the pause time at the end point, the higher the evaluation score is assigned by the SOH calculation value evaluation unit 114. In the example shown in Fig. 8, the length of the pause time at the end point is classified into three categories: "short," "medium," and "long," and the SOH calculation value evaluation unit 114 evaluates the evaluation score as "low" when the pause time at the end point is "short," "medium" when it is "medium," and "high" when it is "long."

[0064] The higher the temperature at the start of discharge, the higher the evaluation score is assigned by SOH calculation value evaluation unit 114. In the example shown in Fig. 8, the temperature at the start of discharge is classified into three categories: "low," "medium," and "high," and SOH calculation value evaluation unit 114 assigns the evaluation score as "low" when the temperature at the start of discharge is "low," "medium" when it is "medium," and "high" when it is "high."

[0065] The higher the temperature at the end of discharge, the higher the evaluation score is assigned by the SOH calculation value evaluation unit 114. In the example shown in Fig. 8, the temperature at the end of discharge is classified into three categories: "low," "medium," and "high," and the SOH calculation value evaluation unit 114 evaluates the evaluation score as "low" when the temperature at the end of discharge is "low," "medium" when it is "medium," and "high" when it is "high."

[0066] The SOH calculation value evaluation unit 114 sums up the evaluation scores of all five items to calculate a total evaluation score. In the example shown in Fig. 8, the SOH calculation value evaluation unit 114 classifies the total evaluation score into three categories: "x", "△", and "circle".

[0067] The regression curve generating unit 115 can assign a weight to each SOH sample data in the range of 0 to 1. Specifically, the regression curve generating unit 115 assigns a larger weight to the SOH sample data with a higher overall evaluation score. For example, the regression curve generating unit 115 may set the weight for an evaluation score of "circle" to 1.0, the weight for "triangle" to 0.5, and the weight for "x" to 0.2. Note that the weight for "x" may also be set to 0.0. In this case, samples with an evaluation score of "x" are excluded from the basic data for generating the regression curve.

[0068] The regression curve generator 115 weights each of the n SOH sample data, and then performs weighted curve regression using the n SOH sample data. The following (Equation 4) is a normal regression, and the regression curve generator 115 derives the deterioration coefficient w1 of the above (Equation 3) so that the sum of squares E of the residuals εi of each sample data is minimized.

[0069] E = Σεi2 (i = 1 to n) ... (Equation 4) The following (Equation 5) is a weighted regression, and the regression curve generation unit 115 derives the deterioration coefficient w1 of the above (Equation 3) so that the sum E of the values ​​obtained by multiplying the square of the residual εi of each sample data by the weight wei is minimized.

[0070] E=Σweiεi2 (i=1 to n) (Equation 5) Figure 9 is a graph summarizing the SOH reliability evaluations calculated from the battery data for six discharge intervals set during the state transition period of battery pack 30 shown in Figure 7. The discharge intervals shown in Figure 8 are extracted only from intervals consisting of a single continuous discharge interval with no intervening pauses. In contrast, the discharge intervals shown in Figure 9 are extracted from intervals consisting of multiple discharge intervals and at least one intervening pause, allowing for at least one intervening pause.

[0071] In addition to the three discharge intervals shown in FIG. 8, the SOH calculation interval setting unit 112 sets a fourth discharge interval between the first pause interval (1) and the third pause interval (3), a fifth discharge interval between the first pause interval (1) and the fourth pause interval (4), and a sixth discharge interval between the second pause interval (2) and the fourth pause interval (4).

[0072] 8 and 9, the ΔSOC values ​​in the first, second, and third discharge intervals are all classified as "small," while the ΔSOC values ​​in the fourth and sixth discharge intervals are classified as "medium," and the ΔSOC value in the fifth discharge interval is classified as "large." As described above, the SOH calculation value evaluation unit 114 increases the evaluation score as the ΔSOC value increases. Therefore, the larger the ΔSOC value, the higher the overall evaluation score tends to be.

[0073] Furthermore, since the internal resistance of the battery increases as the SOC decreases, and the temperature of the battery pack 30 increases, the temperatures at the end of discharge in the third, fifth, and sixth discharge intervals are classified as "high." As described above, the SOH calculation value evaluation unit 114 increases the evaluation score the higher the temperature at the end of discharge. Therefore, the higher the temperature at the end of discharge, the higher the overall evaluation score tends to be.

[0074] From the above perspective, in the example shown in FIG. 9 , the SOH calculation value evaluation unit 114 evaluates the overall evaluation score for the fourth and fifth discharge intervals as "round." Note that the five SOH evaluation items shown in FIGS. 8 and 9 are merely examples, and evaluation items such as the used SOC range and the current SOH level may be added. The SOH calculation value evaluation unit 114 increases the evaluation score the steeper the slope of the used SOC range in the SOC-OCV curve. The SOH calculation value evaluation unit 114 increases the evaluation score the higher the current SOH level. This prevents SOH sample data from being unnecessarily excluded from the basic data for generating the regression curve, making it easier to perform weighted curve regression.

[0075] For simplification, the number of evaluation items for SOH may be reduced. For example, if temperature data is not acquired, (4) the temperature at the start of discharge and (5) the temperature at the end of discharge are excluded from the evaluation items.

[0076] 10 is a flowchart showing the flow of battery analysis processing using SOH by the battery analysis system 10 according to the embodiment. The SOH calculation unit 113 reads time-series battery data of the cells or assembled batteries 31 included in the battery pack 30 to be analyzed, which is stored in the battery data storage unit 121 (S10). The SOH calculation interval setting unit 112 sets an SOH calculation interval that matches a set condition for the time-series battery data (S11).

[0077] The setting conditions can be selected from only the charging section, only the discharging section, or both the charging section and the discharging section. The discharging section can be selected from "only the discharging section without any intervening rest periods" and "a discharging section consisting of multiple discharging periods with intervening rest periods." Note that, to prevent an increase in the number of discharging section combinations, an upper limit can be set for the number of intervening rest periods or the total duration of the discharging section. If an upper limit is set for the total duration of the discharging section, SOH calculation based on battery data for SOH calculation periods longer than the upper limit is skipped. This reduces integration errors associated with longer SOH calculation periods, further improving the accuracy of SOH analysis.

[0078] Similarly, the charging section may be configured to allow selection between "only charging sections with no intervening rest periods" and "charging sections consisting of multiple charging periods with intervening rest periods."

[0079] Furthermore, when both a charging section and a discharging section are selected as SOH calculation sections, a charging / discharging section may also be selected as an SOH calculation section. A charging / discharging section is composed of at least one charging period, at least one discharging period, and at least one intervening rest period. Note that, in order to prevent an increase in the number of combinations of charging / discharging sections, an upper limit may be set on the number of intervening rest periods or on the entire duration of the charging / discharging section.

[0080] The SOH calculation unit 113 calculates the SOH based on the battery data for the set calculation interval (S12). The SOH calculation value evaluation unit 114 evaluates the reliability of the calculated SOH (S13). The SOH calculation unit 113 weights the calculated SOH according to the reliability evaluated by the SOH calculation value evaluation unit 114 (S14). Note that if the weight of the SOH calculated from the battery data for the SOH calculation interval including the discharge period is set to 0, only the SOH calculated from the battery data for the SOH calculation interval not including the discharge period can be left as basic data.

[0081] If SOH calculation has not been completed for all SOH calculation intervals of all patterns that meet the set conditions (N in S15), the processes of steps S11 to S14 are repeated. If SOH calculation has been completed for all SOH calculation intervals of all patterns that meet the set conditions (Y in S15), the regression curve generator 115 performs curve regression using the sample data of the calculated SOHs to generate a degradation regression curve for the cell or battery pack 31 (S16). The life prediction unit 116 predicts the remaining life of the battery pack 30 based on the generated degradation regression curve (S17). Note that the remaining life prediction may be periodically updated by repeating the battery analysis process described above, for example, once a month.

[0082] As described above, according to this embodiment, the accuracy of analysis of the SOH can be improved. By including the SOH based on battery data from the discharging section in the sample data, the number of sample data can be increased, thereby improving the accuracy of the deterioration regression curve using the SOH. Furthermore, the time required to collect the required number of sample data can be shortened, thereby accelerating the timing of generating the deterioration regression curve. Furthermore, by including the SOH based on battery data from the discharging section, charging section, and charge / discharging section, which are separated by rest periods, in the sample data, the number of sample data can be further increased. Furthermore, by evaluating the reliability of the calculated SOH and weighting the SOH, it is possible to prevent a decrease in the accuracy of the deterioration regression curve due to the inclusion of low-quality sample data.

[0083] The present disclosure has been described above based on the embodiments. The embodiments are merely examples, and it will be understood by those skilled in the art that various modifications are possible in the combination of the respective components and processing steps, and that such modifications are also within the scope of the present disclosure.

[0084] The functions of the battery analysis system 10 described above may be implemented in the battery management device 32 of the battery pack 30. When implemented on the battery pack 30 side, the accuracy of the current integrated value of the discharge current or the charge current is improved, and the reliability of the SOH is accordingly increased. Conversely, when implemented on the cloud side, the accuracy of the current integrated value is relatively low, and the reliability evaluation of the SOH described above becomes more important.

[0085] In the above formula (3), a model is used in which the deterioration of the secondary battery progresses in proportion to the square root of time (0.5 power law). In this regard, a model in which the deterioration of the secondary battery progresses in proportion to the square root of the throughput [Ah] may also be used.

[0086] The embodiment may be specified by the following items.

[0087] [Item 1] A battery analysis system (10) comprising: a data acquisition unit (111) that acquires battery data including a voltage and a current of a secondary battery (E1-En); an SOH calculation unit (113) that calculates an SOH based on the voltage of the secondary battery (E1-En) in a first resting state, the voltage of the secondary battery (E1-En) in a second resting state, an SOC difference based on an SOC-OCV curve of the secondary battery (E1-En), a current full charge capacity of the secondary battery (E1-En) calculated based on an integrated current value for a period between the first resting state and the second resting state, and an initial full charge capacity of the secondary battery (E1-En); and a regression curve generation unit (115) that performs curve regression using sample data of a plurality of calculated SOHs to generate a degradation regression curve of the secondary battery (E1-En), wherein the regression curve generation unit (115) assigns a weight to each SOH in a range of 0 to 1.

[0088] This makes it possible to improve the accuracy of analysis regarding SOH.

[0089] [Item 2] The battery analysis system (10) according to Item 1, wherein the SOH calculation unit (113) calculates the SOH based on battery data for a period including a discharging section between the first rest state and the second rest state.

[0090] This allows the number of sample data of the SOH to be increased.

[0091] [Item 3] The battery analysis system (10) according to Item 1, wherein the SOH calculation unit (113) calculates the SOH based on battery data for a period between the first resting state and the second resting state, the period including at least one other resting state and a plurality of discharging intervals.

[0092] This allows the number of sample data of the SOH to be further increased.

[0093] [Item 4] The battery analysis system (10) according to Item 1, wherein, when a period between the first hibernation state and the second hibernation state is longer than a predetermined set time, the SOH calculation unit (113) does not calculate the SOH based on battery data for that period.

[0094] This makes it possible to limit the increase in the number of SOH sample data, and also to prevent the generation of SOH sample data that is significantly affected by self-discharge.

[0095] [Item 5] The battery analysis system (10) according to Item 1, wherein the regression curve generating unit (115) changes a weight of the SOH calculated by the SOH calculating unit (113) depending on the reliability of the SOH.

[0096] This makes it possible to suppress a decrease in accuracy of the degradation regression curve due to the inclusion of low-quality sample data.

[0097] [Item 6] The battery analysis system (10) according to Item 5, wherein the regression curve generating unit (115) assigns a larger weight to sample data of SOH calculated from battery data having a larger value of the SOC difference.

[0098] This can reduce the influence of low quality SOH.

[0099] [Item 7] The battery analysis system (10) according to Item 5, wherein the regression curve generation unit (115) assigns a greater weight to sample data of SOH calculated from battery data having a longer time period in the first rest state or the second rest state.

[0100] This can reduce the influence of low quality SOH.

[0101] [Item 8] The battery analysis system (10) according to Item 5, wherein the data acquisition unit acquires battery data including voltages, currents, and temperatures of the secondary batteries (E1-En), and the regression curve generation unit (115) assigns a greater weight to sample data of SOH calculated from battery data having a higher temperature in the first rest state or a higher temperature in the second rest state.

[0102] This can reduce the influence of low quality SOH.

[0103] [Item 9] The battery analysis system (10) according to Item 5, wherein the regression curve generation unit (115) assigns a greater weight to sample data of SOH calculated from battery data having a steeper slope of an SOC band including an SOC obtained by referring to the SOC-OCV curve with the voltage in the first rest state or the voltage in the second rest state as an OCV.

[0104] This can reduce the influence of low quality SOH.

[0105] [Item 10] The battery analysis system (10) according to Item 1, wherein the regression curve generating unit (115) sets a weight of sample data of SOH calculated from battery data including a discharge section to 0.

[0106] This allows the SOH to be extracted only from the battery data in the charging section.

[0107] [Item 11] A battery analysis method comprising: acquiring battery data including a voltage and a current of a secondary battery (E1-En); calculating an SOH based on the voltage of the secondary battery (E1-En) in a first resting state, the voltage of the secondary battery (E1-En) in a second resting state, an SOC difference based on an SOC-OCV curve of the secondary battery (E1-En), a current full charge capacity of the secondary battery (E1-En) calculated based on an integrated current value for a period between the first resting state and the second resting state, and an initial full charge capacity of the secondary battery (E1-En); performing curve regression using sample data of a plurality of calculated SOHs to generate a degradation regression curve of the secondary battery (E1-En); and weighting each SOH in a range of 0 to 1.

[0108] This makes it possible to improve the accuracy of analysis regarding SOH.

[0109] [Item 12] A battery analysis program that causes a computer to execute the following steps: acquiring battery data including a voltage and a current of a secondary battery (E1-En); calculating an SOH based on the voltage of the secondary battery (E1-En) in a first resting state, the voltage of the secondary battery (E1-En) in a second resting state, an SOC difference based on an SOC-OCV curve of the secondary battery (E1-En), a current full charge capacity of the secondary battery (E1-En) calculated based on an integrated current value for a period between the first resting state and the second resting state, and an initial full charge capacity of the secondary battery (E1-En); performing curve regression using sample data of a plurality of calculated SOHs to generate a degradation regression curve of the secondary battery (E1-En); and weighting each SOH in a range of 0 to 1.

[0110] This makes it possible to improve the accuracy of analysis regarding SOH.

[0111] REFERENCE SIGNS LIST 1 Electric bicycle, 2 Network, 20 Portable communication terminal, 30 Battery pack, 31 Assembled battery, 32 Battery management device, 33 Measurement unit, 34 Control unit, 35 Wireless communication unit, 35a Antenna, E10-En Cell, Rs Shunt resistor, T1 Temperature sensor, SW1 Switch, 10 Battery analysis system, 11 Control unit, 111 Data acquisition unit, 112 SOH calculation interval setting unit, 113 SOH calculation unit, 114 SOH calculation value evaluation unit, 115 Regression curve generation unit, 116 Life prediction unit, 12 Memory unit, 121 Battery data storage unit, 13 Communication unit

Claims

1. A battery analysis system comprising: a data acquisition unit that acquires battery data including a voltage and a current of a secondary battery; a SOH calculation unit that calculates a State of Health (SOH) based on a voltage of the secondary battery in a first rest state and a voltage of the secondary battery in a second rest state included in the battery data, an SOC difference based on a State Of Charge (SOC)-Open Circuit Voltage (OCV) curve of the secondary battery, a current full charge capacity of the secondary battery calculated based on an integrated current value for a period between the first rest state and the second rest state, and an initial full charge capacity of the secondary battery; and a regression curve generation unit that performs curve regression using a plurality of calculated SOH sample data to generate a deterioration regression curve of the secondary battery, wherein the regression curve generation unit weights each SOH in a range of 0 to 1.

2. The battery analysis system according to claim 1, wherein the SOH calculation unit calculates the SOH based on battery data for a period including a discharging section between the first rest state and the second rest state.

3. The battery analysis system according to claim 1, wherein the SOH calculation unit calculates the SOH based on battery data for a period between the first rest state and the second rest state, the period including at least one other rest state and a plurality of discharge intervals.

4. The battery analysis system of claim 1, wherein the SOH calculation unit does not calculate the SOH based on battery data for a period between the first and second hibernation states if the period is longer than a predetermined set time.

5. The battery analysis system according to claim 1, wherein the regression curve generating section changes a weight of the SOH calculated by the SOH calculating section in accordance with a reliability of the SOH.

6. The battery analysis system according to claim 5, wherein the regression curve generating section assigns a larger weight to SOH sample data calculated from battery data having a larger SOC difference value.

7. The battery analysis system according to claim 5, wherein the regression curve generating unit weights SOH sample data calculated from battery data having a longer time in the first rest state or a longer time in the second rest state.

8. The battery analysis system of claim 5, wherein the data acquisition unit acquires battery data including the voltage, current, and temperature of the secondary battery, and the regression curve generation unit assigns a greater weight to SOH sample data calculated from battery data having a higher temperature during the first hibernation state or a higher temperature during the second hibernation state.

9. The battery analysis system according to claim 5, wherein the regression curve generating unit is further configured to weight sample data of SOH calculated from battery data having a steeper slope of an SOC band including an SOC obtained by referencing the SOC-OCV curve with the voltage in the first rest state or the voltage in the second rest state as an OCV.

10. The battery analysis system according to claim 1, wherein the regression curve generating unit sets a weight of sample data of SOH calculated from battery data including a discharge section to zero.

11. A battery analysis method comprising the steps of: acquiring battery data including a voltage and a current of a secondary battery; calculating an SOH based on the voltage of the secondary battery in a first rest state, the voltage of the secondary battery in a second rest state, an SOC difference based on an SOC-OCV curve of the secondary battery, a current full charge capacity of the secondary battery calculated based on an integrated current value for a period between the first rest state and the second rest state, and an initial full charge capacity of the secondary battery; performing curve regression using sample data of a plurality of calculated SOH values ​​to generate a deterioration regression curve of the secondary battery; and weighting each SOH in the range of 0 to 1.

12. A battery analysis program that causes a computer to execute the following processes: acquiring battery data including the voltage and current of a secondary battery; calculating an SOH based on the voltage of the secondary battery in a first rest state, the voltage of the secondary battery in a second rest state, an SOC difference based on an SOC-OCV curve of the secondary battery, a current full charge capacity of the secondary battery calculated based on an integrated current value for the period between the first rest state and the second rest state, and an initial full charge capacity of the secondary battery; performing curve regression using sample data of the calculated multiple SOHs to generate a deterioration regression curve of the secondary battery; and weighting each SOH in the range of zero to one.

13. A non-transitory recording medium on which the battery analysis program according to claim 12 is recorded.

14. A battery analysis system comprising: a data acquisition unit that acquires battery data including a voltage and a current of a secondary battery transitioning from a first hibernation state to a second hibernation state; a SOH calculation unit that calculates a State of Health (SOH) based on the voltage of the secondary battery in the first hibernation state and the voltage of the secondary battery in the second hibernation state included in the battery data, an SOC difference based on a State Of Charge (SOC)-Open Circuit Voltage (OCV) curve of the secondary battery, a current full charge capacity of the secondary battery calculated based on an integrated current value for a period between the first hibernation state and the second hibernation state, and an initial full charge capacity of the secondary battery; and a regression curve generation unit that performs curve regression of the SOH using a plurality of SOH sample data calculated in a time series to generate a deterioration regression curve of the secondary battery, wherein the regression curve generation unit weights each of the plurality of SOHs in a range from zero to one.

Images