Battery management device, computing system, battery degradation prediction method and battery degradation prediction program

Abstract

The SOH estimation unit (4613) estimates the SOH (State of Health) of batteries (E1, 41) based on the measurement data of batteries (E1, 41). The degradation regression curve generation unit (4614) performs curve regression on multiple SOH values ​​of batteries (E1, 41) determined by time series to generate degradation regression curves for batteries (E1, 41). The rapid degradation determination unit (4615) determines whether rapid degradation has occurred in batteries (E1, 41) based on the difference or ratio between the degradation coefficient of the degradation regression curve of batteries (E1, 41) generated based on multiple SOH values ​​in a first data interval and the degradation coefficient of the degradation regression curve of batteries (E1, 41) generated based on multiple SOH values ​​in a second data interval.

Description

Technical Field

[0001] This disclosure relates to a battery management device, a computing system, a battery degradation prediction method, and a battery degradation prediction program for predicting battery degradation. Background Technology

[0002] In recent years, hybrid vehicles (HV), plug-in hybrid vehicles (PHV), and electric vehicles (EV) have become increasingly popular. These electric vehicles utilize secondary batteries, such as lithium-ion batteries, as a key component.

[0003] When secondary batteries such as lithium-ion batteries are repeatedly charged and discharged at low temperatures, rapid capacity degradation (hereinafter referred to as rapid degradation or tertiary degradation) is likely to occur. Furthermore, rapid capacity degradation also easily occurs when secondary batteries are repeatedly charged and discharged at high rates. Rapid capacity degradation occurs due to factors such as the reduction of electrolyte and the decrease in electrode reaction area. After rapid capacity degradation, input and output performance is significantly reduced. In addition, along with rapid degradation, lithium, which dissolves as ions, is prone to precipitate as a metal. When metallic lithium is deposited, it can penetrate the separator, potentially causing a short circuit between the positive and negative electrodes. Thus, after rapid capacity degradation, the stability and safety of the secondary battery decrease, essentially ending its service life.

[0004] As a method for detecting rapid degradation of secondary batteries, the following method is proposed: taking the full charge capacity (FCC) or capacity retention rate (SOH) of the secondary battery as input and using the elapsed time as input, if the change in the slope of the straight line obtained by linear regression of the change in full charge capacity or capacity retention rate with respect to the elapsed time exceeds a threshold, it is determined that rapid degradation has occurred (for example, see Patent Document 1).

[0005] Existing technical documents

[0006] Patent documents

[0007] Patent Document 1: International Publication No. 17 / 098686 Summary of the Invention

[0008] The values ​​of FCC, SOC (State of Charge), and SOH, calculated based on measurement data from electric vehicles in motion, are affected by sensor measurement errors and noise. When the influence of errors and noise is significant, the possibility of misjudgment increases if the aforementioned method of linear regression on changes in FCC or SOC is used.

[0009] This disclosure was made in view of the following circumstances, and its purpose is to provide a technique for detecting rapid degradation of batteries with high precision.

[0010] To address the aforementioned problems, a battery management device according to one aspect of this disclosure includes: a measurement unit that measures at least the voltage and current of a battery; a State of Health (SOH) estimation unit that estimates the SOH of the battery based on the measured data; a degradation regression curve generation unit that performs curve regression on multiple SOH values ​​determined by time series for the battery to generate a degradation regression curve for the battery; and a rapid degradation determination unit that determines whether the battery has experienced rapid degradation based on the difference or ratio between the degradation coefficient of the degradation regression curve generated based on multiple SOH values ​​in a first data interval and the degradation coefficient of the degradation regression curve generated based on multiple SOH values ​​in a second data interval.

[0011] Furthermore, any combination of the above-mentioned constituent elements, or any manner in which the expression of this disclosure is transformed between methods, apparatuses, systems, computer programs, etc., is also valid as a form of this disclosure.

[0012] According to this disclosure, rapid degradation of batteries can be detected with high precision. Attached Figure Description

[0013] Figure 1 This is a diagram used to illustrate the computing system utilized by the operator in the implementation method.

[0014] Figure 2 This is a diagram illustrating the detailed structure of the battery system mounted in an electric vehicle according to the embodiments described.

[0015] Figure 3 This is a diagram showing a structural example of the battery control unit involved in Embodiment 1.

[0016] Figure 4 This is a diagram used to illustrate the estimation method used by the FCC.

[0017] Figure 5 It is a graph that shows the degradation curve of a secondary battery.

[0018] Figure 6 This is an example of a degradation curve showing the rapid deterioration of a secondary battery.

[0019] Figure 7 This is a graph that uses curves to show specific examples of multiple degradation curves for different data ranges.

[0020] Figure 8 This is a diagram illustrating a specific example of the first method for dividing a data interval.

[0021] Figure 9 This is a diagram illustrating a specific example of the second method for dividing the data interval.

[0022] Figure 10 This is a flowchart illustrating the process of rapid degradation determination of battery modules performed by the Battery Management Department.

[0023] Figure 11 This is a diagram illustrating a structural example of the computing system involved in Embodiment 2. Detailed Implementation

[0024] Figure 1 This diagram illustrates the computing system 1 used by the operator in the implementation method. The operator owns multiple electric vehicles 3 and flexibly uses these electric vehicles 3 to operate its business. For example, the operator flexibly uses the multiple electric vehicles 3 to operate a delivery business (courier business), a taxi business, a car rental business, or a car-sharing business. In this embodiment, the electric vehicle 3 is assumed to be a pure EV without an engine.

[0025] The computing system 1 is a system used to manage the operator's business. The computing system 1 consists of one or more information processing devices (e.g., servers, PCs). Some or all of the information processing devices constituting the computing system 1 may also reside in a data center. For example, it may consist of a combination of servers within the data center (company servers, cloud servers, or leased servers) and client PCs within the operator.

[0026] Multiple electric vehicles 3 are parked in the parking lot or garage of the operator's business premises during standby. These electric vehicles 3 are equipped with wireless communication capabilities and can communicate wirelessly with the computing system 1. The multiple electric vehicles 3 transmit driving data, including usage data of their secondary batteries, to the computing system 1. The electric vehicles 3 can wirelessly transmit driving data to the server constituting the computing system 1 via a network during operation. For example, driving data can be transmitted once every 10 seconds. Alternatively, driving data for the entire day can be transmitted in batches once a day at a predetermined time (e.g., at the end of business hours).

[0027] Furthermore, if the computing system 1 consists of the company's server or PC located at the business office, the electric vehicle 3 can also send one day's driving data to the company's server or PC after being returned to the business office at the end of the business day. In this case, the data can be sent wirelessly or via a wired connection to the company's server or PC. Alternatively, the data can be sent to the company's server or PC via a recording medium containing the driving data. Additionally, if the computing system 1 consists of a cloud server and a client PC within the business office, the electric vehicle 3 can also send the driving data to the cloud server via the client PC within the business office.

[0028] Figure 2 This diagram illustrates the detailed structure of the battery system 40 installed in the electric vehicle 3 according to the embodiment. The battery system 40 is connected to the motor 34 via a first relay RY1 and an inverter 35. During power operation, the inverter 35 converts the DC power supplied from the battery system 40 into AC power and supplies it to the motor 34. During regeneration, the inverter 35 converts the AC power supplied from the motor 34 into DC power and supplies it to the battery system 40. The motor 34 is a three-phase AC motor that rotates according to the AC power supplied from the inverter 35 during power operation. During regeneration, the rotational energy generated by deceleration is converted into AC power and supplied to the inverter 35.

[0029] The first relay RY1 is a contactor inserted into the wiring connecting the battery system 40 and the inverter 35. During operation, the vehicle control unit 30 controls the first relay RY1 to be in the ON (closed) state, electrically connecting the battery system 40 to the power system of the electric vehicle 3. When not in operation, the vehicle control unit 30, in principle, controls the first relay RY1 to be in the OFF (open) state, disconnecting the battery system 40 from the power system of the electric vehicle 3. Alternatively, other types of switches, such as semiconductor switches, can be used instead of the relay.

[0030] The battery system 40 can be charged from the commercial power system 9 by connecting to a charger 4 located outside the electric vehicle 3 via a charging cable 38. The charger 4 is connected to the commercial power system 9 and charges the battery system 40 inside the electric vehicle 3 via the charging cable 38. In the electric vehicle 3, a second relay RY2 is inserted in the wiring connecting the battery system 40 and the charger 4. Alternatively, other types of switches, such as semiconductor switches, can be used instead of the relay. Before charging begins, the battery management unit 42 of the battery system 40 controls the second relay RY2 to the ON state (closed state), and after charging is completed, the battery management unit 42 of the battery system 40 controls the second relay RY2 to the OFF state (open state).

[0031] Generally, charging is performed using AC in normal charging and DC in fast charging. When charging with AC, AC power is converted to DC power by an AC / DC converter (not shown) inserted between the second relay RY2 and the battery system 40.

[0032] The battery system 40 includes a battery module 41 and a battery management unit 42. The battery module 41 comprises multiple cells E1-En connected in series. Alternatively, the battery module 41 can be constructed by connecting multiple battery modules in series or in parallel. The cells can be lithium-ion battery cells, nickel-metal hydride battery cells, lead-acid battery cells, etc. In the following specification, we assume the use of lithium-ion battery cells (nominal voltage: 3.6V-3.7V). The number of cells E1-En connected in series is determined by the drive voltage of the motor 34.

[0033] A shunt resistor Rs is connected in series with multiple individual cells E1-En. The shunt resistor Rs functions as a current sensing element. Alternatively, a Hall element can be used instead of the shunt resistor Rs. Furthermore, multiple temperature sensors T1 and T2 are provided within the battery module 41 for detecting the temperature of the multiple individual cells E1-En. Either a single temperature sensor can be provided within the battery module, or a temperature sensor can be provided for each of the multiple individual cells. The temperature sensors T1 and T2 can be, for example, thermistors.

[0034] The battery management unit 42 includes a voltage measurement unit 43, a temperature measurement unit 44, a current measurement unit 45, and a battery control unit 46. Each node of the multiple series-connected cells E1-En is connected to the voltage measurement unit 43 via multiple voltage lines. The voltage measurement unit 43 measures the voltage of each cell E1-En by measuring the voltage between two adjacent voltage lines. The voltage measurement unit 43 sends the measured voltage of each cell E1-En to the battery control unit 46.

[0035] The voltage of the voltage measuring unit 43 is higher than the voltage of the battery control unit 46, therefore the voltage measuring unit 43 and the battery control unit 46 are connected in an insulated state via a communication line. The voltage measuring unit 43 can be constructed from an ASIC (Application Specific Integrated Circuit) or a general-purpose analog front-end IC. The voltage measuring unit 43 includes a multiplexer and an A / D converter. The multiplexer outputs the voltages between two adjacent voltage lines sequentially from top to bottom to the A / D converter. The A / D converter converts the analog voltage input from the multiplexer into a digital value.

[0036] The temperature measurement unit 44 includes voltage divider resistors and an A / D converter. The A / D converter sequentially converts multiple analog voltages obtained by voltage division by multiple temperature sensors T1, T2 and multiple voltage divider resistors into digital values ​​and outputs them to the battery control unit 46. The battery control unit 46 estimates the temperature of multiple individual cells E1-En based on these digital values. For example, the battery control unit 46 estimates the temperature of each individual cell E1-En based on the values ​​measured by the temperature sensor closest to each individual cell E1-En.

[0037] The current measurement unit 45 includes a differential amplifier and an A / D converter. The differential amplifier amplifies the voltage across the shunt resistor Rs and outputs it to the A / D converter. The A / D converter converts the voltage input from the differential amplifier into a digital value and outputs it to the battery control unit 46. The battery control unit 46 estimates the current flowing in the multiple cells E1-En based on this digital value.

[0038] Furthermore, if the battery control unit 46 is equipped with an A / D converter and has an analog input port, the temperature measuring unit 44 and the current measuring unit 45 can also output analog voltage to the battery control unit 46, and the A / D converter in the battery control unit 46 can convert the analog voltage into a digital value.

[0039] The battery control unit 46 manages the state of multiple individual cells E1-En based on the voltage, temperature, and current measured by the voltage measuring unit 43, temperature measuring unit 44, and current measuring unit 45. The battery control unit 46 is connected to the vehicle control unit 30 via an in-vehicle network. For example, CAN (Controller Area Network) or LIN (Local Interconnect Network) can be used as the in-vehicle network.

[0040] Figure 3 This diagram illustrates a structural example of the battery control unit 46 according to Embodiment 1. The battery control unit 46 includes a processing unit 461 and a storage unit 462. The processing unit 461 includes a SOC estimation unit 4611, an FCC estimation unit 4612, a SOH estimation unit 4613, a degradation regression curve generation unit 4614, a rapid degradation determination unit 4615, and a data transmission unit 4616. The functions of the processing unit 461 can be implemented through the cooperation of hardware resources and software resources, or can be implemented solely through hardware resources. As hardware resources, CPUs, ROMs, RAMs, ASICs, FPGAs (Field Programmable Gate Arrays), and other LSIs can be utilized. As software resources, firmware and other programs can be utilized.

[0041] The storage unit 462 includes a SOC-OCV (Open Circuit Voltage) characteristic holding unit 4621, a battery data holding unit 4622, and a time series SOH value holding unit 4623. The storage unit 462 includes non-volatile recording media such as EEPROM (Electrically Erasable Programmable Read-Only Memory) and NAND flash memory to record various programs and data.

[0042] The SOC-OCV characteristic retention section 4621 contains characteristic data of the SOC-OCV curves of multiple individual cells E1-En. These SOC-OCV curves of the multiple individual cells E1-En are pre-made by the battery manufacturer and registered in the SOC-OCV characteristic retention section 4621 at the time of shipment. The battery manufacturer conducts various tests to derive the SOC-OCV curves of the individual cells E1-En.

[0043] The battery data retention unit 4622 records battery data, including the voltage, current, and temperature of multiple individual cells E1-En, in a time-series manner. Furthermore, the battery data may also include the State of Charge (SOC) estimated by the State of Charge (SOC) estimation unit 4611.

[0044] The time-series SOH value holding unit 4623 holds the time-series data of SOH estimated by the SOH estimation unit 4613. The time-series data of SOH is recorded, for example, at a frequency of once a day, once every few days, or once a week. In addition, the time-series SOH value holding unit 4623 and the battery data holding unit 4622 can also be integrated into a single table.

[0045] The SOC estimation unit 4611 estimates the SOC of each of the multiple individual cells E1-En. The SOC estimation unit 4611 estimates the SOC using the OCV method, the current integration method, or a combination of both. The OCV method estimates the SOC based on the OCV of each individual cell E1-En measured by the voltage measuring unit 43 and the characteristic data of the SOC-OCV curve maintained in the SOC estimation unit 4611. The current integration method estimates the SOC based on the OCV of each individual cell E1-En at the start of charging and discharging and the integral value of the current measured by the current measuring unit 45. In the current integration method, the measurement error of the current measuring unit 45 accumulates as the charging and discharging time increases. Therefore, it is preferable to use the SOC estimated by the OCV method and correct the estimated SOC by the current integration method.

[0046] The FCC estimation unit 4612 can estimate the FCC of a cell based on the characteristic data of the SOC-OCV curve maintained in the SOC estimation unit 4611 and the OCV of the cell at two points measured by the voltage measurement unit 43.

[0047] Figure 4 This diagram illustrates the FCC estimation method. The FCC estimation unit 4612 acquires the OCV at two points of the unit. Referring to the SOC-OCV curve, the FCC estimation unit 4612 determines the SOC at the two points corresponding to the voltages at those two points, and calculates the difference ΔSOC between the two points. Figure 4 In the example shown, the SOC at the two points is 20% and 75%, and the ΔSOC is 55%.

[0048] The FCC estimation unit 4612 calculates the current integral (=charge / discharge capacity) Q between the times of the two OCV points obtained from the current measurement unit 45 based on the current shift measured by the current measurement unit 45. The FCC estimation unit 4612 can estimate the FCC by calculating the following (Equation 1).

[0049] FCC=Q / ΔSOC…(Equation 1)

[0050] The SOH estimation unit 4613 estimates the SOH based on the estimated FCC. The SOH is defined by the ratio of the current FCC to the initial FCC; the lower the value (closer to 0%), the more severe the degradation. The SOH estimation unit 4613 can calculate the following (Equation 2) to estimate the SOH.

[0051] SOH = Current FCC / Initial FCC... (Equation 2)

[0052] Furthermore, SOH can be determined either through capacity measurements based on full charge and discharge, or by adding storage degradation and cycle degradation. Storage degradation can be estimated based on SOC, temperature, and storage degradation rate. Cycle degradation can be estimated based on the used SOC range, temperature, current rate, and cycle degradation rate. Storage degradation rate and cycle degradation rate can be derived in advance through experiments and simulations. SOC, temperature, SOC range, and current rate can be determined through measurement.

[0053] Furthermore, SOH can also be estimated based on its relationship with the internal resistance of the monomer. Internal resistance can be estimated by dividing the voltage drop generated when a specified current flows through the monomer for a specified time by that current value. Regarding internal resistance, there is a relationship that the higher the temperature, the lower the internal resistance, and vice versa, that the lower the SOH, the higher the internal resistance.

[0054] The SOH estimation unit 4613 saves the estimated SOH to the time series SOH value holding unit 4623. The SOH estimation unit 4613 estimates the SOH at a frequency of once a day, once every few days, or once a week, and saves the estimated SOH to the time series SOH value holding unit 4623.

[0055] The degradation regression curve generation unit 113 performs curve regression on multiple SOH values ​​of the battery module 41 determined by time series to generate a degradation regression curve for the battery module 41. The curve regression can, for example, use the least squares method.

[0056] Figure 5 This is a graph showing the degradation curve of a secondary battery. It is known that the degradation of a secondary battery increases proportionally to the square root of time (0.5 power), as shown in Equation 3 below.

[0057]

[0058] w0 is the initial value, and w1 is the degradation coefficient.

[0059] The degradation regression curve generation unit 4614 calculates the degradation coefficient w1 in Equation 3 above by performing an exponential curve regression with time t as the independent variable and SOH as the dependent variable, raised to the power of 0.5. w0 is common and is usually set in the range of 1.0 to 1.1. When the actual initial capacity is consistent with the nominal value, w0 is set to 1.0, and the nominal value is set as the minimum guaranteed amount. When the nominal value is set lower than the actual initial capacity, it is set to a value greater than 1.0.

[0060] Figure 6 This is an example of a degradation curve showing the rapid deterioration of a secondary battery. Figure 6 The diagram illustrates an example of rapid degradation occurring at point P1. As described above, rapid degradation is prone to occur when the secondary battery is subjected to repeated charging and discharging at low or high temperatures, or at high rates, which place a heavy burden on the battery. When rapid degradation occurs, the secondary battery becomes essentially unusable, thus shortening its lifespan. The main cause of rapid degradation is the reduction of electrolyte, but directly measuring the amount of electrolyte requires disassembling the secondary battery. It is impractical to disassemble each cell (E1-En) during the use of battery module 41, and a method is sought to determine rapid degradation without disassembling each cell (E1-En). In this embodiment, rapid degradation is detected by detecting a significant deviation of the SOH of battery module 41 from the degradation curve.

[0061] The rapid degradation determination unit 4615 determines whether rapid degradation has occurred in the battery module 41 based on the difference or ratio between the degradation coefficient w1 of the degradation regression curve of the battery module 41 generated based on multiple SOHs in a first data interval and the degradation coefficient w1 of the battery module 41 generated based on multiple SOHs in a second data interval. When the difference or ratio deviates from a predetermined range, the rapid degradation determination unit 4615 determines that rapid degradation has occurred in the battery module 41. That is, when the absolute value of the difference or ratio exceeds a threshold, the rapid degradation determination unit 4615 determines that rapid degradation has occurred; when the absolute value of the difference or ratio is below the threshold, the rapid degradation determination unit 4615 determines that rapid degradation has not occurred. This threshold can be a value derived through experiments or simulations. Furthermore, the determination of rapid degradation can also be performed on a per-unit basis.

[0062] Figure 7 This is a graph that uses curves to show specific examples of multiple degradation curves for different data ranges. Figure 7 The degradation regression curve shown w0 is set to 1.05. This indicates that the actual initial capacity of battery module 41 is greater than the nominal value. Figure 7 In the example shown, the degradation curves based on the SOH of the past 100 points, the past 200 points, the past 300 points, and the SOH of all points are plotted overlaid. The degradation curves based on the past 200 points, the past 300 points, and the all-points are almost identical, and the degradation coefficient w1 of each curve is also almost the same value. In contrast, the degradation coefficient w1 of the degradation curve based on the past 100 points is smaller than the degradation coefficient w1 of the other three curves.

[0063] exist Figure 7 As shown in the example, the SOH in region R1, enclosed by the dashed circle, is significantly lower than the SOH in regions preceding R1. Therefore, it can be estimated that rapid degradation has occurred at a certain point in region R1. If the threshold is set, for example, to the difference between the degradation coefficient w1 based on the degradation curve of the past 100 points and the degradation coefficient w1 based on the degradation curve of the past 200 points in region R1, then region R1 can be detected by comparing the two degradation coefficients w1.

[0064] Figure 8This diagram illustrates a specific example of a first method for dividing data intervals. The first method involves multiple data intervals having the same endpoint but varying the number of data points traced back. For example, the first data interval is defined as including a SOHs from the last determined SOH. The second data interval is defined as including b (b > a) SOHs from the last determined SOH. The third data interval is defined as including c (c > b > a) SOHs from the last determined SOH. Figure 8 In the example shown, a = 100, b = 200, and c = 300.

[0065] Figure 9 This diagram illustrates a specific example of a second method for dividing data intervals. The second method involves dividing multiple data intervals sequentially backwards with the same number of intervals. The first data interval is defined as an interval including *a* SOHs starting from the last determined SOH. The second data interval is defined as an interval including *a* SOHs starting from the last determined SOH excluding those included in the first data interval. The third data interval is defined as an interval including *a* SOHs starting from the last determined SOH excluding those included in the first and second data intervals. Figure 9 In the example shown, a = 100.

[0066] Return to Figure 3 The data transmission unit 4616 of the battery control unit 46 notifies the vehicle control unit 30 of the voltage, current, temperature, SOC, FCC, and SOH of multiple individual cells E1-En via the vehicle network. The vehicle control unit 30 generates driving data that includes battery data and vehicle data. The battery data includes the voltage, current, and temperature of multiple individual cells E1-En. Furthermore, depending on the battery system 40, there are models that can include SOC in addition to voltage, current, and temperature in the battery data. There are also models that can include at least one of FCC and SOH in addition to voltage, current, temperature, and SOC. The vehicle data can include average speed, driving distance, driving route, etc.

[0067] If the rapid degradation determination unit 4615 detects rapid degradation of the battery module 41, the data transmission unit 4616 notifies the vehicle control unit 30 of the rapid degradation detection signal via the vehicle network. Upon receiving the rapid degradation detection signal from the battery module 41, the vehicle control unit 30 illuminates a warning light on the dashboard located in the driver's seat, indicating an abnormality in the battery module 41, to notify the driver of the abnormality. Alternatively, the vehicle control unit 30 can also notify the driver of the abnormality in the battery module 41 via a synthesized voice output.

[0068] The wireless communication unit 36 ​​performs signal processing for wireless connection to the network via antenna 36a. In this embodiment, the wireless communication unit 36 ​​wirelessly transmits driving data acquired from the vehicle control unit 30 to the computing system 1. Additionally, the wireless communication unit 36 ​​wirelessly transmits rapid degradation detection signals of the battery module 41 acquired from the vehicle control unit 30 to the computing system 1. For example, mobile phone networks (cellular networks), wireless LANs, ETC (Electronic Toll Collection System), DSRC (Dedicated Short Range Communications), V2I (Vehicle-to-Infrastructure) communication, and V2V (Vehicle-to-Vehicle) communication can be used as the wireless communication network for the electric vehicle 3.

[0069] Figure 10 This is a flowchart illustrating the process of rapid degradation determination of battery module 41 performed by battery management unit 42. SOH estimation unit 4613 estimates the SOH (S10) of battery module 41 based on the measurement data of battery module 41.

[0070] When determining whether the battery module 41 has experienced rapid degradation, the degradation regression curve generation unit 4614 performs curve regression on multiple SOH values ​​in the first data range of the battery module 41 to generate a first degradation regression curve for the battery module 41 (S11). Simultaneously, the degradation regression curve generation unit 4614 performs curve regression on multiple SOH values ​​in the second data range of the battery module 41 to generate a second degradation regression curve for the battery module 41 (S12).

[0071] The rapid degradation determination unit 4615 calculates the difference between the degradation coefficient w1 of the first degradation regression curve and the degradation coefficient w1 of the second degradation regression curve (S13). When the absolute value of the difference is below a threshold (S14 is "No"), the rapid degradation determination unit 4615 determines that the battery module 41 has not experienced rapid degradation (S15). When the absolute value of the difference exceeds the threshold (S14 is "Yes"), the rapid degradation determination unit 4615 determines that the battery module 41 has experienced rapid degradation (S16).

[0072] In Embodiment 1 described above, an example of rapid degradation determination processing of battery module 41 performed by battery management unit 42 was explained. However, this rapid degradation determination processing of battery module 41 can also be performed by computing system 1.

[0073] Figure 11This diagram illustrates a structural example of the computing system 1 according to Embodiment 2. The computing system 1 includes a processing unit 11, a storage unit 12, a display unit 13, and an operation unit 14. The processing unit 11 includes a data acquisition unit 111, a state of equilibrium (SOH) determination unit 112, a degradation regression curve generation unit 113, a rapid degradation determination unit 114, an operation acceptance unit 115, and a display control unit 116. The functions of the processing unit 11 can be implemented through the cooperation of hardware resources and software resources, or can be implemented solely through hardware resources. As hardware resources, CPUs, GPUs (Graphics Processing Units), ROMs, RAMs, ASICs, FPGAs, and other LSIs can be utilized. As software resources, operating systems, application programs, and other programs can be utilized.

[0074] The storage unit 12 includes a driving data storage unit 121, a driver data storage unit 122, a SOC-OCV characteristic storage unit 123, and a time-series SOH value storage unit 124. The storage unit 12 includes non-volatile recording media such as HDD (Hard Disk Drive) and SSD (Solid State Drive) to record various programs and data.

[0075] The driving data retention unit 121 retains driving data collected from multiple electric vehicles 3 owned by the operator. The driver data retention unit 122 retains data from multiple drivers belonging to the operator. For example, it manages the cumulative driving distance of each electric vehicle 3 driven by each driver.

[0076] The SOC-OCV characteristic retention unit 123 retains the SOC-OCV characteristics of multiple battery modules 41 respectively installed in multiple electric vehicles 3 owned by the operator. The SOC-OCV characteristics of the battery module 41 can be obtained from the SOC-OCV characteristics of each electric vehicle 3, or the SOC-OCV characteristics estimated based on driving data collected from each electric vehicle 3.

[0077] In the latter case, the SOC-OCV characteristic estimation unit (not shown) of the processing unit 11 extracts a group of SOC and voltage (≈OCV) values ​​that can be considered as the period when the battery module 41 is in a resting state from the groups of SOC and voltage values ​​at multiple times contained in the acquired battery data, and approximates the SOC-OCV characteristics based on the extracted groups of SOC and OCV. Furthermore, the SOC-OCV characteristic estimation unit can also generate common SOC-OCV characteristics for that type of battery module 41 based on group data of SOC and OCV values ​​acquired from multiple electric vehicles 3 equipped with the same type of battery module 41. Additionally, the SOC-OCV characteristics can also be maintained on a per-cell basis.

[0078] The time-series SOH value holding unit 124 holds the time-series data of the SOH of each battery module 41. The time-series data of SOH is recorded at a frequency of once a day, once every few days, or once a week, for example.

[0079] The display unit 13 includes a liquid crystal display (LCD), an organic EL display, or other display that shows images generated by the processing unit 11. The operation unit 14 is a user interface such as a keyboard, mouse, or touch panel that handles operations from the user of the computing system 1.

[0080] The data acquisition unit 111 acquires driving data containing battery data of battery modules 41 installed in multiple electric vehicles 3, and saves the acquired driving data to the driving data holding unit 121. The SOH determination unit 112 determines the SOH of each battery module 41 installed in each electric vehicle 3 based on the battery data included in the driving data acquired by the data acquisition unit 111. The SOH determination unit 112 saves the determined SOH to the time series SOH value holding unit 124.

[0081] If the acquired battery data includes State of Health (SOH), the SOH determination unit 112 can directly use the acquired SOH. If the acquired battery data does not include SOH but includes voltage, current, temperature, and State of Charge (SOC), the SOH can be calculated based on Equations 1 and 2 above. That is, the SOH determination unit 112 calculates the current integral Q between the two points of time when the OCV of the two points is acquired based on the current shift included in the battery data, and applies the calculated current integral Q to Equation 1 above to estimate the FCC. The SOH determination unit 112 applies the calculated FCC to Equation 2 above to calculate the SOH.

[0082] If the acquired battery data does not contain either SOC or SOH, the SOH determination unit 112 applies the voltage (≈OCV) during a period when the battery module 41 is in a resting state to the SOC-OCV characteristic to estimate the SOC. Alternatively, the SOH determination unit 112 integrates the current value over a fixed period to estimate the SOC. The SOH determination unit 112 uses the estimated SOC to calculate the SOH in the same way as if the battery data contained SOC.

[0083] The degradation regression curve generation unit 113 performs curve regression on multiple SOH values ​​determined by time series for each battery module 41 to generate a degradation regression curve for each battery module 41. The rapid degradation determination unit 114 calculates the difference or ratio between the degradation coefficient w1 of the degradation regression curve of the battery module 41 generated based on multiple SOH values ​​in a first data interval of the time series SOH values ​​of a specific battery module 41 and the degradation coefficient w1 of the degradation regression curve of the battery module 41 generated based on multiple SOH values ​​in a second data interval. Based on the calculated difference or ratio, the rapid degradation determination unit 114 determines whether the battery module 41 has experienced rapid degradation.

[0084] The operation receiving unit 117 receives user operations on the operation unit 14. The display control unit 118 causes the display unit 13 to display various information. In embodiment 2, the display unit causes the display to show the determination results of rapid degradation of each battery module 41.

[0085] As explained above, according to this embodiment, by referring to the difference or ratio of the degradation coefficient w1 of multiple degradation regression curves generated by changing the data interval, rapid degradation of the battery module 41 can be detected with high precision without disassembling the battery module 41. Like the battery module 41 installed in the electric vehicle 3, reliable detection can be performed even when using data containing estimation errors of SOH (State of Health). Experiments by the inventors show that with approximately 100 points of SOH, the maximum error can be suppressed to approximately 5%. If SOH is estimated once a day, rapid degradation can be detected with high precision after more than three months. Furthermore, the more SOH points there are, the smaller the error becomes.

[0086] Another approach is to determine rapid degradation based on the change in the slope of a straight line obtained by linear regression of the changes in FCC or SOH of battery module 41. This method is considered effective when the error and noise are small, but the determination of rapid degradation may become unstable when the error and noise are large.

[0087] In this embodiment, a degradation regression curve is generated by dividing the SOH data interval of the time series. If it is normal degradation, the degradation coefficient w1 of the degradation regression curve does not change substantially. By using the degradation coefficient w1 of the degradation regression curve as a parameter, changes in the degradation regression curve itself caused by rapid degradation can be detected. Detecting changes in the degradation regression curve itself provides a more reliable detection method compared to detecting changes in the slope of the straight line obtained by linear regression of changes in FCC or SOH.

[0088] When the battery module 41 installed in the electric vehicle 3 deteriorates rapidly, the driving range of the electric vehicle 3 decreases sharply. By detecting rapid deterioration, it is possible to replace the electric vehicle 3 at an appropriate time or change its usage method. By detecting rapid deterioration of the battery module 41 in this way, the safety of using the battery module 41 can be improved.

[0089] Furthermore, if the determination of whether there is rapid degradation is made by the computing system 1 based on the measurement data sent from the electric vehicle 3, rather than by the battery management unit 42 inside the electric vehicle 3, the vehicle management of operators who own a large number of electric vehicles 3 can be made more efficient.

[0090] The present disclosure has been described above based on embodiments. It should be understood by those skilled in the art that the embodiments are illustrative and various modifications can be made to the combination of their constituent elements and processing procedures, and such modifications are also within the scope of the present disclosure.

[0091] In the above embodiment, an example of dividing the data interval by the number of SOHs was described. In this respect, the data interval can also be divided by the number of days (e.g., 100 days). In this case, it is easy to set the confirmation of whether the battery module 41 has rapidly deteriorated as one of the items of regular vehicle inspection.

[0092] In the above embodiment, the degradation regression curve generation unit 4614 compares the degradation coefficient w1 based on the data in the first data interval with the degradation coefficient w1 based on the data in the second data interval. In this regard, the degradation regression curve generation unit 4614 may also compare the degradation coefficient w1 based on the data in the first data interval with values ​​obtained by statistically processing multiple degradation coefficients w1 based on multiple data intervals (e.g., mean, variance, standard deviation).

[0093] When comparing variance values, the degradation regression curve generation unit 4614 compares the squared value of the deviation of the degradation coefficient w1 based on the data in the first data interval with the variance values ​​of multiple degradation coefficients w1 based on multiple data intervals. When comparing standard deviation values, the degradation regression curve generation unit 4614 compares the absolute value of the deviation of the degradation coefficient w1 based on the data in the first data interval with the standard deviation values ​​of multiple degradation coefficients w1 based on multiple data intervals. In these cases, rapid degradation can be detected with higher accuracy.

[0094] The rapid degradation determination method described in the above embodiments can also be combined with other rapid degradation determination methods. For example, one method involves applying an AC signal in a frequency band (e.g., 100Hz to 10kHz) that causes the electrolyte to react from outside the battery module 41, measuring the AC impedance value of the battery module 41, and detecting or predicting rapid degradation of the battery module 41 based on whether the measured AC impedance value is above a threshold. This method requires a circuit for applying the AC signal to the battery module 41 and measuring the AC impedance value. However, the rapid degradation determination method according to this embodiment does not require such a circuit.

[0095] If rapid degradation is determined to have occurred by the rapid degradation determination method according to this embodiment, the electric vehicle 3 may be moved to a facility (e.g., a car dealership) that has a circuit device capable of measuring the AC impedance value of the battery module 41 to perform a rapid degradation determination based on the AC impedance value.

[0096] In order to notify the user of a warning in the early stage after rapid degradation of battery module 41, rapid degradation determination processing needs to be performed frequently. When the rapid degradation determination method according to this embodiment is performed frequently (for example, every time one data point is added), the probability of mistakenly determining that rapid degradation has occurred even if it has not actually occurred increases.

[0097] In this regard, if the rapid degradation determination process involved in the implementation method is defined as a single determination, and the rapid degradation determination process based on AC impedance value is defined as a double determination, then even if the rapid degradation determination method involved in this implementation method is performed at a high frequency, the probability of false determination is reduced. That is, rapid degradation of battery module 41 can be detected at an early stage with high accuracy.

[0098] In the above embodiment, an example is assumed to be used to determine the rapid degradation of the battery module 41 installed in the electric vehicle 3. In this regard, the electric vehicle 3 can also be a two-wheeled electric motorcycle (electric scooter) or an electric bicycle. In addition, the electric vehicle 3 also includes low-speed electric vehicles such as golf carts and land cars used in shopping malls, entertainment facilities, etc.

[0099] Furthermore, the objects equipped with battery module 41 are not limited to electric vehicles 3. For example, they also include electric ships, railway vehicles, multi-rotor helicopters (drones), and other electric mobile bodies. In addition, objects equipped with battery module 41 also include stationary energy storage systems and consumer electronic devices (smartphones, laptops, etc.).

[0100] In addition, the implementation method can also be determined by the following items.

[0101] [Project 1]

[0102] A battery management device (42) is characterized by comprising: a measurement unit (43-45) that measures at least the voltage and current of a battery (E1, 41); a State of Health (SOH) estimation unit (4613) that estimates the SOH of the battery (E1, 41) based on the measurement data of the battery (E1, 41); a degradation regression curve generation unit (4614) that performs curve regression on multiple SOH values ​​of the battery (E1, 41) determined by time series to generate a degradation regression curve of the battery (E1, 41); and a rapid degradation determination unit (4615) that determines whether the battery (E1, 41) has experienced rapid degradation based on the difference or ratio between the degradation coefficient of the degradation regression curve of the battery (E1, 41) generated based on multiple SOH values ​​in a first data interval and the degradation coefficient of the degradation regression curve of the battery (E1, 41) generated based on multiple SOH values ​​in a second data interval.

[0103] The battery (E1, 41) can be either a single E1 or a module 41.

[0104] Therefore, it is possible to detect the rapid degradation of batteries (E1, 41) with high precision.

[0105] [Project 2]

[0106] According to the battery management device (42) of Project 1, the rapid degradation determination unit (4615) determines that the battery (E1, 41) has rapidly degraded when the difference or the ratio deviates from the specified range.

[0107] Therefore, rapid degradation can be detected with high precision by detecting deviations from the usual degradation pattern.

[0108] [Project 3]

[0109] The battery management device (42) according to item 1 or 2 is characterized in that the first data interval is an interval including a SOHs from the last determined SOH to the past, and the second data interval is an interval including b (b>a) SOHs from the last determined SOH to the past.

[0110] Therefore, by repeating the data intervals, stable detection can be achieved.

[0111] [Project 4]

[0112] The battery management device (42) according to item 1 or 2 is characterized in that the first data interval is an interval including a SOHs from the last determined SOH to the past, and the second data interval is an interval including a SOHs from the last determined SOH excluding the first data interval.

[0113] Therefore, by ensuring that the data intervals are not repeated, early detection can be achieved.

[0114] [Project 5]

[0115] The battery management device (42) according to item 1 or 2 is characterized in that the second data interval includes multiple data intervals, and the degradation coefficient of the second data interval is a value obtained by statistically processing the degradation coefficients of the multiple data intervals.

[0116] Therefore, the detection accuracy can be further improved.

[0117] [Project 6]

[0118] A computing system (1) is characterized by comprising: a data acquisition unit (111) for acquiring measurement data of batteries (E1, 41); a SOH determination unit (112) for determining the SOH of batteries (E1, 41) based on the measurement data of batteries (E1, 41); a degradation regression curve generation unit (113) for generating a degradation regression curve of batteries (E1, 41) by performing curve regression on multiple SOHs determined by time series of batteries (E1, 41); and a rapid degradation determination unit (114) for determining whether batteries (E1, 41) have experienced rapid degradation based on the difference or ratio between the degradation coefficient of the degradation curve of batteries (E1, 41) generated based on multiple SOHs in a first data interval and the degradation coefficient of the degradation curve of batteries (E1, 41) generated based on multiple SOHs in a second data interval.

[0119] Therefore, it is possible to detect the rapid degradation of batteries (E1, 41) with high precision.

[0120] [Project 7]

[0121] A method for predicting the degradation of batteries (E1, 41) is characterized by comprising the following steps: determining the state of harm (SOH) of the batteries (E1, 41) based on measurement data of the batteries (E1, 41); performing curve regression on multiple SOH values ​​of the batteries (E1, 41) determined by time series to generate a degradation regression curve of the batteries (E1, 41); and determining whether the batteries (E1, 41) have experienced rapid degradation based on the difference or ratio between the degradation coefficient of the degradation regression curve of the batteries (E1, 41) generated based on multiple SOH values ​​in a first data interval and the degradation coefficient of the degradation regression curve of the batteries (E1, 41) generated based on multiple SOH values ​​in a second data interval.

[0122] Therefore, it is possible to detect the rapid degradation of batteries (E1, 41) with high precision.

[0123] [Project 8]

[0124] A degradation prediction program for batteries (E1, 41) is characterized by having a computer perform the following processing: determining the state of harm (SOH) of the batteries (E1, 41) based on measurement data of the batteries (E1, 41); performing curve regression on multiple SOH values ​​of the batteries (E1, 41) determined by time series to generate degradation regression curves for the batteries (E1, 41); and determining whether the batteries (E1, 41) have experienced rapid degradation based on the difference or ratio between the degradation coefficient of the degradation regression curve of the batteries (E1, 41) generated based on multiple SOH values ​​in a first data interval and the degradation coefficient of the degradation regression curve of the batteries (E1, 41) generated based on multiple SOH values ​​in a second data interval.

[0125] Therefore, it is possible to detect the rapid degradation of batteries (E1, 41) with high precision.

[0126] Explanation of reference numerals in the attached figures

[0127] 1: Computing system; E1-En: Individual units; T1, T2: Temperature sensors; RY1, RY2: Relays; 3: Electric vehicle; 4: Charger; 11: Processing unit; 111: Data acquisition unit; 112: SOH determination unit; 113: Degradation regression curve generation unit; 114: Rapid degradation judgment unit; 115: Operation receiving unit; 116: Display control unit; 12: Storage unit; 121: Driving data retention unit; 122: Driver data retention unit; 123: SOC-OCV characteristic retention unit; 124: Time series SOH value retention unit; 13: Display unit; 14: Operation unit; 30: Vehicle control unit; 34: Motor; 35: Inverter 36: Wireless communication unit; 36a: Antenna; 38: Charging cable; 40: Battery system; 41: Battery module; 42: Battery management unit; 43: Voltage measurement unit; 44: Temperature measurement unit; 45: Current measurement unit; 46: Battery control unit; 461: Processing unit; 4611: SOC estimation unit; 4612: FCC estimation unit; 4613: SOH estimation unit; 4614: Degradation regression curve generation unit; 4615: Rapid degradation judgment unit; 4616: Data transmission unit; 462: Storage unit; 4621: SOC-OCV characteristic retention unit; 4622: Battery data retention unit; 4623: Time series SOH value retention unit.

Claims

1. A battery management device, characterized in that, have: The measuring unit measures at least the battery's voltage and current. The SOH estimation unit estimates the state of health (SOH) of the battery based on the measurement data of the battery. The degradation regression curve generation unit performs curve regression on multiple SOH values ​​determined by the battery according to the time series to generate a degradation regression curve of the battery expressed by the following formula: SOH = W0 + W1√t, Where W0 is the initial value, W1 is the degradation coefficient, and t is time. W1 is obtained through curve regression. as well as The rapid degradation determination unit determines whether the battery has experienced rapid degradation based on the difference or ratio between the W1 value corresponding to the degradation regression curve of the battery generated based on multiple SOHs in the first data interval and the W1 value corresponding to the degradation regression curve of the battery generated based on multiple SOHs in the second data interval.

2. The battery management device according to claim 1, characterized in that, When the difference or the ratio deviates from the specified range, the rapid degradation determination unit determines that the battery has undergone rapid degradation.

3. The battery management device according to claim 1 or 2, characterized in that, The first data interval includes the interval of a SOHs preceding the last determined SOH. The second data interval is the interval that includes b SOH values ​​from the last determined SOH. Where b > a.

4. The battery management device according to claim 1 or 2, characterized in that, The first data interval includes the interval of a SOHs preceding the last determined SOH. The second data interval is an interval that includes a SOHs from the last determined SOH excluding the first data interval.

5. The battery management device according to claim 1, characterized in that, The second data interval includes multiple data intervals. The value of W1 corresponding to the degradation regression curve of the battery generated based on multiple SOH values ​​in the second data interval is obtained by statistically processing each W1 value in the multiple data intervals.

6. A computing system, characterized in that, have: The data acquisition department acquires the battery's measurement data; The SOH determination unit determines the state of health (SOH) of the battery based on the measurement data of the battery. The degradation regression curve generation unit performs curve regression on multiple SOH values ​​determined by the battery according to the time series to generate a degradation regression curve of the battery expressed by the following formula: SOH = W0 + W1√t, Where W0 is the initial value, W1 is the degradation coefficient, and t is time. W1 is obtained through curve regression. as well as The rapid degradation determination unit determines whether the battery has experienced rapid degradation based on the difference or ratio between the W1 value corresponding to the degradation curve of the battery generated based on multiple SOHs in the first data interval and the W1 value corresponding to the degradation curve of the battery generated based on multiple SOHs in the second data interval.

7. A method for predicting battery degradation, characterized in that, Includes the following steps: Based on the battery's measurement data, the battery's state of health (SOH) is determined. Curve regression is performed on multiple SOH values ​​determined by the time series of the battery to generate a degradation regression curve for the battery expressed by the following formula: SOH = W0 + W1√t, Where W0 is the initial value, W1 is the degradation coefficient, and t is time. W1 is obtained through curve regression. as well as The difference or ratio between the W1 value corresponding to the degradation regression curve of the battery generated based on multiple SOHs in the first data interval and the W1 value corresponding to the degradation regression curve of the battery generated based on multiple SOHs in the second data interval is used to determine whether the battery has experienced rapid degradation.

8. A computer-readable storage medium storing a battery degradation prediction program, characterized in that, The computer will perform the following processes: Based on the battery's measurement data, the battery's state of health (SOH) is determined. Curve regression is performed on multiple SOH values ​​determined by the time series of the battery to generate a degradation regression curve for the battery expressed by the following formula: SOH = W0 + W1√t, Where W0 is the initial value, W1 is the degradation coefficient, t is time, and W1 is obtained through curve regression; and The difference or ratio between the W1 value corresponding to the degradation regression curve of the battery generated based on multiple SOHs in the first data interval and the W1 value corresponding to the degradation regression curve of the battery generated based on multiple SOHs in the second data interval is used to determine whether the battery has experienced rapid degradation.

9. A computer program product comprising a battery degradation prediction program, characterized in that, The computer will perform the following processes: Based on the battery's measurement data, the battery's state of health (SOH) is determined. Curve regression is performed on multiple SOH values ​​determined by the time series of the battery to generate a degradation regression curve for the battery expressed by the following formula: SOH = W0 + W1√t, Where W0 is the initial value, W1 is the degradation coefficient, t is time, and W1 is obtained through curve regression; and The difference or ratio between the W1 value corresponding to the degradation regression curve of the battery generated based on multiple SOHs in the first data interval and the W1 value corresponding to the degradation regression curve of the battery generated based on multiple SOHs in the second data interval is used to determine whether the battery has experienced rapid degradation.

Images