Battery analysis system, battery analysis method, battery analysis program, and storage medium on which battery analysis program is recorded

Abstract

A minimum SOH calculation unit calculates a minimum SOH among SOHs of a plurality of unit cells or unit cell blocks included in a battery pack system on the basis of battery data. The SOC difference calculation unit calculates the SOC difference on the basis of the maximum SOC and the minimum SOC among the SOCs of a plurality of unit cells or unit cell blocks included in the battery pack system, and the minimum SOH of the battery pack system, on the basis of the battery data. The statistical processing unit extracts at least one change point from the sample group of the SOC difference, calculates a regression model of the SOC difference for each section divided by the change point, calculates a regression model of the minimum SOH on the basis of the sample group of the minimum SOH, and calculates a regression model of the SOH of the assembled battery system on the basis of the regression model of the minimum SOH and the regression model of the SOC difference of each section.

Description

Technical Field

[0001] The present disclosure relates to a battery analysis system, a battery analysis method, and a battery analysis program for estimating the SOH (State Of Health) of a battery pack system. Background Art

[0002] In a battery pack system in which battery cells are connected in multiple parallel and series configurations, it is required to estimate not only the SOH of each cell but also the SOH (hereinafter, appropriately referred to as system SOH) of the entire battery pack system and predict the system SOH. When estimating and predicting the system SOH, it is necessary to consider the SOH difference and SOC (State Of Charge) difference between the series-connected cell blocks. In this specification, a cell block refers to a cell block formed by connecting multiple cells in parallel.

[0003] The SOC difference between the cell blocks is caused by the difference in leakage current between the cell blocks. Specifically, due to the difference in power consumption between the cell blocks, the presence or absence of micro-shorts inside the cells, equalization processing, etc., the SOC difference between the cell blocks changes. The difference in power consumption and the presence or absence of micro-shorts are factors that expand the SOC difference between the cell blocks, and equalization processing is a factor that reduces the SOC difference between the cell blocks.

[0004] Patent Document 1 discloses a method for estimating the system SOH by setting the capacity difference from the highest SOC at the end of discharge as the SOC difference. However, the deviation in SOH between the cells is not considered, and in the case of a large deviation in SOH, the estimation accuracy of the system SOH sometimes decreases. Patent Document 2 discloses a method for detecting the sharp deterioration point of a cell and predicting the SOH by linear regression after the sharp deterioration. However, the deviation in SOC between the cells in the case of systemization is not considered. When estimating the system SOH in a state where the deviation in SOC is large, the estimation accuracy sometimes decreases.

[0005] Prior Art Documents

[0006] Patent Documents

[0007] Patent Document 1: Japanese Patent Application Laid-Open No. 2019-113414

[0008] Patent Document 2: International Publication No. 17 / 098686 Summary of the Invention

[0009] Problems to be Solved by the Invention

[0010] The present disclosure has been made in view of such circumstances, and an object thereof is to provide a technique for accurately estimating the SOH of a battery pack system.

[0011] To solve the above problems, a battery analysis system according to a certain aspect of the present disclosure includes: a data acquisition unit that acquires battery data including at least the voltage and current of each single battery or each single battery block among the single batteries connected in series or the single battery blocks formed by connecting multiple single batteries in parallel included in a battery pack system; a minimum SOH calculation unit that calculates the minimum SOH among the SOHs of the multiple single batteries or single battery blocks included in the battery pack system based on the battery data; a SOC difference calculation unit that calculates a SOC difference based on the battery data according to the maximum SOC and the minimum SOC among the SOCs of the multiple single batteries or single battery blocks based on at least the voltage of the multiple single batteries or single battery blocks included in the battery pack system, and the minimum SOH of the battery pack system; and a statistical processing unit that extracts at least one change point from a sample group of the SOC differences, calculates a regression model of the SOC differences for each interval obtained by dividing with the change point, calculates a regression model of the minimum SOH based on the sample group of the minimum SOH, and calculates a regression model of the SOH of the battery pack system based on the regression model of the minimum SOH and the regression models of the SOC differences for each of the intervals.

[0012] In addition, any combination of the above components, and a mode obtained by converting the expression of the present disclosure between devices, systems, methods, computer programs, recording media, etc. is also effective as a mode of the present disclosure.

[0013] According to the present disclosure, it is possible to accurately estimate the SOH of a battery pack system. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 is a diagram for explaining the battery analysis system according to the embodiment.

[0015] Figure 2A is a diagram showing a structural example of a battery pack system including a plurality of single batteries connected in series.

[0016] Figure 2B is a diagram showing a structural example of a battery pack system including a plurality of single battery blocks connected in series and each single battery block including a plurality of single batteries connected in parallel.

[0017] Figure 3 is a diagram for explaining the SOC difference according to the embodiment.

[0018] Figure 4 is a diagram showing an example of a chart obtained by performing linear regression on the time series data of the minimum SOH.

[0019] Figure 5It is a diagram showing an example of a graph of the system SOH regression curve derived by showing the regression line obtained by subtracting the SOC difference from the regression curve of the minimum SOH.

[0020] Figure 6 It is a flowchart showing the process of the calculation method of the system SOH related to the embodiment.

[0021] Figure 7 It is a flowchart showing the process of an example algorithm of the change point extraction process.

[0022] Figure 8A It is a diagram showing a sample group of the time series data of the SOC difference.

[0023] Figure 8B It is a diagram showing a sample group of the raw data of the SOC difference and a sample group of the smoothed data of the SOC difference.

[0024] Figure 8C It is based on Figure 8B the smoothed data of the SOC difference to draw the interval linear regression line and obtain the diagram.

[0025] Figure 9A It is a diagram showing a sample group of the time series data of the SOC difference.

[0026] Figure 9B It is a diagram showing a sample group of the raw data of the SOC difference and a sample group of the smoothed data of the SOC difference.

[0027] Figure 9C It is based on Figure 9B the filtered data of the SOC difference to draw the interval linear regression line and obtain the diagram.

[0028] Figure 10A It is a diagram showing a sample group of the time series data of the minimum SOH and the linear regression curve based on the sample group of the minimum SOH.

[0029] Figure 10B It is a diagram showing the regression curve of the system SOH related to the embodiment.

[0030] Figure 11A It is a diagram showing a sample group of the time series data of the minimum SOH and the linear regression curve based on the sample group of the minimum SOH.

[0031] Figure 11B It is a diagram showing the regression curve of the system SOH related to the embodiment. Detailed implementation mode

[0032] Figure 1This is a diagram for explaining the battery analysis system 10 related to the embodiment. The battery analysis system 10 is a system for estimating the SOH of the battery system 21 included in the battery pack mounted on the electric vehicle 20. The battery analysis system 10 can also predict the future SOH trend of the battery system 21, and thus can also provide the user with a rough target including the replacement time of the battery pack containing the battery system 21.

[0033] The battery analysis system 10 can be built, for example, on the company's own server set up in the company's own facility or data center that provides analysis services for the battery pack mounted on the electric vehicle 20. In addition, the battery analysis system 10 can also be built on a cloud server utilized based on cloud services. In addition, the battery analysis system 10 can also be built on multiple servers dispersedly set up at multiple sites (data centers, company facilities). These multiple servers can be any one of a combination of multiple company's own servers, a combination of multiple cloud servers, and a combination of a company's own server and a cloud server.

[0034] The battery system 21 included in the battery pack mounted on the electric vehicle 20 supplies power to a drive motor (not shown). The battery system 21 includes a plurality of single cells or a plurality of single cell blocks connected in series.

[0035] Figure 2A 、 Figure 2B This is a diagram showing a structural example of the battery system 21. Figure 2A The shown battery system 21 includes a plurality of single cells E1-Em connected in series. Figure 2B The shown battery system 21 includes a plurality of single cell blocks Eb1-Ebm connected in series. Each single cell block Eb1-Ebm includes a plurality of single cells E1a~E1n-Ema~Emn connected in parallel.

[0036] The single cell can use a lithium-ion battery cell, a nickel-metal hydride battery cell, a lead battery cell, etc. Hereinafter, an example of using a lithium-ion battery cell (nominal voltage: 3.6V - 3.7V) is assumed in this specification. The number of series-connected single cells or single cell blocks is determined according to the voltage of the drive motor.

[0037] The voltage sensors 22 respectively detect the voltages at both ends of the single cells or single cell blocks connected in series. The plurality of single cells or single cell blocks connected in series are connected in series with a shunt resistor. The current sensor 23 detects the current flowing through the single cells or single cell blocks connected in series based on the voltage at both ends of the shunt resistor. In addition, a Hall element can also be used instead of the shunt resistor. A plurality of temperature sensors 24 are provided in the battery pack including the battery system 21. The temperature sensor 24 can use a thermistor, for example. Regarding the temperature sensor 24, for example, one temperature sensor 24 can be provided for 6 to 8 single cells or single cell blocks.

[0038] The control unit 25 is composed of a BMU (Battery Management Unit) and an ECU (Electronic Control Unit) working in cooperation. The BMU combines the OCV (Open Circuit Voltage) method and the current integration method to estimate the SOC. The OCV method is a method for estimating the SOC based on the measured OCV of a single cell and the SOC-OCV curve of the single cell. The SOC-OCV curve of a single cell is pre-made based on the characteristic tests conducted by the battery manufacturer and registered in the BMU at the time of factory shipment.

[0039] The current integration method is a method for estimating the SOC based on the OCV at the start of charge / discharge of a single cell and the integrated value of the measured current. In the current integration method, the measurement error of the current accumulates continuously as the charge / discharge time becomes longer. 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.

[0040] The BMU periodically (for example, at 10-second intervals) sends battery data including the voltage, current, temperature, and SOC of multiple single cells or single cell blocks to the ECU via the in-vehicle network, whereby the ECU samples the battery data in time series. As the in-vehicle network, for example, CAN (Controller Area Network) or LIN (Local Interconnect Network) can be used.

[0041] The communication unit 26 has a function of performing communication signal processing with the communication unit 33 of the charging pile 30 and a function of performing wireless signal processing for connecting to the network 5. The communication unit 26 can access the network 5 using, for example, a mobile phone network (cellular network), wireless LAN, V2I (Vehicle to Infrastructure), V2V (Vehicle to Vehicle), ETC system (Electronic Toll Collection System), or DSRC (Dedicated Short Range Communications).

[0042] The network 5 is a general term for communication paths such as the Internet, dedicated lines, and VPN (Virtual Private Network), and there are no restrictions on its communication media and protocols. As the communication medium, for example, a mobile phone network, wireless LAN, wired LAN, optical fiber network, ADSL network, CATV network, etc. can be used. As the communication protocol, for example, TCP (Transmission Control Protocol) / IP (Internet Protocol), UDP (User Datagram Protocol) / IP, Ethernet (registered trademark), etc. can be used.

[0043] The ECU can send the sampled battery data to the battery analysis system 10 each time, or can accumulate the sampled battery data in the internal memory and send the battery data accumulated in the memory to the battery analysis system 10 at a specified time. In addition, in a state where the electric vehicle 20 and the charging pile 30 are connected by a charging cable, the ECU can also send the battery data accumulated in the memory to the battery analysis system 10 via the charging pile 30.

[0044] By connecting the electric vehicle 20 to the charging pile 30 using a charging cable, the battery pack system 21 inside the electric vehicle 20 can be charged from the outside. The charging pile 30 is connected to the commercial power system 2 to charge the battery pack system 21.

[0045] Generally speaking, in the case of normal charging, charging is performed with alternating current, and in the case of rapid charging, charging is performed with direct current. In the case of charging with alternating current (for example, single-phase 100 / 200V), the charging voltage or charging current is controlled by a charger (not shown) inside the electric vehicle 20. In the case of charging with direct current, the power supply unit 31 of the charging pile 30 controls the charging voltage or charging current. The power supply unit 31 includes a rectifier circuit, a filter, and a DC / DC converter. The alternating current power supplied from the commercial power system 2 is full-wave rectified by the rectifier circuit and smoothed by the filter, thereby generating direct current power. The DC / DC converter controls the voltage or current of the generated direct current power.

[0046] As a rapid charging standard, for example, CHAdeMO (registered trademark), ChaoJi, GB / T, Combo (Combined Charging System) can be used. In CHAdeMO, ChaoJi, and GB / T, CAN is adopted as the communication method. In Combo, PLC (Power Line Communication) is adopted as the communication method.

[0047] In a charging cable adopting the CAN method, in addition to power lines, communication lines are also included. When the electric vehicle 20 is connected to the charging pile 30 through this charging cable, a communication channel is established between the control unit 25 of the electric vehicle 20 and the control unit 32 of the charging pile 30. In addition, in a charging cable adopting the PLC method, communication signals are superimposed on the power lines for transmission.

[0048] The communication unit 33 of the charging pile 30 has a function of performing communication signal processing between the communication unit 26 of the electric vehicle 20 and a function of performing signal processing for connecting to the network 5. The communication unit 33 can access the network 5 using, for example, a wired LAN, a wireless LAN, or a mobile phone network.

[0049] The battery analysis system 10 includes a control unit 11, a storage unit 12, and a communication unit 13. The communication unit 13 is a communication interface (such as a NIC: Network Interface Card) for connecting to the network 5 by wire or wirelessly.

[0050] The control unit 11 includes a data acquisition unit 111, a minimum SOH calculation unit 112, an SOC difference calculation unit 113, a statistical processing unit 114, and an SOH prediction unit 115. The functions of the control unit 11 can be realized through the cooperation of hardware resources and software resources or only through hardware resources. As hardware resources, a CPU, a ROM, a RAM, a GPU (Graphics Processing Unit), an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array), and other LSIs can be used. As software resources, programs such as an operating system and application programs can be used.

[0051] The storage unit 12 includes non-volatile recording media such as an HDD and an SSD, and is used to store various data. The storage unit 12 includes a battery data holding unit 121. The data acquisition unit 111 acquires battery data from the electric vehicle 20 or the charging pile 30, and accumulates the acquired battery data in the battery data holding unit 121. In addition, the above programs can also be recorded on a recording medium. If this recording medium is used, for example, the above programs can be installed on the above computer. Here, the recording medium on which the above programs are recorded can also be a non-transitory recording medium. There is no particular limitation on the non-transitory recording medium. For example, it can be a recording medium such as a CD-ROM.

[0052] The minimum SOH calculation unit 112 calculates the SOH of each of the plurality of single cells or cell blocks included in the battery pack system 21. For example, the minimum SOH calculation unit 112 calculates the difference (DOD: Depth Of Discharge) between the SOC corresponding to the OCV at the start of charging and the SOC corresponding to the OCV at the end of charging, and calculates the cumulative current value ΣI during charging.

[0053] The minimum SOH calculation unit 112 estimates the FCC (Full Charge Capacity) based on the depth of discharge DOD and the cumulative current value ΣI as shown in the following (Equation 1). The minimum SOH calculation unit 112 estimates the SOH based on the estimated FCC and the initial FCC as shown in the following (Equation 2). The SOH is defined by the ratio of the current FCC to the initial FCC, and the lower the value of the SOH (the closer to 0%), the more progress the deterioration has made.

[0054] FCC = ΣI / DOD ··· (Equation 1)

[0055] SOH = current FCC / initial FCC ··· (Equation 2)

[0056] The minimum SOH calculation unit 112 determines the SOH with the largest DOD among the SOHs of the plurality of single cells or cell blocks included in the battery pack system 21 as the minimum SOH of the battery pack system 21.

[0057] In the present embodiment, the SOH obtained by subtracting the SOH corresponding to the SOC difference of the battery pack system 21 from the minimum SOH is defined as the system SOH (see the following (Equation 3)).

[0058] System SOH = minimum SOH - (ΔSOC × minimum SOH) ··· (Equation 3)

[0059] The SOC difference calculation unit 113 calculates the SOC difference of the battery pack system 21 based on the maximum SOC and the minimum SOC among the SOCs of the multiple single cells or cell blocks included in the battery pack system 21, and the minimum SOH of the battery pack system 21. Specifically, the SOC difference calculation unit 113 calculates the difference between the SOC of the single cell or cell block with the maximum SOC (hereinafter referred to as the maximum SOC single cell) and the SOC of the single cell or cell block with the minimum SOH (hereinafter referred to as the minimum SOH single cell) when the SOC of the minimum SOH single cell is 100%, as the upper SOC difference. The SOC difference calculation unit 113 calculates the difference between the SOC of the single cell or cell block with the minimum SOC (hereinafter referred to as the minimum SOC single cell) and the SOC of the single cell or cell block with the minimum SOH when the SOC of the minimum SOH single cell is 0%, as the lower SOC difference. The SOC difference calculation unit 113 calculates the SOC difference of the battery pack system 21 by summing the upper SOC difference and the lower SOC difference.

[0060] Figure 3 This is a diagram for explaining the SOC difference according to the embodiment. For simplicity, a battery pack system 21 formed by connecting three single cells in series is considered. In Figure 3 this, the DOD of the first single cell with SOC change from the top is 70%, the DOD of the second single cell with SOC change from the top is 85%, the DOD of the third single cell with SOC change from the top is 70%, and the second single cell with SOC change becomes the minimum SOH single cell. The first single cell with SOC change becomes the maximum SOC single cell, and the third single cell with SOC change becomes the minimum SOC single cell.

[0061] After the start of discharge, the discharge stops at the time point when the SOC of the minimum SOC single cell becomes 0%. In the present embodiment, the SOC change of the minimum SOH single cell and the SOC change of the minimum SOC single cell are extended downward, and the SOC difference between the SOC of the minimum SOH single cell and the SOC of the minimum SOC single cell at the time point when the SOC of the minimum SOH single cell reaches 0% is calculated as the lower SOC difference. In Figure 3 the example shown, this SOC difference is 8%. In addition, the SOC difference between the SOC of the minimum SOH single cell and the SOC of the minimum SOC single cell at the time point of discharge stop can be simply set as the lower SOC difference.

[0062] After charging starts, charging stops at the time point when the SOC of the maximum SOC single battery reaches 100%. In this embodiment, the SOC change of the minimum SOH single battery and the SOC change of the maximum SOC single battery are extended upward, and the SOC difference between the SOC of the minimum SOH single battery and the SOC of the maximum SOC single battery at the time point when the SOC of the minimum SOH single battery reaches 100% is calculated as the upper SOC difference. In Figure 3 In the example shown, this SOC difference is 4%. In addition, the SOC difference between the SOC of the minimum SOH single battery and the SOC of the maximum SOC single battery at the time point of charging stop can also be simply set as the upper SOC difference.

[0063] In Figure 3 In the example shown, the SOC difference is 12%. Substitute 12% into ΔSOC in the above (Equation 3) to calculate the system SOH.

[0064] The minimum SOH calculation unit 112 and the SOC difference calculation unit 113 periodically (for example, once a day, once a week) calculate the minimum SOH and the SOC difference of each battery system 21 and accumulate them in the battery data holding unit 121.

[0065] When the statistical processing unit 114 estimates the system SOH of the target battery system 21, it reads the time series data of the minimum SOH and the SOC difference of the battery system 21 from the battery data holding unit 121.

[0066] Figure 4 FIG. is an example of a graph obtained by performing linear regression on the time series data of the minimum SOH. The horizontal axis represents the date, and the vertical axis represents SOH [%]. It is known that the deterioration of a battery generally progresses in proportion to the square root of the elapsed time (square root law, 0.5 - power law), and the deterioration of the battery is obtained by linear regression (linear approximation) with the square root of the elapsed time as the explanatory variable.

[0067] Figure 5 FIG. is an example of a graph showing a regression curve of the system SOH derived from a regression line obtained by subtracting the regression curve of the SOC difference from the regression curve of the minimum SOH. The sample values of the minimum SOH and the regression curve of the minimum SOH are the same as Figure 4 The regression curve of the system SOH plotted continuously with "*" is the regression curve of the system SOH related to the comparative example. The regression curve of the system SOH related to the comparative example is a regression line derived from the regression line of the entire interval of the time series data of the SOC difference subtracted from the regression curve of the minimum SOH.

[0068] The regression curve of the system SOH drawn in thick line is the regression curve of the system SOH involved in the embodiment. The regression curve of the system SOH involved in the embodiment is a regression line derived by subtracting the regression lines of each section of the time series data of the SOC difference from the regression curve of the minimum SOH. As described above, the SOC difference varies greatly due to the difference in power consumption, the presence or absence of micro-short circuits inside the single cells, the balance of the single cells, etc. These events have the characteristic of linearly changing the SOC difference, and the tendency and characteristics of the SOC difference change greatly before and after the occurrence of the events. Therefore, when dividing the SOC difference data at the change points where the tendency and characteristics of the SOC difference change and deriving the linear regression lines for each section, the estimation accuracy of the SOC difference is improved.

[0069] As described in the above (Equation 3), the larger the SOC difference, the more the system SOH decreases, and the smaller the SOC difference, the more the system SOH recovers. As Figure 5 shown, when the SOC difference becomes small due to the balance of the single cells, the system SOH recovers.

[0070] The dotted line part of the regression curve of the system SOH drawn in thick line is a prediction line obtained by extending the regression line of the most recent section of the system SOH. In the embodiment, the future system SOH reflecting the most recent state of the battery pack system 21 can be predicted. Next, a method for calculating the system SOH involved in the embodiment will be described.

[0071] Figure 6 is a flowchart showing the process of the method for calculating the system SOH involved in the embodiment. The statistical processing unit 114 extracts the number of days elapsed and the SOC difference (S10) from the battery data of the battery pack system 21 to be the object. For easy calculation, the date is converted to the number of days elapsed since the initial day of the time series data.

[0072] The statistical processing unit 114 extracts at least one change point from the sample group of the SOC difference (S20). As a preprocessing for extracting the change point, the statistical processing unit 114 applies a filter to the time series data of the SOC difference to smooth it in order to reduce the deviation of the SOC difference. The statistical processing unit 114, for example, calculates the moving average of five markers of the SOC difference and replaces the sample value of each SOC difference with the smoothed SOC difference.

[0073] In addition, in order to shorten the calculation time, in the case of a large number of markers, the statistical processing unit 114 can extract only the SOC difference during the relaxation period after charging, or can extract only the SOC difference of the samples with a large weight in the SOH estimation.

[0074] Figure 7It is a flowchart showing an algorithm example of a change point extraction process. The statistical processing unit 114 calculates the sum of squared residuals (RSS: Residual Sum of Squares) of the regression lines for all combinations of the tag groups (S21). The total number of combinations is ΣNCn (N is the number of tags for the SOC difference, 2 ≤ n ≤ N). When calculating the sum of squared residuals for all combinations independently by the least squares method, the calculation cost becomes large, so the recursive least squares method can also be used to calculate each sum of squared residuals.

[0075] The statistical processing unit 114 calculates the segmentation point (x) of the combination with the minimum total sum of squared residuals at each number of segments while changing the number of segments, and the total sum of squared residuals at this time (S22). The statistical processing unit 114 applies the total sum of squared residuals for each number of segments to the Gaussian error model of the following (Equation 4) to calculate the Bayesian Information Criterion (BIC: Bayesian Information Criterion).

[0076] BIC = Nlog e (RSS / N) + klog e N ··· (Equation 4)

[0077] N is the number of tags for the SOC difference, RSS is the total sum of squared residuals for each number of segments, and k is the number of segments.

[0078] The statistical processing unit 114 determines the number of segments with the minimum Bayesian Information Criterion (S23). Return to Figure 6 . The statistical processing unit 114 divides the number of elapsed days at the position (change point) where the total sum of squared residuals is the minimum according to the calculated number of segments, and calculates the simple linear regression model of the SOC difference for each segmented interval (S30).

[0079] The statistical processing unit 114 derives the linear regression function shown in the following (Equation 5) for each interval divided by the change point.

[0080] f n (x) = a n *(x - X n-1 ) + b ··· (Equation 5)

[0081] b = f n-1 (X n-1 )

[0082] x is the number of elapsed days, n is the interval, and X n is the x value at the end of the linear regression function of interval n.

[0083] Set the limit condition of the intercept b to 0 ≤ b ≤ (f n [0] + 1).

[0084] Set the limit condition of the slope a n as -10 ≤ a n ≤ 1. A negative value indicates a narrowing of the SOC difference. Sometimes the SOC difference rapidly narrows due to cell balancing, so the minimum value is set to -10%. The maximum value represents the maximum SOC difference that expands over a unit number of days. The maximum SOC difference depends on the cell model and the structure of the battery pack system. In this example, the maximum expansion rate of the SOC difference is set to 1%.

[0085] Set the limit condition of the elapsed number of days X n at the end of the interval n n-1 as X n ≤ X n+1 .

[0086] Figure 8A , Figure 8B , Figure 8C are diagrams for explaining the derivation example of the piecewise linear regression function of the SOC difference of the cells involved in Specific Example 1. The horizontal axis represents the elapsed number of days [days], and the vertical axis represents the SOC difference [%]. Figure 8A is a chart showing a sample group of the time series data of the SOC difference. Figure 8B is a chart showing a sample group of the raw data of the SOC difference and a sample group of the smoothed data of the SOC difference. The raw data of the SOC difference is represented by black markers, and the smoothed data of the SOC difference is represented by gray markers. As the smoothing filter, a moving average of five markers is applied. Figure 8C is a chart obtained by plotting the piecewise linear regression line based on Figure 8B the smoothed data of the SOC difference. In the case of the cells involved in Specific Example 1, it is divided into two intervals. The piecewise linear regression function of each interval is represented by the following (Equation 6) and (Equation 7).

[0087] f1(x) = a1*(x) + b ··· (Equation 6)

[0088] f2(x) = a2*(x - X1) + f1(X1) ··· (Equation 7)

[0089] Figure 9A , Figure 9B , Figure 9C are diagrams for explaining the derivation example of the piecewise linear regression function of the SOC difference of the cells involved in Specific Example 2. The horizontal axis represents the elapsed number of days [days], and the vertical axis represents the SOC difference [%]. Figure 9A is a chart showing a sample group of the time series data of the SOC difference. Figure 9BIt is a graph showing a sample group of raw data of the SOC difference and a sample group of smoothed data of the SOC difference. The raw data of the SOC difference is represented by black markers, and the smoothed data of the SOC difference is represented by gray markers. A moving average of five markers is applied to the filter. Figure 9C is based on Figure 9B the filtered data of the SOC difference to draw an interval linear regression line. In the case of the single cell involved in Specific Example 2, it is divided into five intervals.

[0090] Return to Figure 6 The statistical processing unit 114 reads out the sample group with the minimum SOH from the battery data of the target group battery system 21 and calculates the regression model. As described above, it is known that the deterioration of a single cell progresses in proportion to the square root of the elapsed time (square root law, 0.5 - power law). In the present embodiment, the statistical processing unit 114 calculates a linear regression model of the 0.5 - power of the number of elapsed days shown in the following (Equation 8).

[0091] Minimum SOH = w0 + w1√t ··· (Equation 8)

[0092] w0 is the initial value, w1 is the deterioration coefficient, and t is the number of elapsed days.

[0093] 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. When the nominal value is set to the minimum guaranteed amount and is set lower than the actual initial capacity, w0 is set to a value greater than 1.0.

[0094] In addition, a linear regression model using explanatory variables of other powers such as the 0.4 - power, 0.6 - power, and 1 - power of the elapsed days can also be used according to the model of the single cell.

[0095] The statistical processing unit 114 calculates a regression model of the system SOH based on the linear regression model of the minimum SOH and the linear regression model of each interval (S40). Specifically, as shown in the above (Equation 3), the statistical processing unit 114 applies the interval linear regression model to the linear regression model of the minimum SOH, and converts the influence degree of the SOC difference of the group battery system 21 on the system SOH to the SOH based on the single cell with the minimum SOH. At this time, the statistical processing unit 114 interpolates a straight line from the point where SOH = 100% to the starting point of the calculated interval linear regression model. The SOC difference at the point where SOH = 100% of the interpolated straight line is set to 0%. The statistical processing unit 114 calculates the regression model of the system SOH by subtracting the regression model representing the influence degree of the SOC difference on the system SOH from the linear regression model of the minimum SOH.

[0096] The SOH prediction unit 115 predicts the system SOH at a specific future time point based on the regression model of the system SOH. In addition, the minimum value of the SOC difference is set to 0%.

[0097] Figure 10A , Figure 10B is a diagram for explaining a calculation example of the regression model of the minimum SOH of the single battery and the regression model of the system SOH according to the specific example 1. The horizontal axis represents the date, and the vertical axis represents the SOH [%]. Figure 10A is a chart showing a sample group of time series data of the minimum SOH and a linear regression curve based on the sample group of the minimum SOH.

[0098] Figure 10B is a chart showing the regression curve of the system SOH according to the embodiment. Figure 10B The regression curve of the system SOH shown is a regression curve generated based on the linear regression model of the minimum SOH and the linear regression model of each interval. In addition, a data group of the system SOH calculated by the above (Equation 3) based on the data of the minimum SOH and the SOC difference for each date is also included in the chart.

[0099] As the prediction line extending in the future direction from the regression curve of the system SOH, a line (LINEAR) obtained by directly extending the regression curve of the most recent interval according to the regression model can be used, or a line (CLIP) truncated in the middle can be used. The truncation point can be a time point after a specified period has elapsed from the time point where the last sample value of the system SOH exists, or a location of the SOH value after the SOH value has decreased by a specified amount with respect to the last sample value (SOH value) of the system SOH.

[0100] The SOC difference is set to 0 from the truncation point. Therefore, from the truncation point, the prediction model of the system SOH is consistent with the prediction model of the minimum SOH. By truncating the SOC difference at a time point when the prediction period exceeds a specified period, it is possible to limit the period during which the tendency of the current SOC difference is predicted to continue, and to generate a prediction model considering the possibility of a change in the new tendency of the occurrence of the SOC difference.

[0101] Figure 11A , Figure 10B is a diagram for explaining a calculation example of the regression model of the minimum SOH of the single battery and the regression model of the system SOH according to the specific example 2. The horizontal axis represents the date, and the vertical axis represents the SOH [%]. Figure 11A is a chart showing a sample group of time series data of the minimum SOH and a linear regression curve based on the sample group of the minimum SOH.

[0102] Figure 11B is a chart showing the regression curve of the system SOH according to the embodiment. Figure 11BThe regression curve of the system SOH shown is a regression curve generated based on a linear regression model of the minimum SOH and linear regression models for each interval. The prediction line extending from the regression curve of the system SOH in the future direction can be a line (LINEAR) obtained by directly extending the regression curve of the most recent interval according to the regression model, or a line (CLIP) truncated midway.

[0103] As described above, according to the present embodiment, by performing linear regression on the SOC differences between the multiple single cells or single cell blocks constituting the battery pack system 21 for each interval, the current system SOH can be estimated with high accuracy, and the future system SOH can be predicted with high accuracy. By performing linear regression on the SOC differences for each interval, the behavior changes of the SOC differences caused by the loss of balance of single cells, the restoration of the balance of single cells based on equalization processing, the occurrence of internal short circuits, etc. can be modeled with high accuracy.

[0104] The present disclosure has been described based on the embodiments. The embodiments are illustrative, and those skilled in the art can understand that various modifications can be made to the combination of these respective components and processing steps, and such modification examples are also within the scope of the present disclosure.

[0105] In the above embodiment, the statistical processing unit 114 calculated the simple regression model for each interval with the elapsed time as the explanatory variable and the SOC difference as the target variable. In a modification example, it may also be that the statistical processing unit 114 uses the elapsed time and the temperature difference between the maximum temperature and the minimum temperature in the battery pack including the battery pack system 21 as the explanatory variables, and the SOC difference as the target variable, to calculate the multiple regression model for each interval. The SOC difference is also affected by the deviation of the temperature in the battery pack. There is a tendency that the larger the deviation of the temperature in the battery pack, the larger the SOC difference. In this modification example, by taking into account the deviation of the temperature in the battery pack, the SOC difference can be modeled with higher accuracy.

[0106] In the case where the number of series connections of the battery pack system 21 is large, it may also be that the maximum voltage single cell and the minimum voltage single cell are simply determined from the multiple single cells or single cell blocks constituting the battery pack system 21, and regarded as two sets of systems to calculate the minimum SOH and the SOC difference. In this case, the ΔSOC in the above (Equation 3) can directly use the SOC difference between the maximum voltage single cell and the minimum voltage single cell. In this modification example, as the voltage data included in the battery data sent from the electric vehicle 20 to the battery analysis system 10, it is only necessary to include the maximum voltage and the minimum voltage of the multiple single cells or single cell blocks in the battery data, thereby reducing the data volume.

[0107] In addition, when the battery analysis system 10 obtains battery data including voltage data of all series-connected cells, the minimum SOH calculation unit 112 and the SOC difference calculation unit 113 can calculate the minimum SOH and the SOC difference by extracting only the battery data of the maximum voltage single cell and the minimum voltage single cell from the battery data of multiple single cells or single cell blocks. In this case, the amount of calculation can be reduced.

[0108] Figure 7 The change point extraction algorithm shown is an example, and other change point extraction algorithms can also be used. In addition, the person in charge of analysis can also visually divide the intervals of the time series data of the SOC difference.

[0109] In the above embodiment, as the electric vehicle 20, a four-wheeled electric vehicle is assumed. In this regard, it can also be an electric motorcycle (scooter), an electric bicycle, or an electric kick scooter. In addition, the electric vehicle includes not only full-specification electric vehicles but also low-speed electric vehicles such as golf carts and electric scooters. In addition, the device equipped with the battery pack system 21 is not limited to the electric vehicle 20. Devices equipped with the battery pack system 21 also include electric moving bodies such as electric ships, railway vehicles, and multi-rotor aircraft (drones), stationary energy storage systems, and civilian electronic devices (smartphones, laptop PCs, etc.).

[0110] In addition, the embodiment can also be determined by the following items.

[0111] [Item 1]

[0112] A battery analysis system (10), characterized by comprising:

[0113] A data acquisition unit (111) that acquires battery data including at least the voltage and current of each single cell (E1-Em) or each single cell block (Eb1-Ebm) of the series-connected single cells (E1-Em) included in the battery pack system (21) or the single cell blocks (Eb1-Ebm) formed by parallel connection of multiple single cells;

[0114] A minimum SOH calculation unit (112) that calculates the minimum SOH in the SOH (State Of Health) of the multiple single cells (E1-Em) or single cell blocks (Eb1-Ebm) included in the battery pack system (21) based on the battery data;

[0115] The SOC difference calculation unit (113) calculates the SOC difference based on the battery data, according to the maximum SOC and the minimum SOC among the SOCs (State Of Charge) of the plurality of single cells (E1-Em) or cell blocks (Eb1-Ebm) based on at least the voltage of the plurality of single cells (E1-Em) or cell blocks (Eb1-Ebm) included in the battery system (21), and the minimum SOH of the battery system (21); and

[0116] The statistical processing unit (114) extracts at least one change point from the sample group of the SOC differences, calculates a regression model of the SOC differences for each interval obtained by dividing at the change points, calculates a regression model of the minimum SOH based on the sample group of the minimum SOH, and calculates a regression model of the SOH of the battery system (21) based on the regression model of the minimum SOH and the regression models of the SOC differences for the respective intervals.

[0117] Accordingly, the SOH of the battery system (21) can be estimated with high accuracy.

[0118] [Item 2]

[0119] The battery analysis system (10) according to Item 1, characterized in that

[0120] It further includes an SOH prediction unit (115), and the SOH prediction unit (115) predicts the SOH of the battery system (21) at a specific future time point based on the regression model of the SOH of the battery system (21).

[0121] Accordingly, the future SOH of the battery system (21) can be predicted with high accuracy.

[0122] [Item 3]

[0123] The battery analysis system (10) according to Item 1, wherein

[0124] The statistical processing unit (114) calculates a simple regression model for each interval with the elapsed time as an explanatory variable and the SOC difference as a target variable.

[0125] Accordingly, the behavior of the SOC difference can be modeled with high accuracy.

[0126] [Item 4]

[0127] The battery analysis system (10) according to Item 1, wherein

[0128] The data acquisition unit (111) acquires battery data including the voltage, current, and temperature of each single battery (E1-Em) or each single battery block (Eb1-Ebm).

[0129] The statistical processing unit (114) calculates a multiple regression model for each of the intervals, using the elapsed time and the temperature difference between the maximum temperature and the minimum temperature in the battery pack including the battery system (21) as explanatory variables and the SOC difference as the target variable.

[0130] Accordingly, the behavior of the SOC difference can be modeled with higher accuracy.

[0131] [Item 5]

[0132] The battery analysis system (10) according to any one of Items 1 to 4, characterized in that

[0133] The SOC difference calculation unit (113) calculates the SOC difference for each sample by summing the SOC difference between when the SOC of the single battery (E1-Em) or single battery block (Eb1-Ebm) with the minimum SOH is 100 and the SOC of the single battery (E1-Em) or single battery block (Eb1-Ebm) with the maximum SOC, and the SOC difference between when the SOC of the single battery (E1-Em) or single battery block (Eb1-Ebm) with the minimum SOH is 0 and the SOC of the single battery (E1-Em) or single battery block (Eb1-Ebm) with the minimum SOC.

[0134] Accordingly, the SOC difference of the battery system (21) that affects the SOH of the battery system (21) can be precisely defined.

[0135] [Item 6]

[0136] The battery analysis system (10) according to any one of Items 1 to 4, characterized in that

[0137] The minimum SOH calculation unit (112) calculates the minimum SOH based on the two single batteries or single battery blocks with the maximum voltage and the minimum voltage among the multiple single batteries (E1-Em) or single battery blocks (Eb1-Ebm) included in the battery system (21).

[0138] The SOC difference calculation unit (113) calculates the SOC difference based on the SOC of the two single batteries or single battery blocks and the minimum SOH of the battery system (21).

[0139] Accordingly, the amount of data or the amount of computation can be reduced.

[0140] [Item 7]

[0141] A battery analysis method, characterized by comprising the following steps:

[0142] Obtain battery data including at least the voltage and current of each single battery (E1-Em) or each single battery block (Eb1-Ebm) formed by connecting multiple single batteries in parallel, where the single batteries (E1-Em) or the single battery blocks (Eb1-Ebm) are connected in series and included in the battery system (21);

[0143] Based on the battery data, calculate the minimum SOH (State Of Health) among the SOHs of the multiple single batteries (E1-Em) or single battery blocks (Eb1-Ebm) included in the battery system (21);

[0144] Based on the battery data, calculate the SOC difference according to the maximum SOC and the minimum SOC among the SOCs (State Of Charge) of the multiple single batteries (E1-Em) or single battery blocks (Eb1-Ebm) based on at least the voltage of the multiple single batteries (E1-Em) or single battery blocks (Eb1-Ebm) included in the battery system (21), and the minimum SOH of the battery system (21); and

[0145] Extract at least one change point from the sample group of the SOC difference, calculate the regression model of the SOC difference for each interval obtained by dividing with the change point, calculate the regression model of the minimum SOH based on the sample group of the minimum SOH, and calculate the regression model of the SOH of the battery system (21) based on the regression model of the minimum SOH and the regression models of the SOC differences of the respective intervals.

[0146] Accordingly, the SOH of the battery system (21) can be estimated with high precision.

[0147] [Item 8]

[0148] A battery analysis program, characterized by causing a computer to perform the following processing:

[0149] Obtain battery data including at least the voltage and current of each single battery (E1-Em) or each single battery block (Eb1-Ebm) formed by connecting multiple single batteries in parallel, where the single batteries (E1-Em) or the single battery blocks (Eb1-Ebm) are connected in series and included in the battery system (21);

[0150] Based on the battery data, calculate the minimum SOH (State Of Health) among the SOHs of the multiple single batteries (E1-Em) or single battery blocks (Eb1-Ebm) included in the battery system (21);

[0151] Based on the battery data, calculate the SOC difference according to the maximum SOC and the minimum SOC among the SOCs (State Of Charge) of the plurality of single cells (E1-Em) or single cell blocks (Eb1-Ebm) based on at least the voltage of the plurality of single cells (E1-Em) or single cell blocks (Eb1-Ebm) included in the battery system group (21), and the minimum SOH of the battery system group (21); and

[0152] Extract at least one change point from the sample group of the SOC differences, calculate the regression model of the SOC difference for each interval obtained by dividing with the change point, calculate the regression model of the minimum SOH based on the sample group of the minimum SOH, and calculate the regression model of the SOH of the battery system group (21) based on the regression model of the minimum SOH and the regression models of the SOC differences of the respective intervals.

[0153] Accordingly, the SOH of the battery system group (21) can be estimated with high accuracy.

[0154] Description of Reference Numerals

[0155] 2: Commercial power system; 5: Network; 10: Battery analysis system; 11: Control unit; 12: Storage unit; 13: Communication unit; 20: Electric vehicle; 21: Battery system group; 22: Voltage sensor; 23: Current sensor; 24: Temperature sensor; 25: Control unit; 26: Communication unit; 30: Charging pile; 31: Power supply unit; 32: Control unit; 33: Communication unit; 111: Data acquisition unit; 112: Minimum SOH calculation unit; 113: SOC difference calculation unit; 114: Statistical processing unit; 115: SOH prediction unit; 121: Battery data holding unit.

Claims

1. A battery analysis system, characterized in that, Comprising: a data acquisition unit that acquires at least one set of battery data including at least the voltage and current of each single battery or each battery block formed by connecting a plurality of single cells in parallel among the plurality of series-connected single cells included in the battery pack system; a minimum SOH calculation unit that calculates the minimum SOH among the SOHs of the plurality of single cells or the battery blocks included in the battery pack system based on the at least one set of battery data, where SOH is State Of Health; a SOC difference calculation unit that calculates a SOC difference based on the battery data, according to the maximum SOC and minimum SOC among the SOCs of the plurality of single cells or the battery blocks based on at least the voltage of the plurality of single cells or the battery blocks included in the battery pack system, and the minimum SOH of the battery pack system, where SOC is State Of Charge; and a statistical processing unit that extracts at least one change point from a sample group of the SOC differences, calculates a regression model of the SOC differences for each interval obtained by dividing at the change point, calculates a regression model of the minimum SOH based on a sample group of the minimum SOH, and calculates a regression model of the SOH of the battery pack system based on the regression model of the minimum SOH and the regression models of the SOC differences for each of the intervals.

2. The battery analysis system according to claim 1, wherein: it further comprises an SOH prediction unit that predicts the SOH of the battery pack system at a specific future time point based on the regression model of the SOH of the battery pack system.

3. The battery analysis system according to claim 1, wherein: the statistical processing unit calculates a simple regression model for each of the intervals, with the elapsed time as an explanatory variable and the SOC difference as a target variable.

4. The battery analysis system according to claim 1, wherein: the data acquisition unit acquires battery data including the voltage, current, and temperature of each single battery or each battery block, and the statistical processing unit calculates a multiple regression model for each of the intervals, with the elapsed time and the temperature difference between the maximum temperature and the minimum temperature in the battery pack including the battery pack system as explanatory variables and the SOC difference as a target variable.

5. The battery analysis system according to any one of claims 1 to 4, wherein: the SOC difference calculation unit calculates the SOC difference for each sample by summing the SOC difference between the SOC of 100 of the single battery or battery block with the minimum SOH among the plurality of single batteries or the battery blocks and the SOC of the single battery or battery block with the maximum SOC, and the SOC difference between the SOC of 0 of the single battery or battery block with the minimum SOH and the SOC of the single battery or battery block with the minimum SOC.

6. The battery analysis system according to any one of claims 1 to 4, wherein: The minimum SOH calculation unit calculates the minimum SOH based on two single cells or two cell blocks having the maximum voltage and the minimum voltage among the plurality of single cells or the cell blocks included in the battery pack system. The SOC difference calculation unit calculates the SOC difference based on the SOCs of the two single cells or two cell blocks and the minimum SOH of the battery pack system.

7. A battery analysis method, characterized in that, The method includes the following steps: Obtaining battery data including at least the voltage and current of each single cell or each cell block of a plurality of single cells connected in series or cell blocks formed by connecting a plurality of single cells in parallel, which are included in a battery pack system; Based on the at least one set of battery data, calculating the minimum SOH among the SOHs of the plurality of single cells or the cell blocks included in the battery pack system, where SOH is State Of Health; Based on the at least one set of battery data, calculating the SOC difference according to the maximum SOC and the minimum SOC among the SOCs of the plurality of single cells or the cell blocks based on at least the voltage of the plurality of single cells or the cell blocks included in the battery pack system, and the minimum SOH of the battery pack system, where SOC is State Of Charge; and Extracting at least one change point from the sample group of the SOC difference, calculating a regression model of the SOC difference for each interval obtained by dividing with the change point, calculating a regression model of the minimum SOH based on the sample group of the minimum SOH, and calculating a regression model of the SOH of the battery pack system based on the regression model of the minimum SOH and the regression models of the SOC differences of the respective intervals.

8. A battery analysis program, characterized in that, Causing a computer to execute the following processing: Obtaining at least one set of battery data including at least the voltage and current of each single cell or each cell block of a plurality of single cells connected in series or cell blocks formed by connecting a plurality of single cells in parallel, which are included in a battery pack system; Based on the at least one set of battery data, calculating the minimum SOH among the SOHs of the plurality of single cells or the cell blocks included in the battery pack system, where SOH is State Of Health; Based on the battery data, calculating the SOC difference according to the maximum SOC and the minimum SOC among the SOCs of the plurality of single cells or the cell blocks based on at least the voltage of the plurality of single cells or the cell blocks included in the battery pack system, and the minimum SOH of the battery pack system, where SOC is State Of Charge; and Extracting at least one change point from the sample group of the SOC difference, calculating a regression model of the SOC difference for each interval obtained by dividing with the change point, calculating a regression model of the minimum SOH based on the sample group of the minimum SOH, and calculating a regression model of the SOH of the battery pack system based on the regression model of the minimum SOH and the regression models of the SOC differences of the respective intervals.

9. A transient storage medium storing a battery analysis program according to claim 8, characterized in that it causes a computer to execute.

Images