Battery deterioration estimation system and battery deterioration estimation program

Abstract

A battery deterioration estimation system (101) comprising: a use history acquisition unit (41) that acquires a use history (A) of a plurality of batteries (1a) included in a battery pack (1); a simple measurement data acquisition unit (42) that acquires simple measurement data (B) of the plurality of batteries (1a); and a deterioration calculation unit (44) that calculates battery deterioration performance (D) on the basis of the use history (A) acquired by the use history acquisition unit (41) and the simple measurement data (B) acquired by the simple measurement data acquisition unit (42). The deterioration calculation unit (44) calculates a reference deterioration performance (Da) from the use history (A), and corrects the reference deterioration performance (Da) to the battery deterioration performance (D) on the basis of the simple measurement data (B).

Description

Battery degradation estimation system, battery degradation estimation program CROSS-REFERENCE TO RELATED APPLICATIONS

[0001] This application is based on Japanese Application No. 2024-114280, filed on July 17, 2024, the contents of which are incorporated herein by reference.

[0002] The present disclosure relates to a technique for estimating battery degradation performance.

[0003] Patent Document 1 below discloses a secondary battery capacity estimation system for estimating battery degradation performance. This secondary battery capacity estimation system is configured to divide secondary batteries into trend groups based on information indicating their usage history and resistance index values, and calculate a capacity maintenance rate from the resistance index values ​​using a model for each trend group. This secondary battery capacity estimation system attempts to estimate battery degradation performance, i.e., the capacity of a secondary battery, from the state of its internal resistance, even when the usage history of the secondary battery is different.

[0004] Patent No. 7276928

[0005] The secondary battery capacity estimation system configured as described above has the problem that the usage history of the secondary battery is only used for grouping, and is unable to classify minute differences in usage history. Therefore, this secondary battery capacity estimation system has the disadvantage of low accuracy in estimating battery degradation performance.

[0006] The present disclosure aims to provide a technique that is effective for estimating battery degradation performance with high accuracy.

[0007] One aspect of the present disclosure is a battery degradation estimation system comprising: a usage history acquisition unit that acquires usage histories of multiple batteries included in a battery pack; a simple measurement data acquisition unit that acquires simple measurement data of the multiple batteries; and a degradation calculation unit that calculates battery degradation performance based on the usage history acquired by the usage history acquisition unit and the simple measurement data acquired by the simple measurement data acquisition unit, wherein the degradation calculation unit calculates a standard degradation performance from the usage history and corrects the standard degradation performance to the battery degradation performance based on the simple measurement data.

[0008] Another aspect of the present disclosure is a battery degradation estimation program that causes a processor to acquire usage history of multiple batteries included in a battery pack, acquire simple measurement data for the multiple batteries, calculate a reference degradation performance from the usage history, and correct the reference degradation performance to a battery degradation performance based on the simple measurement data.

[0009] In each of the above-described aspects, the reference degradation performance calculated from the usage histories of a plurality of batteries is corrected to the battery degradation performance based on simplified measurement data of the battery, thereby improving the accuracy of estimating the battery degradation performance compared to when the battery degradation performance is estimated from the usage history alone.

[0010] Therefore, according to the above-described aspects, it is possible to estimate the battery degradation performance with high accuracy.

[0011] Note that the symbols in parentheses in the claims indicate the correspondence with the specific means described in the embodiments described below, and do not limit the technical scope of the present disclosure.

[0012] The above and other objects, features, and advantages of the present disclosure will become more apparent from the following detailed description taken in conjunction with the accompanying drawings, in which Fig. 1 is a block diagram showing the configuration of a battery degradation estimation system according to a first embodiment, Fig. 2 is a flowchart of a model generation process performed by a model generation unit in Fig. 1, Fig. 3 is a diagram for explaining the procedure for generating a first degradation model formula and a second degradation model formula in the model generation process of Fig. 2, Fig. 4 is a flowchart of a battery degradation estimation process performed by the battery degradation estimation system according to the first embodiment, Fig. 5 is a diagram for explaining the processing from the fourth step to the sixth step in Fig. 4, Fig. 6 is a diagram for explaining correction of SOH_E to SOH_C in a correlation diagram between SOH, which is an example of battery degradation performance, and battery load, and Fig. 7 is a block diagram showing the configuration of a battery degradation estimation system according to a second embodiment.

[0013] The battery degradation estimation technology according to each of the above aspects will be described in detail below with reference to the drawings. (First embodiment) Fig. 1 shows the configuration of a battery degradation estimation system 101 according to the first embodiment. This battery degradation estimation system 101 is a system for estimating the battery degradation performance of a battery 1a in a battery pack 1 mounted on a vehicle 23 such as an electric vehicle or a hybrid vehicle.

[0014] The functions of the battery degradation estimation system 101 are executed by a processor 10. The term "processor 10" as used herein broadly encompasses, among the components of a computer, processing devices that perform operations such as data calculations and conversions, program execution, and control of other devices. The processor 10 includes a CPU (Central Processing Unit) that controls the entire computer, or an MPU (Micro Processing Unit) that integrates some of the functions of the CPU.

[0015] 1. Configuration of battery degradation estimation system 101 As shown in Fig. 1, the battery degradation estimation system 101 includes, as its components, a battery pack 1, a simplified measurement unit 20, a battery diagnosis unit 40, and an output unit 50. On the other hand, the battery degradation estimation system 101 can include at least the battery diagnosis unit 40 as a component, if necessary.

[0016] 1-1. Configuration of Battery Pack 1 The battery pack 1 includes multiple battery modules 1a and a main battery management unit 2. The battery module 1a includes an assembled battery made up of multiple battery cells, and a satellite battery management unit 3. The battery cells are secondary batteries that can be recharged and reused. The main battery management unit 2 stores battery information for the multiple battery modules 1a. The functions of the main battery management unit 2 and the satellite battery management unit 3 are executed by a processor 10A. This processor 10A is contained in a processor 10 together with processors 10B, 10C, and 10D, which will be described later.

[0017] In this embodiment, the plurality of battery modules 1a included in the battery pack 1, or the battery cells included in the battery modules 1a, are also referred to simply as "batteries 1a" for convenience. The battery pack 1 may be in a state where it has been removed from a device such as a vehicle 23 and collected, or may be in a state where it is mounted on the device.

[0018] 1-2. Configuration of the simplified measurement unit 20 The simplified measurement unit 20 is connected to the main battery management unit 2 of the battery pack 1 either constantly or when necessary. This simplified measurement unit 20 has the function of measuring simplified measurement data B of multiple batteries 1a. This function is executed by the processor 10B. For example, a charging / discharging device that charges and discharges the batteries 1a can be used as this simplified measurement unit 20. In this case, the simplified measurement data B can be measured from sensing data such as current, voltage, and temperature during charging and discharging of the charging / discharging device. The simplified measurement data B is sent to the simplified measurement data acquisition unit 42 of the battery diagnosis unit 40. The simplified measurement unit 20 is provided appropriately in, for example, a dedicated measuring device 21, a charging station 22, or a vehicle 23.

[0019] 1-3. Configuration of Battery Diagnostic Unit 40 The battery diagnostic unit 40 has the function of diagnosing the battery pack 1. The battery diagnostic unit 40 includes, as its components, a usage history acquisition unit 41, a simplified measurement data acquisition unit 42, a model generation unit 43, and a deterioration calculation unit 44. The functions of each of these components are executed by the processor 10C. The battery diagnostic unit 40 is appropriately provided, for example, in a dedicated device such as a data server 30 or the cloud 31.

[0020] The usage history acquisition unit 41 acquires usage histories A of the multiple batteries 1a from the main battery management unit 2 of the battery pack 1. The usage history A is history information of parameters related to the current use of each of the multiple batteries 1a. Examples of this usage history A include histories such as the usage period of the battery 1a, the mileage of the vehicle 23 in which the battery 1a is mounted, the number of times the vehicle 23 has been started, and accumulated charge amount, accumulated discharge amount, temperature frequency distribution, SOC (State Of Charge) frequency distribution, SOC fluctuation frequency, current frequency, current continuation frequency, and multidimensional frequency of each feature amount (for example, two-dimensional frequency of SOC-temperature) generated from sensing data such as the usage period of the battery 1a, the mileage of the vehicle 23 in which the battery 1a is mounted, the number of times the vehicle 23 has been started, current, voltage, and temperature.

[0021] The simplified measurement data acquisition unit 42 acquires simplified measurement data B of the multiple batteries 1a from the simplified measurement unit 20. The simplified measurement data B is data obtained by measuring the battery characteristics of each of the multiple batteries 1a using a simplified method. Examples of this simplified measurement data B include the resistance values ​​(DC resistance value, AC resistance value) and section capacity of the batteries 1a. The DC resistance value is a resistance value obtained using the so-called "direct current method," and is the resistance value calculated from the changes in voltage and current when a load is connected to the battery 1a and a direct current is passed through it. The AC resistance value is a resistance value obtained using the so-called "alternating current method," and is the internal resistance value determined from the voltage value of an AC voltmeter when a measurement current of a measurement frequency is applied to the battery 1a. DC resistance values ​​and AC resistance values ​​have the advantage of being relatively easy to measure.

[0022] The model generation unit 43 generates a first degradation model formula M1 and a second degradation model formula M2 by machine learning. The first degradation model formula M1 is an estimation formula for estimating a reference degradation performance Da from the usage history A acquired by the usage history acquisition unit 41. The second degradation model formula M2 is an estimation formula for estimating a degradation error ΔDa from the simplified measurement data B acquired by the simplified measurement data acquisition unit 42.

[0023] The deterioration calculation unit 44 calculates battery deterioration performance D based on the usage history A acquired by the usage history acquisition unit 41 and the simplified measurement data B acquired by the simplified measurement data acquisition unit 42. At this time, the deterioration calculation unit 44 calculates a reference deterioration performance Da from the usage history A and corrects the reference deterioration performance Da to battery deterioration performance D based on the simplified measurement data B. More specifically, the deterioration calculation unit 44 corrects the reference deterioration performance Da estimated from the first deterioration model formula M1 to battery deterioration performance D based on the deterioration error ΔDa estimated from the second deterioration model formula M2. Then, the deterioration calculation unit 44 transmits the calculated battery deterioration performance D to the output unit 50.

[0024] The battery degradation performance D may be the degradation performance of all of the multiple batteries 1a in the battery pack 1, or may be the degradation performance of only the representative battery 1b among the multiple batteries 1a. For example, as shown in FIG. 1, the first battery 1a and the last battery 1a can be designated as representative batteries 1b. When all of the multiple batteries 1a are targeted, accurate degradation estimation is possible. On the other hand, when only the representative battery 1b is targeted, the time required for degradation estimation can be shortened.

[0025] 1-4. Configuration of the Output Unit 50 The output unit 50 has a function of outputting information transmitted from the deterioration calculation unit 44 of the battery diagnosis unit 40. The function of this output unit 50 is executed by the processor 10D. The output unit 50 is mounted on a terminal device 51. Note that "output" here broadly encompasses not only output by screen display but also output by printing, audio, etc. Typical examples of the terminal device 51 include desktop or notebook personal computers (PCs), tablet terminals, mobile terminals, etc.

[0026] The output unit 50 may be configured to output the battery degradation performance D, which is the calculation result of the degradation calculation unit 44, as is, or may have a conversion output function that converts the battery degradation performance D into another index (for example, the lifespan of the battery pack 1 or the cruising range of the vehicle 23) and outputs it in response to a user's request. This conversion output function may be provided in the degradation calculation unit 44. This allows the output unit 50 to output battery information in the form of another index in response to a user's request.

[0027] The functions of each component of the battery degradation estimation system 101 are realized by having the processor 10 execute a battery degradation estimation program P. Therefore, the battery degradation estimation program P is a program that causes the processor 10 to realize the functions of each component. The processor 10 may be configured by all or part of the aforementioned processors 10A, 10B, 10C, and 10D. The battery degradation estimation program P is stored in a non-transitory storage medium 11. The non-transitory storage medium 11 is depicted separately in the drawings to avoid clutter. However, the non-transitory storage medium 11 is included in the battery pack 1, the measuring device 21, the charging station 22, the vehicle 23, the battery diagnostic unit 40, and the terminal device 51. At least a portion of the battery degradation estimation program P is stored in the non-transitory storage medium 11 included in each of these components. Various types of non-transitory storage medium 11, such as a memory type, a disk type, or a tape type, can be used.

[0028] The battery degradation estimation program P may be stored in the data server 30 or the cloud 31. In this case, a configuration may be adopted in which at least a portion of the battery degradation estimation program P is downloaded from the data server 30 or the cloud 31 to the battery pack 1, the measuring device 21, the charging station 22, the vehicle 23, the battery diagnosis unit 40, and the terminal device 51.

[0029] The allocation of the multiple components (functional components) of the battery degradation estimation system 101 to each device, etc. is not limited to that shown in Fig. 1 and can be changed as needed. In addition, other devices or facilities may be used as the allocation destinations of the components (functional components).

[0030] 2. Model Generation Process Next, the model generation process executed primarily by the model generation unit 43 in FIG. 1 will be described with reference to FIGS. 2 and 3. In this model generation process, the first step S101 to the fourth step S104 in the flowchart shown in FIG. 2 are executed sequentially. One or more steps may be added to these steps as needed, or multiple steps may be appropriately integrated. This model generation process can be executed for each of the multiple batteries 1a, or for a representative battery 1b among the multiple batteries 1a. In this case, the model generation process may be executed only once for each battery 1a, or may be executed multiple times.

[0031] In the following, a case where the SOH (State Of Health) which is the degree of deterioration of the battery 1a is estimated as an example of the battery deterioration performance D will be described.

[0032] The first step S101 is a step of acquiring the usage history A and simplified measurement data B used in generating the deterioration model formulas M1 and M2.

[0033] The second step S102 is a step of generating a first deterioration model formula M1 using the usage history A acquired in the first step S101. As shown in Fig. 3, the first deterioration model formula M1 is a model formula obtained by linear regression analysis in which one or more usage histories A are used as explanatory variables and SOH_T, which is the true value of the SOH, is used as the objective variable. According to this first deterioration model formula M1, when the usage history A is input, SOH_E, which is the estimated value of the SOH, is output as a result.

[0034] The third step S103 is a step of generating a second degradation model formula M2 using the simplified measurement data B acquired in the first step S101. As shown in FIG. 3 , the second degradation model formula M2 is a model formula obtained by linear regression analysis using one or more pieces of simplified measurement data B as explanatory variables and ΔSOH as a response variable. ΔSOH in this case is defined as the difference between the true value SOH_T and the estimated value SOH_E. According to this second degradation model formula M2, when the simplified measurement data B is input, ΔSOH_E, which is an estimated value of ΔSOH, is output as a result.

[0035] The fourth step S104 is a step of storing the first deterioration model formula M1 generated in the second step S102 and the second deterioration model formula M2 generated in the third step S103. Note that the deterioration model formulas M1 and M2 may be updated as appropriate.

[0036] As the linear regression analysis used in the second step S102 and the third step S103, an appropriate algorithm such as lasso regression, ridge regression, or elastic net can be used.

[0037] 3. Battery Deterioration Estimation Process Next, with reference to FIGS. 4 to 6, the battery deterioration estimation process executed primarily by the deterioration calculation unit 44 in FIG. 1 will be described. In this battery deterioration estimation process, the first step S201 to the eighth step S208 of the flowchart shown in FIG. 4 are executed sequentially. One or more steps may be added to these steps as needed, or multiple steps may be appropriately combined. This battery deterioration estimation process can be executed for each of the multiple batteries 1a or for a representative battery 1b among the multiple batteries 1a. In this case, the battery deterioration estimation process may be executed only once for each battery 1a, or may be executed multiple times. The battery deterioration estimation process is typically executed at times such as when the vehicle 23 undergoes inspection or when the battery pack 1 is charged.

[0038] The first step S201 is a step of selecting a battery 1a to be estimated. The second step S202 is a step of acquiring a usage history A and simplified measurement data B to be used in estimating battery deterioration of the battery 1a selected in the first step S201. The third step S203 is a step of reading out the deterioration model formulas M1 and M2 previously stored in the fourth step S104 of FIG. 2 .

[0039] The fourth step S204 is a step for calculating an estimated value, SOH_E. As shown in FIG. 5, in this fourth step S204, the usage history A acquired in the second step S202 is input into the first deterioration model formula M1 read out in the third step S203. As a result, SOH_E is output from the first deterioration model formula M1. SOH_E is the reference value (median value) of SOH before correction. SOH_E at this time is an example of the "reference deterioration performance Da."

[0040] The fifth step S205 is a step for calculating an estimated value ΔSOH_E. As shown in FIG. 5, in this fifth step S205, the simplified measurement data B acquired in the second step S202 is input to the second deterioration model formula M2 read out in the third step S203. As a result, ΔSOH_E is output from the second deterioration model formula M2. ΔSOH_E represents variation in SOH_E (individual differences between batteries 1a). ΔSOH_E at this time is an example of a "deterioration error ΔDa."

[0041] The sixth step S206 is a step of calculating a corrected value, SOH_C. As shown in FIG. 5, in this sixth step S206, the SOH_E calculated in the fourth step S204 is corrected to SOH_C based on ΔSOH_E calculated in the fifth step S205. This makes it possible to calculate SOH_C in which the variation in SOH_E has been corrected. The SOH_C at this time is an example of "battery degradation performance D."

[0042] The seventh step S207 is a step of outputting the SOH_C derived by the calculation in the sixth step S206 to the output unit 50 (see FIG. 1).

[0043] In the eighth step S208, it is determined whether or not to terminate the battery deterioration estimation process. If the battery deterioration estimation process is to be terminated ("Yes" in the eighth step S208), the process is terminated. If not ("No" in the eighth step S208), the process returns to the first step S201.

[0044] As shown in FIG. 6 , it is known that the SOH decreases as the battery load (e.g., conditions such as time, charge / discharge amount, and temperature frequency) increases. In this case, it is expected that the SOH_E calculated in the fourth step S204 may deviate from the true value SOH_T due to variations in the manufacturing process of the battery 1a or variations in the manufacturing of constituent materials that contribute to battery degradation. Therefore, in this embodiment, the fifth step S205 and the sixth step S206 are performed in addition to the fourth step S204 to correct the SOH_E to SOH_C, which is closer to SOH_T. This makes it possible to correct for individual differences between batteries 1a, thereby improving the accuracy of estimating SOH_C relative to SOH_T.

[0045] 4. Effects According to the first embodiment, SOH_E, which is the reference degradation performance Da calculated from the usage histories A of multiple batteries 1a, is corrected to SOH_C, which is the battery degradation performance D, based on the simplified measurement data B of the batteries 1a. That is, by performing the fifth step S205 and the sixth step S206 in addition to the fourth step S204 in Fig. 4, the accuracy of estimating the battery degradation performance D is improved compared to when estimating the battery degradation performance D from only the usage histories A (only the fourth step S204). Therefore, it becomes possible to estimate the battery degradation performance D with high accuracy.

[0046] In particular, according to the first embodiment, by correcting the reference degradation performance Da to the battery degradation performance D using machine learning, it is possible to further improve the estimation accuracy of the battery degradation performance D.

[0047] Hereinafter, other embodiments related to the above-described embodiment 1 will be described with reference to the drawings. In the other embodiments, the same elements as those in embodiment 1 are denoted by the same reference numerals, and the description of the same elements will be omitted.

[0048] 7 , a battery degradation estimation system 102 of embodiment 2 differs from the battery degradation estimation system 101 of embodiment 1 in that a battery information storage unit 45 is provided in the battery diagnosis unit 40. The battery information storage unit 45 readably stores a plurality of pieces of battery information, such as a usage history A output from the main battery management unit 2, simplified measurement data B measured by the simplified measurement unit 20, degradation model formulas M1 and M2 generated by the model generation unit 43, and battery degradation performance D calculated by the degradation calculation unit 44.

[0049] The other configurations are the same as those in the first embodiment.

[0050] According to the second embodiment, a plurality of pieces of battery information relating to the estimation of battery deterioration can be temporarily stored in the data server 30 of a dedicated device or the battery information storage unit 45 of the cloud 31 .

[0051] In addition, the same effects as those of the first embodiment are achieved.

[0052] Although the present disclosure has been described based on the above-described embodiments, it is understood that the present disclosure is not limited to these forms and structures. The present disclosure also encompasses various modifications and modifications within the scope of equivalents. In addition, various combinations and forms, as well as other combinations and forms including only one element, more than one element, or less than one element, are also within the scope and spirit of the present disclosure. For example, the following forms can be implemented by applying the above-described forms.

[0053] In the above embodiment, an example has been given of a case where machine learning is used to correct the reference degradation performance Da to the battery degradation performance D, but the correction method is not limited to machine learning. For example, instead of correction using machine learning, offset correction that corrects the reference degradation performance Da to the battery degradation performance D based on a specified offset amount, map correction that uses mapping data created based on the usage history of the battery 1a, or the like may be adopted.

[0054] In the above embodiment, an example is given of estimating battery deterioration of a battery 1a installed in a vehicle 23, but this diagnostic technique may also be applied to estimating battery deterioration of a battery 1a installed in consumer equipment or industrial equipment other than a vehicle 23.

Claims

1. A battery degradation estimation system (101, 102) comprising: a usage history acquisition unit (41) that acquires usage history (A) of multiple batteries (1a) included in a battery pack (1); a simplified measurement data acquisition unit (42) that acquires simplified measurement data (B) of the multiple batteries; and a degradation calculation unit (44) that calculates battery degradation performance (D) based on the usage history acquired by the usage history acquisition unit and the simplified measurement data acquired by the simplified measurement data acquisition unit, wherein the degradation calculation unit calculates a standard degradation performance (Da) from the usage history and corrects the standard degradation performance to the battery degradation performance based on the simplified measurement data.

2. A battery degradation estimation system as described in claim 1, comprising a model generation unit (43) that generates, by machine learning, a first degradation model formula (M1) that estimates the reference degradation performance from the usage history and a second degradation model formula (M2) that estimates a degradation error (ΔDa) from the simplified measurement data, and the degradation calculation unit corrects the reference degradation performance estimated from the first degradation model formula to the battery degradation performance based on the degradation error estimated from the second degradation model formula.

3. A battery degradation estimation system according to claim 1 or 2, wherein the simplified measurement data is the DC resistance value or AC resistance value of the battery.

4. A battery degradation estimation system according to claim 1 or 2, wherein the battery degradation performance is the degradation performance of all of the plurality of batteries in the battery pack.

5. A battery degradation estimation system according to claim 1 or 2, wherein the battery degradation performance is the degradation performance of only a representative battery (1b) among the plurality of batteries of the battery pack.

6. A battery degradation estimation program (P) that causes a processor (10) to acquire usage history (A) of multiple batteries (1a) included in a battery pack (1), acquire simple measurement data (B) of the multiple batteries, calculate a standard degradation performance (Da) from the usage history, and correct the standard degradation performance to a battery degradation performance (D) based on the simple measurement data.

7. A battery degradation estimation program as described in claim 6, which causes a processor (10) to generate, by machine learning, a first degradation model formula (M1) that estimates the reference degradation performance from the usage history and a second degradation model formula (M2) that estimates a degradation error (ΔDa) from the simplified measurement data, and corrects the reference degradation performance estimated from the first degradation model formula to the battery degradation performance based on the degradation error estimated from the second degradation model formula.

8. The battery degradation estimation program according to claim 6 or 7, wherein the simplified measurement data is a DC resistance value or an AC resistance value of the battery.

9. The battery degradation estimation program according to claim 6 or 7, wherein the battery degradation performance is the degradation performance of all of the plurality of batteries in the battery pack.

10. A battery degradation estimation program according to claim 6 or 7, wherein the battery degradation performance is the degradation performance of only a representative battery (1b) among the plurality of batteries of the battery pack.

Images