System for estimating the state of a secondary battery, method for estimating the state of a secondary battery, and program

Abstract

This invention provides a secondary battery state estimation system that can suppress the decrease in the accuracy of internal state estimation. [Solution] System 1, which includes a secondary battery state estimation device 10, includes an OCV estimation unit 24 that acquires a machine learning model and estimates the open-circuit voltage using the machine learning model. The OCV estimation unit 24 acquires first data having time-series data of the current, closed-circuit voltage, and open-circuit voltage of the secondary battery 11 in a predetermined state including at least one of charging and discharging. Based on the first data, the OCV estimation unit 24 acquires a machine learning model that outputs the overvoltage of the secondary battery 11 at any given time. The OCV estimation unit 24 estimates the overvoltage of the secondary battery 11 at any given time by inputting second data having current and closed-circuit voltage data detected at a predetermined time in the secondary battery 11 to be estimated into the machine learning model.

Description

[Technical Field]

[0001] This invention relates to a secondary battery state estimation system, a secondary battery state estimation method, and a program. [Background technology]

[0002] In recent years, research and development has been conducted on rechargeable batteries that contribute to energy efficiency, in order to ensure that more people have access to affordable, reliable, sustainable, and advanced energy. Conventionally, there are known devices that obtain an OCV curve showing the change in open circuit voltage (OCV) according to the discharge capacity by filtering the historical data of the closed circuit voltage (CCV) of a battery (see, for example, Patent Document 1). This device extracts data that can be considered as open circuit voltage from the historical data of closed circuit voltage by filtering based on predetermined conditions related to the battery's current, voltage, and temperature, for example. [Prior art documents] [Patent Documents]

[0003] [Patent Document 1] International Publication No. 2023 / 054443 [Overview of the project] [Problems that the invention aims to solve]

[0004] Incidentally, in the field of secondary battery technology, a challenge is to suppress the decrease in the accuracy of estimating the internal state, including the OCV curve, even when the secondary battery deteriorates. For example, in the conventional device described above, if the overvoltage increases due to the increase in internal resistance caused by the deterioration of the secondary battery, the accuracy of estimating the open-circuit voltage (OCV) may decrease. Furthermore, even when estimating the open-circuit voltage (OCV) based on, for example, an electrochemical model for estimating internal resistance or the correspondence between the open-circuit voltage (OCV) at the start of discharge and the discharge capacity, the accuracy of the OCV estimation may decrease as the internal resistance increases due to the degradation of the secondary battery.

[0005] This invention aims to solve the above-mentioned problems by suppressing the decrease in the accuracy of estimating the internal state of a secondary battery. [Means for solving the problem]

[0006] In order to solve the above problems and achieve the above objectives, the present invention employs the following embodiments. (1) A secondary battery state estimation system according to one aspect of the present invention (for example, system 1 in the embodiment) comprises a learning model constructed to acquire first data having time-series data of current, closed-circuit voltage and open-circuit voltage of a secondary battery (for example, secondary battery 11 in the embodiment) in a predetermined state including at least one of charging and discharging, and to output the open-circuit voltage of the secondary battery at any given time or an overvoltage which is the difference between the open-circuit voltage and the closed-circuit voltage of the secondary battery based on the first data, and a voltage estimation unit (for example, OCV estimation unit 24 in the embodiment) that estimates the open-circuit voltage or overvoltage of the secondary battery to be estimated at any given time by inputting second data having data of current and closed-circuit voltage at a predetermined time detected in the secondary battery to be estimated to the learning model.

[0007] (2) In the secondary battery state estimation system described in (1) above, the second data may include time-series data of the current and closed-circuit voltage detected in the secondary battery to be estimated.

[0008] (3) In the secondary battery state estimation system described in (2) above, the voltage estimation unit may estimate the open-circuit voltage or overvoltage of the secondary battery to be estimated at the predetermined time by inputting the preceding interval data, which has current and closed-circuit voltage data for a predetermined preceding time interval prior to the predetermined time detected in the secondary battery to be estimated, into the learning model.

[0009] (4) In the secondary battery state estimation system described in (2) or (3) above, the voltage estimation unit may estimate the open-circuit voltage or overvoltage of the secondary battery to be estimated at the predetermined time by inputting to the learning model data of current and closed-circuit voltage in a predetermined time interval after a predetermined time detected in the secondary battery to be estimated.

[0010] (5) In the secondary battery state estimation system described in (1) or (2) above, the second data does not need to include data relating to the temperature of the secondary battery to be estimated.

[0011] (6) In the secondary battery state estimation system described in (1) or (2) above, the learning model may output an overvoltage of the secondary battery.

[0012] (7) In the secondary battery state estimation system described in (1) or (2) above, the learning model may output the open-circuit voltage of the secondary battery.

[0013] (8) In the secondary battery state estimation system described in (1) or (2) above, the first data may include time-series data of the current, closed-circuit voltage and open-circuit voltage of the degraded secondary battery.

[0014] (9): A method for estimating the state of a secondary battery according to one aspect of the present invention is a method executed by an electronic device (e.g., system 1 in the embodiment) equipped with a processing unit (e.g., OCV estimation unit 24 in the embodiment) for estimating the open-circuit voltage or overvoltage of a secondary battery to be estimated, and includes a model acquisition step (e.g., steps S01 and S02 in the embodiment) for acquiring first data having time-series data of the current, closed-circuit voltage and open-circuit voltage of a secondary battery in a predetermined state including at least one of charging and discharging, and acquiring a learning model that outputs the open-circuit voltage or overvoltage, which is the difference between the open-circuit voltage and the closed-circuit voltage of the secondary battery at an arbitrary time based on the first data, and a voltage estimation step (e.g., steps S03, S04 and S05 in the embodiment) for estimating the open-circuit voltage or overvoltage of the secondary battery to be estimated at an arbitrary time by inputting second data having data of the current and closed-circuit voltage at a predetermined time detected in the secondary battery to be estimated to the learning model acquired by the model acquisition step.

[0015] (10): A program according to one aspect of the present invention causes a computer in an electronic device (e.g., system 1 in the embodiment) equipped with a processing unit (e.g., OCV estimation unit 24 in the embodiment) for estimating the open-circuit voltage or overvoltage of a secondary battery to be estimated to execute a model acquisition step (e.g., steps S01 and S02 in the embodiment) which acquires first data having time-series data of the current, closed-circuit voltage and open-circuit voltage of a secondary battery in a predetermined state including at least one of charging and discharging, and acquires a learning model that outputs the open-circuit voltage or overvoltage, which is the difference between the open-circuit voltage and the closed-circuit voltage of the secondary battery at an arbitrary time based on the first data, and a voltage estimation step (e.g., steps S03, S04 and S05 in the embodiment) which estimates the open-circuit voltage or overvoltage of the secondary battery to be estimated to execute a model acquisition step (e.g., steps S03, S04 and S05 in the embodiment) which has second data having data of the current and closed-circuit voltage at a predetermined time detected in the secondary battery to be estimated to execute a voltage estimation model acquisition step [Effects of the Invention]

[0016] According to (1) above, by using a learning model that outputs an open-circuit voltage or an overvoltage in response to the input of data on current and closed-circuit voltage, it is possible to suppress a decrease in the estimation accuracy of the internal state of the secondary battery.

[0017] In the case of (2) above, since the second data has time-series data, the estimation accuracy of the open-circuit voltage by the learning model can be improved.

[0018] In the case of (3) above, the open-circuit voltage can be estimated from data in the past relative to the target time, and the versatility of the learning model can be improved.

[0019] In the case of (4) above, the open-circuit voltage can be estimated from data in the future relative to the target time, and the versatility of the learning model can be improved.

[0020] In the case of (5) above, the open-circuit voltage can be estimated without requiring temperature data. For example, even when the learning model is obtained only from data before the deterioration of the secondary battery, the estimation accuracy of the open-circuit voltage for the deteriorated secondary battery can be improved.

[0021] In the case of (6) or (7) above, the open-circuit voltage of the secondary battery can be easily and accurately estimated.

[0022] According to (8) or (9) above, by using a learning model that outputs an open-circuit voltage or an overvoltage in response to the input of data on current and closed-circuit voltage, it is possible to suppress a decrease in the estimation accuracy of the internal state of the secondary battery.

Brief Description of the Drawings

[0023] [Figure 1] A block diagram showing the functional configuration of a system including a state estimation device for a secondary battery in an embodiment of the present invention. [Figure 2]This figure shows an example of the information flow in the processing by the estimation unit of the secondary battery state estimation device in an embodiment of the present invention. [Figure 3] This figure shows an example of the correspondence between the open-circuit voltage (OCV) and the closed-circuit voltage (CCV) obtained by the estimation unit of the secondary battery state estimation device in an embodiment of the present invention. [Figure 4] This figure shows an example of an OCV curve obtained based on the OCP curves of the positive electrode and the negative electrode by the estimation unit of the secondary battery state estimation device in an embodiment of the present invention. [Figure 5] A flowchart showing the process performed by the secondary battery state estimation device in an embodiment of the present invention. [Modes for carrying out the invention]

[0024] Hereinafter, a secondary battery state estimation system, a secondary battery state estimation method, and a program according to embodiments of the present invention will be described with reference to the attached drawings. The secondary battery according to this embodiment is, for example, attached to or permanently installed in various electrical devices. Various electrical devices include, for example, electric vehicles, electric mobile devices, electric machinery, and power supply devices. Electric vehicles include, for example, electric automobiles, saddle-type vehicles, and kick scooters equipped with a rotating electric machine powered by the secondary battery, hybrid vehicles combining a rotating electric machine and an internal combustion engine, and fuel cell vehicles combining a secondary battery and a fuel cell. Electric mobile devices include, for example, robots, mobile work machines, aircraft, and mobile devices on and underwater. Electric machinery includes, for example, construction machinery equipped with a rotating electric machine as a power source. Power supply devices include, for example, stationary or mobile power supply devices that discharge and charge secondary batteries, or exchange devices that provide and receive secondary batteries to users in a so-called battery sharing service.

[0025] Furthermore, various electrical devices may be equipped with an external charging function, such as in PHV (Plug-in Hybrid Vehicle) or PHEV (Plug-in Hybrid Electric Vehicle), which allows them to be charged by an external power source (external DC power source and external AC power source). Various electrical devices may also be equipped with a function to supply power to an external source using the power of a secondary battery. In addition, a rotating electric machine mounted on an electric vehicle may exchange power with a secondary battery, for example, through regenerative operation using rotational power input from the wheels, or through power generation using power input from an internal combustion engine, in addition to the traction operation.

[0026] Figure 1 is a block diagram showing the functional configuration of a system 1 equipped with a secondary battery state estimation device 10 according to an embodiment. As shown in Figure 1, the system 1 of the embodiment comprises, for example, a vehicle 2 and a server 3. The vehicle 2 and the server 3 are connected, for example, via a wired or wireless communication network 4. The network 4 is, for example, the Internet, a mobile communication network, a LAN (Local Area Network), and a WAN (Wide Area Network). For example, the LAN is a wired LAN (Local Area Network) of a predetermined standard such as Ethernet, or a wireless LAN of various standards such as Wi-Fi and Bluetooth (registered trademark). The secondary battery state estimation device 10 of this embodiment is configured, for example, by a server 3.

[0027] Vehicle 2 includes, for example, a secondary battery 11, a battery sensor 12, a battery control unit 13, a power control unit 14, a rotating electric machine 15, a drive mechanism 16, and an overall processing unit 17. The secondary battery 11 is one of various types of batteries that undergo repeated charging and discharging, such as lithium-ion batteries, sodium-ion batteries, or nickel-metal hydride batteries. The electrolyte of the secondary battery 11 is a non-aqueous electrolyte such as a liquid, solid, or polymer.

[0028] The positive electrode active material that constitutes the positive electrode of the secondary battery 11 is, for example, in the case of a lithium ion battery, a metal oxide containing lithium ions or the like. The metal oxide containing lithium ions is, for example, a simple substance of a composite oxide composed of lithium and metals such as nickel, cobalt, manganese, and aluminum, or a mixture of a plurality of different composite oxides. The composite oxide is classified, for example, from the viewpoint of the crystal structure, into a layered rock salt type, a spinel type, and an olivine type. The layered rock salt type composite oxide is, for example, lithium cobalt oxide (LCO: LiCoO2), nickel-cobalt-manganese oxide (NCM: Li(Ni , x , 12 ,

[0030] , Co y Mn z )O2), nickel-cobalt-aluminum oxide (NCA: LiNi x Co y Al x O2), etc. The spinel type composite oxide is, for example, lithium manganese oxide (LMO: LiMn2O4) and lithium nickel manganese oxide (LNMO: LiNi x Mn y O4), etc. The olivine type composite oxide is, for example, lithium iron phosphate (LFP: LiFePO4) and lithium manganese iron phosphate (LMFP: LiMn x [[ID=2​​​​​​​​​​​​​The battery sensor 12 includes, for example, various sensors for detecting the state of the secondary battery 11. The battery sensor 12 includes, for example, a voltage sensor, a current sensor, and a temperature sensor. The battery sensor 12 outputs signals of various detected values ​​such as voltage, current, and temperature related to the state of the secondary battery 11.

[0031] The battery control unit 13 is, for example, a so-called BMU (Battery Management Unit) that monitors and controls the state of the secondary battery 11. The battery control unit 13 is a software function unit that functions when a predetermined program is executed by a processor such as a CPU (Central Processing Unit). The software function unit is an ECU (Electronic Control Unit) that includes an ECU (Electronic Control Unit) equipped with a processor such as a CPU, a ROM (Read Only Memory) for storing programs, a RAM (Random Access Memory) for temporarily storing data, and electronic circuits such as a timer. At least a part of the battery control unit 13 may be an integrated circuit such as an LSI (Large Scale Integration).

[0032] The battery control unit 13 stores, for example, information relating to the secondary battery 11 and a predetermined program. The information relating to the secondary battery 11 includes, for example, identification information such as an ID (IDentifier) ​​exclusively assigned to the secondary battery 11, the date and time of manufacture, the initial capacity, and information relating to the state of the secondary battery 11 based on the output of the battery sensor 12. The information relating to the state of the secondary battery 11 includes, for example, the charge status such as the charge rate, remaining capacity (SOC: State Of Charge) or energy amount, the charge and discharge history such as the number of charge cycles, information relating to the current state such as voltage and temperature, information relating to the current degradation state such as the degree of degradation, and information relating to the presence or absence of abnormalities.

[0033] The power control unit 14 is connected to the secondary battery 11 and the rotating electric machine 15. The power control unit 14 includes, for example, a voltage converter such as a DC-DC converter that converts DC voltage and a power converter such as a DC-AC converter that converts power between DC and AC. The power control unit 14 controls the power transfer between the secondary battery 11 and the rotating electric machine 15 based on control signals received from, for example, the control unit 17.

[0034] The rotating electric machine 15 is, for example, a three-phase AC brushless DC motor. The rotating electric machine 15 generates rotational power by performing a motoring operation using power supplied from the power control unit 14. When the rotating electric machine 15 is connected to the wheels of a vehicle 2, for example, it generates driving force by performing a motoring operation using power supplied from the power control unit 14. The rotating electric machine 15 may also generate power by performing a regenerative operation using rotational power input from the wheels of the vehicle 2. When the rotating electric machine 15 is connected to the internal combustion engine of the vehicle 2, it may also generate power using the power of the internal combustion engine.

[0035] The drive mechanism 16 is a power transmission mechanism connected to the rotor of the rotating electric machine 15. The drive mechanism 16 includes, for example, gears, belts, and chains. The drive mechanism 16 transmits power between, for example, the rotating electric machine 15 and the wheels of the vehicle 2. The drive mechanism 16 may also include a regulating mechanism to restrict power transmission, such as an electric parking brake and parking lock mechanism that stops the rotation of the wheels or drive shafts.

[0036] The integrated processing unit 17 comprehensively controls the operation of the vehicle 2. The integrated processing unit 17 includes, for example, a software function unit. At least a part of the integrated processing unit 17 may include an integrated circuit. The integrated processing unit 17 includes, for example, an input / output unit and a communication unit. The input / output unit includes, for example, various operating devices such as a keyboard, touch panel, mouse, and buttons; a display device such as a liquid crystal display or an organic EL (Electro-Luminescence) display; and various input / output devices such as a microphone for voice input and a speaker for sound output. The input / output unit receives input operations, such as operations by an operator such as a user or voice input, and outputs a signal corresponding to the input operation. The communications unit transmits and receives various types of information to and from the server 3 via the network 4. For example, the communications unit transmits information to the server 3 that is a combination of information such as the date and time, identification information of the vehicle 2 or secondary battery 11, and information about the secondary battery 11 received from the battery control unit 13.

[0037] Server 3 includes, for example, a software function unit. At least a portion of the integrated processing unit 17 may include an integrated circuit. Server 3 includes, for example, a storage unit 21, an acquisition unit 22, a pre-processing unit 23, an OCV estimation unit 24, an optimization unit 25, and a diagnostic unit 26. The memory unit 21 stores various types of information, such as information about the secondary battery 11 that the server 3 acquires in advance or receives from the vehicle 2 at an appropriate time, and information generated by the server 3, as well as a predetermined program.

[0038] The acquisition unit 22 acquires, for example, time-series data such as voltage, current, and temperature of the secondary battery 11 from the vehicle 2. The voltage of the secondary battery 11 is, for example, the closed-circuit voltage (CCV). The acquisition unit 22 acquires the discharge capacity (discharge amount) by, for example, integrating the time-series data of the current.

[0039] The preprocessing unit 23 performs processing such as cleansing and filtering of the time-series data acquired by the acquisition unit 22. For example, the preprocessing unit 23 excludes data that is missing or abnormal from the time-series data.

[0040] The OCV estimation unit 24 performs, for example, the acquisition of a machine learning model and the estimation of the open-circuit voltage (OCV) using the machine learning model. The OCV estimation unit 24 acquires a machine learning model based, for example, on tests performed on a secondary battery 11 whose degradation state is known or simulations performed on a predetermined model of the secondary battery 11. The OCV estimation unit 24 acquires time-series data (first data) of the current, closed-circuit voltage (CCV), and open-circuit voltage (OCV) of the secondary battery 11 in a predetermined state including at least one of charging and discharging, for example, by tests or simulations. Based on the acquired time-series data, the OCV estimation unit 24 acquires a machine learning model that outputs the overvoltage of the secondary battery 11 at an arbitrary time (i.e., the difference between the open-circuit voltage and the closed-circuit voltage) in response to input of current and closed-circuit voltage (CCV) data (second data) at an appropriate time. The machine learning model is, for example, a regression model such as a random forest, support vector machine, and neural network.

[0041] Figure 2 shows an example of the information flow in the processing performed by the OCV estimation unit 24 of the secondary battery state estimation device 10 of the embodiment. As shown in Figure 2, the OCV estimation unit 24 includes, for example, an overvoltage output unit 31 and an open-circuit voltage (OCV) calculation unit 32. The overvoltage output unit 31 outputs an overvoltage at a predetermined time t, which is the target variable, in response to input data of current and closed-circuit voltage (CCV) at appropriate times, which are the explanatory variables, using, for example, a pre-acquired machine learning model. For example, the current and closed-circuit voltage (CCV) data includes data at the target predetermined time t and data at at least one of a predetermined first time interval (predetermined pre-time interval) before the predetermined time t and a predetermined second time interval (predetermined post-time interval) after the predetermined time t (at least one of the pre-interval data and the post-interval data). The predetermined first time interval is, for example, a time interval from a time (tn) at any time n to the predetermined time t. The predetermined second time interval is, for example, a time interval from a predetermined time t to a time (t+k) at any time k. The overvoltage at the predetermined time t is, for example, the difference between the open-circuit voltage (OCV(t)) and the closed-circuit voltage (CCV(t)) at the predetermined time t (=(OCV(t)-CCV(t)).

[0042] The open-circuit voltage (OCV) calculation unit 32 outputs the open-circuit voltage (OCV(t)) at a predetermined time t by adding, for example, the overvoltage at a predetermined time t output from the overvoltage output unit 31 and the closed-circuit voltage (CCV(t)) at a predetermined time t, which is an explanatory variable. Figure 3 shows an example of the correspondence between open-circuit voltage (OCV) and closed-circuit voltage (CCV) obtained by the OCV estimation unit 24 of the secondary battery state estimation device 10 of the embodiment. As shown in Figure 3, for example, the open-circuit voltage (OCV(t)) at a predetermined time t can be obtained based on the current and closed-circuit voltage (CCV) data for the time interval from time (tn) to time (t+k) that includes a predetermined time t.

[0043] Figure 4 shows an example of an OCV curve obtained by the optimization unit 25 of the secondary battery state estimation device 10 in the embodiment, based on the OCP curves of the positive electrode and the negative electrode, respectively. As shown in Figure 4, the optimization unit 25 obtains an OCP curve (OCP curve) that shows the change in open-circuit potential (OCP) according to the discharge capacity x (Ah) of the positive and negative electrodes of the secondary battery 11, based on, for example, several parameters relating to the state of the secondary battery 11. The optimization unit 25 obtains a positive electrode OCP curve (=fca(x)) and a negative electrode OCP curve (=fan(x)) by, for example, applying several parameters to an OCP curve (reference OCP curve) that has been stored in the memory unit 21 in advance.

[0044] The reference OCP curves stored in the memory unit 21 are obtained, for example, through pre-conducted tests or simulations using appropriate models. The reference OCP curves are, for example, the individual OCP curves for each active material constituting the positive electrode and negative electrode of the secondary battery 11. The optimization unit 25 estimates an OCV curve (=fca(x)-fan(x)) that shows the change in open-circuit voltage (OCV) according to the discharge capacity x (Ah), based, for example, the difference between the positive electrode OCP curve (=fca(x)) and the negative electrode OCP curve (=fan(x)).

[0045] The optimization unit 25 optimizes (resets) several parameters related to the state of the secondary battery 11 based on, for example, historical data obtained by the OCV estimation unit 24 from the open-circuit voltage (OCV(t)) at a predetermined time t, and the OCV curve estimated based on the OCP curve. The optimization unit 25 performs a predetermined optimization process based on an error function that shows the error between the estimated OCV curve and the historical data of the secondary battery 11. The error function is, for example, the weighted mean squared error (Weighted RMSE) or the weighted mean absolute error (Weighted MAE). The predetermined optimization process is, for example, a local optimization algorithm such as the BFGS method, the conjugate gradient method and the COBYLA method, or a global optimization algorithm such as a genetic algorithm, the differential evolution method, the SHGO method and the simulated annealing method. In a series of processes including the predetermined optimization process, for example, the optimization unit 25 repeatedly resets multiple parameters, obtains the positive electrode OCP curve and the negative electrode OCP curve, and estimates the OCV curve so that the value of the error function is less than or equal to a predetermined value. Note that the historical data of the open-circuit voltage (OCV(t)) is not limited to a series of data such as time-series data, but rather data acquired over appropriate periods.

[0046] The diagnostic unit 26 obtains diagnostic values ​​related to the degradation state of the secondary battery 11 based on the OCV curve estimated based on the OCP curve after the optimization of multiple parameters by the optimization unit 25. The diagnostic unit 26 takes the fully charged capacity of the secondary battery 11 in its initial state as 100%, and uses the percentage of the fully charged capacity at the time of degradation as the State of Health (SOH) diagnostic value. The fully charged capacity at the time of degradation is, for example, the difference between the discharge capacity at the fully charged voltage and the discharge capacity at the completely discharged voltage, which are obtained based on the OCV curve. The diagnostic unit 26 stores the history data of the SOH diagnostic values ​​in the storage unit 21 by associating the acquired SOH diagnostic values ​​with the date and time on which the OCV curve was obtained by the optimization unit 25.

[0047] (Operation of the secondary battery state estimation device) The operation of the secondary battery state estimation device 10 of this embodiment, in particular the process by which the OCV estimation unit 24 obtains the open-circuit voltage (OCV(t)) at a predetermined time t, will be described below. Figure 5 is a flowchart showing the process performed by the secondary battery state estimation device 10 in the embodiment.

[0048] As shown in Figure 5, first, the OCV estimation unit 24 acquires time-series data of the current, closed-circuit voltage (CCV), and open-circuit voltage (OCV) of the secondary battery 11 as training data for acquiring a machine learning model (step S01). Next, the OCV estimation unit 24 acquires a machine learning model that outputs the overvoltage of the secondary battery 11 at any given time, based on the acquired time-series data, in response to inputs of current and closed-circuit voltage (CCV) data at appropriate times (step S02).

[0049] Next, the OCV estimation unit 24 acquires data on current and closed-circuit voltage (CCV) at appropriate times, which are explanatory variables input to a machine learning model (step S03). Next, the OCV estimation unit 24 obtains the target variable, which is the overvoltage at a predetermined time t, by inputting explanatory variables to a machine learning model, for example (step S04). Next, the OCV estimation unit 24 obtains the open-circuit voltage (OCV(t)) at a predetermined time t by adding, for example, the overvoltage at a predetermined time t output from the machine learning model and the closed-circuit voltage (CCV(t)) at a predetermined time t, which is an explanatory variable (step S05). Then, the OCV estimation unit 24 proceeds to the end of the process.

[0050] As described above, according to the system 1 equipped with the secondary battery state estimation device 10 of the embodiment, by using a machine learning model that outputs an overvoltage in response to input data of current and closed-circuit voltage (CCV), it is possible to suppress a decrease in the accuracy of estimating the internal state of the secondary battery 11, even when the overvoltage increases due to an increase in internal resistance due to the degradation of the secondary battery 11. The change in the shape of the OCV curve due to the degradation of the secondary battery 11 is reflected in the closed-circuit voltage (CCV) detected in the secondary battery 11, so that the accuracy of estimating the open-circuit voltage (OCV) can be improved for various diverse degradation modes, not limited to a specific degradation mode.

[0051] By using time-series data as explanatory variables, the accuracy of machine learning models in estimating open-circuit voltage (OCV) can be improved. The open-circuit voltage (OCV) can be estimated using data from before or after a predetermined target time t, thereby improving the versatility of the machine learning model. The open-circuit voltage (OCV) can be estimated without requiring temperature data as an explanatory variable. For example, even if a machine learning model is acquired using only data from the secondary battery 11 before degradation, the accuracy of the OCV estimation can be improved for the secondary battery 11 after degradation.

[0052] (modified version) Modified examples of the embodiments are described below. Note that parts identical to those in the embodiments described above are denoted by the same reference numerals, and their descriptions are omitted or simplified. In the embodiment described above, the secondary battery state estimation device 10 is assumed to be composed of a server 3, but it is not limited to this. For example, at least one of the processes performed by the server 3 may be performed by the battery control unit 13 of the vehicle 2. In other words, the secondary battery state estimation device 10 may be composed of the server 3 and the battery control unit 13, or the battery control unit 13 alone.

[0053] In the embodiment described above, the machine learning model of the OCV estimation unit 24 outputs the overvoltage at a predetermined time t, which is the target variable, in response to input data of current and closed-circuit voltage (CCV) at appropriate times, which are the explanatory variables. However, it is not limited to this. For example, the machine learning model may output the open-circuit voltage (OCV(t)) at a predetermined time t, which is the target variable. In the embodiment described above, the machine learning model of the OCV estimation unit 24 uses current and closed-circuit voltage (CCV) data as explanatory variables, but is not limited to this. For example, the machine learning model may also use temperature data of the secondary battery 11 as an explanatory variable in addition to current and closed-circuit voltage (CCV). In the embodiment described above, the explanatory variables of the machine learning model may be data from different time intervals. In the embodiments described above, the machine learning model is acquired by the OCV estimation unit 24, but this is not limited to this. For example, the model may be acquired by the OCV estimation unit 24 or another device other than the server 3 and then stored in the storage unit 21 of the server 3 or the OCV estimation unit 24.

[0054] In the embodiments described above, the OCP curve and OCV curve represent the change in open-circuit potential (OCP) or open-circuit voltage (OCV) according to the discharge capacity x (Ah), but the invention is not limited to this. For example, instead of discharge capacity (Ah), a capacity-related parameter such as remaining capacity (SOC: State Of Charge) or depth of discharge (DOD: Depth Of Discharge) may be used.

[0055] Furthermore, a program for realizing all or part of the functions of System 1 equipped with the secondary battery state estimation device 10 in this invention may be recorded on a computer-readable recording medium, and all or part of the processing performed by System 1 may be performed by having the computer system read and execute the program recorded on this recording medium. Herein, "computer system" includes hardware such as the OS and peripheral devices. Furthermore, "computer system" also includes a WWW system equipped with a homepage provisioning environment (or display environment). Furthermore, "computer-readable recording medium" refers to portable media such as flexible disks, magneto-optical disks, ROMs and CD-ROMs, and storage devices such as hard disks built into a computer system. Furthermore, "computer-readable recording medium" also includes volatile memory (RAM) inside a computer system that acts as a server or client when a program is transmitted via a network such as the Internet or a communication line such as a telephone line, which holds the program for a certain period of time.

[0056] Furthermore, the above program may be transmitted from a computer system that stores the program in a memory device or the like to another computer system via a transmission medium or by transmission waves within the transmission medium. Here, the "transmission medium" for transmitting the program refers to a medium that has the function of transmitting information, such as a network (communication network) such as the Internet or a communication line (communication line) such as a telephone line. Furthermore, the above program may be for the purpose of realizing a part of the functions described above. Moreover, it may be a so-called differential file (differential program) that can realize the functions described above in combination with a program already recorded in the computer system.

[0057] The embodiments of the present invention are presented as examples and are not intended to limit the scope of the invention. These embodiments can be carried out in various other forms, and various omissions, substitutions, and modifications can be made without departing from the spirit of the invention. These embodiments and their variations are included in the scope and spirit of the invention, as well as in the claims and their equivalents. [Explanation of Symbols]

[0058] 1...System (secondary battery state estimation system), 2...Vehicle, 3...Server, 4...Network, 10...Secondary battery state estimation device, 11...Secondary battery, 12...Battery sensor, 13...Battery control unit, 14...Power control unit, 15...Rotating electric machine, 16...Drive mechanism, 17...Integration processing unit, 21...Storage unit, 22...Acquisition unit, 23...Preprocessing unit, 24...OCV estimation unit (voltage estimation unit, processing unit), 25...Optimization unit, 26...Diagnostic unit.

Claims

1. A learning model constructed to acquire first data having time-series data of the current, closed-circuit voltage, and open-circuit voltage of a secondary battery in a predetermined state including at least one of charging and discharging, and to output the open-circuit voltage of the secondary battery at any given time or the overvoltage which is the difference between the open-circuit voltage and the closed-circuit voltage, based on the first data, A voltage estimation unit estimates the open-circuit voltage or overvoltage of the secondary battery at any given time by inputting second data, which includes data on the current and closed-circuit voltage detected at a predetermined time in the secondary battery to be estimated, into the learning model. Equipped with A system for estimating the state of a secondary battery.

2. The second data mentioned above is The data includes time-series data of the current and closed-circuit voltage detected in the secondary battery being estimated. The secondary battery state estimation system according to claim 1.

3. The voltage estimation unit, By inputting data from a previous time interval, which includes current and closed-circuit voltage data for a predetermined time interval prior to a predetermined time detected in the secondary battery to be estimated, into the learning model, the open-circuit voltage or overvoltage of the secondary battery to be estimated at the predetermined time is estimated. The secondary battery state estimation system according to claim 2.

4. The voltage estimation unit, By inputting the subsequent interval data, which includes current and closed-circuit voltage data for a predetermined time interval after a predetermined time detected in the secondary battery to be estimated, into the learning model, the open-circuit voltage or overvoltage of the secondary battery to be estimated at the predetermined time is estimated. The secondary battery state estimation system according to claim 2 or claim 3.

5. The second data mentioned above is We do not have data on the temperature of the secondary battery being estimated. A secondary battery state estimation system according to claim 1 or claim 2.

6. The aforementioned learning model, Outputting the overvoltage of the aforementioned secondary battery A secondary battery state estimation system according to claim 1 or claim 2.

7. The aforementioned learning model, Outputting the open-circuit voltage of the secondary battery. A secondary battery state estimation system according to claim 1 or claim 2.

8. The first data is, The data includes time-series data of the current, closed-circuit voltage, and open-circuit voltage of the degraded secondary battery. A secondary battery state estimation system according to claim 1 or claim 2.

9. A method for estimating the state of a secondary battery, which is performed by an electronic device equipped with a processing unit that estimates the open-circuit voltage or overvoltage of a secondary battery to be estimated, A model acquisition step involves acquiring first data having time-series data of the current, closed-circuit voltage, and open-circuit voltage of a secondary battery in a predetermined state including at least one of charging and discharging, and acquiring a learning model that outputs the open-circuit voltage of the secondary battery at an arbitrary time or the overvoltage which is the difference between the open-circuit voltage and the closed-circuit voltage, based on the first data. A voltage estimation step is performed by inputting second data, which has current and closed-circuit voltage data detected at a predetermined time in the secondary battery to be estimated, into the learning model acquired in the model acquisition step, thereby estimating the open-circuit voltage or overvoltage of the secondary battery to be estimated at any given time. A method for estimating the state of a secondary battery, including the following.

10. A computer in an electronic device equipped with a processing unit for estimating the open-circuit voltage or overvoltage of a secondary battery under estimation, A model acquisition step involves acquiring first data having time-series data of the current, closed-circuit voltage, and open-circuit voltage of a secondary battery in a predetermined state including at least one of charging and discharging, and acquiring a learning model that outputs the open-circuit voltage of the secondary battery at an arbitrary time or the overvoltage which is the difference between the open-circuit voltage and the closed-circuit voltage, based on the first data. A voltage estimation step is performed by inputting second data, which has current and closed-circuit voltage data detected at a predetermined time in the secondary battery to be estimated, into the learning model acquired in the model acquisition step, thereby estimating the open-circuit voltage or overvoltage of the secondary battery to be estimated at any given time. A program that executes the command.

Images