Storage battery degradation estimation device and storage battery degradation estimation method

Abstract

This degradation estimation device comprises: a first arithmetic unit (voltage change amount calculation unit (22)) for calculating the voltage change amount of a storage battery (1) obtained from a first voltage value and a second voltage value that are measured when the storage battery (1) is charged / discharged; a second arithmetic unit (electric charge change amount calculation unit (25)) for calculating the electric charge change amount of the storage battery (1) obtained from a first current value and a second current value that are measured when the storage battery (1) is charged / discharged; a measurement data storing unit (31) for storing a plurality of measurement data sets including the voltage change amounts and the electric charge change amounts; an estimation model storing unit (34) for storing an estimation model for outputting a degradation state of the storage battery (1) using, as inputs, at least one of the voltage change amounts and at least one of the electric charge change amounts; and an estimation unit (degradation state estimation unit (321)) for estimating, by use of the estimation model, the degradation state of the storage battery (1) using, as inputs, at least one of the voltage change amounts and at least one of the electric charge change amounts.

Description

Battery deterioration estimation device and battery deterioration estimation method

[0001] The present disclosure relates to a storage battery deterioration estimation device and a storage battery deterioration estimation method.

[0002] In recent years, devices that run on battery power sources such as storage batteries have been used in a wide range of fields. To use these devices safely, it is necessary to prevent situations where the storage battery runs out of power or breaks down, and it is therefore necessary to accurately understand the deterioration state of the storage battery.

[0003] The state of health (SOH) is used as an index to represent the deterioration state of a storage battery. A storage battery deteriorates with continued use, and has the characteristic that the full charge capacity after deterioration is smaller than the initial full charge capacity. The state of health (SOH) is defined as the ratio of the full charge capacity after deterioration to the initial full charge capacity, and the smaller the value of the state of health (SOH), the more the deterioration of the storage battery progresses.

[0004] One method for calculating the state of health (SOH) of a storage battery is the coulomb counting method, in which a fully discharged storage battery is charged up to full charge, and the full charge capacity is calculated by integrating the charging current, and then dividing the calculated value by the initial full charge capacity.

[0005] Another method for calculating the capacity maintenance rate (SOH) of a storage battery is the SOC-OCV curve method, which measures the open circuit voltage (OCV) voltage when fully charged, calculates the full charge capacity from the state of charge (SOC)-OCV curve, and divides it by the initial full charge capacity.

[0006] Furthermore, as another method for calculating the state of health (SOH) of a storage battery, a machine learning method has been developed in recent years that applies artificial intelligence (AI) technology and calculates the state of health (SOH) of a storage battery using an estimation model trained by machine learning, with various states of the storage battery as input and the state of health (SOH) of the storage battery as output. Patent Documents 1 and 2 disclose devices for estimating deterioration of a storage battery using a machine learning method.

[0007] JP 2021-185354 A International Publication No. 2021 / 241115

[0008] However, the above-described method for calculating the capacity maintenance rate (SOH) of a storage battery using the coulomb counting method, the method for calculating the capacity maintenance rate (SOH) of a storage battery using the SOC-OCV curve method, and the method for calculating the capacity maintenance rate (SOH) of a storage battery using the machine learning method disclosed in Patent Document 1 and Patent Document 2 have the following problems.

[0009] The coulomb counting method for calculating the state of health (SOH) of a battery requires the battery to be fully charged from a completely discharged state, which is inconvenient to perform in a device equipped with a battery. Furthermore, calculations using this method are difficult when the device equipped with a battery is in constant operation.

[0010] Furthermore, in the method of calculating the capacity maintenance rate (SOH) of a storage battery using the SOC-OCV method, the characteristics of the SOC-OCV curve fluctuate depending on the temperature and the deterioration state of the storage battery, resulting in calculation errors due to these characteristic fluctuations. To address the characteristic fluctuations of the SOC-OCV curve, it is necessary to switch between multiple SOC-OCV curves, but even in this case, calculation errors occur due to incorrect judgment of the switching. Furthermore, in addition to the fact that the calculation can only be performed when the battery is fully charged, it is necessary to measure the OCV voltage, making it difficult to calculate when the device equipped with the storage battery is constantly operating.

[0011] Furthermore, the methods of calculating the state of health (SOH) of a storage battery using machine learning methods disclosed in Patent Documents 1 and 2 require measurement data of the storage battery when the storage battery has been fully charged multiple times or until the storage battery has been fully charged multiple times, making it difficult to instantly calculate the state of health (SOH) of the storage battery while it is in use.

[0012] Therefore, the present disclosure has been made in consideration of the above-mentioned problems, and provides a storage battery deterioration estimation device and a storage battery deterioration estimation method that can estimate the deterioration state even when the storage battery is in use and can estimate the deterioration state with higher accuracy.

[0013] The storage battery degradation estimation device disclosed herein includes a first calculation unit that calculates a voltage change amount of the storage battery obtained from a first voltage value measured when the storage battery is charged and discharged and a second voltage value that is further measured; a second calculation unit that calculates a charge change amount of the storage battery obtained from a first current value measured when the storage battery is charged and discharged and a second current value that is further measured; a measurement data memory unit that stores a plurality of measurement data including the voltage change amount and the charge change amount; an estimation model memory unit that stores an estimation model learned by machine learning, which receives one or more voltage change amounts and one or more charge change amounts as input and outputs the degradation state of the storage battery; and an estimation unit that uses the estimation model stored in the estimation model memory unit to receive one or more voltage change amounts and one or more charge change amounts stored in the measurement data memory unit as input and estimates the degradation state of the storage battery.

[0014] The storage battery degradation estimation device disclosed herein also includes a first calculation unit that calculates a voltage change amount of the storage battery obtained from a first voltage value measured when the storage battery is charged and discharged and a second voltage value that is further measured; a second calculation unit that calculates a charge change amount of the storage battery obtained from a first current value measured when the storage battery is charged and discharged and a second current value that is further measured; a measurement data memory unit that stores a plurality of measurement data including the voltage change amount and the charge change amount; an estimation model memory unit that stores an estimation model learned by machine learning, which receives one or more voltage change amounts and one or more charge change amounts as input and outputs a degradation state of the storage battery; an estimation unit that receives one or more voltage change amounts and one or more charge change amounts stored in the measurement data memory unit as input and estimates the degradation state of the storage battery using the estimation model stored in the estimation model memory; and a low-pass filter unit that receives the degradation state of the storage battery output from the estimation unit as input and smooths and outputs the degradation state of the storage battery using a low-pass filter.

[0015] The deterioration estimation device for a storage battery according to the present disclosure includes a first low-pass filter unit that receives as input a voltage measured when the storage battery is charged and discharged, smooths the voltage using a low-pass filter, and outputs a smoothed voltage; a first calculation unit that calculates a smoothed voltage change amount of the storage battery obtained from a first smoothed voltage value output from the first low-pass filter unit and a second smoothed voltage value that is further output; a second low-pass filter unit that receives as input a current measured when the storage battery is charged and discharged, smooths the current using a low-pass filter, and outputs a smoothed current; and a first calculation unit that calculates a smoothed voltage change amount of the storage battery obtained from a first smoothed voltage value output from the first low-pass filter unit and a second smoothed voltage value that is further output. the second smoothed current value calculated based on the smoothed voltage change amount and the smoothed charge change amount; a measurement data storage unit that stores a plurality of measurement data including the smoothed voltage change amount and the smoothed charge change amount; an estimation model storage unit that stores an estimation model learned by machine learning, which receives one or more smoothed voltage change amounts and one or more smoothed charge change amounts as input and outputs the degradation state of the storage battery; and an estimation unit that receives one or more smoothed voltage change amounts and one or more smoothed charge change amounts stored in the measurement data storage unit as input and estimates the degradation state of the storage battery using the estimation model stored in the estimation model storage unit.

[0016] The storage battery degradation estimation device of the present disclosure also includes a first low-pass filter unit that receives as input a voltage measured when the storage battery is charged or discharged, smooths the voltage using a low-pass filter, and outputs a smoothed voltage; a first calculation unit that calculates a smoothed voltage change amount of the storage battery obtained from a first smoothed voltage value output from the first low-pass filter unit and a second smoothed voltage value that is further output; a second low-pass filter unit that receives as input a current measured when the storage battery is charged or discharged, smooths the current using a low-pass filter, and outputs a smoothed current; and a calculation unit that calculates a smoothed charge change amount of the storage battery obtained from the first smoothed current value output from the second low-pass filter unit and a second smoothed current value that is further output. The storage battery includes a second calculation unit, a measurement data storage unit that stores a plurality of measurement data including a smoothed voltage change amount and a smoothed charge change amount, an estimation model storage unit that stores an estimation model learned by machine learning, which receives one or more smoothed voltage change amounts and one or more of the charge change amounts as input and outputs a degradation state of the storage battery, an estimation unit that receives one or more smoothed voltage change amounts and one or more smoothed charge change amounts stored in the measurement data storage unit as input using the estimation model stored in the estimation model storage unit, and estimates the degradation state of the storage battery, and a third low-pass filter unit that receives the degradation state of the storage battery output from the estimation unit as input, smooths the degradation state of the storage battery using a low-pass filter, and outputs the result.

[0017] Furthermore, the deterioration state of the storage battery in the storage battery deterioration estimation device of the present disclosure is preferably the capacity maintenance rate of the storage battery or the full charge capacity value of the storage battery.

[0018] It is also preferable to further include a receiving and communicating unit that receives a new estimation model from outside, and an updating unit that updates the estimation model stored in the estimation model storage unit to the new estimation model received from the receiving and communicating unit.

[0019] It is also preferable to further include an estimated data storage unit that stores the deterioration state of the storage battery estimated by the estimation unit, and a transmission and communication unit that transmits to the outside the multiple measurement data stored in the measurement data storage unit and the deterioration state of the storage battery stored in the estimated data storage unit.

[0020] In addition, the method for estimating deterioration of a storage battery disclosed herein includes a calculation step of calculating a voltage change amount of the storage battery obtained from a first voltage value measured when the storage battery is charged and discharged and a second voltage value further measured; a calculation step of calculating a charge change amount of the storage battery obtained from a first current value measured when the storage battery is charged and discharged and a second current value further measured; and an estimation step of estimating the deterioration state of the storage battery by inputting one or more voltage change amounts and one or more charge change amounts using an estimation model trained by machine learning, which inputs one or more voltage change amounts and one or more charge change amounts and outputs the deterioration state of the storage battery.

[0021] In addition, the storage battery deterioration estimation method disclosed herein includes a calculation step of calculating a voltage change amount of the storage battery obtained from a first voltage value measured when the storage battery is charged and discharged and a second voltage value further measured; a calculation step of calculating a charge change amount of the storage battery obtained from a first current value measured when the storage battery is charged and discharged and a second current value further measured; an estimation step of estimating the deterioration state of the storage battery by inputting one or more voltage change amounts and one or more charge change amounts using an estimation model trained by machine learning, which inputs one or more voltage change amounts and one or more charge change amounts and outputs the deterioration state of the storage battery; and a calculation step of inputting the deterioration state of the storage battery and calculating the deterioration state of the storage battery smoothed using a low-pass filter.

[0022] Furthermore, the method for estimating deterioration of a storage battery disclosed herein includes a calculation step of taking as input a voltage measured when the storage battery is charged or discharged, and calculating a first smoothed voltage value obtained by smoothing the voltage using a low-pass filter, and a second smoothed voltage value obtained therefrom; a calculation step of calculating a smoothed voltage change amount of the storage battery obtained from the first smoothed voltage value and the second smoothed voltage value; a calculation step of taking as input a current measured when the storage battery is charged or discharged, and calculating a first smoothed current value obtained by smoothing the current using a low-pass filter, and a second smoothed current value obtained therefrom; a calculation step of calculating a smoothed charge change amount obtained from the first smoothed current value and the second smoothed current value; and an estimation step of taking as input one or more smoothed voltage change amounts and one or more smoothed charge change amounts, and estimating the deterioration state of the storage battery, using an estimation model trained by machine learning, which takes as input one or more smoothed voltage change amounts and one or more smoothed charge change amounts, and outputs the deterioration state of the storage battery.

[0023] The method for estimating deterioration of a storage battery according to the present disclosure includes a calculation step of taking a voltage measured when the storage battery is charged and discharged as an input, and calculating a first smoothed voltage value obtained by smoothing the voltage using a low-pass filter, and a second smoothed voltage value obtained therefrom; a calculation step of calculating a smoothed voltage change amount of the storage battery obtained from the first smoothed voltage value and the second smoothed voltage value; and a calculation step of taking a current measured when the storage battery is charged and discharged as an input, and calculating a first smoothed current value obtained by smoothing the current using a low-pass filter, and a second smoothed current value obtained therefrom. the first smoothed current value and the second smoothed current value; an estimation step of estimating the degradation state of the storage battery by using an estimation model trained by machine learning, which receives as input one or more smoothed voltage changes and one or more smoothed charge changes and outputs the degradation state of the storage battery; and a calculation step of receiving as input the degradation state of the storage battery and calculating the degradation state of the storage battery smoothed using a low-pass filter.

[0024] In addition, the deterioration state of the storage battery in the method for estimating deterioration of a storage battery according to the present disclosure is preferably the capacity maintenance rate of the storage battery or the full charge capacity value of the storage battery.

[0025] According to the storage battery degradation estimation device and degradation estimation method disclosed herein, it is possible to estimate the degradation state of the storage battery even when it is in use, and to estimate the degradation state with higher accuracy.

[0026] FIG. 1 is a diagram illustrating the overall configuration of a storage battery degradation estimation device according to a first embodiment of the present disclosure. FIG. 2 is a diagram illustrating the capacity maintenance rate (SOH) of a storage battery. FIG. 3 is a diagram illustrating calculation of a voltage measured when a storage battery is charged and discharged, and an amount of voltage change. FIG. 4 is a diagram illustrating calculation of a current measured when a storage battery is charged and discharged, and an amount of charge change. FIG. 5 is a diagram illustrating a configuration of a degradation state estimation unit according to an embodiment of the present disclosure. FIG. 6 is a diagram illustrating another configuration of a degradation state estimation unit according to an embodiment of the present disclosure. FIG. 7 is a flowchart illustrating a storage battery degradation estimation method according to the first embodiment of the present disclosure. FIG. 8 is a diagram illustrating a degradation estimation result of a storage battery according to the first embodiment of the present disclosure. FIG. 9 is a diagram illustrating the overall configuration of a storage battery degradation estimation device according to a second embodiment of the present disclosure. FIG. 10 is a flowchart illustrating a storage battery degradation estimation method according to the second embodiment of the present disclosure. FIG. 11 is a diagram illustrating a degradation estimation result of a storage battery according to the second embodiment of the present disclosure. FIG. 12 is a diagram illustrating the overall configuration of a storage battery degradation estimation device according to a third embodiment of the present disclosure. Fig. 13 is a diagram showing voltages measured when a storage battery according to a third embodiment of the present disclosure is charged and discharged, and a smoothed voltage smoothed using a low-pass filter. Fig. 14 is a flowchart showing a method for estimating deterioration of a storage battery according to a third embodiment of the present disclosure. Fig. 15 is a diagram showing the overall configuration of a storage battery deterioration estimation device according to a fourth embodiment of the present disclosure. Fig. 16 is a flowchart showing a method for estimating deterioration of a storage battery according to the fourth embodiment of the present disclosure.

[0027] The storage battery degradation estimation device and storage battery degradation estimation method disclosed herein use an estimation model trained by machine learning, which inputs voltage change and charge change amounts obtained from voltage and current values ​​that can be measured while the storage battery is in use, and outputs the storage battery's degradation state. This allows the storage battery's degradation state to be directly calculated from the voltage change and charge change amounts. This eliminates the need for the prior art to calculate the charge capacity required to fully charge the storage battery from a fully discharged state, or to measure the OCV voltage after fully charging the storage battery, making it possible to estimate the degradation state even when the storage battery is in use. Furthermore, since it is not necessary to calculate the full charge capacity using an SOC-OCV curve, which is prone to calculation errors due to characteristic fluctuations, it is possible to estimate the degradation state with higher accuracy. Additionally, by preparing a large amount of training data that inputs voltage change and charge change amounts and outputs the storage battery's degradation state, and training the estimation model using machine learning, it is possible to estimate the degradation state under various storage battery usage conditions, making it possible to estimate the degradation state with higher accuracy.

[0028] Furthermore, by configuring the system so that the deterioration state of the storage battery calculated using the estimation model is input and the deterioration state of the storage battery is smoothed and output using a low-pass filter, even if the deterioration state of the storage battery calculated using the estimation model has a large error compared to the actual deterioration state of the storage battery at a certain time, the output data is smoothed and output by filter processing so that it follows the output data that was previously output, making it possible to estimate the deterioration state with less error.

[0029] Furthermore, the degradation state of the battery is calculated using an estimation model trained by machine learning, which takes as input the voltage and current measured when the storage battery is charged and discharged, smooths the voltage and current of the storage battery using a low-pass filter to output a smoothed voltage and smoothed current, and takes as input the smoothed voltage change and smoothed charge change obtained from the smoothed voltage and smoothed current, and outputs the degradation state of the storage battery.Even if the voltage and current measured from the storage battery are affected by sudden fluctuations in the charging and discharging current of the storage battery, measurement noise, etc., and discrete voltage and current fluctuations that are not continuous for each measurement occur, the discrete voltage and current fluctuations are smoothed using the low-pass filter to follow the previously measured voltage and current.As a result, the voltage change and charge change amounts input to the estimation model that estimates the degradation state of the storage battery become continuous data, making it possible to eliminate the effects of noise, etc. from the voltage change and charge change amounts, and enable the degradation state of the storage battery to be estimated with less error.

[0030] Furthermore, by configuring the degradation estimation device to include a receiving and communication unit that receives an estimation model that estimates the degradation state of the storage battery from the outside, and an update unit that updates the estimation model stored in the degradation estimation device to a new estimation model received from the receiving and communication unit, even after the degradation estimation device has been installed in a product, it is possible to update the estimation model to one that has been learned by adding new learning data thereafter and can estimate the degradation state of the storage battery with higher accuracy, thereby making it possible to estimate the degradation state with higher accuracy.

[0031] Furthermore, by configuring the degradation estimation device to include a measurement data storage unit that stores multiple pieces of measurement data measured and calculated by the degradation estimation device, an estimation data storage unit that stores the degradation state of the storage battery estimated by the degradation estimation device, and a transmission and communication unit that transmits the multiple pieces of measurement data stored in the measurement data storage unit and the degradation state of the storage battery stored in the estimation data storage unit to an external device, it becomes possible to collect measurement data of the storage battery when the degradation estimation device is used and estimated data of the degradation state of the storage battery even after the degradation estimation device is installed in a product, thereby making it possible to add learning data for various usage states of the storage battery.This makes it possible to generate and update a new estimation model with higher accuracy, thereby making it possible to estimate the degradation state with higher accuracy.

[0032] Hereinafter, embodiments of the present disclosure will be described with reference to the drawings.

[0033] Note that the embodiments described below are all comprehensive or specific examples. The numerical values, shapes, components, component placement and connection configurations, steps, and step orders shown in the following embodiments are merely examples and are not intended to limit the present disclosure. For example, numerical values ​​are not expressions that express only strict meanings, but expressions that include a substantially equivalent range, for example, a difference of a few percent. Furthermore, among the components in the following embodiments, components not described in independent claims are described as optional components.

[0034] Furthermore, each figure is a schematic diagram and is not necessarily an exact illustration. Therefore, for example, the scales of the figures do not necessarily match. Furthermore, in each figure, substantially the same components are given the same reference numerals, and redundant explanations are omitted or simplified.

[0035] Furthermore, in this specification, numerical values ​​and numerical ranges are not expressions that express only the strict meaning, but are expressions that mean that they also include a substantially equivalent range, for example, a difference of about several percent (e.g., about 10%).

[0036] First Embodiment [Overall Configuration of a Storage Battery Deterioration Estimation Device] FIG. 1 is a diagram showing the overall configuration of a storage battery deterioration estimation device according to a first embodiment of the present disclosure.

[0037] The storage battery 1 is configured by connecting a plurality of battery modules in series and in parallel.

[0038] The battery management device 10 is connected to the battery 1 and is a device for measuring the voltage, current, and temperature of the battery 1. The battery management device 10 has an ADC (Analog-Digital Converter) unit 11 that converts analog signal values ​​corresponding to the voltage, current, and temperature of the battery 1 into digital signal values, a calculation unit 12 that processes the digital signal values ​​output from the ADC unit to calculate voltage values, current values, and temperatures, and a communication I / F 13 that transmits the voltage values, current values, and temperatures calculated by the calculation unit 12 to the outside.

[0039] The battery state estimation device 100a includes a communication I / F 41 that receives the voltage value, current value, and temperature of the battery 1 from the battery management device 10, a battery measurement unit 20a that processes the data received from the communication I / F 41 to calculate various measurement values ​​of the battery 1, a battery state estimation unit 30a that receives the various measurement values ​​calculated by the battery measurement unit 20a and estimates the degradation state of the battery 1, and a communication I / F 42 that transmits and receives the degradation state of the battery 1 calculated by the battery state estimation unit 30a, an estimation model that estimates the degradation state of the battery 1, and measurement data and estimation data of the battery 1 to and from an external device. The battery state estimation device 100a is an example of a degradation estimation device.

[0040] The storage battery measuring unit 20a includes a measurement quantity acquiring unit 21 that acquires voltage values, current values, and temperature values ​​from the communication I / F 41; a voltage change amount calculating unit 22 that uses data stored in the measurement quantity acquiring unit 21 to calculate a voltage change amount from a start voltage value which is a first voltage value and an end voltage value which is a second voltage value; a current change amount calculating unit 23 that calculates a current change amount from a start current value which is a first current value and an end current value which is a second current value; a charge amount calculating unit 24 that calculates the charge amount of the storage battery 1 by integrating the current values; a charge change amount calculating unit 25 that calculates a charge change amount from the start charge amount and end charge amount of the storage battery 1 using the charge amount calculated by the charge amount calculating unit 24; and a measurement start / stop control unit 26 that controls the start conditions and end conditions required to calculate the above-mentioned voltage change amount, current change amount, and charge change amount. That is, the battery measurement unit 20a calculates various measurement parameters such as the voltage change amount, the current change amount, and the charge change amount of the battery 1 using the voltage value and the current value, which are measurement data of the battery 1. This measurement data is transmitted to the battery state estimation unit 30a, which will be described later.

[0041] The battery state estimation unit 30a includes a measurement data memory unit 31 that stores measurement data acquired and calculated by the battery measurement unit 20a, a degradation state calculation unit 32a that estimates the degradation state of the storage battery 1 using the measurement data stored in the measurement data memory unit 31 as input, a degradation state estimation unit 321 that constitutes the degradation state calculation unit 32a, an estimation data memory unit 33 that stores the degradation state of the storage battery 1 estimated by the degradation state calculation unit 32a, an estimation model memory unit 34 that stores an estimation model learned by machine learning and used when the degradation state estimation unit 321 estimates the degradation state of the storage battery 1, and an estimation model update unit 35 that receives a new estimation model from outside and updates the estimation model stored in the estimation model memory unit 34.

[0042] The communication I / F 51 transmits and receives measurement data and estimation data of the storage battery 1, and an estimation model for estimating the degradation state of the storage battery 1, via the communication I / F 42 of the storage battery state estimation device 100a.

[0043] The degradation state display unit 52 is a device that displays the degradation state of the storage battery 1 estimated by the battery state estimation device 100a. The estimation model transmission unit 53 is a device that transmits a new estimation model that estimates the degradation state of the storage battery 1 to the battery state estimation device 100a. The measurement data / estimated data reception unit 54 is a device that receives data measured and estimated by the battery state estimation device 100a.

[0044] The deterioration estimation device for a storage battery 1 according to the first embodiment of the present disclosure is characterized in that the deterioration state estimation unit 321 of the storage battery state estimation unit 30a estimates the state of health (SOH) of the storage battery 1 using an estimation model learned by machine learning, in which the voltage change amount and charge change amount of the storage battery 1 are input and the state of health (SOH), which represents the deterioration state of the storage battery 1, is output.

[0045] [Capacity Maintenance Rate (SOH) of Storage Battery] FIG. 2 is a diagram illustrating the capacity maintenance rate (SOH) of the storage battery 1. As shown in FIG.

[0046] If the initial full charge capacity of the storage battery 1 is Cinit, continued use of the storage battery 1 causes the storage battery 1 to deteriorate, and the full charge capacity Ct after deterioration becomes smaller than the initial full charge capacity Cinit. Figure 2 shows an example in which the full charge capacity Ct after deterioration has decreased to 80% of the initial full charge capacity Cinit. The capacity maintenance rate (SOH) (%) of the storage battery 1 is the ratio of the full charge capacity Ct after deterioration to the initial full charge capacity Cinit, and is expressed by the following equation 1.

[0047] SOH=Ct / Cinit×100 (Formula 1)

[0048] In an embodiment of the present disclosure, a deterioration estimation device and a deterioration estimation method for a storage battery 1 are described that can estimate the deterioration state even when the storage battery 1 is in use and can estimate the capacity maintenance rate (SOH) of the storage battery 1 with higher accuracy.

[0049] [Calculation of Voltage Change and Charge Change of Storage Battery] A method for calculating the voltage change and charge change of the storage battery 1, which are inputs when calculating the state of health (SOH) of the storage battery 1 using an estimation model learned by machine learning, will be described.

[0050] 3 is a diagram illustrating the voltage measured when the storage battery 1 according to an embodiment of the present disclosure is charged and discharged, and the calculation of the voltage change amount. By measuring the voltage of the storage battery 1 from time to time, it is possible to grasp the voltage fluctuation of the storage battery 1. The voltage value V1 at time T1 and the voltage value V2 at a subsequent time T2 are obtained, and the voltage change amount ΔV is obtained by calculating the difference V2-V1 between the voltage value V1 and the voltage value V2. By repeating this operation, multiple voltage change amounts according to the usage state of the storage battery 1 can be obtained.

[0051] 4 is a diagram illustrating the current measured when the storage battery 1 according to an embodiment of the present disclosure is charged and discharged, and the calculation of the charge change amount. By measuring the current of the storage battery 1 from time to time, it is possible to grasp the current fluctuation of the storage battery 1. The current value I1 at time T1 and the current value I2 at a subsequent time T2 are acquired, and the charge change amount ΔQ is obtained by integrating the current values ​​as shown in the shaded area in FIG. 4. By repeating this operation, it is possible to obtain multiple charge fluctuation amounts according to the usage state of the storage battery 1.

[0052] The calculation of the voltage change amount ΔV and charge change amount ΔQ of the storage battery 1 described above is performed by the voltage change amount calculation unit 22 and charge change amount calculation unit 25 of the storage battery measurement unit 20a in the deterioration estimation device for the storage battery 1 shown in Fig. 1. The voltage change amount calculation unit 22 is an example of a first calculation unit, and the charge change amount calculation unit 25 is an example of a second calculation unit.

[0053] [Configuration of the storage battery degradation state estimation unit] An example will be described below regarding the storage battery 1 degradation state estimation unit that calculates the capacity maintenance rate (SOH) of the storage battery 1 using an estimation model learned by machine learning, i.e., the degradation state estimation unit 321 of the degradation state calculation unit 32a of the storage battery state estimation unit 30a in the storage battery 1 degradation estimation device shown in Figure 1.

[0054] 5 is a diagram illustrating a configuration of a degradation state estimation unit according to an embodiment of the present disclosure. The degradation state estimation unit 321 receives one or more voltage change amounts and one or more charge change amounts as input, and calculates the state of health (SOH) of the storage battery 1 using an estimation model trained by machine learning, which outputs the state of health (SOH) of the storage battery 1. The degradation state estimation unit 321 includes a neural network calculation unit 3211, which includes a fully connected neural network (FCN).

[0055] FIG. 6 is a diagram illustrating another configuration of a degradation state estimation unit according to an embodiment of the present disclosure. The degradation state estimation unit 321 in FIG. 5 described above is configured to include only a neural network calculation unit 3211, and one or more voltage change amounts and one or more charge change amounts are directly input to the neural network calculation unit 3211. However, the degradation state estimation unit 321 in FIG. 6 differs in that a feature calculation unit 3212 is added before the neural network calculation unit 3211. That is, the degradation state estimation unit 321 in FIG. 6 is characterized in that one or more voltage change amounts and one or more charge change amounts are input to the feature calculation unit 3212, which calculates new feature amounts, such as average values ​​and difference values ​​of the one or more voltage change amounts and one or more charge change amounts, and inputs the output to the neural network calculation unit 3211. The feature calculation unit 3212 calculates feature amounts optimal for machine learning, and the feature amounts are used as input to estimate the capacity maintenance ratio (SOH) of the storage battery 1 using machine learning.

[0056] In the above, in this embodiment, the case where the estimation model using machine learning of the neural network calculation unit 3211 is a fully connected neural network (FCN) has been described. However, a machine learning estimation model other than a fully connected neural network (FCN), such as a recurrent neural network (RNN), a decision tree type machine learning estimation model, or a multiple regression analysis type machine learning estimation model, may also be used.

[0057] [Flowchart of Storage Battery Deterioration Estimation Method] FIG. 7 is a flowchart showing a storage battery deterioration estimation method according to the first embodiment of the present disclosure.

[0058] First, the voltage change amount calculation unit 22 calculates a voltage change amount from a first voltage value and a second voltage value measured when the storage battery 1 is charged and discharged (S11). Next, the charge change amount calculation unit 25 calculates a charge change amount from a first current value and a second current value measured when the storage battery 1 is charged and discharged (S12). Finally, the degradation state estimation unit 321 estimates a capacity maintenance rate (SOH) of the storage battery 1 from one or more voltage change amounts and one or more charge change amounts using an estimation model (S13). The second voltage value is a voltage value acquired after the first voltage value, and may be a voltage value acquired consecutively after the first voltage value or a voltage value acquired after a predetermined time. Furthermore, the second current value is a current value acquired after the first current value, and may be a current value acquired consecutively after the first current value or a current value acquired after a predetermined time. Furthermore, the first voltage value and the first current value may be acquired at the same position and time, or at different times; similarly, the second voltage value and the second current value may be acquired at the same position and time, or at different times.

[0059] As described above, in this embodiment, the input data for estimating the state of health (SOH) of the storage battery 1 are the voltage (amount of voltage change) and current (amount of charge change) measured when the storage battery 1 is in use, and measurement data of the storage battery 1 in a fully charged state is not required, so it is possible to estimate the state of degradation even when the storage battery 1 is in use. Furthermore, because the state of health (SOH) of the storage battery 1 is calculated using an estimation model trained by machine learning, it is possible to estimate the state of degradation in various usage situations of the storage battery 1 by preparing a large amount of training data and training the estimation model using machine learning, and therefore it is possible to estimate the state of degradation of the storage battery 1 with higher accuracy.

[0060] [Estimation Results of Capacity Maintenance Rate (SOH) of Storage Battery] Applying the first embodiment of the present disclosure, learning data for an estimation model that actually estimates the capacity maintenance rate (SOH) of the storage battery 1 is prepared, and the estimation model is trained. Then, the results of estimating the capacity maintenance rate (SOH) of the storage battery 1 using the estimation model trained by machine learning will be described.

[0061] 8 is a diagram showing the degradation estimation result of the storage battery 1 according to the first embodiment of the present disclosure. In Fig. 8, the number of pieces of training data is approximately 5000 pieces of data, which is an extremely small number of pieces of training data compared to the number of pieces of training data used when developing an estimation model for the degradation state of the storage battery 1 to be installed in an actual product, and this result was obtained for simple verification of the effects of the present disclosure.

[0062] 8 , the horizontal axis represents the actual capacity maintenance factor (SOH) of the storage battery 1, and the vertical axis represents the capacity maintenance factor (SOH) of the storage battery 1 estimated using the estimation model learned by machine learning. That is, in FIG. 8 , when the horizontal axis value and the vertical axis value are plotted in the same place, this indicates that the estimation result of the capacity maintenance factor (SOH) of the storage battery 1 estimated using the estimation model learned by machine learning correctly estimates the capacity maintenance factor (SOH) of the actual storage battery 1.

[0063] As shown in Figure 8, although the number of learning data is approximately 5,000 data items, it can be seen that the capacity maintenance factor (SOH) of the storage battery 1 estimated using the estimation model trained by machine learning is able to estimate the capacity maintenance factor (SOH) of the actual storage battery 1. However, in some cases, the capacity maintenance factor (SOH) of the storage battery 1 estimated using the estimation model trained by machine learning has a maximum error of up to approximately 8% compared to the actual capacity maintenance factor (SOH) of the storage battery 1, and it is possible to reduce this error by optimizing the learning method of the estimation model or by further increasing the amount of learning data and developing an estimation model with higher accuracy.

[0064] The learning data is data in which the voltage (amount of voltage change) and the current (amount of charge change) are input information and the state of health (SOH) at that time is output information.

[0065] [Receiving and updating a new estimation model for estimating the deterioration state of a storage battery from outside] The storage battery deterioration estimation device in an embodiment of the present disclosure is capable of receiving a new estimation model for estimating the capacity maintenance rate (SOH) of the storage battery 1 from outside and updating the estimation model stored in the deterioration estimation device for the storage battery 1.

[0066] 1 , if a new estimation model capable of estimating the capacity maintenance rate (SOH) of the storage battery 1 with higher accuracy is developed, the new estimation model is transmitted from the estimation model transmission unit 53 to the estimation model update unit 35 of the storage battery state estimation unit 30a of the storage battery state estimation device 100a via the communication I / F 51 and the communication I / F 42. Thereafter, the estimation model update unit 35 can update the estimation model stored in the estimation model storage unit 34 to the new estimation model.

[0067] As described above, in this embodiment, even after the deterioration estimation device for the storage battery 1 is installed in a product, it is possible to add new learning data and update the estimation model to one that can estimate the deterioration state of the storage battery 1 with higher accuracy, thereby making it possible to estimate the deterioration state of the storage battery 1 with higher accuracy.

[0068] [Storage of measurement data and estimated data of storage battery and transmission to the outside] The deterioration estimation device for storage battery 1 in an embodiment of the present disclosure is capable of transmitting measurement data of storage battery 1 and estimated data of the deterioration state of storage battery 1 to the outside.

[0069] In the deterioration estimation device for the storage battery 1 shown in Figure 1, the measurement data of the storage battery 1 stored in the measurement data memory unit 31 and the estimated data of the deterioration state of the storage battery 1 stored in the estimated data memory unit 33 can be transmitted to an external measurement data / estimated data receiving unit 54 via the communication I / F 42 and the communication I / F 51.

[0070] As described above, in this embodiment, even after the deterioration estimation device for the storage battery 1 is installed in a product, it is possible to collect measurement data of the storage battery 1 and estimated data on the deterioration state of the storage battery 1. Therefore, it is possible to collect learning data for various usage states of the storage battery 1, and it is possible to learn and update a new estimation model with higher accuracy, making it possible to estimate the deterioration state with higher accuracy.

[0071] Second Embodiment [Overall Configuration of Storage Battery Degradation Estimation Device] Fig. 9 is a diagram showing the overall configuration of a storage battery degradation estimation device according to a second embodiment of the present disclosure. In Fig. 9, the same components as those in Fig. 1, which is the storage battery degradation estimation device according to the first embodiment described above, are given the same names, symbols, and numbers, and detailed descriptions thereof will be omitted.

[0072] 9 , the deterioration estimation device for a storage battery 1 according to the second embodiment differs from that according to the first embodiment in the configuration of a storage battery state estimation unit 30b of a storage battery state estimation device 100b. Specifically, in a deterioration state calculation unit 32b of the storage battery state estimation unit 30b, a deterioration state low-pass filter unit 322 is newly installed at the output of a deterioration state estimation unit 321, and the output of the deterioration state low-pass filter unit 322 is connected to a deterioration state display unit 52 via communication I / F 42 and communication I / F 51, and the output of the deterioration state low-pass filter unit 322 is further connected to an estimation data storage unit 33.

[0073] [Configuration of the Deterioration State Low-Pass Filter Unit of the Storage Battery] The deterioration state low-pass filter unit 322 is characterized by performing filter processing to smooth input data, which is time-series data, and output the smoothed data as output data. Here, where n is an integer value, the nth output data y[n] of the deterioration state low-pass filter unit 322 is determined by the previously output (n-1)th output data y[n-1] and the nth input data x[n], and is expressed by the following equation 2:

[0074] y[n]=k×x[n]+(1-k)×y[n-1]...(Formula 2)

[0075] Here, the constant k is called a time constant, and adopts a value such as k=0.1, 0.01, 0.001, etc. In other words, the degradation state low-pass filter unit 322 functions as a filter processing unit that smooths the input data, which is time-series data, and outputs the smoothed data as output data, in which the nth output data y[n] follows the previously output (n-1)th output data y[n-1] and also reflects the nth input data x[n].

[0076] [Flowchart of Storage Battery Deterioration Estimation Method] FIG. 10 is a flowchart showing a storage battery deterioration estimation method according to the second embodiment of the present disclosure.

[0077] First, the voltage change amount calculation unit 22 calculates a voltage change amount from a first voltage value and a second voltage value measured when the storage battery 1 is charged and discharged (S21). Next, the charge change amount calculation unit 25 calculates a charge change amount from a first current value and a second current value measured when the storage battery 1 is charged and discharged (S22). Next, the degradation state estimation unit 321 estimates a capacity maintenance rate (SOH) of the storage battery 1 from one or more voltage change amounts and one or more charge change amounts using an estimation model (S23). Finally, the degradation state low-pass filter unit 322 calculates a smoothed capacity maintenance rate (SOH) of the storage battery 1 from the capacity maintenance rate (SOH) of the storage battery 1 calculated in step S23 using a low-pass filter (S24).

[0078] As described above, in this embodiment, the degradation state of the storage battery 1 calculated using the estimation model is input, and the degradation state of the storage battery 1 is smoothed and output using a low-pass filter.Therefore, even if the degradation state of the storage battery 1 calculated using the estimation model has a large error compared to the actual degradation state of the storage battery 1 at a certain time, the output data is smoothed and output by filter processing so that it follows the output data that was previously output, making it possible to estimate the degradation state of the storage battery 1 with less error.

[0079] [Estimation Results of Capacity Maintenance Rate (SOH) of Storage Battery] Fig. 11 is a diagram showing the degradation estimation results of a storage battery according to the second embodiment of the present disclosure. In Fig. 11, the learning data used when training the estimation model and the input data used when estimating the capacity maintenance rate (SOH) of the storage battery 1 are the same as those used in the first embodiment. The difference between the degradation estimation result of the storage battery 1 according to the second embodiment shown in Fig. 11 and the degradation estimation result of the storage battery 1 according to the first embodiment shown in Fig. 8 is due to the filtering performed by the degradation state low-pass filter unit 322 in the degradation state calculation unit 32b of the battery state estimation unit 30b in the second embodiment of the present disclosure shown in Fig. 9.

[0080] 11 , the horizontal axis represents the actual capacity maintenance ratio (SOH) of the storage battery 1, and the vertical axis represents the capacity maintenance ratio (SOH) of the storage battery 1 estimated using the estimation model learned by machine learning and smoothed using a low-pass filter. That is, in FIG. 11 , when the horizontal axis value and the vertical axis value are plotted in the same place, this indicates that the capacity maintenance ratio (SOH) of the storage battery 1 smoothed using filter processing correctly estimates the actual capacity maintenance ratio (SOH) of the storage battery 1 compared to the capacity maintenance ratio (SOH) of the storage battery 1 estimated using the estimation model learned by machine learning.

[0081] As shown in FIG. 11 , although the number of training data is the same as in the first embodiment, approximately 5,000 data, the capacity maintenance rate (SOH) of the storage battery 1 estimated using the estimation model trained by machine learning can be seen to be accurate to the actual capacity maintenance rate (SOH) of the storage battery 1. Furthermore, compared to the estimation result of the first embodiment shown in FIG. 8 , the estimation result of the second embodiment shown in FIG. 11 shows a significantly reduced maximum error with respect to the actual capacity maintenance rate (SOH) of the storage battery 1. That is, while the estimation result of the first embodiment shown in FIG. 8 has a maximum error of approximately 8% with respect to the actual capacity maintenance rate (SOH) of the storage battery 1, the estimation result of the second embodiment shown in FIG. 11 shows a reduced maximum error with respect to the actual capacity maintenance rate (SOH) of the storage battery 1 to approximately 5% or less. Even in the second embodiment, some estimations have errors. It is possible to reduce these errors by optimizing the learning method for the estimation model or by further increasing the amount of training data to develop a more accurate estimation model.

[0082] Third Embodiment [Overall Configuration of Storage Battery Degradation Estimation Device] Fig. 12 is a diagram showing the overall configuration of a storage battery degradation estimation device according to a third embodiment of the present disclosure. In Fig. 12, the same components as those in Fig. 1, which is the storage battery degradation estimation device according to the first embodiment described above, are given the same names, symbols, and numbers, and detailed descriptions thereof will be omitted.

[0083] 12 , the deterioration estimation device for a storage battery 1 according to the third embodiment differs from the first embodiment in the configuration of a storage battery measurement unit 20b of a storage battery state estimation device 100c. Specifically, the deterioration estimation device for a storage battery 1 according to the third embodiment newly includes a voltage low-pass filter unit 27 that smooths a voltage value input via a low-pass filter and outputs the smoothed voltage value, and a current low-pass filter unit 28 that smooths a current value input via a low-pass filter and outputs the smoothed current value. In this embodiment, the filtering processes of the voltage low-pass filter unit 27 and the current low-pass filter unit 28 are similar to the filtering processes of the degradation state low-pass filter unit 322 in the second embodiment, and detailed description thereof will be omitted.

[0084] In this embodiment, in the battery measurement unit 20b, a voltage value acquired by the measurement quantity acquisition unit 21 is input to the voltage low-pass filter unit 27, which smooths and outputs a smoothed voltage. The smoothed voltage output from the voltage low-pass filter unit 27 is input to the voltage change amount calculation unit 22, which calculates a smoothed voltage change amount and stores it in the measurement data storage unit 31 of the battery state estimation unit 30a. Similarly, a current value acquired by the measurement quantity acquisition unit 21 is input to the current low-pass filter unit 28, which smooths and outputs a smoothed current. The smoothed current output from the current low-pass filter unit 28 is input to the current change amount calculation unit 23 and the charge amount calculation unit 24, and the output of the charge amount calculation unit 24 is input to the charge change amount calculation unit 25, which calculates a smoothed charge change amount. The smoothed charge change amount calculated by the charge change amount calculation unit 25 is stored in the measurement data storage unit 31 of the battery state estimation unit 30a.

[0085] The estimation model for calculating the capacity maintenance rate (SOH) of the storage battery 1 stored in the estimation model memory unit 34 of the storage battery state estimation unit 30a is an estimation model learned by machine learning that takes a smoothed voltage value and a smoothed current value as input and outputs the capacity maintenance rate (SOH) of the storage battery 1.

[0086] A deterioration estimation device for a storage battery 1 according to a third embodiment of the present disclosure is characterized in that it calculates smoothed voltage values ​​and smoothed current values ​​using a voltage low-pass filter and a current low-pass filter for voltage values ​​and current values ​​measured from the storage battery 1, calculates smoothed voltage change amounts and smoothed charge change amounts from the smoothed voltage values ​​and smoothed current values, and calculates the capacity maintenance rate (SOH) of the storage battery 1 using an estimation model learned by machine learning, which uses the smoothed voltage change amounts and smoothed charge change amounts as inputs and outputs the capacity maintenance rate (SOH) of the storage battery 1.

[0087] [Voltage measured from storage battery and smoothed voltage smoothed using a low-pass filter] FIG. 13 is a diagram illustrating the voltage measured when the storage battery according to the third embodiment of the present disclosure is charged and discharged, and the smoothed voltage smoothed using a low-pass filter.

[0088] In Figure 13, the horizontal axis represents time and the vertical axis represents voltage value, and the voltage measured from storage battery 1 and the smoothed voltage obtained by smoothing the voltage measured from storage battery 1 using a voltage low-pass filter (voltage low-pass filter unit 27) are shown.

[0089] 13 , the voltage measured from the storage battery 1 may exhibit discrete voltage fluctuations rather than continuity with each measurement due to influences such as abrupt fluctuations in the charge / discharge current of the storage battery 1 and measurement noise. However, by using a low-pass filter, the discrete voltage values ​​measured from the storage battery 1 are smoothed so that they follow the previously measured voltage value. This allows the voltage change amount input to the estimation model that estimates the degradation state of the storage battery 1 to also become a smoothed continuous value. This makes it possible to eliminate the influence of noise, etc. from the voltage value and voltage fluctuation amount of the storage battery 1, making it possible to estimate the degradation state of the storage battery 1 with less error. Here, while FIG. 13 relates to the voltage and smoothed voltage measured from the storage battery 1, the same applies to the current and smoothed current measured from the storage battery 1.

[0090] [Flowchart of Storage Battery Deterioration Estimation Method] FIG. 14 is a flowchart showing a storage battery deterioration estimation method according to the third embodiment of the present disclosure.

[0091] First, the voltage low-pass filter unit 27 calculates a first smoothed voltage value and a second smoothed voltage value using a low-pass filter for the voltage measured when the storage battery 1 is charged and discharged (S31). Next, the voltage change amount calculation unit 22 calculates a smoothed voltage change amount from the first smoothed voltage value and the second smoothed voltage value calculated in step S31 (S32). Next, the current low-pass filter unit 28 calculates a first smoothed current value and a second smoothed current value using a low-pass filter for the current measured when the storage battery 1 is charged and discharged (S33). Next, the charge change amount calculation unit 25 calculates a smoothed charge change amount from the first smoothed current value and the second smoothed current value calculated in step S33 (S34). Finally, the degradation state estimation unit 321 estimates a capacity maintenance rate (SOH) of the storage battery 1 from one or more smoothed voltage change amounts and one or more smoothed charge change amounts using an estimation model (S35).

[0092] As described above, in this embodiment, the voltage and current measured when the storage battery 1 is charged and discharged are input, the voltage and current of the storage battery 1 are smoothed using a low-pass filter to output a smoothed voltage and a smoothed current, and the smoothed voltage change amount and smoothed charge change amount obtained from the smoothed voltage and smoothed current are input, and the degradation state of the storage battery 1 is output, and an estimation model trained by machine learning is used to calculate the degradation state of the storage battery 1. As a result, even if the voltage and current measured from the storage battery 1 are affected by abrupt fluctuations in the charge and discharge current of the storage battery 1, measurement noise, and the like, and thus discrete voltage and current fluctuations that are not continuous occur for each measurement, the discrete voltage and current fluctuations are smoothed using the low-pass filter to follow the previously measured voltage and current, and therefore the voltage change amount and charge change amount input to the estimation model that estimates the degradation state of the storage battery 1 become continuous data. It is possible to eliminate the influence of noise and the like from the continuous data of the voltage change amount and the charge change amount, and it is possible to estimate the deterioration state of the storage battery 1 with less error.

[0093] <Fourth embodiment> [Overall configuration of a storage battery deterioration estimation device] Fig. 15 is a diagram showing the overall configuration of a storage battery deterioration estimation device according to a fourth embodiment of the present disclosure. In Fig. 15, components that are the same as those in any of Figs. 1, 9, and 12, which are the storage battery deterioration estimation devices according to the first, second, and third embodiments, are given the same names, symbols, and numbers, and detailed descriptions thereof will be omitted.

[0094] In the deterioration estimation device for a storage battery 1 according to the fourth embodiment, the configuration of a storage battery measurement unit 20b of a storage battery state estimation device 100d is the same as that of the third embodiment, and the configuration of a storage battery state estimation unit 30b is the same as that of the second embodiment. That is, as described in the second and third embodiments, the deterioration estimation device for a storage battery 1 according to the fourth embodiment is characterized in that it calculates a smoothed voltage and a smoothed current that are obtained by smoothing a voltage and a current measured from the storage battery 1 using a low-pass filter, calculates a smoothed voltage change amount and a smoothed charge change amount from the smoothed voltage and the smoothed current, calculates the capacity maintenance rate (SOH) of the storage battery 1 using an estimation model learned by machine learning, which receives the smoothed voltage change amount and the smoothed charge change amount as input and outputs the capacity maintenance rate (SOH) of the storage battery 1, and further calculates the capacity maintenance rate (SOH) of the storage battery 1 that is smoothed using a low-pass filter for the capacity maintenance rate (SOH) of the storage battery 1 calculated using the estimation model.

[0095] [Flowchart of Storage Battery Deterioration Estimation Method] FIG. 16 is a flowchart showing a storage battery deterioration estimation method according to the fourth embodiment of the present disclosure.

[0096] First, the voltage low-pass filter unit 27 calculates a first smoothed voltage value and a second smoothed voltage value using a low-pass filter for the voltage measured when the storage battery 1 is charged and discharged (S41). Next, the voltage change amount calculation unit 22 calculates a smoothed voltage change amount from the first smoothed voltage value and the second smoothed voltage value calculated in step S41 (S42). Next, the current low-pass filter unit 28 calculates a first smoothed current value and a second smoothed current value using a low-pass filter for the current measured when the storage battery 1 is charged and discharged (S43). Next, the charge change amount calculation unit 25 calculates a smoothed charge change amount from the first smoothed current value and the second smoothed current value calculated in step S43 (S44). Next, the degradation state estimation unit 321 estimates a capacity maintenance rate (SOH) of the storage battery 1 from one or more smoothed voltage change amounts and one or more smoothed charge change amounts using an estimation model (S45). Finally, the degradation state low-pass filter unit 322 uses a low-pass filter to calculate a smoothed capacity maintenance rate (SOH) of the storage battery 1 from the capacity maintenance rate (SOH) of the storage battery 1 calculated in step S45 (S46).

[0097] As described above, in this embodiment, a filter process is performed on the voltage and current measured from the storage battery 1 to calculate a smoothed voltage value and a smoothed current value that are smoothed so as to follow the previously output voltage value and current value, and the state of degradation of the storage battery 1 is estimated using an estimation model trained by machine learning, which inputs the smoothed voltage change amount and smoothed charge change amount obtained from the smoothed voltage value and smoothed current value and outputs the state of degradation of the storage battery 1. This configuration makes it possible to eliminate the influence of noise, etc. from the voltage change amount and charge change amount, making it possible to estimate the state of degradation of the storage battery 1 with less error.

[0098] Furthermore, by using the degradation state of the storage battery 1 calculated using the estimation model as input and performing filter processing, the degradation state of the storage battery 1 is smoothed and output.Even if the degradation state of the storage battery 1 calculated using the estimation model has a large error compared to the actual degradation state of the storage battery 1 at a certain time, the output data is smoothed by filter processing so that it follows the output data previously output, making it possible to estimate the degradation state of the storage battery 1 with less error.

[0099] <Conclusion> As described above, the storage battery degradation estimation device and degradation estimation method disclosed herein are configured to calculate the amount of voltage change and amount of charge change from the voltage and current measured when the storage battery is charged and discharged, and output the degradation state of the storage battery using an estimation model learned by machine learning, which uses the amount of voltage change and amount of charge change as input and the degradation state of the storage battery as output.Therefore, it is possible to estimate the degradation state even when the storage battery is in use, and to estimate the degradation state with higher accuracy.

[0100] Furthermore, by configuring the storage battery degradation state calculated by the estimation model to be output as a storage battery degradation state smoothed by filtering, it becomes possible to estimate the storage battery degradation state with less error. Furthermore, by configuring the storage battery degradation state to be estimated by calculating a smoothed voltage value and a smoothed current value smoothed by filtering the voltage and current measured from the storage battery, and calculating a smoothed voltage change amount and a smoothed charge change amount from the smoothed voltage value and the smoothed current value, it becomes possible to eliminate the influence of noise, etc., in the voltage and current measured from the storage battery, and it becomes possible to estimate the storage battery degradation state with less error.

[0101] Furthermore, by configuring the storage battery deterioration estimation device so that the estimation model for estimating the deterioration state of the storage battery stored therein can be updated to a new estimation model, it becomes possible to update the estimation model to one that can estimate the deterioration state of the storage battery with higher accuracy even after the storage battery deterioration estimation device has been installed in a product, thereby making it possible to estimate the deterioration state of the storage battery with higher accuracy.

[0102] Furthermore, by configuring the storage battery degradation estimation device to be able to transmit multiple measurement data and multiple estimation data measured and estimated by the storage battery degradation estimation device to an external device, it becomes possible to collect measurement data of the storage battery when it is used and estimation data of the storage battery's degradation state even after the storage battery degradation estimation device is installed in a product. This makes it possible to learn and update a new estimation model with higher accuracy using learning data that includes various usage states of the storage battery, thereby making it possible to estimate the degradation state of the storage battery with higher accuracy.

[0103] The above describes embodiments of the present disclosure, but the storage battery degradation estimation device and degradation estimation method of the present disclosure are not limited to the above examples, and are also effective for embodiments in which various modifications that a person skilled in the art can think of are made within the scope of the gist of the present disclosure, or for forms constructed by combining components of different embodiments.

[0104] For example, the specific configuration of the storage battery state estimation device in the storage battery degradation estimation device of this embodiment is not limited to the configuration shown in this embodiment. In this embodiment, in order for the storage battery measurement unit of the storage battery state estimation device to calculate the voltage change amount and charge change amount, or the smoothed voltage change amount and smoothed charge change amount, the storage battery state estimation device is described as being configured with a measurement amount acquisition unit, a voltage change amount calculation unit, a current change amount calculation unit, a charge amount calculation unit, a charge change amount calculation unit, a measurement start / stop control unit, a voltage low-pass filter unit, and a current low-pass filter unit. However, the configuration is not limited to the above, and any other configuration may be used as long as it can calculate the voltage change amount and charge change amount, or the smoothed voltage change amount and smoothed charge change amount, of the storage battery.

[0105] Furthermore, in this embodiment, a calculation formula for filtering and smoothing the voltage and current of the storage battery and the deterioration state of the storage battery has been described, but this is not limited to the above, and another calculation formula capable of smoothing may also be adopted.

[0106] In addition, in this embodiment, a fully connected neural network is used as an estimation model learned by machine learning, but this is not limited to the above, and it is also possible to adopt an estimation model that can estimate the deterioration state of a storage battery, such as a recurrent neural network, a decision tree type machine learning estimation model, or a multiple regression analysis type machine learning estimation model.

[0107] In the above embodiments, each component may be configured with dedicated hardware, or may be realized by executing a software program suitable for each component. Each component may be realized by a program execution unit such as a CPU or processor reading and executing a software program recorded on a recording medium such as a hard disk or semiconductor memory.

[0108] The order in which the steps in the flowchart are executed is merely an example for specifically explaining the present disclosure, and other orders may be used. Some of the steps may be executed simultaneously (in parallel) with other steps, or some of the steps may not be executed.

[0109] The division of functional blocks in the block diagram is an example, and multiple functional blocks may be realized as a single functional block, one functional block may be divided into multiple blocks, or some functions may be moved to another functional block.Furthermore, the functions of multiple functional blocks having similar functions may be processed in parallel or in time-sharing by a single piece of hardware or software.

[0110] Furthermore, each component described in the above embodiment may be realized as software, or may be realized as an LSI, which is typically an integrated circuit. These may be individually integrated into a single chip, or some or all of them may be integrated into a single chip. Here, we refer to an LSI, but depending on the level of integration, it may also be called an IC, VLSI, ULSI, or system LSI. Furthermore, the method of integration is not limited to an LSI, and may be realized using a dedicated circuit or a general-purpose processor.

[0111] Furthermore, one aspect of the present disclosure may be a computer program that causes a computer to execute each of the characteristic steps included in the method for estimating deterioration of a storage battery shown in any of Figures 7, 10, 14, and 16.

[0112] Furthermore, for example, the program may be a program to be executed by a computer. Another aspect of the present disclosure may be a computer-readable non-transitory recording medium on which such a program is recorded. For example, such a program may be recorded on a recording medium and distributed or circulated. For example, the distributed program may be installed in a device having another processor, and the program may be executed by the processor, thereby causing the device to perform each of the above processes.

[0113] The storage battery degradation estimation device and degradation estimation method disclosed herein are capable of estimating the degradation state even when the storage battery is in use, and are capable of estimating the degradation state with higher accuracy. Therefore, they are useful for devices that operate continuously for long periods of time using a storage battery as a power source, such as electric bicycles, electric motorcycles, electric cars, electric airplanes, electric ships, electric agricultural machinery, drones, automatic transport devices, industrial robots, and energy storage systems, and that have a significant impact on business operations or system operations if they run out of power or break down.

[0114] REFERENCE SIGNS LIST 1 Storage battery 10 Storage battery management device 11 ADC unit 12 Calculation unit 13 Communication I / F 20a, 20b Storage battery measurement unit 21 Measurement quantity acquisition unit 22 Voltage change amount calculation unit 23 Current change amount calculation unit 24 Charge amount calculation unit 25 Charge change amount calculation unit 26 Measurement start / stop control unit 27 Voltage low-pass filter unit 28 Current low-pass filter unit 30a, 30b Storage battery state estimation unit 31 Measurement data storage unit 32a, 32b Deterioration state calculation unit 33 Estimation data storage unit 34 Estimation model storage unit 35 Estimation model update unit 41, 42, 51 Communication I / F 52 Deterioration state display unit 53 Estimation model transmission unit 54 Measurement data / estimated data reception unit 100a, 100b, 100c, 100d Storage battery state estimation device 321 Deterioration state estimation unit 322 Deterioration state low-pass filter unit 3211 Neural network calculation unit 3212 Feature amount calculation unit

Claims

1. An apparatus for estimating a deterioration state of a storage battery from a measurement result of the storage battery, a first calculation unit that calculates a voltage change amount of the storage battery obtained from a first voltage value measured when the storage battery is charged and discharged and a second voltage value measured further; a second calculation unit that calculates an amount of charge change of the storage battery obtained from a first current value measured when the storage battery is charged and discharged and a second current value measured further; a measurement data storage unit for storing a plurality of measurement data including the voltage change amount and the charge change amount; an estimation model storage unit that stores an estimation model learned by machine learning, the estimation model having one or more of the voltage change amounts and one or more of the charge change amounts as input and a deterioration state of the storage battery as output; an estimation unit that estimates a degradation state of the storage battery by using an estimation model stored in the estimation model storage unit and receiving as input one or more of the voltage change amounts and one or more of the charge change amounts stored in the measurement data storage unit. A device for estimating battery deterioration.

2. An apparatus for estimating a deterioration state of a storage battery from a measurement result of the storage battery, a first calculation unit that calculates a voltage change amount of the storage battery obtained from a first voltage value measured when the storage battery is charged and discharged and a second voltage value measured further; a second calculation unit that calculates an amount of charge change of the storage battery obtained from a first current value measured when the storage battery is charged and discharged and a second current value measured further; a measurement data storage unit for storing a plurality of measurement data including the voltage change amount and the charge change amount; an estimation model storage unit that stores an estimation model learned by machine learning, the estimation model having one or more of the voltage change amounts and one or more of the charge change amounts as input and a deterioration state of the storage battery as output; an estimation unit that estimates a degradation state of the storage battery by using an estimation model stored in the estimation model storage unit and receiving as input one or more of the voltage change amounts and one or more of the charge change amounts stored in the measurement data storage unit; a low-pass filter unit that receives the deterioration state of the storage battery output from the estimation unit, smooths the deterioration state of the storage battery using a low-pass filter, and outputs the smoothed deterioration state of the storage battery. A device for estimating battery deterioration.

3. An apparatus for estimating a deterioration state of a storage battery from a measurement result of the storage battery, a first low-pass filter unit that receives a voltage measured when the storage battery is charged and discharged, smoothes the voltage using a low-pass filter, and outputs a smoothed voltage; a first calculation unit that calculates a smoothed voltage change amount of the storage battery obtained from a first smoothed voltage value output from the first low-pass filter unit and a second smoothed voltage value output from the first low-pass filter unit; a second low-pass filter unit that receives a current measured when the storage battery is charged and discharged, smooths the current using a low-pass filter, and outputs a smoothed current; a second calculation unit that calculates a smoothed charge change amount of the storage battery obtained from a first smoothed current value output from the second low-pass filter unit and a second smoothed current value output from the second low-pass filter unit; a measurement data storage unit for storing a plurality of measurement data including the smoothed voltage change amount and the smoothed charge change amount; an estimation model storage unit that stores an estimation model learned by machine learning, the estimation model having one or more of the smoothed voltage change amounts and one or more of the smoothed charge change amounts as input and a deterioration state of the storage battery as output; an estimation unit that estimates a degradation state of the storage battery by using an estimation model stored in the estimation model storage unit and receiving as input one or more of the smoothed voltage change amounts and one or more of the smoothed charge change amounts stored in the measurement data storage unit. A device for estimating battery deterioration.

4. An apparatus for estimating a deterioration state of a storage battery from a measurement result of the storage battery, a first low-pass filter unit that receives a voltage measured when the storage battery is charged and discharged, smoothes the voltage using a low-pass filter, and outputs a smoothed voltage; a first calculation unit that calculates a smoothed voltage change amount of the storage battery obtained from a first smoothed voltage value output from the first low-pass filter unit and a second smoothed voltage value output from the first low-pass filter unit; a second low-pass filter unit that receives a current measured when the storage battery is charged and discharged, smooths the current using a low-pass filter, and outputs a smoothed current; a second calculation unit that calculates a smoothed charge change amount of the storage battery obtained from a first smoothed current value output from the second low-pass filter unit and a second smoothed current value output from the second low-pass filter unit; a measurement data storage unit for storing a plurality of measurement data including the smoothed voltage change amount and the smoothed charge change amount; an estimation model storage unit that stores an estimation model learned by machine learning, the estimation model having one or more of the smoothed voltage change amounts and one or more of the smoothed charge change amounts as input and a deterioration state of the storage battery as output; an estimation unit that estimates a degradation state of the storage battery by using an estimation model stored in the estimation model storage unit and receiving as input one or more of the smoothed voltage change amounts and one or more of the smoothed charge change amounts stored in the measurement data storage unit; a third low-pass filter unit that receives the deterioration state of the storage battery output from the estimation unit, smooths the deterioration state of the storage battery using a low-pass filter, and outputs the smoothed deterioration state of the storage battery. A device for estimating battery deterioration.

5. The deterioration state of the storage battery is a capacity maintenance rate of the storage battery. The battery deterioration estimation device according to any one of claims 1 to 4.

6. The deterioration state of the storage battery is a full charge capacity value of the storage battery. The battery deterioration estimation device according to any one of claims 1 to 4.

7. A receiving and communicating unit that receives a new estimation model from an external device; and an update unit that updates the estimation model stored in the estimation model storage unit to the new estimation model received from the receiving communication unit. The battery deterioration estimation device according to any one of claims 1 to 4.

8. an estimated data storage unit that stores the deterioration state of the storage battery estimated by the estimation unit; a transmission and communication unit that transmits to an outside the plurality of measurement data stored in the measurement data storage unit and the deterioration state of the storage battery stored in the estimated data storage unit. The battery deterioration estimation device according to any one of claims 1 to 4.

9. A method executed by a deterioration estimation device that estimates a deterioration state of a storage battery from a measurement result of the storage battery, comprising: a calculation step of calculating a voltage change amount of the storage battery obtained from a first voltage value measured when the storage battery is charged and discharged and a second voltage value measured further; a calculation step of calculating an amount of charge change of the storage battery obtained from a first current value measured when the storage battery is charged and discharged and a second current value measured further; an estimation step of estimating a degradation state of the storage battery by using an estimation model trained by machine learning, the estimation model having an input of one or more of the voltage change amounts and one or more of the charge change amounts and an output of a degradation state of the storage battery, the estimation step including: A method for estimating battery deterioration.

10. A method executed by a deterioration estimation device that estimates a deterioration state of a storage battery from a measurement result of the storage battery, comprising: a calculation step of calculating a voltage change amount of the storage battery obtained from a first voltage value measured when the storage battery is charged and discharged and a second voltage value measured further; a calculation step of calculating an amount of charge change obtained from a first current value measured when the storage battery is charged and discharged and a second current value measured therefrom; an estimation step of estimating a degradation state of the storage battery by using an estimation model trained by machine learning, the estimation model having an input of one or more of the voltage change amounts and one or more of the charge change amounts and an output of a degradation state of the storage battery; A calculation step of inputting the deterioration state of the storage battery and calculating the deterioration state of the storage battery smoothed using a low-pass filter. A method for estimating battery deterioration.

11. A method executed by a deterioration estimation device that estimates a deterioration state of a storage battery from a measurement result of the storage battery, comprising: a calculation step of calculating a first smoothed voltage value obtained by smoothing a voltage measured when the storage battery is charged and discharged using a low-pass filter, and a second smoothed voltage value obtained by the smoothing using the low-pass filter; a calculation step of calculating a smoothed voltage change amount of the storage battery obtained from the first smoothed voltage value and the second smoothed voltage value; a calculation step of calculating a first smoothed current value obtained by smoothing a current measured when the storage battery is charged and discharged using a low-pass filter, and a second smoothed current value obtained by the smoothing using the low-pass filter; a calculation step of calculating a smoothed charge change amount obtained from the first smoothed current value and the second smoothed current value; an estimation step of estimating a degradation state of the storage battery by using an estimation model trained by machine learning, the estimation model having an input of one or more of the smoothed voltage changes and one or more of the smoothed charge changes, and an output of the degradation state of the storage battery, the estimation step including: A method for estimating battery deterioration.

12. A method executed by a deterioration estimation device that estimates a deterioration state of a storage battery from a measurement result of the storage battery, comprising: a calculation step of calculating a first smoothed voltage value obtained by smoothing a voltage measured when the storage battery is charged and discharged using a low-pass filter, and a second smoothed voltage value obtained by the smoothing using the low-pass filter; a calculation step of calculating a smoothed voltage change amount of the storage battery obtained from the first smoothed voltage value and the second smoothed voltage value; a calculation step of calculating a first smoothed current value obtained by smoothing a current measured when the storage battery is charged and discharged using a low-pass filter, and a second smoothed current value obtained by the smoothing using the low-pass filter; a calculation step of calculating a smoothed charge change amount of the storage battery obtained from the first smoothed current value and the second smoothed current value; an estimation step of estimating a degradation state of the storage battery by using an estimation model trained by machine learning, the estimation model having an input of one or more of the smoothed voltage change amounts and one or more of the smoothed charge change amounts and an output of a degradation state of the storage battery; A calculation step of inputting the deterioration state of the storage battery and calculating the deterioration state of the storage battery smoothed using a low-pass filter. A method for estimating battery deterioration.

13. The deterioration state of the storage battery is a capacity maintenance rate of the storage battery. The method for estimating deterioration of a storage battery according to any one of claims 9 to 12.

14. The deterioration state of the storage battery is a full charge capacity value of the storage battery. The method for estimating deterioration of a storage battery according to any one of claims 9 to 12.