JP2019040845A - State prediction method of secondary battery, charge control method, system and program - Google Patents

Abstract

A state prediction method for a secondary battery that can be used in actual use in which charging and discharging are performed randomly is provided. The method includes a step of obtaining an actual measurement value corresponding to a charge / discharge capacity Q and a deformation amount T of a secondary battery 2, and a charge / discharge capacity Q of the secondary battery from time series data of the actual measurement value. Step S6 for extracting at least one inflection point P1 (P2) in the characteristic curve Ln indicating the relationship with the deformation amount T, and among the characteristic curve L1 indicating the relationship between the charge / discharge capacity of the secondary battery and the deformation amount, Step S8 for obtaining past characteristic data having at least a full charge time point Pf, a remaining amount zero time point Pe, and stage inflection points P1 and P2, and a characteristic curve L1 indicated by the past characteristic data on the basis of the extracted inflection points Step S9 for fitting to the characteristic curve Ln indicated by the time series data of the actual measurement value to generate predicted characteristic data indicating the characteristic curve L2 interpolated with a portion not included in the time series data of the actual measurement value. . [Selection] Figure 6C

Description

  The present invention relates to a method for predicting a state of a secondary battery, a charge control method, a system, and a program.

  In recent years, sealed secondary batteries represented by lithium ion secondary batteries (hereinafter sometimes referred to simply as “secondary batteries”) are not only mobile devices such as mobile phones and laptop computers, but also electric vehicles and hybrids. It is also used as a power source for electric vehicles such as cars. The secondary battery deteriorates by repeating the charge / discharge cycle, and as the deterioration progresses, it becomes difficult to accurately grasp the state of the remaining capacity, the deteriorated capacity of the battery, and the like.

  Patent Document 1 describes that the remaining capacity is estimated from the relationship between the pressure between the batteries and the SOC (State of charge). However, in this method, it is necessary to obtain a charge behavior of a cycle close to full discharge from full charge, and it is difficult to estimate in actual use in which charging and discharging are performed randomly.

  Patent Document 2 describes that the thickness of a secondary battery is measured and the remaining capacity is measured from the thickness. However, changes in remaining capacity due to deterioration are not taken into consideration.

JP, 2006-101048, A JP 2004-14462 A

  The present invention has been made paying attention to such problems, and its purpose is to provide a state prediction method, a charging method, a system, and a program for a secondary battery that can be used in actual use in which charging and discharging are performed randomly. Is to provide.

The state prediction method for a secondary battery according to the present invention includes:
Obtaining actual measurement values corresponding to the charge / discharge capacity and deformation amount of the secondary battery;
Extracting from the time series data of the measured values at least one inflection point in a characteristic curve indicating the relationship between the charge / discharge capacity and deformation amount of the secondary battery;
Of the characteristic curve showing the relationship between the charge / discharge capacity of the secondary battery and the deformation amount, obtaining past characteristic data having at least a full charge time, a remaining time zero, and a stage inflection point;
Using the extracted inflection point as a reference, the characteristic curve indicated by the past characteristic data is fitted to the characteristic curve indicated by the time series data of the actual measurement value, and a portion not included in the time series data of the actual measurement value is interpolated. Generating predicted characteristic data indicating the characteristic curve,
including.

The characteristic curve indicating the relationship between the charge / discharge capacity of the secondary battery and the deformation amount has two inflection points at which the slope of the characteristic curve changes greatly as the stage changes. According to this method, the inflection point in the characteristic curve is extracted from the time series data of the actually measured value, and the past characteristic data and the actually measured value data are fitted on the basis of the extracted inflection point. Even if there is no long-term measurement data from full charge to complete discharge, if there is measured data for a short period of time, for example, from full charge to inflection point, from one inflection point to the other inflection point, Predictive characteristic data with a certain degree of accuracy can be generated.
Therefore, it is possible to generate predictive characteristic data with a certain degree of accuracy in actual use in which charging and discharging are performed randomly rather than a series of charging and discharging from full charge to complete discharge, and predict the state of the secondary battery It becomes possible.

The block diagram which shows an example of the system for performing the state prediction method of a secondary battery. FIG. 3 is a perspective view schematically showing a sealed secondary battery. AA sectional drawing of FIG. 2A. The graph which shows the characteristic curve which the past characteristic data shows. The graph which shows the characteristic curve which the time-series data of measured value and past characteristic data show. The graph which shows the time series data of a measured value, and the differential value of a measured value. Explanatory drawing regarding the process which produces | generates prediction characteristic data. Explanatory drawing regarding the process which produces | generates prediction characteristic data. Explanatory drawing regarding the process which produces | generates prediction characteristic data. Explanatory drawing regarding the process which produces | generates prediction characteristic data. Explanatory drawing regarding the process which produces | generates prediction characteristic data. Explanatory drawing regarding the process which produces | generates prediction characteristic data. Explanatory drawing regarding the process which produces | generates prediction characteristic data. Explanatory drawing regarding the process which produces | generates prediction characteristic data. Explanatory drawing regarding the calculation process of remaining capacity. Explanatory drawing regarding the fitting process of initial characteristic data and prediction characteristic data. Explanatory drawing regarding the fitting process of initial characteristic data and prediction characteristic data. Explanatory drawing regarding the calculation process of a degradation state. Explanatory drawing regarding the calculation process of a degradation state. Explanatory drawing regarding the calculation process of a degradation state. Explanatory drawing regarding the process which calculates thickness variation | change_quantity until lithium precipitation. Explanatory drawing regarding the process which calculates thickness variation | change_quantity until lithium precipitation. Explanatory drawing regarding charge control. The flowchart which shows the state prediction process routine performed with a system.

  Hereinafter, embodiments of the present invention will be described.

  FIG. 1 shows a system mounted on an electric vehicle such as an electric vehicle or a hybrid vehicle. This system includes a battery module 1 in which an assembled battery composed of a plurality of sealed secondary batteries 2 is housed in a casing. In the present embodiment, four secondary batteries 2 are connected in two parallel two series, but the number of batteries and the connection form are not limited to this. Although only one battery module 1 is shown in FIG. 1, the battery pack 1 actually includes a plurality of battery modules 1. In the battery pack, a plurality of battery modules 1 are connected in series, and they are housed in a casing together with various devices such as a controller. The casing of the battery pack is formed in a shape suitable for in-vehicle use, for example, a shape that matches the underfloor shape of the vehicle.

  The secondary battery 2 shown in FIG. 2 is configured as a cell (single cell) in which an electrode group 22 is accommodated inside a sealed exterior body 21. The electrode group 22 has a structure in which a positive electrode 23 and a negative electrode 24 are laminated or wound through a separator 25 therebetween, and the separator 25 holds an electrolytic solution. The secondary battery 2 of the present embodiment is a laminated battery using a laminated film such as an aluminum laminated foil as the outer package 21, and is specifically a laminated lithium ion secondary battery having a capacity of 1.44 Ah. The secondary battery 2 is formed in a thin rectangular parallelepiped shape as a whole, and the X, Y, and Z directions correspond to the length direction, the width direction, and the thickness direction of the secondary battery 2, respectively. The Z direction is also the thickness direction of the positive electrode 23 and the negative electrode 24.

  A detection sensor 5 for detecting deformation of the secondary battery 2 is attached to the secondary battery 2. The detection sensor 5 includes a polymer matrix layer 3 attached to the secondary battery 2 and a detection unit 4. The polymer matrix layer 3 contains a filler that disperses the external field according to deformation of the polymer matrix layer 3 in a dispersed manner. The polymer matrix layer 3 of the present embodiment is formed in a sheet shape from an elastomer material that can be flexibly deformed. The detector 4 detects a change in the external field. When the secondary battery 2 swells and deforms, the polymer matrix layer 3 is deformed accordingly, and a change in the external field accompanying the deformation of the polymer matrix layer 3 is detected by the detection unit 4. In this way, deformation of the secondary battery 2 can be detected with high sensitivity.

  In the example of FIG. 2, since the polymer matrix layer 3 is attached to the outer package 21 of the secondary battery 2, the polymer matrix layer 3 can be deformed according to deformation (mainly swelling) of the outer package 21. it can. On the other hand, the polymer matrix layer 3 may be affixed to the electrode group 22 of the secondary battery 2. According to such a configuration, the polymer matrix layer 3 is deformed in accordance with deformation (mainly swelling) of the electrode group 22. be able to. The deformation of the secondary battery 2 to be detected may be any deformation of the outer package 21 and the electrode group 22.

  A signal detected by the detection sensor 5 is transmitted to the control device 6, whereby information regarding deformation of the secondary battery 2 is supplied to the control device 6.

<Secondary battery state prediction system>
As shown in FIG. 1, the secondary battery state prediction system includes a past characteristic data acquisition unit 60, an actual measurement value acquisition unit 61, an inflection point extraction unit 62, a prediction characteristic data generation unit 63, and a remaining capacity calculation. Unit 64, initial characteristic data acquisition unit 65, and deterioration information generation unit 66. In this embodiment, each part 60-66 is implement | achieved by the control apparatus 6, However, It is not limited to this. For example, it may be realized by an information processing apparatus located far away via a communication network. The remaining capacity calculation unit 64, the initial characteristic data acquisition unit 65, and the deterioration information generation unit 66 can be omitted as necessary.

<Acquisition of past characteristic data>
The past characteristic data acquisition unit 60 acquires past characteristic data of the secondary battery 2. As shown in FIG. 3, the past characteristic data includes at least a full charge point Pf, a remaining amount zero point Pe, and a stage change in a characteristic curve L1 indicating the relationship between the charge / discharge capacity Q of the secondary battery and the deformation amount T. It has music points P1 and P2. This is because with these four points, the characteristic curve L1 can be reproduced to some extent. The state of the secondary battery 2 can be expressed by a characteristic curve L1.

  In the graph of FIG. 3, the horizontal axis is the discharge capacity Q with the origin as the fully charged time point Pf, and the vertical axis is the detected deformation amount T of the secondary battery 2. As the discharge capacity Q from the full charge point Pf increases, the deformation amount T of the secondary battery 2 changes in a direction in which the thickness of the secondary battery 2 decreases. This is because, in the charged secondary battery 2, the electrode group 22 swells (hereinafter sometimes referred to as “electrode swell”) due to the volume change of the negative electrode active material, and the electrode swell is accompanied by discharge. This is because it becomes smaller. Due to the stage change of the electrode, the shape includes two inflection points P1 and P2 (change in gradient) as shown in FIG. For example, in the case of a lithium ion secondary battery using graphite (graphite) for the negative electrode, it is known that the crystalline state of the graphite sequentially changes in stages as it is discharged from the full charge point Pf. This is because the negative electrode active material expands as the distance between the graphene layers increases stepwise with the insertion amount of lithium ions. In short, the volume of the active material changes stepwise due to the stage change, which is reflected in the characteristic curve L1. When the discharge capacity Q further increases, the remaining time point Pe is reached.

  In the present embodiment, data is also included between the four points of the full charge point Pf, the remaining amount zero point Pe, and the inflection points P1 and P2. The characteristic data may be discrete data in which data of a plurality of points is provided between four points, or may be continuous data in which four points are connected by a curve. The past characteristic data is not the initial characteristics set at the time of shipment of the secondary battery 2, but is past data from the start of use until the present time. The most recent charge cycle data is more preferable because there is little difference between the current data and past data. Past characteristic data is stored in the memory of the control device 6, and the past characteristic data acquisition unit 60 acquires data from the memory.

<Acquisition of actual measurement data>
The actual measurement value acquisition unit 61 acquires actual measurement values corresponding to the charge / discharge capacity and deformation amount of the secondary battery 2. The value corresponding to the deformation amount T of the secondary battery is a value detected by the detection sensor 5. In this embodiment, it is expressed by a change amount of voltage or magnetic flux density or the like, but is not limited to this, and can be variously changed as long as it is a physical quantity corresponding to the thickness or the deformation amount. The charge / discharge capacity of the secondary battery 2 is a generic term for a discharge capacity and a charge capacity. In this embodiment, the discharge capacity, specifically, the discharge capacity from the point of full charge, is used, but the present invention is not limited to this. The value corresponding to the charge / discharge capacity can be expressed by current, charge amount or discharge amount, and can be variously changed. The actual measurement values corresponding to the charge / discharge capacity and deformation amount of the secondary battery 2 are repeatedly acquired and become time-series data. FIG. 4 shows a characteristic curve Ln represented by time-series data of actual measurement values with a solid line. The example of FIG. 4 shows an example in which actual measurement is started from the fully charged time point Pf, discharging is continued, and the inflection point P1 has not yet been reached. In FIG. 4, a characteristic curve L1 indicated by a broken line is exaggerated so that the characteristic curve L1 indicated by the past characteristic data shown in FIG. 3 is different for comparison.

  In FIG. 4, the characteristic curve Ln is indicated by a solid line. However, in reality, the raw data that is actually measured values often include large noise that vibrates on the vertical axis, and data per unit discharge capacity. Exist in a form that vibrates up and down multiple points. In order to remove noise on the vertical axis and facilitate data processing, filter processing such as moving average or median is performed on the vertical axis so that the data per unit discharge capacity becomes one plot. Yes. Thereby, a differential value indicating the slope of the characteristic curve Ln can be calculated.

<Inflection point extraction>
The inflection point extraction unit 62 extracts at least one inflection point (P1 or P2) in the characteristic curve Ln indicating the relationship between the charge / discharge capacity and the deformation amount of the secondary battery 2 from the time-series data of the actually measured values. . Specifically, the inflection point is extracted based on the differential value (ΔT / ΔQ) of the deformation amount T with respect to the charge / discharge capacity Q determined based on the actual value. This is because the differential value is the slope of the characteristic curve Ln, and the portion where the change in the differential value is large is an inflection point. It is possible to extract an inflection point on the condition that the amount of change in the differential value is larger than a certain value.

  In the present embodiment, in order to increase the extraction accuracy, an extraction condition is set according to the charge / discharge capacity Q indicated by the acquired actual measurement value, and the differential value (the differential value of the deformation amount T with respect to the charge / discharge capacity Q determined based on the acquired actual measurement value ( When ΔT / ΔQ) satisfies the extraction condition, the actual measurement value is extracted as an inflection point. Examples will now be described.

  As shown in FIG. 5, if the acquisition of the actual measurement values is continued, discrete data indicating the relationship between the discharge capacity Q and the deformation amount (here, the voltage value of the Hall element) is obtained as shown in the upper graph of FIG. It is done. If this discrete data is differentiated, a differential value [ΔmV / ΔmAh] of the deformation amount relating to the discharge capacity is obtained as shown in the lower graph of FIG. There is one differential value for each unit charge / discharge capacity (unit discharge capacity), and if it is illustrated as shown in the lower graph of FIG.

Setting the extraction conditions, as shown in FIG. 4, the charge-discharge capacity Q _Pn showing actual measurement values acquired in a certain time Pn is the charge-discharge capacity Q _P1 inflection point P1 (P2) in the past of the characteristic data When entering within a predetermined range centered on (Q _P2 ), as shown in the lower graph of FIG. 5, a threshold value Th _P1 (Th _P2 ) of the differential value is set based on past characteristic data. For example, as shown in FIG. 4, when the charge / discharge capacity Q _Pn indicated by the actually measured value data at a certain point Pn is near the inflection point P1, as shown in the lower graph of FIG. A threshold value Th _P1 is set. Then, the acquisition (measurement) of the actual measurement value further proceeds, and as shown in FIG. 5, the inflection point P1 is extracted when the differential value based on the actual measurement value passes the threshold value Th _P1 . In the lower graph of FIG. 5, the extraction condition for extracting the inflection point P1 during discharge is that the threshold Th _P1 has passed downward, and the extraction for extracting the inflection point P1 during charging is performed. The condition is that the threshold Th _P1 is passed upward. Similarly, the extraction condition for extracting the inflection point P2 during discharge is that the threshold Th _P2 has passed upward, and the extraction condition for extracting the inflection point P2 during charging is the threshold value. It has passed through Th _P2 downward.

In the method using the threshold value Th _P1, Th _P2 of the differential value, there is a fear that even when passed through the the primary threshold value Th _P1, Th _P2 due to the effects of noise become inflection point is extracted. Therefore, in addition to the differential value passing through the threshold values Th _P1 and Th _P2 , it is set as an extraction condition that the differential value continuously increases or decreases for a certain period during which the charge / discharge capacity changes. ing. In the present embodiment, since the differential value is calculated for each unit discharge capacity, the extraction condition is that at least three differential values continue to increase or decrease continuously. In this case, a certain period during which the charge / discharge capacity changes is equivalent to three points, and becomes unit discharge capacity [ΔmAh] × 3. This period can be changed as appropriate.

<Generation of predicted characteristic data>
As shown in FIGS. 6C and 7C, the predicted characteristic data generation unit 63 converts the characteristic curve L1 indicated by the past characteristic data into the characteristic curve Ln indicated by the time-series data of the actual measurement values with the extracted inflection point as a reference. A fitting process is performed to generate predicted characteristic data indicating a characteristic curve L2 in which a portion not included in the time-series data of the actual measurement values is interpolated. The prediction characteristic data generation unit 63 stores the generated prediction characteristic data in a memory. In the present embodiment, the method described below is adopted in order to improve the prediction accuracy of the characteristic curve.

<Prediction characteristic data generation process 1>
FIG. 6A shows an example in which two inflection points P1 and P2 are extracted. The inflection point P2 is an inflection point extracted most recently. When the inflection point extraction unit 62 extracts two inflection points P1 and P2, the predicted characteristic data generation unit 63 uses the two inflection points P1 and P2 in the past characteristic data as time-series data of measured values. A coefficient for matching the corresponding two inflection points P1 and P2 with the expansion and contraction of the characteristic curve is calculated. As an example of the coefficient, the horizontal axis enlargement ratio Xr = a ′ / a, the vertical axis enlargement ratio Yr = b ′ / b, and the slope.
Next, as shown in FIG. 6B, the entire characteristic curve L1 indicated by the past characteristic data is expanded / contracted using a coefficient to generate a scaled characteristic curve L1 ′.
Next, as shown in FIG. 6C, the adjusted characteristic curve L1 ′ is moved, and the inflection point P2 extracted most recently in the characteristic curve Ln indicated by the time-series data of the actually measured values and the adjusted characteristic curve L1 ′ are obtained. Are matched with the corresponding inflection point P2 to generate prediction characteristic data. Then, as shown in the figure, predicted characteristic data indicating the characteristic curve L2 in which a portion (indicated by a broken line in the figure) not included in the time series data of the actual measurement values is interpolated is generated. The characteristic curve L2 indicated by the predicted characteristic data includes a characteristic curve Ln and a characteristic curve L1 ′.

<Prediction characteristic data generation process 2>
FIG. 7A shows an example in which one inflection point P1 is extracted and the measured value corresponding to one of the fully charged time point Pf or the remaining charge zero time point Pe is included in the time-series data of the actually measured values. When the inflection point extraction unit 62 extracts one inflection point P1, the prediction characteristic data generation unit 63 inflections with either one of the full charge time point Pf or the zero remaining time point Pe in the past characteristic data. A coefficient for matching the two points P1 with the corresponding two points in the time-series data of the actually measured values by expansion / contraction of the characteristic curve is calculated. In the example of FIG. 7A, a coefficient for matching the full charge time point Pf and the inflection point P1 by vertical and horizontal expansion and contraction is calculated. As an example of the coefficient, the horizontal axis enlargement ratio Xr = a ′ / a, the vertical axis enlargement ratio Yr = b ′ / b, and the slope.
Next, as shown in FIG. 7B, the entire characteristic curve L1 indicated by the past characteristic data is expanded / contracted using a coefficient to generate a scaled characteristic curve L1 ′.
Next, as shown in FIG. 7C, the adjusted characteristic curve L1 ′ is moved, and the inflection point P1 extracted most recently in the characteristic curve Ln indicated by the time-series data of the actual measurement values and the adjusted characteristic curve L1 ′. Are matched with the corresponding inflection point P1 to generate prediction characteristic data. Then, as shown in the figure, predicted characteristic data indicating the characteristic curve L2 in which a portion (indicated by a broken line in the figure) not included in the time series data of the actual measurement values is interpolated is generated. The characteristic curve L2 indicated by the predicted characteristic data includes a characteristic curve Ln and a characteristic curve L1 ′.

<Prediction characteristic data generation process 3>
FIG. 7D shows an example in which one inflection point P1 is extracted, but the actual measurement value corresponding to one of the full charge time point Pf or the remaining zero time point Pe is not included in the time series data of the actual measurement value. Show. As shown in FIG. 7D and FIG. 7E, the predicted characteristic data generation unit 63 uses the inflection point most recently extracted as the coefficient 1 (same size) without expanding or reducing the characteristic curve L1 indicated by the past characteristic data. The characteristic curve L1 indicated by the past characteristic data is moved and interpolated so that P1 and the inflection point P1 in the past characteristic data match, thereby generating predicted characteristic data.

  The predicted characteristic data includes a fully charged time point Pf, a remaining time zero point Pe, and two inflection points P1 and P2. 6A to 6C and FIGS. 7A to 7B show the idea, and the procedure is not limited to this.

  Although the reproduction accuracy of the characteristic curve of the secondary battery 2 may be inferior to the generation process 1 shown in FIGS. 6A to 6C and the generation process 2 shown in FIGS. 7A to 7C, the least square method is used as the fitting process. Other methods can also be executed. For example, commercially available software such as Solver (registered trademark) in Excel (registered trademark) manufactured by Microsoft Corporation is also possible.

<Calculation of remaining capacity>
The remaining capacity is calculated using prediction characteristic data and an actual measurement value at the time of predicting the remaining capacity. Specifically, as shown in FIG. 8, the remaining capacity calculation unit 64 includes a charge / discharge capacity Q _Pn indicated by an actual measurement value at the time when the remaining capacity is predicted, and a charge / discharge capacity Q at the remaining time zero point Pe in the predicted characteristic data. The difference from _Pe is calculated as the remaining capacity Qr.

<Generation of degradation information>
In order to generate the deterioration information, the predicted characteristic data obtained above and the initial characteristic data are used. In some cases, it is possible to generate the degradation information only with the prediction characteristic data, which will be described later. Deterioration information includes the side reaction balance of the electrode, the degree of change in the amount of active material contributing to charge and discharge, and the amount of change in thickness until lithium deposition.

  As shown in FIG. 9A, the initial characteristic data acquisition unit 65 includes at least a full charge point Pf and a remaining charge zero point Pe among characteristic curves L0 indicating the relationship between the charge / discharge capacity Q and the deformation amount T of the secondary battery 2. And initial characteristic data having stage inflection points P1 and P2. The initial characteristic data acquisition unit 65 acquires data from the memory. The initial characteristic data is based on the secondary battery 2 in the initial stage that has not deteriorated, and is obtained using, for example, the secondary battery 2 at the time of manufacture or before shipment. Information on the characteristic curve L0 is obtained from the control device. 6 is stored in advance in a memory (not shown). In the charging / discharging process for obtaining the characteristic curve L0, the secondary battery 2 before shipment is placed in a constant temperature bath at 25 ° C., left for 120 minutes, and then charged at a constant current of 4.34 V with a charging current of 0.144 A. After reaching .32 V, constant voltage charging was performed until the current value decreased to 0.07 A, and then the open circuit state was maintained for 10 minutes, and constant current discharging was performed to 3.0 V with a current of 0.144 A. The discharge capacity from the fully charged state to the fully discharged state at this time was 1.44 Ah.

<Degree of change in the amount of active material contributing to charge / discharge>
As shown in FIG. 9A, the deterioration information generation unit 66 uses two inflection points P1 and P2 in the initial characteristic data (L0) and two corresponding inflection points P1 and P2 in the prediction characteristic data (L2). Then, the expansion rate (a ′ / a) of the charge / discharge capacity for matching with the expansion / contraction of the characteristic curve is calculated as the degree of change in the amount of active material contributing to charge / discharge. In FIG. 9A, the expansion rate on the horizontal axis = the expansion rate of charge / discharge capacity (a ′ / a).

  Further, as shown in FIG. 9B, the expansion rate of the charge / discharge capacity may be calculated. The deterioration information generation unit 66 has a ratio (a ′ / a) between two inflection points P1 and P2, a ratio (b ′ / b) between one inflection point P1 and a full charge point Pf, and the other. By using the ratio (c ′ / c) between the inflection point P2 and the remaining amount zero point Pe, the weighting between the two inflection points P1 and P2 is set larger than the weighting of other sections, and the total calculate. For example, a case where data is tabulated with a weight of b: a: c = 1: 8: 1 will be described. In this case, the expansion ratio of the charge / discharge capacity is {b ′ / b × 1 + a ′ / a × 8 + c ′ / c × 1} / {1 + 8 + 1}. In this way, compared to the ratio between the two inflection points, the full charge time point Pf and the remaining zero time point Pe are close as fitting results, so the degree of coincidence in the entire characteristic curve is increased, and the calculation accuracy is improved. I think it will improve.

<Deterioration state>
As illustrated in FIGS. 10A to 10C, the deterioration information generation unit 66 uses two inflection points P1 and P2 in the initial characteristic data (L0) and two inflection points P1 and P2 in the prediction characteristic data (L2). After the initial characteristic data is adjusted, the entire characteristic curve (L0) indicated by the initial characteristic data is stretched and adjusted using coefficients (expansion rate of charge / discharge capacity, expansion rate of deformation amount) for matching with the expansion and contraction of the characteristic curve. The two inflection points P1 and P2 in the characteristic curve (L0 ′) and the characteristic curve (L2) of the predicted characteristic data are made to coincide with each other, and the left / right shift amount D1 and the remaining time zero point of the charge / discharge capacity between the full charge points The average value [(D1 + D2) / 2] of the left-right deviation amount D2 of the charge / discharge capacity between Pe is calculated as the deterioration state of the battery.
In FIGS. 9A and 9B, only the horizontal axis expansion rate (charge / discharge capacity expansion rate) is calculated, but here the vertical axis expansion rate (deformation amount expansion rate) is also calculated.

The specific steps are as follows.
First, as shown in FIG. 10A, the deterioration information generation unit 66 uses two inflection points P1 and P2 in the predicted characteristic data (L2) and two inflection points P1 and P2 in the initial characteristic data (L0). The coefficient for matching the expansion and contraction of the characteristic curve is calculated. The calculation of the coefficient is the same as the method shown in FIGS.

Next, as illustrated in FIG. 10B, the deterioration information generation unit 66 adjusts the entire characteristic curve (L0) indicated by the initial characteristic data using a coefficient, and generates data indicating the adjusted characteristic curve (L0 ′). Generate.
Next, as shown in FIGS. 10B and 10C, the deterioration information generation unit 66 uses the most recently extracted inflection point P2 in the predicted characteristic data (L2) and the initial characteristic data (L0 ′) after the expansion / contraction adjustment. The characteristic curve (L2, L0 ′) of at least one of the predicted characteristic data and the initial characteristic data is moved so that the corresponding inflection point P2 matches.

  Next, as illustrated in FIG. 10C, the deterioration information generation unit 66 uses the adjusted initial characteristic curve as a reference, and the charge / discharge capacity shift amount D1 between the fully charged time points Pf and the charge amount between the remaining charge time points Pe. The average value [(D1 + D2) / 2] of the right and left displacement amount D2 of the discharge capacity is calculated as the deterioration state of the battery.

  If this average value is large, when the battery deteriorates due to the use of the battery, the side reaction amount is biased in either the positive electrode or the negative electrode. It can be determined which side reaction of the negative electrode is increasing.

  10A to 10C, the adjusted initial characteristic curve L0 ′ is moved after the expansion / contraction adjustment of the initial characteristic curve L0. However, the present invention is not limited to this. For example, instead of the adjusted initial characteristic curve L0 ', the predicted characteristic curve L2 may be moved, or both curves may be moved. In addition, the initial characteristic curve L0 may be expanded / contracted after moving both curves first and the inflection point P2 detected most recently is matched, and the movement and expansion / contraction adjustment may be performed simultaneously.

  The coefficient used when adjusting the expansion / contraction of the initial characteristic curve L0 in FIGS. 10A to 10C can be calculated by the same method as the calculation method of the expansion rate of the charge / discharge capacity shown in FIG. 9B. That is, the deterioration information generation unit 66 has a ratio between the two inflection points P1 and P2, a ratio between the one inflection point P1 and the full charge point Pf, and the other inflection point P2 and the remaining amount zero point Pe. The weight between the two inflection points P1 and P2 is set to be larger than the weights of the other sections using the ratio between the two points, and calculation is performed by aggregation. Here, the coefficients on both the horizontal axis and the vertical axis are calculated.

<Thickness change until lithium deposition>
If the amount of thickness change until lithium deposition can be known, charge control can be performed so that lithium does not precipitate. Here, as shown in FIG. 11A, a maximum thickness point PL is set in the initial characteristic data. The maximum thickness point PL is a limit point where lithium is deposited if the battery is charged more than this.
If sufficient safety can be ensured, the fully charged time point Pf may be set to the maximum thickness point PL. In this case, the amount of change from the fully charged time point Pf to the zero remaining time point Pe is T1 below, and the fully charged time point Pf in the predicted characteristic data can be set as the maximum thickness point without using the initial characteristic data.

  The degradation information generation unit 66 uses the method shown in FIGS. 9A and 9B to perform a coefficient for fitting the characteristic curve L0 indicated by the initial characteristic data to the characteristic curve L2 indicated by the predicted characteristic data by the expansion and contraction of the characteristic curve. (Charge / discharge capacity expansion rate, deformation rate expansion rate) are calculated. In FIGS. 9A and 9B, only the horizontal axis expansion rate (charge / discharge capacity expansion rate) is calculated, but the vertical axis expansion rate (deformation amount expansion rate) is also calculated.

  Next, as illustrated in FIG. 11B, the deterioration information generation unit 66 identifies the maximum thickness point PL ′ in the predicted characteristic data from the maximum thickness point PL in the initial characteristic data using the calculated coefficient. Next, as shown in the figure, the deterioration information generation unit 66 uses the deformation amount T1 from the maximum thickness point PL ′ to the remaining zero point Pe in the predicted characteristic data (L2) as the base point at the remaining zero point. The thickness change amount T1 until lithium deposition is calculated.

  Although the above-described state prediction method is described by exemplifying discharging, it can be realized even during charging.

<Charge control>
The control device 6 may be provided with a charge control unit 67 that controls charging of the secondary battery 2 using the thickness change amount T1 until lithium deposition predicted by the deterioration information generation unit 66. That is, the charging control unit 67 performs charging so that the deformation amount of the secondary battery 2 detected during charging by the detection sensor 5 does not exceed the thickness change amount T1 until lithium deposition starting from the time point when the remaining amount is zero. Control. For example, as shown in FIG. 12, a threshold value T2 smaller than the thickness change amount T1 may be set, and the current may be controlled so as to maintain the threshold value T2. As the current control method, on / off control, P control, I control, D control, PD control, PI control, PID control, pulse control, PWM control, etc., with T2 as a target value may be used. In addition, T1 may be set as a target value as long as it can be controlled so as not to exceed T1.

  The operation of the system will be described with reference to FIG.

  First, in step S <b> 1, the actual measurement value acquisition unit 61 acquires actual measurement values corresponding to the charge / discharge capacity Q and the deformation amount T of the secondary battery 2. Since this step is repeatedly executed, time-series data of actual measurement values is obtained.

  In the next step S2, it is determined whether or not prediction characteristic data exists. If it is determined that the predicted characteristic data does not exist (S2: NO), the process proceeds to step S6. If it is determined that the predicted characteristic data exists (S2: YES), the process proceeds to step S3. The process of step S3 will be described later.

In step S6, the inflection point extraction unit 62 obtains at least one inflection point P1 (in the characteristic curve Ln indicating the relationship between the charge / discharge capacity Q and the deformation amount T of the secondary battery 2 from the time series data of the actually measured values. P2) is extracted. In the present embodiment, the inflection point extraction unit 62 sets extraction conditions according to the charge / discharge capacity _QPn indicated by the acquired actual measurement value, and the differential value of the deformation amount related to the charge / discharge capacity determined based on the acquired actual measurement value [ When ΔmV / ΔmAh] satisfies the extraction condition, the actual measurement value is extracted as an inflection point. Specifically, when the charge / discharge capacity Q _Pn indicated by the acquired actual measurement value falls within a predetermined range centered on the charge / discharge capacity Q _P1 at the inflection point P1 in the past characteristic data, based by setting a threshold Th _P1 of the differential value is set as the extraction condition that the differential value passes the threshold Th _P1. Furthermore, in addition to the differential value passing through the threshold values Th _P1 and Th _P2 , it is set as an extraction condition that the differential value continuously increases or decreases for a certain period during which the charge / discharge capacity Q changes. doing. In the present embodiment, the differential value is calculated for each unit discharge capacity, and the extraction condition is that at least three differential values continue to increase or decrease continuously.

  Next, in step S7, it is determined whether or not an inflection point has been extracted. If the inflection point cannot be extracted (S7: NO), it is determined in step S13 whether the end condition is satisfied, and the process returns to step S1 until the end condition is satisfied.

  If the inflection point can be extracted in step S7 (S7: YES), in the next step S8, the past characteristic data acquisition unit 60 shows the relationship between the charge / discharge capacity of the secondary battery and the deformation amount. Of the characteristic curve L1, the past characteristic data having at least the full charge point Pf, the remaining amount zero point Pe, and the stage inflection points P1 and P2 are acquired.

  In the next step S9, the predicted characteristic data generation unit 63 performs a fitting process on the characteristic curve L1 indicated by the past characteristic data to the characteristic curve Ln indicated by the time-series data of the actual measurement values with the extracted inflection point as a reference. Then, the prediction characteristic data indicating the characteristic curve L2 in which the part not included in the time series data of the actual measurement values is interpolated is generated.

  When the actual measurement value acquisition unit 61 extracts one inflection point in step S6, the predicted characteristic data generation unit 63 changes to either one of the full charge time point Pf or the zero remaining time point Pe in the past characteristic data. A coefficient for matching the two inflection points P1 to the corresponding two points in the time-series data of the actual measurement values by expansion / contraction of the characteristic curve is calculated, and the entire characteristic curve is adjusted by using the coefficient to adjust the inflection point P1. Are matched to generate prediction characteristic data.

  When the actual measurement value acquisition unit 61 extracts two inflection points in step S6, the predicted characteristic data generation unit 63 converts the two inflection points P1 and P2 in the past characteristic data into the time series data of the actual measurement values. A coefficient for matching the corresponding two inflection points by expansion / contraction of the characteristic curve is calculated, the entire characteristic curve is adjusted using the coefficient, the inflection point P2 extracted most recently is matched, and the predicted characteristic data is obtained. Generate.

  In the next step S10, it is determined whether or not two different inflection points are extracted from the time-series data of the actually measured values. The determination of whether the inflection point is different can be made based on the SOC and charge / discharge capacity. In step S12, when generating the deterioration information based on the prediction characteristic data, the prediction characteristic data generated based on two different inflection points is more accurate than the data generated based on only one inflection point. This is because it is considered high. Of course, the deterioration information may be generated in a state where only one inflection point is extracted.

  If it is determined in step S10 that two different inflection points have been extracted (S10: YES), in step S11, the initial characteristic data acquisition unit 65 determines the charge / discharge capacity and deformation amount of the secondary battery. Among the characteristic curve L0 indicating the relationship, initial characteristic data having at least a maximum thickness point PL, a fully charged time point Pf, a remaining amount zero time point Pe, and stage inflection points P1 and P2 are acquired. The maximum thickness point PL can be omitted.

  In the next step S12, the deterioration information generation unit 66 obtains at least one of the deterioration capacity, the degree of change in the amount of active material that contributes to charging / discharging, or the amount of change in thickness until the lithium deposition starting from the time when the remaining amount is zero. .

  In step S12, when obtaining the degree of change in the amount of active material that contributes to charging / discharging, the deterioration information generation unit 66 has two inflection points P1 and P2 in the initial characteristic data and two corresponding inflections in the predicted characteristic data. The expansion rate of the charge / discharge capacity for matching the points P1 and P2 with the expansion and contraction of the characteristic curve is calculated.

  In step S12, when the deterioration state is obtained, the deterioration information generation unit 66 includes two inflection points P1 and P2 in the initial characteristic data (L0) and two inflection points P1 and P2 in the prediction characteristic data (L2). The initial characteristic data is adjusted by expanding / contracting the entire characteristic curve (L0) indicated by the initial characteristic data by using coefficients (expansion rate of charge / discharge capacity, enlargement rate of deformation amount) for matching the characteristic curve by expansion / contraction of the characteristic curve. The two inflection points P1 and P2 in the subsequent characteristic curve (L0 ′) and the characteristic curve (L2) of the predicted characteristic data are made to coincide with each other, and the left / right deviation amount D1 and the remaining amount of charge / discharge capacity between the full charge points Pf The average value [(D1 + D2) / 2] of the left-right deviation amount D2 of the charge / discharge capacity between the time points Pe is calculated as the battery deterioration state.

In step S12, when obtaining the amount of change in thickness from the time when the remaining amount is zero until the lithium deposition, the deterioration information generation unit 66 uses the two inflection points P1 and P2 in the initial characteristic data and the predicted characteristic data. A coefficient for matching the two inflection points P1 and P2 with the expansion and contraction of the characteristic curve is calculated, and the maximum thickness point PL ′ in the predicted characteristic data is specified from the maximum thickness point PL in the initial characteristic data using the coefficient. Then, a deformation amount T1 from the maximum thickness point PL ′ to the remaining time zero point Pe in the predicted characteristic data is calculated.
When Pf = PL is set, the deterioration information generation unit 66 calculates the deformation amount T1 from the maximum thickness point PL (Pf) to the remaining zero point Pe in the predicted characteristic data.

  When the process of step S12 ends, the process proceeds to step S13.

In step S3, the remaining capacity calculation unit 64 calculates the difference between the charge / discharge capacity Q _Pn indicated by the actual measurement value at the time of predicting the remaining capacity and the charge / discharge capacity Q _Pe at the remaining time zero Pe in the predicted characteristic data. Calculated as Qr. That is, once the prediction characteristic data is generated, the remaining capacity is calculated every time the actual measurement data is acquired.

Steps S4 to S5 are processes for reviewing the predicted characteristic data. Specifically, in step S4, whether the predetermined regeneration condition is satisfied when the inflection points P1 and P2 are included in the time series data of the actually measured values and the predicted characteristic data has already been generated. Determine whether or not. In the present embodiment, as a predetermined regeneration condition, it is determined whether or not the difference between the once generated prediction characteristic data and the actually measured value exceeds a threshold value. When the difference exceeds the threshold value, in step S5, the prediction characteristic data generation unit 63 regenerates the prediction characteristic data. This is because once the predicted characteristic data is generated based on the actual measurement value, it does not match the actual measurement value. Here, attention is paid to the error as the regeneration condition, but there are various examples such as the fact that a predetermined charge / discharge capacity has been reached and that a predetermined time has elapsed from the predicted characteristic data, and can be changed as appropriate.
As with the method shown in FIG. 9B, the method for generating the predicted characteristic data calculates a coefficient for making the characteristic curve of the predicted characteristic data coincide with the characteristic curve of the time-series data of the actual measurement value by expansion and contraction of the characteristic curve. Is used to adjust the expansion and contraction of the entire characteristic curve, match the inflection points extracted most recently, and regenerate the predicted characteristic data. The coefficient is at least the ratio between the two inflection points, the ratio between the latest measured value and the inflection point closer to the latest measured value, and the weight between the two inflection points is set to another interval. It is preferable that the weight is set to be greater than the weighting and calculated by aggregation. In this way, fitting is performed so that the latest measured values are matched using not only one section between two inflection points but also the ratio of the two sections. It becomes a shape.
Further, when the time series data of the actual measurement values includes the full charge time point Pf and the remaining zero time point Pe which is far from the latest actual measurement value, the full charge time point Pf and the remaining zero time point Pe Use the ratio between the one farthest from the latest measured value and one inflection point, the ratio between the two inflection points, the ratio between the latest measured value and the other inflection point close to the latest measured value. It is preferable that the weighting between two inflection points is set larger than the weighting of other sections and is calculated by aggregation. In this way, since the ratio of the three sections is used and fitting is performed so that either the fully charged time point Pf or the remaining amount zero time point Pe matches the latest measured value, the predicted characteristic data is in a form corresponding to the actually measured value. become.

  The detection unit 4 is disposed at a location where a change in the external field can be detected, and is preferably attached to a relatively rigid location that is not easily affected by the swelling of the secondary battery 2. In this embodiment, as shown in FIG. 2B, the detection unit 4 is attached to the inner surface of the casing 11 of the battery module facing the wall 28a. The casing 11 of the battery module is formed of, for example, metal or plastic, and a laminate film may be used. In the drawing, the detection unit 4 is disposed close to the polymer matrix layer 3, but may be disposed away from the polymer matrix layer 3.

  In the present embodiment, an example is shown in which the polymer matrix layer 3 contains a magnetic filler as the filler, and the detection unit 4 detects a change in the magnetic field as the external field. In this case, the polymer matrix layer 3 is preferably a magnetic elastomer layer in which a magnetic filler is dispersed in a matrix made of an elastomer component.

  Examples of the magnetic filler include rare earth-based, iron-based, cobalt-based, nickel-based, and oxide-based materials, but a rare earth-based material that can obtain higher magnetic force is preferable. The shape of the magnetic filler is not particularly limited, and may be spherical, flat, needle-like, columnar, or indefinite. The average particle size of the magnetic filler is preferably 0.02 to 500 μm, more preferably 0.1 to 400 μm, and still more preferably 0.5 to 300 μm. When the average particle size is smaller than 0.02 μm, the magnetic properties of the magnetic filler tend to be lowered, and when the average particle size exceeds 500 μm, the mechanical properties of the magnetic elastomer layer tend to be lowered and become brittle.

  The magnetic filler may be introduced into the elastomer after magnetization, but is preferably magnetized after being introduced into the elastomer. Magnetization after introduction into the elastomer facilitates control of the polarity of the magnet and facilitates detection of the magnetic field.

  As the elastomer component, a thermoplastic elastomer, a thermosetting elastomer, or a mixture thereof can be used. Examples of the thermoplastic elastomer include styrene-based thermoplastic elastomer, polyolefin-based thermoplastic elastomer, polyurethane-based thermoplastic elastomer, polyester-based thermoplastic elastomer, polyamide-based thermoplastic elastomer, polybutadiene-based thermoplastic elastomer, polyisoprene-based thermoplastic elastomer, A fluororubber-based thermoplastic elastomer can be used. Examples of the thermosetting elastomer include polyisoprene rubber, polybutadiene rubber, styrene-butadiene rubber, polychloroprene rubber, nitrile rubber, ethylene-propylene rubber and other diene synthetic rubbers, ethylene-propylene rubber, butyl rubber, acrylic rubber, Non-diene synthetic rubbers such as polyurethane rubber, fluorine rubber, silicone rubber, epichlorohydrin rubber, and natural rubber can be mentioned. Among these, a thermosetting elastomer is preferable because it can suppress the sag of the magnetic elastomer accompanying heat generation and overload of the battery. More preferred is polyurethane rubber (also referred to as polyurethane elastomer) or silicone rubber (also referred to as silicone elastomer).

  The polyurethane elastomer is obtained by reacting a polyol and a polyisocyanate. When using a polyurethane elastomer as an elastomer component, an active hydrogen-containing compound and a magnetic filler are mixed, and an isocyanate component is mixed here to obtain a mixed solution. Moreover, a liquid mixture can also be obtained by mixing a magnetic filler with an isocyanate component and mixing an active hydrogen-containing compound. The mixed liquid is poured into a mold subjected to a release treatment, and then heated to a curing temperature and cured to produce a magnetic elastomer. When a silicone elastomer is used as an elastomer component, a magnetic elastomer can be produced by adding a magnetic filler to a silicone elastomer precursor, mixing it, putting it in a mold, and then heating and curing it. In addition, you may add a solvent as needed.

  As the isocyanate component that can be used in the polyurethane elastomer, compounds known in the field of polyurethane can be used. For example, 2,4-toluene diisocyanate, 2,6-toluene diisocyanate, 2,2′-diphenylmethane diisocyanate, 2,4′-diphenylmethane diisocyanate, 4,4′-diphenylmethane diisocyanate, 1,5-naphthalene diisocyanate, p-phenylene Aromatic diisocyanates such as diisocyanate, m-phenylene diisocyanate, p-xylylene diisocyanate, m-xylylene diisocyanate, aliphatic diisocyanates such as ethylene diisocyanate, 2,2,4-trimethylhexamethylene diisocyanate, 1,6-hexamethylene diisocyanate 1,4-cyclohexane diisocyanate, 4,4′-dicyclohexylmethane diisocyanate, isophorone diisocyanate, nor It can be mentioned alicyclic diisocyanates such as Renan diisocyanate. These may be used alone or in combination of two or more. The isocyanate component may be modified such as urethane modification, allophanate modification, biuret modification, and isocyanurate modification. Preferred isocyanate components are 2,4-toluene diisocyanate, 2,6-toluene diisocyanate, 4,4'-diphenylmethane diisocyanate, more preferably 2,4-toluene diisocyanate, 2,6-toluene diisocyanate.

  As the active hydrogen-containing compound, those usually used in the technical field of polyurethane can be used. For example, polytetramethylene glycol, polypropylene glycol, polyethylene glycol, polyether polyols typified by copolymers of propylene oxide and ethylene oxide, polybutylene adipates, polyethylene adipates, and 3-methyl-1,5-pentane adipates Polyester polyol such as polyester polyol such as polyester polyol, polycaprolactone polyol, reaction product of polyester glycol such as polycaprolactone glycol and alkylene carbonate, and ethylene carbonate are reacted with polyhydric alcohol. Polyester polycarbonate polyol reacted with organic dicarboxylic acid, polyhydroxyl compound and aryl carbonate It can be mentioned a high molecular weight polyol and polycarbonate polyols obtained by ester exchange reaction. These may be used alone or in combination of two or more.

  In addition to the high molecular weight polyol component described above as the active hydrogen-containing compound, ethylene glycol, 1,2-propylene glycol, 1,3-propylene glycol, 1,4-butanediol, 1,6-hexanediol, neopentyl glycol, 1,4-cyclohexanedimethanol, 3-methyl-1,5-pentanediol, diethylene glycol, triethylene glycol, 1,4-bis (2-hydroxyethoxy) benzene, trimethylolpropane, glycerin, 1,2,6- Hexanetriol, pentaerythritol, tetramethylolcyclohexane, methylglucoside, sorbitol, mannitol, dulcitol, sucrose, 2,2,6,6-tetrakis (hydroxymethyl) cyclohexanol, and triethanolamine Low molecular weight polyol component equal, ethylenediamine, tolylenediamine, diphenylmethane diamine, may be used low molecular weight polyamine component of diethylenetriamine. These may be used alone or in combination of two or more. Furthermore, 4,4′-methylenebis (o-chloroaniline) (MOCA), 2,6-dichloro-p-phenylenediamine, 4,4′-methylenebis (2,3-dichloroaniline), 3,5-bis ( Methylthio) -2,4-toluenediamine, 3,5-bis (methylthio) -2,6-toluenediamine, 3,5-diethyltoluene-2,4-diamine, 3,5-diethyltoluene-2,6- Diamine, trimethylene glycol-di-p-aminobenzoate, polytetramethylene oxide-di-p-aminobenzoate, 1,2-bis (2-aminophenylthio) ethane, 4,4'-diamino-3,3 ' -Diethyl-5,5'-dimethyldiphenylmethane, N, N'-di-sec-butyl-4,4'-diaminodiphenylmethane, 4,4'-diamy −3,3′-diethyldiphenylmethane, 4,4′-diamino-3,3′-diethyl-5,5′-dimethyldiphenylmethane, 4,4′-diamino-3,3′-diisopropyl-5,5′- Dimethyldiphenylmethane, 4,4′-diamino-3,3 ′, 5,5′-tetraethyldiphenylmethane, 4,4′-diamino-3,3 ′, 5,5′-tetraisopropyldiphenylmethane, m-xylylenediamine, Polyamines exemplified by N, N′-di-sec-butyl-p-phenylenediamine, m-phenylenediamine, p-xylylenediamine and the like can also be mixed. Preferred active hydrogen-containing compounds are polytetramethylene glycol, polypropylene glycol, a copolymer of propylene oxide and ethylene oxide, 3-methyl-1,5-pentane adipate, more preferably a copolymer of polypropylene glycol, propylene oxide and ethylene oxide. It is a coalescence.

  Preferred combinations of the isocyanate component and the active hydrogen-containing compound include, as the isocyanate component, one or more of 2,4-toluene diisocyanate, 2,6-toluene diisocyanate, and 4,4′-diphenylmethane diisocyanate, active hydrogen The containing compound is polytetramethylene glycol, polypropylene glycol, a copolymer of propylene oxide and ethylene oxide, and a combination of one or more of 3-methyl-1,5-pentaneadipate. More preferably, it is a combination of 2,4-toluene diisocyanate and / or 2,6-toluene diisocyanate as the isocyanate component and polypropylene glycol and / or a copolymer of propylene oxide and ethylene oxide as the active hydrogen-containing compound. is there.

  The polymer matrix layer 3 may be a foam containing dispersed fillers and bubbles. A general resin foam can be used as the foam, but it is preferable to use a thermosetting resin foam in consideration of characteristics such as compression set. Examples of the thermosetting resin foam include a polyurethane resin foam and a silicone resin foam. Among these, a polyurethane resin foam is preferable. The above-mentioned isocyanate component and active hydrogen-containing compound can be used for the polyurethane resin foam.

  The amount of the magnetic filler in the magnetic elastomer is preferably 1 to 450 parts by weight, more preferably 2 to 400 parts by weight with respect to 100 parts by weight of the elastomer component. If it is less than 1 part by weight, it tends to be difficult to detect a change in the magnetic field, and if it exceeds 450 parts by weight, the magnetic elastomer itself may become brittle.

  For the purpose of preventing rust of the magnetic filler, a sealing material for sealing the polymer matrix layer 3 may be provided to the extent that the flexibility of the polymer matrix layer 3 is not impaired. As the sealing material, a thermoplastic resin, a thermosetting resin, or a mixture thereof can be used. Examples of the thermoplastic resin include styrene-based thermoplastic elastomers, polyolefin-based thermoplastic elastomers, polyurethane-based thermoplastic elastomers, polyester-based thermoplastic elastomers, polyamide-based thermoplastic elastomers, polybutadiene-based thermoplastic elastomers, polyisoprene-based thermoplastic elastomers, Fluorine-based thermoplastic elastomer, ethylene / ethyl acrylate copolymer, ethylene / vinyl acetate copolymer, polyvinyl chloride, polyvinylidene chloride, chlorinated polyethylene, fluororesin, polyamide, polyethylene, polypropylene, polyethylene terephthalate, polybutylene terephthalate, polystyrene, polybutadiene Etc. Examples of the thermosetting resin include polyisoprene rubber, polybutadiene rubber, styrene / butadiene rubber, polychloroprene rubber, diene-based synthetic rubber such as acrylonitrile / butadiene rubber, ethylene / propylene rubber, ethylene / propylene / diene rubber, butyl rubber, Non-diene rubbers such as acrylic rubber, polyurethane rubber, fluorine rubber, silicone rubber, epichlorohydrin rubber, natural rubber, polyurethane resin, silicone resin, epoxy resin and the like can be mentioned. These films may be laminated, or may be a film including a metal foil such as an aluminum foil or a metal vapor deposition film in which a metal is vapor deposited on the film.

  The polymer matrix layer 3 may be one in which fillers are unevenly distributed in the thickness direction. For example, the polymer matrix layer 3 may have a structure composed of two layers of a region on one side with a relatively large amount of filler and a region on the other side with a relatively small amount of filler. In the region on one side containing a large amount of filler, the change in the external field with respect to the small deformation of the polymer matrix layer 3 becomes large, so that the sensor sensitivity to a low internal pressure can be enhanced. Further, the region on the other side with relatively little filler is relatively flexible and easy to move. By attaching this region, the polymer matrix layer 3 (especially the region on one side) is likely to be deformed.

  The filler uneven distribution ratio in the region on one side is preferably more than 50, more preferably 60 or more, and still more preferably 70 or more. In this case, the filler uneven distribution rate in the other region is less than 50. The filler uneven distribution rate in the region on one side is 100 at the maximum, and the filler uneven distribution rate in the region on the other side is 0 at the minimum. Therefore, a laminate structure of an elastomer layer containing a filler and an elastomer layer not containing a filler may be used. For the uneven distribution of the filler, after introducing the filler into the elastomer component, it can be allowed to stand at room temperature or at a predetermined temperature, and then spontaneously settled according to the weight of the filler, by changing the temperature and time of standing. The filler uneven distribution rate can be adjusted. The filler may be unevenly distributed using a physical force such as centrifugal force or magnetic force. Alternatively, the polymer matrix layer may be constituted by a laminate composed of a plurality of layers having different filler contents.

  The filler uneven distribution rate is measured by the following method. That is, the cross section of the polymer matrix layer is observed at 100 times using a scanning electron microscope-energy dispersive X-ray analyzer (SEM-EDS). The area of the entire cross section in the thickness direction and the two areas obtained by dividing the cross section into two in the thickness direction are each subjected to elemental analysis of a metal element specific to the filler (for example, Fe element in the case of the magnetic filler of this embodiment). Find the abundance. For this abundance, the ratio of one area to the entire area in the thickness direction is calculated, and this is used as the filler uneven distribution rate in the one area. The filler uneven distribution rate in the other region is the same as this.

  The other side region with relatively little filler may have a structure formed of a foam containing bubbles. Thereby, the polymer matrix layer 3 is further easily deformed and the sensor sensitivity is enhanced. Moreover, the area | region of one side may be formed with the foam with the area | region of the other side, and the polymer matrix layer 3 in that case becomes a foam entirely. Such a polymer matrix layer in which at least a part in the thickness direction is a foam is composed of a laminate composed of a plurality of layers (for example, a non-foamed layer containing a filler and a foamed layer not containing a filler). It doesn't matter.

  For example, a reed switch, a magnetoresistive element, a Hall element, a coil, an inductor, an MI element, a fluxgate sensor, or the like can be used as the detection unit 4 that detects a change in the magnetic field. Examples of the magnetoresistive element include a semiconductor compound magnetoresistive element, an anisotropic magnetoresistive element (AMR), a giant magnetoresistive element (GMR), and a tunnel magnetoresistive element (TMR). Among these, the Hall element is preferable because it has high sensitivity over a wide range and is useful as the detection unit 4. As the Hall element, for example, EQ-430L manufactured by Asahi Kasei Electronics Corporation can be used.

  Since the secondary battery 2 in which the gas expansion has progressed may lead to troubles such as ignition and rupture, in this embodiment, when the expansion amount when the secondary battery 2 is deformed is greater than or equal to a predetermined amount, charging and discharging are performed. It is configured to be blocked. Specifically, the signal detected by the detection sensor 5 is transmitted to the control device 6, and the control device 6 transmits a signal to the switching circuit 7 when a change in the external field exceeding the set value is detected by the detection sensor 5. Then, the current from the power generation device (or charging device) 8 is cut off, and charging / discharging to the battery module 1 is cut off. Thereby, the trouble resulting from gas bulging can be prevented beforehand.

  In the above-described embodiment, an example in which the secondary battery is a lithium ion secondary battery is shown, but the present invention is not limited to this. The secondary battery used is not limited to a non-aqueous electrolyte secondary battery such as a lithium ion battery, and may be an aqueous electrolyte secondary battery such as a nickel metal hydride battery.

  In the above-described embodiment, the example in which the change in the magnetic field due to the deformation of the polymer matrix layer is detected by the detection unit has been described. However, another change in the external field may be detected. For example, the polymer matrix layer may contain conductive fillers such as metal particles, carbon black, and carbon nanotubes as fillers, and the detector may detect changes in the electric field (changes in resistance and dielectric constant) as external fields. It is done.

As described above, the secondary battery state prediction method of the present embodiment is
Step S1 for obtaining an actual measurement value corresponding to the charge / discharge capacity Q and the deformation amount T of the secondary battery 2,
Step S6 for extracting at least one inflection point P1 (P2) in the characteristic curve Ln indicating the relationship between the charge / discharge capacity Q of the secondary battery and the deformation amount T from the time series data of the actual measurement values;
A step of acquiring past characteristic data having at least a full charge time point Pf, a remaining amount zero time point Pe, and stage inflection points P1 and P2 among the characteristic curve L1 indicating the relationship between the charge / discharge capacity and deformation amount of the secondary battery. S8,
Using the extracted inflection point as a reference, the characteristic curve L1 indicated by the past characteristic data is fitted to the characteristic curve Ln indicated by the time-series data of the actual measurement values, and a portion not included in the time-series data of the actual measurement values is interpolated. Step S9 for generating predicted characteristic data indicating the characteristic curve L2,
including.

The state prediction system of the secondary battery of this embodiment is
An actual value acquisition unit 61 for acquiring actual values corresponding to the charge / discharge capacity Q and the deformation amount T of the secondary battery 2;
An inflection point extraction unit 62 for extracting at least one inflection point P1 (P2) in the characteristic curve Ln indicating the relationship between the charge / discharge capacity Q of the secondary battery and the deformation amount T from the time series data of the actual measurement values;
In the past, the characteristic curve L1 indicating the relationship between the charge / discharge capacity of the secondary battery and the deformation amount is obtained by acquiring past characteristic data having at least the full charge point Pf, the remaining amount zero point Pe, and the stage inflection points P1 and P2. A characteristic data acquisition unit 60;
Using the extracted inflection point as a reference, the characteristic curve L1 indicated by the past characteristic data is fitted to the characteristic curve Ln indicated by the time-series data of the actual measurement values, and a portion not included in the time-series data of the actual measurement values is interpolated. A predicted characteristic data generating unit 63 that generates predicted characteristic data indicating the characteristic curve L2,
Is provided.

The characteristic curve indicating the relationship between the charge / discharge capacity of the secondary battery and the deformation amount has two inflection points P1 and P2 at which the inclination of the characteristic curve greatly changes with the stage change. According to this method, the inflection point in the characteristic curve is extracted from the time series data of the actually measured value, and the past characteristic data and the actually measured value data are fitted on the basis of the extracted inflection point. Even if there is no long-term measurement data from full charge to complete discharge, if there is measured data for a short period of time, for example, from full charge to inflection point, from one inflection point to the other inflection point, Predictive characteristic data with a certain degree of accuracy can be generated.
Therefore, in the actual use in which charging and discharging are performed at random, instead of a series of charging and discharging from full charge to complete discharge, the prediction characteristic data can be detected, and the state of the secondary battery can be predicted.

  In the present embodiment, the inflection point extraction unit 62 sets the extraction condition according to the charge / discharge capacity indicated by the acquired actual measurement value, and the differential value of the deformation amount related to the charge / discharge capacity determined based on the acquired actual measurement value is the extraction condition. In the case where the condition is satisfied, the actually measured value is extracted as an inflection point (step S7).

  The two inflection points P1 and P2 included in the characteristic curve can be identified by the charge / discharge capacity Q. In the present embodiment, since the inflection point extraction unit 62 sets the extraction condition according to the charge / discharge capacity Q indicated by the actual measurement value, the extraction accuracy is higher than when the inflection point is extracted only by the change of the differential value. Can be improved.

In the present embodiment, the inflection point extraction unit 62 has the charge / discharge capacity Q _Pn indicated by the acquired actual measurement value centered on the charge / discharge capacities Q _P1 and Q _P2 of the inflection points P1 and P2 in the past characteristic data. When entering the predetermined range, the threshold values Th _P1 and Th _P2 of the differential value are set based on the past characteristic data, and the fact that the differential value has passed the threshold values Th _P1 and Th _P2 is set as the extraction condition (step S7). .

  In this way, the inflection points P1 and P2 can be extracted with a simple calculation.

In the present embodiment, the inflection point extraction unit 62 continuously increases the differential value during a certain period in which the charge / discharge capacity changes in addition to the differential value passing through the threshold values Th _P1 and Th _P2. Or to keep falling is set as an extraction condition (step S7).

With such an extraction condition, even if the differential value temporarily passes the threshold values Th _P1 and Th _P2 , the extraction condition is not satisfied, so that the extraction accuracy can be further improved.

In the present embodiment, the prediction characteristic data generation unit 63
When one inflection point P1 is extracted and the measured value time-series data includes the measured value corresponding to any one of the fully charged time point Pf or the remaining amount zero time point Pe, in the past characteristic data Calculates a coefficient for matching one of the full charge time point Pf or the zero remaining time point Pe and the inflection point P1 with the corresponding two points in the time-series data of the actual measurement value by expansion and contraction of the characteristic curve. The entire characteristic curve is adjusted using coefficients, and the inflection point P1 is matched to generate predicted characteristic data (step S9).
Or
When one inflection point is extracted and the actual measurement value time-series data does not include the actual measurement value corresponding to any one of the fully charged time point Pf or the remaining charge zero time point Pe, the past characteristic data indicates The characteristic curve L1 indicated by the past characteristic data is moved so that the most recently extracted inflection point P1 matches the inflection point P1 in the past characteristic data without enlarging or reducing the characteristic curve L1. To generate prediction characteristic data (step S9),
Or
When two inflection points P1 and P2 are extracted, the two inflection points P1 and P2 in the past characteristic data are converted into two corresponding inflection points in the time-series data of the actual measurement values by expanding and contracting the characteristic curve. A coefficient for matching is calculated, the entire characteristic curve is stretched and adjusted using the coefficient, the inflection point P2 extracted most recently is matched, and predicted characteristic data is generated (step S9).

  According to this method, since the inflection points or the inflection points and the end of the characteristic curve are matched with priority, the shape of the characteristic curve accompanying the stage change is less likely to be collapsed and predicted compared to general fitting. The accuracy can be improved.

  In the present embodiment, the prediction characteristic data generation unit 63 includes two inflection points in the time-series data of actual measurement values and the prediction characteristic data has already been generated, and a predetermined regeneration condition is satisfied (step S4: YES), a coefficient for matching the characteristic curve L2 of the predicted characteristic data with the characteristic curve Ln of the time-series data of the actual measurement value is calculated by the expansion / contraction of the characteristic curve, and the entire characteristic curve L2 is adjusted by using the coefficient Then, the inflection points extracted most recently are matched, and the prediction characteristic data is regenerated. The coefficient is at least the ratio between the two inflection points, the ratio between the latest measured value and the inflection point closer to the latest measured value, and the weight between the two inflection points is set to another interval. It is set to be larger than the weight of and is calculated by aggregation.

  In this way, fitting is performed so that the latest measured values are matched using not only one section between two inflection points but also the ratio of the two sections. It becomes a shape.

In the present embodiment, the remaining capacity calculation unit 64 stores the difference between the charge / discharge capacity Q _Pn indicated by the actual measurement value at the time of predicting the remaining capacity and the charge / discharge capacity Q _Pe at the remaining time zero Pe in the predicted characteristic data. It calculates as capacity | capacitance Qr (step S3).

  Thus, since the prediction characteristic data includes data of the remaining time zero Pe, the remaining amount can be accurately calculated.

  In the present embodiment, an initial characteristic data acquisition unit 65 and a deterioration information generation unit 66 are included. The initial characteristic data acquisition unit 65 has at least a full charge point Pf, a remaining zero point Pe, and stage inflection points P1 and P2 among the characteristic curve L0 indicating the relationship between the charge / discharge capacity of the secondary battery and the deformation amount. Initial characteristic data is acquired (step S11). The deterioration information generation unit 66 has a charge / discharge capacity for matching the two inflection points P1 and P2 in the initial characteristic data with the corresponding two inflection points P1 and P2 in the prediction characteristic data by expansion and contraction of the characteristic curve. Is calculated as the degree of change in the amount of active material that contributes to charge / discharge (step S12).

  In this way, it becomes possible to know the degree of change in the amount of active material that contributes to charging and discharging.

  In the present embodiment, the deterioration information generation unit 66 has a ratio between two inflection points, a ratio between one inflection point and a full charge point Pf, and the other inflection point with respect to the expansion rate of the charge / discharge capacity. The weight between the two inflection points is set to be larger than the weights of the other sections using the ratio between the remaining amount zero point Pe and the remaining amount zero point Pe (step S12).

  In this way, compared to the ratio between the two inflection points, the full charge time point Pf and the remaining zero time point Pe are close as fitting results, so the degree of coincidence in the entire characteristic curve is increased, and the calculation accuracy is improved. I think it will improve.

  In the present embodiment, an initial characteristic data acquisition unit 65 and a deterioration information generation unit 66 are included. The initial characteristic data acquisition unit 65 has at least a full charge point Pf, a remaining zero point Pe, and stage inflection points P1 and P2 among the characteristic curve L0 indicating the relationship between the charge / discharge capacity of the secondary battery and the deformation amount. Initial characteristic data is acquired (step S11). The deterioration information generation unit 66 calculates a coefficient for matching the two inflection points P1 and P2 in the initial characteristic data and the two inflection points P1 and P2 in the prediction characteristic data by expansion and contraction of the characteristic curve, The entire characteristic curve L0 indicated by the initial characteristic data is adjusted by using the coefficient, and the two inflection points P1 and P2 in the characteristic curve L0 ′ after adjustment of the initial characteristic data and the characteristic curve L2 of the predicted characteristic data are matched. And calculating the average value [(D1 + D2) / 2] of the right / left deviation amount D1 of the charge / discharge capacity between the full charge points Pf and the right / left deviation amount D2 of the charge / discharge capacity between the remaining time points Pe. (Step S12).

  In this way, it is possible to calculate the deterioration state of the battery as to how much the one side reaction of the positive electrode or the negative electrode is more than the other side reaction.

  In the present embodiment, an initial characteristic data acquisition unit 65 and a deterioration information generation unit 66 are included. The initial characteristic data acquisition unit 65 includes at least a maximum thickness point PL, a full charge point Pf, a remaining zero point Pe, and a stage inflection point in the characteristic curve L0 indicating the relationship between the charge / discharge capacity of the secondary battery and the deformation amount. Initial characteristic data having P1 and P2 is acquired (step S11). The deterioration information generation unit 66 calculates a coefficient for matching the two inflection points P1 and P2 in the initial characteristic data and the two inflection points P1 and P2 in the prediction characteristic data by expansion and contraction of the characteristic curve, Using the coefficient, the maximum thickness point PL ′ in the predicted characteristic data is identified from the maximum thickness point PL in the initial characteristic data, and the deformation amount T1 from the maximum thickness point PL ′ in the predicted characteristic data to the remaining zero point Pe is calculated as the remaining amount. It is calculated as the amount of change in thickness from the zero point in time until lithium deposition.

  In this way, it is possible to calculate the amount of change in thickness from the time when the remaining amount is zero until the lithium deposition. This value is an index of charge control and is useful.

  In the present embodiment, the deterioration information generation unit 66 has a coefficient between a ratio between two inflection points, a ratio between one inflection point and a fully charged time point Pf, and the other inflection point and the remaining zero time point. Using the ratio between Pe, the weighting between two inflection points is set larger than the weighting of the other sections, and calculation is performed by tabulation (step S12).

  In this way, compared to the ratio between the two inflection points, the full charge time point Pf and the remaining zero time point Pe are close as fitting results, so the degree of coincidence in the entire characteristic curve is increased, and the calculation accuracy is improved. I think it will improve.

  In the present embodiment, the fully charged time point Pf in the predicted characteristic data is the maximum thickness point PL, and the deterioration information generation unit 66 determines the deformation amount T1 from the maximum thickness point PL (Pf) in the predicted characteristic data to the zero remaining time point Pe. Is calculated as the amount of change in thickness from the time when the remaining amount is zero until the lithium deposition.

  In this way, it is possible to calculate the amount of change in thickness from the time when the remaining amount is zero until the lithium deposition. This value is an index of charge control and is useful.

  In the charge control system of the present embodiment, the deformation amount T of the secondary battery detected by the deterioration information generation unit 66 and the detection sensor 5 does not exceed the thickness change amount T1 generated by the deterioration information generation unit 66. And a charging control unit 67 that controls charging.

  According to this configuration, charge control is performed so that lithium does not precipitate, and therefore, there is a case where charging can be performed rapidly without causing deterioration due to lithium precipitation.

  In the present embodiment, the polymer matrix layer 3 is attached directly or indirectly to the secondary battery 2, and the polymer matrix layer 3 is a filler that changes the external field according to the deformation of the polymer matrix layer 3. The deformation amount T of the secondary battery 2 is detected by detecting the change in the external field in accordance with the deformation of the polymer matrix layer 3 contained.

  In this way, an actual measurement value corresponding to the deformation amount of the secondary battery 2 can be detected appropriately.

The program according to the present embodiment is a program that causes a computer to execute the above method.
By executing these programs, it is possible to obtain the operational effects of the above method.

  The present invention is not limited to the embodiment described above, and various improvements and modifications can be made without departing from the spirit of the present invention. For example, the execution order of each process such as operation, procedure, step, and stage in the apparatus, system, program, and method shown in the claims, the description, and the drawings is the output of the previous process. Unless used in processing, it can be realized in any order. Concerning the flow in the claims, the specification, and the drawings, even if “first”, “next”, etc. are used for convenience, it does not mean that it is essential to execute in this order. .

In the embodiment described above, inflection points are extracted from time-series data of actual measurement values, and predicted characteristic data is generated by fitting past characteristic data. However, if the accuracy is not considered, the prediction characteristic data generation process 4 is also exemplified as follows. That is, the predicted characteristic data generation process 4 _{fits the} characteristic curve L1 indicated by the past characteristic data as it is without being enlarged or reduced based on the charge / discharge capacity _QPn indicated by the actual measurement value at the present time. According to this method, it is possible to generate prediction characteristic data without considering capacity degradation and balance shift.
It is possible to implement at least one generation process among the generation processes 1 to 4 of the prediction characteristic data. Combinations of the generation processes 1 to 4 can be arbitrarily performed. However, generation processing 1, generation processing 2, generation processing 3, and generation processing 4 are arranged in descending order of accuracy. The generation process 1 is the highest and the generation process 4 is the least accurate. On the other hand, the generation process 4 takes the shortest time for generating the prediction characteristic data, and the generation process 1 takes the longest time.

L0: Characteristic curve indicated by initial characteristic data Ln: Characteristic curve indicated by actual measurement value L1: Characteristic curve indicated by past characteristic data L2: Characteristic curve indicated by predicted characteristic data P1, P2: Inflection point Pf: Point of full charge Pe ... When the remaining amount is zero PL: Maximum thickness point Th _P1 , Th _P2 : Threshold Qr: Remaining capacity

Claims (33)

Obtaining actual measurement values corresponding to the charge / discharge capacity and deformation amount of the secondary battery;
Extracting from the time series data of the measured values at least one inflection point in a characteristic curve indicating the relationship between the charge / discharge capacity and deformation amount of the secondary battery;
Of the characteristic curve showing the relationship between the charge / discharge capacity of the secondary battery and the deformation amount, obtaining past characteristic data having at least a full charge time, a remaining time zero, and a stage inflection point;
Using the extracted inflection point as a reference, the characteristic curve indicated by the past characteristic data is fitted to the characteristic curve indicated by the time series data of the actual measurement value, and a portion not included in the time series data of the actual measurement value is interpolated. Generating predicted characteristic data indicating the characteristic curve,
A method for predicting the state of a secondary battery. Set the extraction conditions according to the charge / discharge capacity indicated by the actual measured value,
The method according to claim 1, wherein the measured value is extracted as an inflection point when a differential value of a deformation amount related to the charge / discharge capacity determined based on the acquired measured value satisfies the extraction condition.   When the charge / discharge capacity indicated by the acquired actual measurement value falls within a predetermined range centered on the charge / discharge capacity at the inflection point in the past characteristic data, the threshold value of the differential value is set based on the past characteristic data, The method according to claim 2, wherein an extraction condition is set that the differential value has passed the threshold value.   The extraction condition is set such that the differential value continues to increase or decrease continuously during a certain period when the charge / discharge capacity changes in addition to the differential value passing through the threshold value. 3. The method according to 3. When one inflection point is extracted and the time series data of the actual measurement value includes an actual measurement value corresponding to any one of the full charge time or the remaining time zero, the past characteristic data A coefficient for matching any one point of the full charge time or zero remaining time and the inflection point in FIG. 2 to the corresponding two points in the time-series data of the actual measurement values by expansion / contraction of the characteristic curve Calculating, adjusting the expansion and contraction of the entire characteristic curve using the coefficient, matching the inflection points, generating the predicted characteristic data,
Or
If one inflection point is extracted and the time-series data of the actual measurement value does not include the actual measurement value corresponding to any one of the full charge time point or the zero remaining time point, the past characteristic data The characteristic curve indicated by the past characteristic data is moved so that the inflection point extracted most recently coincides with the inflection point in the past characteristic data without enlarging or reducing the characteristic curve indicated by Interpolating to generate the prediction characteristic data,
Or
When two inflection points are extracted, the two inflection points in the past characteristic data are matched with the corresponding two inflection points in the time-series data of the actual measurement values by expansion and contraction of the characteristic curve. The method according to claim 1, wherein a coefficient is calculated, the entire characteristic curve is stretched and adjusted using the coefficient, the inflection points extracted most recently are matched, and the predicted characteristic data is generated. When the time series data of the actual measurement value includes two inflection points, the prediction characteristic data has already been generated, and a predetermined regeneration condition is satisfied, the characteristic curve of the prediction characteristic data is expressed as the actual measurement value. The coefficient for matching the characteristic curve of the time series data by the expansion and contraction of the characteristic curve is calculated, the entire characteristic curve is adjusted using the coefficient, the inflection points extracted most recently are matched, and the predicted characteristic data Including the step of regenerating
The coefficient uses at least the ratio between the two inflection points, the ratio between the latest measured value and the inflection point on the side closer to the latest measured value, and other weights between the two inflection points. The method according to any one of claims 1 to 5, wherein the calculation is performed by aggregation by setting the weight to be greater than the weight of the section.   The difference between the charge / discharge capacity indicated by the actual measurement value at the time of predicting the remaining capacity and the charge / discharge capacity at the time of the remaining capacity zero in the prediction characteristic data is calculated as the remaining capacity. Method. Of the characteristic curve indicating the relationship between the charge / discharge capacity of the secondary battery and the deformation amount, obtaining initial characteristic data having at least a full charge time, a remaining time zero, and a stage inflection point;
The expansion rate of the charge / discharge capacity for matching the two inflection points in the initial characteristic data and the corresponding two inflection points in the prediction characteristic data by expansion / contraction of the characteristic curve is an activity that contributes to charge / discharge. And calculating the degree of change in the amount of substance.   The expansion ratio of the charge / discharge capacity uses a ratio between the two inflection points, a ratio between one inflection point and a fully charged time point, and a ratio between the other inflection point and the remaining amount zero time point. The method according to claim 8, wherein the weighting between the two inflection points is set to be larger than the weighting of other sections, and is calculated by aggregation. Of the characteristic curve indicating the relationship between the charge / discharge capacity of the secondary battery and the deformation amount, obtaining initial characteristic data having at least a full charge time, a remaining time zero, and a stage inflection point;
A coefficient for matching the two inflection points in the initial characteristic data and the two inflection points in the predicted characteristic data by the expansion and contraction of the characteristic curve is calculated, and the initial characteristic data indicates by using the coefficient While adjusting the expansion and contraction of the entire characteristic curve, the two inflection points in the characteristic curve after the adjustment of the initial characteristic data and the characteristic curve of the predicted characteristic data are matched, and the amount of deviation of the charge / discharge capacity between the full charge points And calculating an average value of deviation amounts of charge / discharge capacities between the remaining time points of zero as a battery deterioration state.   The coefficient uses the ratio between the two inflection points, the ratio between one inflection point and a fully charged time point, and the ratio between the other inflection point and the remaining zero time point. The method according to claim 10, wherein the weighting between the music points is set to be larger than the weighting of other sections and is calculated by aggregation. Among the characteristic curves indicating the relationship between the charge / discharge capacity and deformation amount of the secondary battery, obtaining initial characteristic data having at least the maximum thickness point, the full charge time, the remaining time zero, and the stage inflection point;
A coefficient for matching the two inflection points in the initial characteristic data and the two inflection points in the predicted characteristic data by expansion and contraction of a characteristic curve is calculated, and the coefficient in the initial characteristic data is calculated using the coefficient. The maximum thickness point in the predicted characteristic data is identified from the maximum thickness point, and the deformation amount from the maximum thickness point to the remaining zero time point in the predicted characteristic data is determined until the lithium deposition based on the remaining zero time point. The method according to claim 1, further comprising a step of calculating the thickness change amount.   The coefficient uses the ratio between the two inflection points, the ratio between one inflection point and a fully charged time point, and the ratio between the other inflection point and the remaining zero time point. The method according to claim 12, wherein the weighting between the music points is set to be larger than the weighting of the other sections and is calculated by aggregation. The full charge point in the predicted characteristic data is the maximum thickness point,
8. The method according to claim 1, further comprising: calculating a deformation amount from the maximum thickness point to the remaining zero time point in the predicted characteristic data as a thickness change amount from the remaining zero time point to lithium deposition. The method of crab. A method according to any one of claims 12 to 14 is executed to calculate in advance the amount of change in thickness from the time when the remaining amount is zero to the lithium deposition.
A charge control method for controlling charging so that a deformation amount of a secondary battery detected by a detection sensor does not exceed the thickness change amount.   A polymer matrix layer is directly or indirectly attached to the secondary battery, and the polymer matrix layer contains a filler that changes an external field according to deformation of the polymer matrix layer. The method according to claim 1, wherein a deformation amount of the secondary battery is detected by detecting a change in the external field according to a deformation of the matrix layer. An actual value acquisition unit for acquiring actual values corresponding to the charge / discharge capacity and deformation amount of the secondary battery;
An inflection point extraction unit for extracting at least one inflection point in a characteristic curve indicating a relationship between a charge / discharge capacity and a deformation amount of the secondary battery from the time series data of the actual measurement values;
Of the characteristic curve indicating the relationship between the charge / discharge capacity and deformation amount of the secondary battery, the past characteristic data acquisition unit for acquiring past characteristic data having at least a full charge time, a remaining time zero, and a stage inflection point;
Using the extracted inflection point as a reference, the characteristic curve indicated by the past characteristic data is fitted to the characteristic curve indicated by the time series data of the actual measurement value, and a portion not included in the time series data of the actual measurement value is interpolated. A predicted characteristic data generation unit that generates predicted characteristic data indicating the characteristic curve;
A state prediction system for a secondary battery. The inflection point extraction unit
Set the extraction conditions according to the charge / discharge capacity indicated by the actual measured value,
The system according to claim 17, wherein the measured value is extracted as an inflection point when a differential value of a deformation amount related to the charge / discharge capacity determined based on the acquired measured value satisfies the extraction condition.   When the charge / discharge capacity indicated by the acquired actual measurement value falls within a predetermined range centered on the charge / discharge capacity at the inflection point in the past characteristic data, the inflection point extraction unit is based on the past characteristic data. The system according to claim 18, wherein a threshold value of a differential value is set, and an extraction condition is set that the differential value has passed the threshold value.   In addition to the differential value passing through the threshold value, the inflection point extraction unit is configured to continuously increase or decrease the differential value for a certain period when the charge / discharge capacity changes. The system according to claim 19, wherein the system is set as an extraction condition. The prediction characteristic data generation unit
When one inflection point is extracted and the time series data of the actual measurement value includes an actual measurement value corresponding to any one of the full charge time or the remaining time zero, the past characteristic data A coefficient for matching any one point of the full charge time or zero remaining time and the inflection point in FIG. 2 to the corresponding two points in the time-series data of the actual measurement values by expansion / contraction of the characteristic curve Calculating, adjusting the expansion and contraction of the entire characteristic curve using the coefficient, matching the inflection points, generating the predicted characteristic data,
Or
If one inflection point is extracted and the time-series data of the actual measurement value does not include the actual measurement value corresponding to any one of the full charge time point or the zero remaining time point, the past characteristic data The characteristic curve indicated by the past characteristic data is moved so that the inflection point extracted most recently coincides with the inflection point in the past characteristic data without enlarging or reducing the characteristic curve indicated by Interpolating to generate the prediction characteristic data,
Or
When two inflection points are extracted, the two inflection points in the past characteristic data are matched with the corresponding two inflection points in the time-series data of the actual measurement values by expansion and contraction of the characteristic curve. The system according to claim 17, wherein a coefficient is calculated, the entire characteristic curve is stretched and adjusted using the coefficient, the inflection points extracted most recently are matched, and the predicted characteristic data is generated. The prediction characteristic data generation unit
When the time series data of the actual measurement value includes two inflection points, the prediction characteristic data has already been generated, and a predetermined regeneration condition is satisfied, the characteristic curve of the prediction characteristic data is expressed as the actual measurement value. The coefficient for matching the characteristic curve of the time series data by the expansion and contraction of the characteristic curve is calculated, the entire characteristic curve is adjusted using the coefficient, the inflection points extracted most recently are matched, and the predicted characteristic data Is configured to regenerate,
The coefficient uses at least the ratio between the two inflection points, the ratio between the latest measured value and the inflection point on the side closer to the latest measured value, and other weights between the two inflection points. The system according to any one of claims 17 to 21, wherein the system is set so as to be larger than the weight of the section and calculated by aggregation.   The remaining capacity calculation part which calculates the difference between the charging / discharging capacity which the measured value at the time of predicting a remaining capacity and the charging / discharging capacity at the time of the remaining capacity zero in the said prediction characteristic data as a remaining capacity is provided. A system according to any of the above. Of the characteristic curve showing the relationship between the charge / discharge capacity and deformation amount of the secondary battery, an initial characteristic data acquisition unit that acquires at least the initial characteristic data having the full charge time, the remaining time zero, and the stage inflection point;
The expansion rate of the charge / discharge capacity for matching the two inflection points in the initial characteristic data and the corresponding two inflection points in the prediction characteristic data by expansion / contraction of the characteristic curve is an activity that contributes to charge / discharge. A deterioration information generation unit that calculates the degree of change in the amount of substance;
The system according to claim 17, comprising:   The expansion ratio of the charge / discharge capacity uses a ratio between the two inflection points, a ratio between one inflection point and a fully charged time point, and a ratio between the other inflection point and the remaining amount zero time point. 25. The system according to claim 24, wherein the weighting between the two inflection points is set larger than the weighting of other sections and is calculated by aggregation. Of the characteristic curve showing the relationship between the charge / discharge capacity and deformation amount of the secondary battery, an initial characteristic data acquisition unit that acquires at least the initial characteristic data having the full charge time, the remaining time zero, and the stage inflection point;
A coefficient for matching the two inflection points in the initial characteristic data and the two inflection points in the predicted characteristic data by the expansion and contraction of the characteristic curve is calculated, and the initial characteristic data indicates by using the coefficient While adjusting the expansion and contraction of the entire characteristic curve, the two inflection points in the characteristic curve after the adjustment of the initial characteristic data and the characteristic curve of the predicted characteristic data are matched, and the amount of deviation of the charge / discharge capacity between the full charge points And a deterioration information generation unit that calculates an average value of the deviation amount of the charge / discharge capacity between the remaining time zero points as a battery deterioration state,
The system according to claim 17, comprising:   The coefficient uses the ratio between the two inflection points, the ratio between one inflection point and a fully charged time point, and the ratio between the other inflection point and the remaining zero time point. 27. The system according to claim 26, wherein the weighting between the music points is set to be larger than the weighting of other sections and is calculated by aggregation. Initial characteristic data acquisition for acquiring initial characteristic data having at least the maximum thickness point, the full charge point, the remaining amount zero point, and the stage inflection point among the characteristic curves indicating the relationship between the charge / discharge capacity and the deformation amount of the secondary battery And
A coefficient for matching the two inflection points in the initial characteristic data and the two inflection points in the predicted characteristic data by expansion and contraction of a characteristic curve is calculated, and the coefficient in the initial characteristic data is calculated using the coefficient. The maximum thickness point in the predicted characteristic data is identified from the maximum thickness point, and the deformation amount from the maximum thickness point to the remaining zero time point in the predicted characteristic data is determined until the lithium deposition based on the remaining zero time point. A deterioration information generation unit that calculates the amount of change in thickness;
The system according to claim 17, comprising:   The coefficient uses the ratio between the two inflection points, the ratio between one inflection point and a fully charged time point, and the ratio between the other inflection point and the remaining zero time point. The system according to claim 28, wherein the weighting between music points is set larger than the weighting of other sections and is calculated by aggregation. The full charge point in the predicted characteristic data is the maximum thickness point,
The deterioration information generation part which computes the amount of deformation from the maximum point of thickness in the prediction characteristic data to the time point of the remaining zero as a thickness change amount from the time point of the remaining amount zero to the lithium deposition. 24. The system according to any one of 23. The deterioration information generation unit according to any one of claims 28 to 30,
A charge control system comprising: a charge control unit that controls charging so that a deformation amount of the secondary battery detected by the detection sensor does not exceed a thickness change amount generated by the deterioration information generation unit.   A polymer matrix layer is directly or indirectly attached to the secondary battery, and the polymer matrix layer contains a filler that changes an external field according to deformation of the polymer matrix layer. The system according to any one of claims 17 to 31, wherein a deformation amount of the secondary battery is detected by detecting a change in the external field according to a deformation of the matrix layer. The program which makes a computer perform the method in any one of Claims 1-16.

Images