JP7697200B2 - Estimation device, estimation method, and computer program - Google Patents

Description

The present invention relates to an estimation device, an estimation method, and a computer program for estimating the state of a storage element.

Energy storage elements are used to store electrical energy and supply it as a power source when needed. Energy storage elements are used in portable devices, power supply units, transportation equipment including automobiles and railways, and industrial equipment including aviation, space, and construction equipment.

In order to realize a high capacity of an electric storage element (hereinafter, also referred to as a battery), a high capacity negative electrode is required. In the lithium secondary battery of Patent Document 1, it is disclosed that Li (lithium) metal is contained as a negative electrode active material. By using Li metal in the negative electrode, the discharge capacity of the negative electrode is significantly improved from 372 mAh/g in the case of using graphite to 3860 mAh/g. When a negative electrode containing Li metal is used, the potential of the negative electrode is always 0 V vs. Li/Li ⁺ , which is lower than the potential in the case of using graphite, and the battery voltage can be increased. Accordingly, the positive electrode can also exhibit a capacity at a lower potential, and the energy density of the battery is dramatically improved.

However, batteries gradually deteriorate as they are repeatedly charged and discharged. Therefore, when deciding whether or not to use a battery and how to use it, it is important to easily and accurately estimate the battery's SOH (State of Health: capacity maintenance rate, etc.) and whether or not there are any abnormalities.

On the other hand, in recent years, the electrification of the aircraft industry has progressed rapidly against the backdrop of environmental issues. Batteries for aircraft applications are required to have not only high energy density, but also high safety and reliability. HAPS (High-altitude platform station) has been proposed as a new communication system. Aircraft used for such purposes do not return to the ground for a long period of time once they take off. Therefore, if an abnormality occurs in the battery, it is necessary to deal with it appropriately. However, it is difficult to deal with the abnormality after it occurs. In addition, batteries containing Li metal as the negative electrode active material have high energy density, and are therefore promising as one of the batteries for such aircraft.

JP 2001-243957 A

Batteries containing Li metal as the negative electrode active material have a high energy density as described above, but Li may precipitate in a dendrite shape during charging, breaking through the separator and causing a short circuit. In other words, this may result in an "abnormal battery state" that is problematic in terms of safety and life characteristics. Unlike conventional lithium-ion secondary batteries that use graphite as the negative electrode active material, it is difficult to determine battery abnormalities based on current, voltage, and temperature. It is necessary to detect abnormalities at an early stage in order to take measures such as remotely controlling the charging and discharging of the battery.
An object of the present invention is to provide an estimation device, an estimation method, and a computer program for estimating a state of a storage element.

The estimation device according to one aspect of the present invention includes a first derivation unit that derives estimated information including changes in shape and/or pressure of an energy storage element having a negative electrode on which Li is deposited during charging and which releases the Li into an electrolyte during discharging and an exterior body, and an estimation unit that estimates the state of the energy storage element based on the derived estimated information.

The estimation method according to one aspect of the present invention derives estimated information including changes in the shape and/or pressure of an energy storage element having an exterior body and a negative electrode that deposits Li during charging and releases the Li into an electrolyte during discharging, and estimates the state of the energy storage element based on the derived estimated information.

A computer program according to one aspect of the present invention derives estimated information including changes in shape and/or pressure of an energy storage element having an exterior body and a negative electrode that deposits Li during charging and releases the Li into an electrolyte during discharging, and causes a computer to execute a process of estimating the state of the energy storage element based on the derived estimated information.

According to the above aspect, the state of the storage element can be estimated.

13 is a graph showing the relationship between the cumulative amount of current flowing through an energy storage element and the rate of increase in thickness. FIG. 2 is an explanatory diagram for explaining the process of increasing the thickness of a Li metal negative electrode. FIG. 13 is an image diagram showing the relationship between time and the thickness of an energy storage element. 1 is a block diagram showing a configuration of a charge/discharge system and a server according to a first embodiment. FIG. FIG. 2 is an explanatory diagram illustrating an example of a record layout of a history DB. FIG. 11 is an explanatory diagram illustrating an example of a rule-based model for estimating an abnormality degree. FIG. 2 is a cross-sectional view of a battery module. FIG. 10 is a flowchart showing the procedure of a process for estimating the degree of abnormality of a cell by a control unit. 4 is a flowchart showing a procedure of a process for controlling current flow to a cell by a control unit. FIG. 11 is a schematic circuit diagram showing the battery module when power supply to the second cell is stopped and charging is performed. 10 is a flowchart showing a procedure of a cell discharge control process performed by a control unit. FIG. 2 is a perspective view of the exterior of the HAPS aircraft. FIG. 2 is a block diagram showing the configuration of a HAPS aircraft. FIG. 2 is an explanatory diagram illustrating an example of a record layout of a history DB. FIG. 11 is a block diagram showing the configuration of a charge/discharge system and a server according to a third embodiment. FIG. 13 is a schematic diagram showing the configuration of a learning model in embodiment 3. 13 is a flowchart showing the steps of a learning model generation process performed by a control unit. 10 is a flowchart showing the procedure of a process for estimating the degree of abnormality of a cell by a control unit.

(Overview of the embodiment)
In this specification, the term "electricity storage element" refers to at least a battery, a battery cell (or a cell), a battery module, a battery pack, etc. The electricity storage element may be an electric double layer capacitor, a capacitor, etc.
In the case of a carbon anode containing graphite or the like as an active material, the active material is layered, and the expansion of the active material due to the absorption of Li (lithium) during charging and the contraction of the active material due to the release of Li during discharging occur as expected. As the number of cycles increases, a new interface is generated due to the expansion and contraction of the active material, and a SEI (Solid Electrolyte Interphase) coating is generated at this interface, increasing the thickness of the anode and the thickness of the storage element (cell), but the thickness increase rate at the end of life is a maximum of about 10%.

1 is a graph showing the relationship between the cumulative amount of current and the thickness increase rate of a storage element. In FIG. 1, a is a storage element having a negative electrode containing Li metal as an active material (capacity per unit area of positive and negative electrodes: 6 mAh/cm ² ), b is a storage element having a negative electrode containing an active material of 50% graphite and 50% Si (silicon) (capacity per unit area of positive and negative electrodes: 4.7 mAh/cm ² ), c is a storage element having a negative electrode containing an active material of 20% graphite and 80% SiO (capacity per unit area of positive and negative electrodes: 7.5 mAh/cm ² ), and d is a storage element having a negative electrode containing an active material of 40% graphite and 60% SiO (capacity per unit area of positive and negative electrodes: 7.4 mAh/cm ² ). Here, O is oxygen.
In FIG. 1, the horizontal axis represents the accumulated amount of electricity (mAh), and the vertical axis represents the thickness increase rate (%).
1, it can be seen that the energy storage element having the Li metal negative electrode (a) has the largest rate of thickness increase over time. The thickness of the energy storage element is due to the increase in the thickness of the negative electrode.

FIG. 2 is an explanatory diagram for explaining the process of increasing the thickness of the Li metal negative electrode.
The negative electrode plate is formed by forming an active material layer 62 made of Li metal on a substrate foil 61, which is a plate-shaped metal foil made of a metal such as copper or a copper alloy.

During charging, Li is precipitated on the active material layer 62 to form a precipitated Li layer 63. A coating 64 made of decomposition products of components of the electrolyte, called an SEI coating, is formed on the surface of the precipitated Li layer 63 ( FIG. 2A ).
During discharge, Li from the precipitated Li layer 63 is released into the electrolyte, but not all of it dissolves, and some Li 63a remains on the active material layer 62, making the surface of the active material layer 62 uneven (FIG. 2B).

When Li precipitates on the active material layer 62 during charging, it is formed on the non-uniform surface of the active material layer 62, resulting in an uneven precipitated Li layer 63. A coating 64 is formed on the uneven precipitated Li layer 63 (FIG. 2C).
When charging and discharging are repeated, the degree of unevenness increases, and a precipitated Li layer 63 and a coating 64 are formed (FIG. 2D).

Li is surrounded by the coating 64, and this Li becomes dead Li 63b that does not participate in the battery reaction (Figure 2E). As shown in Figure 2E, the coating 64 containing the dead Li 63b has a gap between it and the surface of the active material layer 62 and the other coatings 64. During discharge, a portion of the coating 64 containing the dead Li is released into the electrolyte while containing the dead Li 63b. During discharge, a portion of the coating 64 containing the dead Li remains on the surface of the active material layer 62 with its root portion attached to the active material layer 62. A porous coating 64 is formed on the surface of the active material layer 62.

During charging, Li ions in the electrolyte enter through gaps in the coating 64, which have low resistance, and are precipitated on the surface of the active material layer 62. Li metal is precipitated non-uniformly on the active material layer 62, forming dendrites 63c.
The thickness of the deposit of porous coating 64, which contains dendrites 63c and dead Li 63b, increases (FIG. 2F).

As described above, in the case of a carbon negative electrode storage element, the thickness of the storage element increases by a maximum of 10%, but in the case of a Li metal negative electrode, the increase in the thickness of the porous coating 64 deposit causes the dendrites 63c to grow, which further accelerates the increase in the thickness of the porous coating 64 deposit. In this way, with the sudden increase in the thickness of the negative electrode, the dendrites 63c grow rapidly and break through the separator, causing a short circuit. Since the thickness increases rapidly and a short circuit occurs, unlike conventional lithium-ion secondary batteries that use graphite as the negative electrode active material, it is not possible to detect an abnormality from current, voltage, and temperature information and predict a short circuit.

The present inventors have found that by monitoring changes in the shape, i.e., changes in thickness, of an electricity storage element, it is possible to estimate the state of the electricity storage element and detect (determine) an abnormality, i.e., a sign of a short circuit.
FIG. 3 is an image diagram showing the relationship between time and the thickness of the energy storage element.
1a, etc., if there is a deviation from the first relationship between the amount of electricity and the rate of thickness increase, i.e., if the rate of thickness increase with respect to time is large, then it is an abnormal state, and there is a very high possibility of a short circuit occurring due to the growth of Li dendrites. By monitoring the thickness of the energy storage element, the state of the energy storage element can be estimated and a sign of an abnormality can be detected.

The estimation device according to the embodiment includes a first derivation unit that derives estimated information including changes in the shape and/or pressure of an energy storage element having a negative electrode on which Li is deposited during charging and which releases the Li into an electrolyte during discharging and an exterior body, and an estimation unit that estimates the state of the energy storage element based on the derived estimated information.

Here, examples of the negative electrode include metal negative electrodes containing metals such as Li metal and Na (sodium).
Here, the change in shape and/or pressure refers to the amount of displacement related to the change in shape of the power generating element, such as the thickness, the amount of displacement of the exterior body that houses the power generating element, the change in the force (reaction force) of the exterior body spreading outward, the change in pressure applied to at least one surface of the energy storage element, or the change in pressure applied to at least a part of the exterior body, or a combination of these. The power generating element is formed by laminating a positive electrode plate, a separator, and a negative electrode plate, or by winding a positive electrode plate and a negative electrode plate in a flat shape with a separator interposed therebetween. The amount of displacement may be the amount of displacement in the thickness direction (direction intersecting the long side) of the power generating element or the exterior body, or may be the amount of displacement of a tapered portion of the laminate film exterior body, for example, where a terminal protrudes.

The change in shape and/or pressure of the energy storage element can be obtained by a thickness sensor, a pressure sensor, imaging by a camera, etc.
Examples of the thickness sensor include an X-ray CT (computed tomography) device, a laser displacement sensor, and a strain sensor (strain gauge).
The pressure sensor may measure a compressive force (restraining force) of a pair of end plates that sandwich the stacked energy storage elements in a state of compressing the stacked energy storage elements in the stacking direction. When the internal pressure of the energy storage elements increases, the compressive force between the end plates increases. The pressure sensor may measure a reaction force of each energy storage element or a pressure applied to at least one surface of the energy storage element.
Regardless of whether the storage element is in a completely constrained state, a free state, or an intermediate state, it is possible to acquire the shape change and detect signs of abnormality in the storage element.

The exterior body may include a laminate film or may be made of a metal can. A laminate film is a film in which a resin film layer is laminated and bonded (laminated) to both sides of a metal layer, and aluminum, nickel, etc. can be used for the metal layer.

According to the above configuration, by deriving estimated information including changes in the shape and/or pressure of the storage element, it is possible to estimate changes in the shape and/or pressure of the negative electrode, and based on the estimated information, it is possible to estimate the state (degree of abnormality) of the storage element. Based on the state of the storage element, it is possible to accurately detect signs of abnormality.

In the above-mentioned estimation device, the estimated information may be information on changes in the shape and/or pressure of the energy storage element caused by deposits on the negative electrode, including an SEI coating containing dead Li that is not involved in the reaction and Li dendrites.

As mentioned above, the thickness of the deposits of the porous SEI coating, which contain dendrites and dead Li, increases rapidly, so information based on these deposits can be used to accurately estimate the state of the energy storage element.

In the above-mentioned estimation device, the negative electrode may have an active material containing Li metal.

In the case of a negative electrode having an active material containing Li metal, the amount of change in shape is large and the change can occur suddenly. By deriving the change in shape and/or pressure of the energy storage element, the amount of change in the shape of the negative electrode can be grasped, and the state of the energy storage element can be accurately estimated.

In the above-mentioned estimation device, the estimation unit may input the estimation information derived by the first derivation unit to a learning model that outputs information related to the state of the storage element when the estimation information is input, and estimate the state of the storage element.

The above configuration allows the state of the storage element to be estimated easily and accurately.

The above-mentioned estimation device may include a determination unit that determines whether there are signs of an abnormality based on the state of the storage element estimated by the estimation unit.

The above configuration allows for accurate detection of abnormalities such as short circuits based on the state of the storage element.

The estimation device according to the embodiment includes a first derivation unit that derives estimated information including changes in the shape and/or pressure of an energy storage element having an exterior body and a negative electrode that deposits Li during charging and releases the Li into an electrolyte during discharging, and a determination unit that inputs the estimated information derived by the first derivation unit into a learning model that outputs information related to signs of abnormality in the energy storage element when the estimated information is input, and determines whether or not there are signs of abnormality in the energy storage element.

The above configuration makes it possible to easily and accurately determine whether or not there are signs of a problem with the storage element.

The above-mentioned estimation device may include a current control unit that stops current flow at least during charging to a storage element identified based on the state estimated by the estimation unit or the sign of abnormality determined by the determination unit.

According to the above configuration, when a sign of an abnormality such as a short circuit is detected in a storage element, the occurrence of the abnormality can be prevented by not passing electricity through the storage element at least during charging.

The above-mentioned estimation device may include a discharge control unit that discharges the storage element to which the current control unit has stopped current flow.

With the above configuration, if it becomes difficult to operate using only the storage element in use, the operation can be performed by restoring and discharging the storage element that is not currently energized.

In the above-mentioned estimation device, the current control unit may include a second derivation unit that derives the resistance of the multiple storage elements to which current has been stopped, and the discharge control unit may discharge the storage elements selected based on the resistance derived by the second derivation unit.

It is presumed that the rapid increase in thickness of the negative electrode, which is the main cause of change, remains unchanged among energy storage elements in which current flow has been stopped based on changes in the shape and/or pressure of the energy storage element.
The resistance of the storage element is not only the amount of dendrites but also the sum of the resistances of the positive electrode, negative electrode, electrolyte, current collector, etc., and varies from one storage element to another. When storage elements are connected in series, it is considered that the greater the resistance of the storage element, the greater the amount of heat generated and the higher the temperature of the storage element. Dendrites formed when the temperature of the storage element is high tend to be less likely to cause a short circuit. When the temperature of the storage element is high, the form of deposition of Li metal changes, that is, the dendrites become granular rather than resinous, and it is considered that they are less likely to grow tall. Among several possible methods for selecting a storage element, for example, by selecting a storage element with a high resistance from among those with no current flowing, the possibility of a short circuit occurring during use is reduced, and operation can be performed safely.

The estimation method according to the embodiment derives estimated information including changes in the shape and/or pressure of an energy storage element having an exterior body and a negative electrode in which Li precipitates during charging and releases the Li into an electrolyte during discharging, and estimates the state of the energy storage element based on the derived estimated information.

The above configuration allows the state of the storage element to be estimated with high accuracy based on the estimation information.

The computer program according to the embodiment derives estimated information including changes in the shape and/or pressure of an energy storage element having a negative electrode that deposits Li during charging and releases the Li into an electrolyte during discharging, and an exterior body, and causes a computer to execute a process of estimating the state of the energy storage element based on the derived estimated information.

The above configuration allows the state of the storage element to be detected with high accuracy based on the estimated information.

(Embodiment 1)
4 is a block diagram showing the configuration of the charge/discharge system 30 and the server 40 according to the embodiment 1. The charge/discharge system 30 is provided in, for example, transportation equipment including automobiles and railways, industrial equipment including aviation, space, and construction equipment, and the like, and examples of the charge/discharge system include those in which the depth of one discharge of the battery module 1 is large.

The charging/discharging system 30 includes a battery module 1, a control device 2, a voltage sensor 6, a current sensor 7, a temperature sensor 8, a thickness sensor 9, a pressure sensor 10, and a camera 20.

The battery module 1 has a plurality of lithium secondary batteries (hereinafter referred to as cells) 12 connected in series as power storage elements.
The control device 2 controls the entire charge/discharge system 30. The control device 2 includes a control unit 21, a storage unit 22, an input unit 23, and a communication unit 24.
The server 40 includes a control unit 41 and a communication unit 42 .
The control unit 21 of the control device 2 is connected to the control unit 41 via the communication unit 24, the network NW, and the communication unit 42.
The load 3 is connected to the battery module 1 via terminals 31 and 32. When charging, a charger is connected to the battery module 1.

The control units 21 and 41 are configured, for example, with a CPU (Central Processing Unit), a ROM (Read Only Memory), and a RAM (Random Access Memory), and control the operation of the control device 2 and the server 40, respectively.

The memory unit 22 stores various programs and data. The memory unit 22 stores an abnormality degree estimation program 221 (hereinafter referred to as program 221), a current flow control program 222 (hereinafter referred to as program 222), and a discharge control program 223 (hereinafter referred to as program 223). The programs 221, 222, and 223 are provided in a state stored in computer-readable recording media 25, 26, and 27, such as a CD-ROM, DVD-ROM, or USB (universal serial bus) memory, and are stored in the memory unit 22 by installing them in the control device 2. The programs 221, 222, and 223 may also be obtained from an external computer (not shown) connected to a communication network and stored in the memory unit 22.

The storage unit 22 also stores a history DB (database) 224 .
FIG. 5 is an explanatory diagram showing an example of a record layout of the history DB 224. As shown in FIG.
The history DB 224 stores, for each cell 12 and for each estimated time point in charging, the current, voltage, temperature, SOC (State of Charge), pressure, thickness, image data, the amount of displacement obtained by processing the image data, and the degree of abnormality. Fig. 5 shows the history of the cell 12 with ID No. 1. The history DB 224 stores a No. column, a current column, a voltage column, a temperature column, an SOC column, a pressure column, a thickness column, an image data column, an amount of displacement column, and an abnormality degree column. The history DB 224 may further store a time (date and time) column.

The No. column stores the number (number of times) of the estimation process of the degree of abnormality. Since a short circuit occurs due to the growth of dendrites during charging, the estimation process may be performed at a predetermined interval during charging. The current column stores the history of the current acquired from the current sensor 7 at the time of estimation. The voltage column stores the voltage acquired from the voltage sensor 6 at the time of estimation. The temperature column stores the temperature acquired from the temperature sensor 8 at the time of estimation. The SOC column stores the SOC at the time of estimation. The SOC is derived by a current integration method or the like. The pressure column stores the pressure acquired from the pressure sensor 10 at the time of estimation. The thickness column stores the thickness acquired from the thickness sensor 9 at the time of estimation. The image data column stores the image data acquired by the camera 20 at the time of estimation. The displacement amount column stores the amount of displacement from the initial state obtained by processing the image data by an image processing program (not shown) stored in the memory unit 22. The abnormality degree column stores the degree of abnormality derived based on a rule base described later. The degree of abnormality is expressed in four stages from 0 to 3, with "0" being a normal state with no signs of abnormality, and the thickness of the cell 12 increasing as the value increases. When it reaches "3," measures must be taken to prevent the occurrence of an abnormal state. The degree of abnormality is estimated based on thresholds for pressure, thickness, displacement, etc. The thresholds are set to be smaller than the pressure, thickness, and displacement when there is a high probability that a short circuit will occur due to dendrites, and in order to reliably prevent a short circuit. The degree of abnormality is estimated based on the pressure, thickness, and displacement derived by actual measurement at the time of estimation, using a rule base to be described later. The degree of abnormality is not limited to four stages. Instead of the abnormality degree column, the history DB 224 may store a sign presence/absence column that stores the presence or absence of a sign, or may store a sign presence/absence column in addition to the abnormality degree column.

The relationship DB 225 stores the rule-based model.
FIG. 6 is an explanatory diagram showing an example of a rule-based model for estimating the degree of abnormality. In FIG. 6A, for convenience, four cases No. 1 to No. 4 are described. In the case of No. 1, the pressure is less than the threshold, the thickness is less than the threshold, the amount of displacement is less than the threshold, and the degree of abnormality is "0". Similarly, in the case of No. 2, the degree of abnormality is "1", in the case of No. 3, the degree of abnormality is "2", and in the case of No. 4, the degree of abnormality is "3". Here, a case in which four levels of abnormality degrees are derived using pressure, thickness, and amount of displacement will be described, but is not limited to this case. Any one of pressure, thickness, and amount of displacement may be used, or current, voltage, temperature, SOC, time, etc. may be used. Furthermore, a value calculated from these amounts (for example, the rate of change of thickness over time, etc.) may be used. The amount of displacement may be, for example, the amount of deformation of one cell 12 measured by the camera 20, or the amount of deformation of the entire battery module 1 including the cell 12. The thickness may be, for example, the thickness of one cell 12 measured by the thickness sensor 9, or the thickness of the entire battery module 1 including that cell 12.

Figure 6B shows an example of a rule-based model that uses thickness, thickness time rate of change, pressure, and pressure time rate of change. In case No. 1 shown in Figure 6B, the thickness, thickness time rate of change, pressure, and pressure time rate of change are all less than the threshold, and the degree of abnormality is "0". In cases No. 2 to 5, one of the thickness, thickness time rate of change, pressure, and pressure time rate of change is equal to or greater than the threshold, and the others are less than the threshold, and the degree of abnormality is "1". In cases No. 6 to 11, two of the thickness, thickness time rate of change, pressure, and pressure time rate of change are equal to or greater than the threshold, and the others are less than the threshold, and the degree of abnormality is "2". In cases No. 12 to 15, one of the thickness, thickness time rate of change, pressure, and pressure time rate of change is less than the threshold, and the others are equal to or greater than the threshold, and the degree of abnormality is "3". In case 16, the thickness, the time rate of change of the thickness, the pressure, and the time rate of change of the pressure are all equal to or greater than the threshold value, and the degree of abnormality is "3." In this way, the degree of abnormality can be determined based on a comparison of the thickness, the time rate of change of the thickness, the pressure, and the time rate of change of the pressure with the threshold value.

The relationship DB 225 may store a first relationship between the time in a normal state and the thickness of the cell as shown in FIG. 3. In this case, the state of the cell 12 is estimated based on whether or not the second relationship between the time and the estimated change in the thickness of the cell 12 deviates from the first relationship, and the presence or absence of a warning sign is determined based on whether or not the thickness has exceeded a warning sign threshold, etc.

The communication units 24 and 42 have the function of communicating with other devices via the network NW, and can send and receive required information.

In the present embodiment, either the control device 2 or the server 40 functions as the estimation device of the present invention. Alternatively, the entire system including the control device 2 and the server 40 may be the estimation device of the present invention, with the control device 2 functioning as the first derivation unit and the server 40 functioning as the estimation unit. Note that, when the server 40 does not function as the estimation device, the charging/discharging system 30 does not need to be connected to the server 40.
The control device 2 may be a battery ECU (Electronic Control Unit).

The voltage sensor 6 is connected in parallel to the battery module 1, and outputs a detection result corresponding to the overall voltage of the battery module 1. The voltage sensor 6 is connected to a positive terminal 12b and a negative terminal 12c (described later) of each cell 12, and measures a voltage _V1 between the positive terminal 12b and the negative terminal 12c of each cell 12 to detect a voltage V which is the sum of the _V1 of each cell 12.

The current sensor 7 is connected in series to the battery module 1 and outputs a detection result corresponding to the current of the battery module 1 .
The thickness sensor 9 is composed of an X-ray CT device, a laser displacement sensor, a strain sensor, etc. The thickness sensor 9 measures the thickness of the cells 12 of the battery module 1 in the stacking direction.

The pressure sensor 10 measures the compressive force of a pair of end plates 17 (described later) that sandwich the battery module 1 while compressing it in the stacking direction. When the internal pressure of the cells 12 increases due to deterioration, the compressive force between the end plates 17 increases. The pressure sensor 10 may measure the force spreading outward of the battery module 1, or the pressure applied to at least one surface of the battery module 1. The pressure sensor 10 may also measure the internal pressure of each cell 12.
The swelling of the cells 12 arranged at the outermost position on one side of the battery module 1 may be used. The swelling (warping) of the end plates 17 may also be measured. The gap between the cells 12 may also be measured.
Alternatively, the compression force of the end plate 17 of the cell 12 may be measured and used as a substitute for the amount of swelling. Methods for measuring the compression force include pressing the end plate 17 to measure the repulsive force, removing the fixing bolts, and measuring the repulsive force.
The camera 20 captures an image of the tapered portion of the exterior body 12f (described later) of each cell 12, where the terminals protrude from the power generating element 12a.

Although FIG. 4 shows a case where one battery module 1 is provided, a plurality of battery modules 1 may be connected in series.
The battery module 1 includes a plurality of battery circuits 11. Each battery circuit 11 includes one cell 12, and is a circuit through which current flows when the battery module 1 charges and discharges. The plurality of battery circuits 11 are connected in series. When the battery module 1 charges and discharges, current flows through the plurality of battery circuits 11, and the plurality of cells 12 charge and discharge.

The battery circuit 11 includes a first electrical path 13, a cell 12 provided on the first electrical path 13, and a first switch 14 provided on the first electrical path 13 and connected in series to the cell 12. The cell 12 is connected to the adjacent battery circuit 11 via the first electrical path 13 and the first switch 14. The first switch 14 is configured using a field effect transistor (FET). The first switch 14 is provided on the first electrical path 13 so that the forward direction of its body diode is the charging direction of the battery module 1. The first switch 14 makes the first electrical path 13 conductive when on, and makes the first electrical path 13 non-conductive in the discharging direction of the battery module 1 when off. Specifically, when the first switch 14 is on, the cell 12 is connected to another battery circuit 11 through the first electrical path 13, and can be charged and discharged. When the first switch 14 is off, no current flows through the cell 12 in the discharging direction, but current flows through the body diode of the first switch 14 in the charging direction.

The battery circuit 11 further includes a second electrical circuit 15 connected to another battery circuit 11 in parallel with the cell 12 and the first switch 14, and a second switch 16 provided on the second electrical circuit 15. The second switch 16 is configured using a FET. The first switch 14 and the second switch 16 are connected in the battery circuit 11 so that they have opposite polarity. The second switch 16 is provided on the second electrical circuit 15 so that the forward direction of its body diode is the discharge direction of the battery module 1. In a certain battery circuit 11, when the first switch 14 is turned off and the second switch 16 is turned on, charging current does not flow to the cell 12, and the second electrical circuit 15 becomes a bypass and the charging current flows to the other battery circuit while bypassing the cell 12.

When the second switch 16 is on, it puts the second electrical circuit 15 into a conductive state, and when it is off, it puts the second electrical circuit 15 into a non-conductive state in the charging direction of the battery module 1. When the first switch 14 is on and the second switch 16 is off, the cell 12 is connected to the other battery circuit 11 through the first electrical circuit 13.

FIG. 7 is a cross-sectional view of the battery module 1, and FIG.
The cell 12 includes a power generating element 12a, a positive electrode terminal 12b, a negative electrode terminal 12c, and an exterior body 12f made of laminate films 12d and 12e. The exterior body 12f is formed by overlapping two laminate films 12d and 12e and welding the peripheral parts. The positive electrode terminal 12b and the negative electrode terminal 12c protruding from the power generating element 12a are pulled out to the outside through the welding part of the exterior body 12f. For example, the laminate films 12d and 12e are each formed by stacking a PE welding layer, a PET layer, an aluminum layer, a PET layer, and a PE welding layer in this order from the inside where the power generating element 12a is arranged to the outside. The stacking configuration is not limited to this case.
The multiple cells 12 are sandwiched between a pair of end plates 17, 17 in a state where they are compressed in the stacking direction.

The power generating element 12a is formed by laminating a positive electrode plate, a separator, and a negative electrode plate, and is housed in an exterior body 12f. The negative electrode plate corresponds to the negative electrode of the present invention.
The power generating element 12a may be obtained by winding a positive electrode plate and a negative electrode plate in a flat shape with a separator between them, or may be obtained by stacking a plurality of positive electrode plates and negative electrode plates with separators between them.

The positive electrode plate is a positive electrode substrate foil, which is a plate-shaped (sheet-shaped) or long strip-shaped metal foil made of metals such as aluminum, titanium, tantalum, stainless steel, etc., or alloys thereof, on which an active material layer is formed.The negative electrode plate is a negative electrode substrate foil, which is a plate-shaped (sheet-shaped) or long strip-shaped metal foil made of metals such as copper, nickel, stainless steel, nickel-plated steel, etc., or alloys thereof, on which an active material layer is formed.The separator is a microporous sheet made of synthetic resin.

The positive electrode active material can be appropriately selected from, for example, known positive electrode active materials. A material capable of absorbing and releasing Li ions is usually used as the positive electrode active material for a lithium secondary battery. Examples of the positive electrode active material include lithium transition metal composite oxides having an α-NaFeO ₂ type crystal structure and lithium transition metal oxides having a spinel type crystal structure. Here, Fe is iron. Examples of lithium transition metal composite oxides having α-NaFeO type ₂ crystal structure include Li[Li _x Ni _1-x ]O ₂ (0≦x<0.5), Li[Li _x Ni _γ Co _(1-x-γ) ]O ₂ (0≦x<0.5, 0<γ<1), Li[Li _x Co _(1-x) ]O ₂ (0≦x<0.5), Li[Li _x Ni _γ Mn _(1-x-γ) ]O ₂ (0≦x<0.5, 0<γ<1), Li[Li _x Ni _γ Mn _β Co _(1-x-γ-β) ]O ₂ (0≦x<0.5, 0<γ, 0<β, 0.5<γ+β≦1), Li[Li _x Ni _γ Co _β Al _(1-x-γ-β) ]O ₂ (0≦x<0.5, 0<γ, 0<β, 0.5<γ+β<1), etc. are included. Here, Ni is nickel, Co is cobalt, and Al is aluminum. Examples of lithium transition metal oxides having a spinel crystal structure include Li _x Mn ₂ O ₄ and Li _x Ni _γ Mn _(2-γ) O _4. The surfaces of these materials may be coated with other materials. In the positive electrode active material layer, one of these materials may be used alone, or two or more of them may be mixed and used. In the positive electrode active material layer, one of these compounds may be used alone, or two or more of them may be mixed and used. The content of the positive electrode active material in the positive electrode active material layer is not particularly limited, but the lower limit is preferably 50% by mass, more preferably 80% by mass, and even more preferably 90% by mass. The upper limit of the content is preferably 99% by mass, more preferably 98% by mass.

The positive electrode mixture forming the active material layer of the positive electrode plate contains optional components such as a conductive agent, binder, thickener, and filler as necessary. Examples of conductive agents include carbonaceous materials such as carbon black, metals, and conductive ceramics. Examples of binders include thermoplastic resins such as fluororesins (polytetrafluoroethylene (PTFE), polyvinylidene fluoride (PVDF), etc.), polyethylene, polypropylene, and polyimide. Examples of thickeners include polysaccharide polymers such as carboxymethylcellulose (CMC) and methylcellulose. Examples of fillers include polyolefins such as polypropylene and polyethylene.

The negative electrode active material used in the negative electrode active material layer preferably contains Li metal. Since the negative electrode active material contains Li metal, it has excess Li ions and can transfer Li ions from the negative electrode plate to the positive electrode plate to achieve the required discharge capacity. Li metal includes Li simple substance as well as Li alloys. Examples of Li alloys include LiAl alloys. A negative electrode plate containing Li metal can be manufactured by cutting Li metal into a predetermined shape or forming it into a predetermined shape. In this case, a stainless steel negative electrode substrate may be connected only to the end of the Li metal plate.

Furthermore, the negative electrode active material layer may contain elements such as Na, K (potassium), Ca (calcium), Fe, Mg (magnesium), Si, and N (nitrogen). The negative electrode active material may contain Li metal and Si or SiO.
The lower limit of the Li metal content in the negative electrode active material is preferably 80 mass %, more preferably 90 mass %, and even more preferably 95 mass %, and the upper limit of the content may be 100 mass %.

An electrolyte solution is injected into the exterior body 12f. The electrolyte solution includes a non-aqueous solvent, as well as a sulfur-based cyclic compound, a fluorinated cyclic carbonate, a chain carbonate, and an electrolyte salt dissolved in the non-aqueous solvent. Examples of the sulfur-based cyclic compound include compounds having a sultone structure or a cyclic sulfate structure. Examples of the fluorinated cyclic carbonate include fluoroethylene carbonate. Examples of the chain carbonate include ethyl methyl carbonate, dimethyl carbonate, and the like.

The negative electrode active material of cell 12 may contain Li metal, and the positive electrode active material may be lithium-excess type. Such a cell 12 can exhibit a high discharge capacity.

The positive electrode active material may be a three _- component system such as NCM (Ni ₊ Co+Mn-based mixed positive electrode active material) represented by _Lix ( _NiaCobMnc ) _O2 (a+b+c=1, 0<x<1.1), or may be a Li-excess type.

The positive electrode terminals 12b and negative electrode terminals 12c of adjacent cells 12 in the battery module 1 are electrically connected by a bus bar (not shown), so that the multiple cells 12 are connected in series.
The positive and negative terminals 12b, 12c of the cells 12 at both ends of the battery module 1 are provided with positive and negative leads (not shown) for extracting electric power.
The exterior body 12f of the cell 12 may be made of a metal can.

A specific example of the cell 12 will be described.
[Preparation of positive electrode plate]
The positive electrode active material used is a lithium transition metal composite oxide having an α-NaFeO ₂ type crystal structure and expressed as Li _1+α Me 1 _-α O ₂ (Me is a transition metal), where the molar ratio Li/Me of Li to Me is 1.33, and Me is composed of Ni and Mn, with the molar ratio of Ni:Mn being 1:2.

A positive electrode paste containing N-methylpyrrolidone (NMP) as a dispersion medium, the positive electrode active material, acetylene black (AB) as a conductive agent, and polyvinylidene fluoride (PVDF) as a binder in a mass ratio of 92.5:4.5:3.0 is prepared. The positive electrode paste is applied to both sides of an aluminum foil having a thickness of 15 μm as a positive electrode substrate, dried, pressed, and cut to prepare a positive electrode plate in which a positive electrode active material layer is arranged in a rectangular shape having a width of 30 mm and a length of 40 mm. The thickness of the positive electrode active material layer is about 110 μm on one side, and the positive electrode mixture is contained in an amount of 26 mg/cm ² per unit area. The positive electrode plate is used after drying under reduced pressure at 120 ° C. for 14 hours or more.

[Preparation of negative electrode plate]
Metallic lithium foil (average thickness 100 μm: pure metallic lithium containing 100 mass% metallic lithium) was laminated as a negative electrode active material layer on both sides of a copper foil having a thickness of 8 μm, which was a negative electrode substrate, and pressed. This resulted in a negative electrode plate having a negative electrode active material layer laminated on both sides of the negative electrode substrate. The negative electrode plate was rectangular with a width of 30 mm and a length of 42 mm.

[Preparation of electrolyte solution]
A solution is prepared by dissolving _LiPF6 at a concentration of 1 mol dm ^-3 in a mixed solvent in which fluoroethylene carbonate (FEC) and dimethyl carbonate (DMC) are mixed in a volume ratio of FEC:DMC = 30:70. Furthermore, 2 mass% of 1,3-propene sultone is added to the solution to obtain an electrolyte solution.

[Preparation of cell 12]
A plurality of the positive electrode plates and a plurality of the negative electrode plates are alternately stacked one by one with a polyolefin microporous film as a separator interposed therebetween to prepare a stacked power generating element 12a.

The power generating element 12a is covered with two laminate films 12d and 12e so that the open ends of the positive electrode terminal 12b and the negative electrode terminal 12c, which have been previously connected to the positive electrode plate and the negative electrode plate, respectively, are exposed to the outside, and the film is sealed except for the part that will become the liquid injection hole. After the electrolyte is injected, the liquid injection hole is hermetically sealed. In this manner, the cell 12 is produced.

The configuration of the cell 12 is not limited to the above. Lithium may be attached to the negative electrode plate.

A method for estimating the degree of abnormality of the cell 12 will be specifically described below.
FIG. 9 is a flowchart showing the procedure of the process of estimating the degree of abnormality of the cell 12 by the control unit 21.
The control unit 21 derives history information (S1). The control unit 21 stores in the history DB 224 the current acquired from the current sensor 7 at the time of estimation. The control unit 21 stores in the history DB 224 the voltage acquired from the voltage sensor 6 at the time of estimation. The control unit 21 stores in the history DB 224 the temperature acquired from the temperature sensor 8 at the time of estimation. The control unit 21 stores in the history DB 224 the SOC derived at the time of estimation. The control unit 21 stores in the history DB 224 the pressure acquired from the pressure sensor 10 at the time of estimation. The control unit 21 stores in the history DB 224 the thickness acquired from the thickness sensor 9 at the time of estimation. The control unit 21 stores in the history DB 224 the image data acquired by the camera 20 at the time of estimation. The control unit 21 stores in the history DB 224 the amount of displacement from the initial state obtained by processing the image data using an image processing program stored in the storage unit 22.

The control unit 21 reads out estimation information required for estimating the degree of abnormality from the history DB 224 (S2).
The control unit 21 reads the relationship DB 225 (S3).

The control unit 21 estimates the degree of abnormality (S4) and judges whether there is a sign (S5). When the rule base of FIG. 6 is stored in the relationship DB 225, the control unit 21 estimates the degree of abnormality based on whether the pressure, thickness, and displacement amount read from the history DB 224 are each less than a threshold value. When the degree of abnormality is "3", the control unit 21 judges that there is a sign. When the relationship DB 225 stores a first relationship between the time of a normal state and the thickness of the cell, the control unit 21 estimates the state of the cell 12 based on whether the second relationship between the time and the change in thickness deviates from the first relationship (S4), and judges whether there is a sign based on whether the thickness exceeds the threshold value of the sign, etc. (S5). The control unit 21 stores the degree of abnormality or the presence or absence of a sign in the history DB 224.
When the control unit 21 determines that there is a sign (S5: YES), it notifies the user that there is a sign (S6). The control unit 21 transmits information indicating that there is a sign of an abnormality to the server 40 or to the user's terminal device via the communication unit 24. When there is no sign (S5: NO), the control unit 21 ends the process.
The control unit 21 may estimate the presence or absence of a sign of an abnormality in the battery module 1 as a whole, and then identify the cell 12 that has the sign.

According to this embodiment, by deriving estimated information including changes in the shape and/or pressure of the cell 12, changes in the shape and/or pressure of the negative electrode are obtained, and the state of the cell 12 can be accurately estimated based on the estimated information, making it possible to accurately detect signs of abnormalities such as short circuits.

FIG. 10 is a flowchart showing the procedure of the process of controlling the energization of the cells 12 by the control unit 21.
The control unit 21 determines whether or not there is a sign of an abnormality in the target cell 12 (S11). The control unit 21 determines whether or not the abnormality degree is "3." If there is no sign (S11: NO), the control unit 21 ends the process.

If there is a warning sign (S11: YES), the control unit 21 determines whether the SOC of the cell 12 with the warning sign at the time of estimation is equal to or greater than the threshold value a (S12). If the SOC is not equal to or greater than the threshold value a (S12: NO), the control unit 21 ends the process. As described below, there are cases where a cell 12 that has been stopped from supplying current during charging is restored and discharged, and the SOC needs to be equal to or greater than the threshold value a when stopped.

When the SOC is equal to or greater than the threshold value a (S12: YES), the control unit 21 identifies the cell 12 as a cell to which power supply is to be stopped, and turns off the first switch 14 corresponding to the cell 12 (S13).
The control unit 21 turns on the second switch 16 (S14) and ends the process. The second electric circuit 15 becomes conductive, and the cell 12 with the warning sign becomes disconnected from the other battery circuits 11 and cannot be charged.

Figure 11 is a schematic circuit diagram showing the battery module 1 when the power supply to the second cell 12 is stopped and charging is performed. In Figure 11, the current flowing during charging is indicated by solid arrows. Assume that the second cell 12 from the top in Figure 11 shows signs of an abnormality. In the battery circuit 11 including this cell 12, the first switch 14 is off, and the cell 12 is not connected to the other battery circuits 11. The second switch 16 is on, and the second electrical circuit 15 is in a conductive state. In the battery circuit 11 including a normal cell 12, the first switch 14 is on, and the cell 12 is connected to the other battery circuits 11 via the first electrical circuit 13. The second switch 16 is off, and the second electrical circuit 15 is in a non-conductive state.

In a battery circuit 11 including a normal cell 12, current flows through the cell 12 and the first electric circuit 13, and the cell 12 charges. In a battery circuit 11 including a cell 12 with a warning sign, current does not flow through the cell 12 and the first electric circuit 13, but flows through the second electric circuit 15. In other words, no current flows through the cell 12 with a warning sign, and the second electric circuit 15 becomes a bypass and current flows through the second electric circuit 15. Current flows throughout the battery module 1, and normal cells 12 charge.
The affected cell 12 is not allowed to charge, thereby preventing further dendrite growth and creating a short circuit.

When discharging the cell 12 that has a warning sign, the second switch 16 is turned off and the first switch 14 is turned on. Hereinafter, a case where the cell 12 that has stopped being energized is restored and discharged will be described.
FIG. 12 is a flowchart showing the procedure of the process of controlling the discharge of the cells 12 by the control unit 21.
The control unit 21 determines whether or not to restore (S21). The control unit 21 determines whether or not it is difficult to operate using only the cells 12 in use. If the control unit 21 does not want to restore (S21: NO), the process ends.

When the control unit 21 is to be restored (S21: YES), it reads the history DB 224 and derives the resistance of the cell 12 to which current has been stopped (S22). The control unit 21 reads the initial current and voltage of the number of times the target cell 12 has been charged, which are stored in the history DB 224, and derives the resistance.

The control unit 21 selects the cell 12 to be discharged (S23). The control unit 21 selects the cell 12 with the highest resistance from among the cells 12 whose resistances have been derived. As described above, when the resistance of a cell 12 is high, the amount of heat generated is large and a short circuit is unlikely to occur. The control unit 21 may select the cell 12 with a high SOC when the power supply is stopped. When the SOC is high, sufficient discharge can be achieved.

The control unit 21 judges whether or not the load can be disconnected (S24). If the load cannot be disconnected (S24: NO), the control unit 21 ends the process.
If disconnection is possible (S24: YES), the control unit 21 disconnects the load (S25).The control unit 21 turns off the second switch 16 corresponding to the cell 12 (S26).
The control unit 21 turns on the first switch 14 (S27). The second electric circuit 15 is in a non-conductive state, the first electric circuit 13 is in a conductive state, the cell 12 is connected to the other battery circuit 11, and discharging is possible.

The control unit 21 connects the load to the battery module 1 (S28).
The control unit 21 starts discharging (S29) and ends the process.

According to this embodiment, when it becomes difficult to operate using only the cells 12 in use, the operation can be performed by restoring the cells 12 that are not energized and discharging them.

(Embodiment 2)
The second embodiment will be described using a HAPS aircraft 50 as an example of an aircraft. The aircraft may be, but is not limited to, an eVTOL (electric vertical takeoff and landing aircraft). The aircraft includes a power generation device and a storage element, and preferably does not include an internal combustion engine.
Fig. 13 is a perspective view of the exterior of the HAPS airframe 50, and Fig. 14 is a block diagram showing the configuration of the HAPS airframe 50. In Fig. 13, the same parts as in Fig. 4 are given the same reference numerals and detailed explanations will be omitted. In Fig. 14, the solar panel 54 is omitted.
The HAPS airframe 50 includes a wing section 51, a plurality of propellers 52, a plurality of legs 53, a plurality of solar panels 54, a battery module 1, a control device 2, a motor 4, a radio relay station 5, and wheels 55. The propeller 52 is connected to the motor 4. The configuration of the HAPS airframe 50 is not limited to this example. The battery module 1, the control device 2, and the radio relay station 5 are housed in the legs 53. Alternatively, they may be provided on the wing section 51.

The HAPS aircraft 50 equipped with the wireless relay station 5 can simultaneously connect to a large number of terminal devices over a wide area, and by linking the HAPS aircraft 50 with artificial satellites and ground stations, a high-speed communications infrastructure can be constructed. A stable communications environment can be maintained even during disasters.

The solar panel 54 is formed by arranging and connecting a plurality of modules, each module having a plurality of silicon-based solar cells arranged therein.
The battery module 1 is formed by connecting, for example, a plurality of cells 12 in series and/or in parallel. A plurality of battery modules 1 may be connected in series to form a bank, or the bank may be connected in parallel. The cells 12 and the battery module 1 have the same configuration as the cells 12 and the battery module 1 according to the first embodiment.

The control device 2 has a configuration similar to that of the control device 2 of the first embodiment, and further includes a motor drive unit 28. The memory unit 22 stores an abnormality degree estimation program 221, a current control program 222, a discharge control program 223, a history DB 224, and a relationship DB 225.

Figure 15 is an explanatory diagram showing an example of the record layout of the history DB 224. The history DB 224 of embodiment 2 stores a solar panel power generation amount column and a consumed energy column in addition to the columns of the history DB 224 of embodiment 1. The solar panel power generation amount column stores the amount of power generated by the solar panel 54. The consumed energy column stores the amount of energy consumed by the entire HAPS aircraft 50.

The input unit 23 receives the detection results of the current and voltage of the solar panel 54 and the battery module 1 .
The communication unit 24 has a function of communicating with other devices such as the wireless relay station 5, and transmits and receives required information.
The motor drive unit 28 controls the rotational drive of each motor 4 for each propeller 52 .

An inverter circuit and a switching unit (not shown) are connected to the load 3. The battery module 1 and the solar panel 54 are connected to the switching unit. The inverter circuit converts the input direct current into alternating current by switching between the battery module 1 and the solar panel 54 using the switching unit, and outputs the converted alternating current to the load 3.
FIG. 14 shows a state in which the battery module 1 is connected to a load 3 .

The wireless relay station 5 includes a communication unit 56. The communication unit 56 has an antenna, a duplexer, an amplifier, etc., and transmits and receives wireless signals to and from a relay station on land or sea. The wireless relay station 5 is connected to a network NW of a mobile communication network via the relay station. A server 40 and a terminal device 60 are connected to the network NW.
The motor 4 rotates and drives the propeller 52. Alternatively, the motor 4 may drive an air vehicle propulsion device or an air vehicle ascent device of a configuration other than that shown in FIG.

The HAPS aircraft 50 configured as described above leaves the ground diagonally upward, then rotates within a specified horizontal area, using lift to rise to airspace A. Airspace A may be, for example, an airspace in the stratosphere at an altitude of 11 km to 50 km. Of these, an airspace at an altitude of 20 km is preferable. After rising to airspace A, the HAPS aircraft 50 moves horizontally to position B and remains at position B.

At night, power cannot be generated by the solar panel 54, so power is supplied by the battery module 1. The battery module 1 supplies power when flying by driving the motor 4, and also supplies power to the control unit 21, input unit 23, communication unit 24, wireless relay station 5, etc., which are required when the HAPS airframe 50 glides within airspace A.
The depth of discharge increases at night because the battery module 1 supplies power until electricity is generated by the solar panel 54. For example, the battery module 1 performs charging and discharging in the range of SOC 10% to 100%.

In this embodiment, similar to the first embodiment, the degree of abnormality of the cell 12 is estimated using a rule base stored in the relationship DB 225 based on the pressure, thickness, and displacement of the cell 12. The degree of abnormality may be estimated based on any one of the pressure, thickness, and displacement, or the degree of abnormality may be estimated based on at least one of the current, voltage, temperature, solar panel power generation, and energy consumption in addition to the pressure, thickness, and displacement.

If it is determined that there is a sign of an abnormality based on the degree of abnormality in the cell 12, the first switch 14 of that cell 12 is turned off and the second switch 16 is turned on to stop the flow of electricity during charging. This prevents a short circuit in that cell 12. The HAPS aircraft 50 is required to simultaneously connect to a large number of terminal devices over a wide area for a specified period of time, such as six months, to build a high-speed communications infrastructure. According to this embodiment, the life is extended and stable discharge can be performed at night for a longer period of time, making it possible to perform stable flight operations and maintain a stable communications environment.

When the control unit 21 determines that it is difficult to continue flying using only the cells 12 in use, it selects the cell 12 with high resistance and high SOC from among the cells 12 for which current flow during charging has been stopped, turns off the second switch 16 of that cell 12, turns on the first switch 14, and restores current flow during discharging. This allows the HAPS aircraft 50 to land, avoiding a crash.

(Embodiment 3)
FIG. 16 is a block diagram showing the configuration of a charge/discharge system 30 and a server 40 according to the third embodiment.
The server 40 of the third embodiment includes a storage unit 43 .
The storage unit 43 includes a storage device such as a hard disk drive. The storage unit 43 stores various computer programs executed by the control unit 41, various data used by the computer programs, data acquired from the outside, and the like. An example of a computer program stored in the storage unit 43 is a model generation program for generating a learning model 430. The storage unit 43 may also store training data including image data of the cells 12 and label data indicating the degree of abnormality. These data may be acquired from the control device 2 via the communication unit 42.
The history DB 224 of the control device 2 in the third embodiment does not store the amount of displacement based on image data. Since the image data is input to the learning model 430 to estimate the degree of abnormality, it is not necessary to derive the amount of displacement by performing image processing.

Figure 17 is a schematic diagram showing the configuration of the learning model 430 in embodiment 3. The learning model 430 in embodiment 3 is a learning model using CNN (Convolutional Neural Networks), R-CNN (Region-based CNN), etc., and includes an input layer 431, an intermediate layer 432, and an output layer 433. The learning model 430 is trained to output the degree of anomaly in response to input of image data of cell 12. The learning model 430 is generated by, for example, the server 40, and stored in the memory unit 22 of the control device 2.

Image data is input to the input layer 431 of the learning model 430.
The intermediate layer 432 is composed of, for example, a convolution layer 432a, a pooling layer 432b, and a fully connected layer 432c. The convolution layer 432a and the pooling layer 432b may be provided alternately in a plurality of layers. The convolution layer 432a and the pooling layer 432b extract features of the image data input through the input layer 431 by calculation using the nodes of each layer. The fully connected layer 432c combines the data from which the feature parts are extracted by the convolution layer 432a and the pooling layer 432b into one node, and outputs feature variables transformed by the activation function. The feature variables are output to the output layer 433 through the fully connected layer 432c.

The output layer 433 has one or more nodes. Based on the feature variables input from the fully connected layer 432c of the intermediate layer 432, the output layer 433 converts them into probabilities using a softmax function, and outputs probabilities indicating each degree of abnormality, for example, from 0 to 3, from each node.

For example, the output layer 433 may be composed of four nodes, from the first node to the fourth node, and the probability that the degree of abnormality is 0 may be output from the first node, the probability that the degree of abnormality is 1 from the second node, the probability that the degree of abnormality is 2 from the third node, and the probability that the degree of abnormality is 3 from the fourth node.
The output layer 433 is
For example, abnormality level 0...0.04
Abnormality level 1...0.90
Abnormality level 2...0.05
Abnormality level 3...0.01
The output will be as follows.
The number of nodes constituting the output layer 433 and the calculation results assigned to each node are not limited to the above example, and can be designed as appropriate.
The output layer 433 may include one node and output the probability that there is a sign. In this case, the training data includes image data of the cells 12 and label data indicating the presence or absence of a sign.

The control unit 21 of the control device 2 can estimate the degree of abnormality by referring to the calculation results obtained from the learning model 430 and selecting, for example, the node with the highest probability.

The internal parameters (weights and biases between nodes) that determine the configuration of the learning model 430 are learned, for example, in the server 40 by using an appropriate learning algorithm.

FIG. 18 is a flowchart showing the procedure of the process of generating the learning model 430 by the control unit 41.
As a preparatory step for generating the learning model 430, the control unit 41 collects image data of the tapered portion of the cell 12 and label data indicating the degree of abnormality of the cell 12, and stores the collected data as training data in the storage unit 43 (S31). By collecting a sufficient number of image data and label data in this preparatory step, the accuracy of estimating the degree of abnormality can be improved. As the learning progresses, the estimation result by the learning model 430 and the image data used in the estimation process may be obtained, and the obtained data may be used as training data.

The control unit 41 uses the training data to generate a learning model 430 that outputs the probability of the degree of abnormality when image data is input (S32). Specifically, the control unit 41 inputs the training data to the input layer 431, and acquires the probability of the degree of abnormality from the output layer 433 after arithmetic processing in the intermediate layer 432.
The control unit 41 compares the estimation results of the degree of abnormality output from the output layer 433 with the information of the degree of abnormality labeled with the image data in the training data, i.e., the correct answer value, and optimizes the parameters used in the calculation process in the intermediate layer 432 so that the output value from the output layer 433 approaches the correct answer value. The parameters are, for example, weights (coupling coefficients), activation function coefficients, etc. The method of optimizing the parameters is not particularly limited, but for example, the control unit 41 optimizes various parameters using an error backpropagation method.
The control unit 41 performs the above-mentioned processing on the image data of each training data, and generates a learning model 430. The control unit 41 stores the generated learning model 430 in the storage unit 43, and ends the series of processing.
In addition to image data, pressure and thickness may be input for learning to the input layer 431. Furthermore, current, voltage, temperature, SOC, etc. may be input. If a charge/discharge system is provided in the HAPS aircraft 50, the amount of power generated by the solar panel and energy consumption may be input.

The control device 2 downloads the learning model 430 generated in the server 40. The control device 2 can estimate the degree of abnormality in the cell 12 by inputting the image data acquired during estimation into the learning model 430.

FIG. 19 is a flowchart showing the procedure of the process of estimating the degree of abnormality of the cell 12 by the control unit 21.
The control unit 21 derives history information (S41).
The control unit 21 reads out the estimated information from the history DB 224 (S42). Here, image data is read out as the estimated information.

The control unit 21 inputs the image data as the estimated information to the learning model 430 (S43).
The control unit 21 estimates the degree of abnormality based on the probability of the degree of abnormality output by the learning model 430 (S44). The control unit 21 obtains the degree of abnormality with the highest probability and estimates a sign.
The control unit 21 determines whether or not there is a sign (S45). If the abnormality degree is "3", the control unit 21 determines that there is a sign.
When the control unit 21 determines that there is a sign of an abnormality (S45: YES), the control unit 21 notifies the user of the presence of the sign of an abnormality (S46). The control unit 21 transmits information indicating that there is a sign of an abnormality to the server 40 or to the user's terminal device via the communication unit 24.
When the learning model 430 outputs the probability that there is a precursor, the control unit 21 determines that there is a precursor when the probability that there is a precursor output by the learning model 430 is, for example, 50% or more. When there is no precursor (S45: NO), the control unit 21 ends the process.

According to this embodiment, the state of the cell 12 can be easily and accurately estimated, and signs of abnormality in the cell 12 can be determined. In the third embodiment, an example of supervised learning using CNN is shown, but it is also possible to estimate the state of the cell 12 using unsupervised learning or regression, etc.

The above embodiments are not intended to be limiting, and the scope of the present invention is intended to include all modifications within the meaning and scope of the claims.
The power storage element is not limited to a lithium secondary battery, and may be another type of secondary battery or a capacitor.

1 Battery module (energy storage element)
12 Cell (energy storage element)
REFERENCE SIGNS LIST 12a Power generating element 12b Positive electrode terminal 12c Negative electrode terminal 12f Exterior body 14 First switch 16 Second switch 2 Control device 25, 26, 27 Recording medium 21, 41 Control unit 22, 43 Storage unit 221 Abnormality degree estimation program 222 Current flow control program 223 Discharge control program 224 History DB
23 Input section 24, 42 Communication section 28 Motor drive section 3 Load 4 Motor 5 Wireless relay station 30 Charging/discharging system 40 Server 50 HAPS airframe 51 Wing section 52 Propeller 53 Leg section 54 Solar panel 55 Wheel

Claims (11)

A first derivation unit that derives estimated information including a thickness or a time rate of change in thickness of an energy storage element having an exterior body and a negative electrode in which Li (lithium) is deposited during charging and which releases the Li into an electrolyte during discharging;
an estimation unit that estimates a state of the energy storage element, including the presence or absence of a sign of an abnormality in the energy storage element, based on the derived estimation information;
a determination unit that determines whether or not there is a sign of an abnormality in the storage element based on the state of the storage element estimated by the estimation unit ;
a current supply control unit that stops current supply at least during charging to a storage element identified based on the state estimated by the estimation unit;
a discharge control unit that discharges the storage element to which the current supply control unit has stopped supplying current;
a second derivation unit that derives resistances of the plurality of storage elements to which the current supply control unit has stopped supplying current,
The discharge control unit discharges a storage element selected based on the resistance derived by the second derivation unit . A first derivation unit that derives estimated information including a thickness or a time rate of change in thickness of an energy storage element having an exterior body and a negative electrode in which Li (lithium) is deposited during charging and which releases the Li into an electrolyte during discharging;
an estimation unit that estimates a state of the energy storage element, including the presence or absence of a sign of an abnormality in the energy storage element, based on the derived estimation information;
a determination unit that determines whether or not there is a sign of an abnormality in the storage element based on the state of the storage element estimated by the estimation unit ;
a current supply control unit that stops current supply at least during charging to a storage element identified based on the abnormality sign determined by the determination unit;
a discharge control unit that discharges the storage element to which the current supply control unit has stopped supplying current;
a second derivation unit that derives resistances of the plurality of storage elements to which the current supply control unit has stopped supplying current,
The discharge control unit discharges a storage element selected based on the resistance derived by the second derivation unit . The estimation device according to claim 1 or 2, wherein the estimated information is information on a change in thickness of the energy storage element caused by a SEI (Solid Electrolyte Interphase) coating containing dead Li not involved in the reaction and/ or deposits on the negative electrode including Li dendrites. The estimation device according to claim 1 , wherein the negative electrode has an active material containing Li metal. The estimation unit is
An estimation device as described in any one of claims 1 to 4, wherein the estimation information derived by the first derivation unit is input into a learning model that outputs information related to the state of a storage element when the estimation information is input, thereby estimating the state of the storage element. The estimation device according to any one of claims 1 to 4, wherein the estimation unit estimates the state of the storage element based on whether a second relationship between the change in thickness of the storage element in accordance with the estimation information and time deviates from a first relationship between a normal thickness of the storage element and time . a first derivation unit that derives estimated information including a thickness of an energy storage element having a negative electrode in which Li is deposited during charging and which releases the Li into an electrolyte during discharging and an exterior body;
a determination unit that inputs the estimation information derived by the first derivation unit into a learning model that outputs information related to a sign of abnormality in an energy storage element when the estimation information is input, and determines whether or not there is a sign of abnormality in the energy storage element ;
a current supply control unit that stops current supply at least during charging to a storage element identified based on the abnormality sign determined by the determination unit;
a discharge control unit that discharges the storage element to which the current supply control unit has stopped supplying current;
a second derivation unit that derives resistances of the plurality of storage elements to which the current supply control unit has stopped supplying current,
The discharge control unit discharges a storage element selected based on the resistance derived by the second derivation unit . Deriving estimated information including a change in thickness or a time rate of change in thickness of an electricity storage element having an exterior body and a negative electrode that deposits Li during charging and releases the Li into an electrolyte during discharging;
Estimating a state of the energy storage element, including the presence or absence of a sign of an abnormality in the energy storage element, based on the derived estimation information;
determining whether or not there is a sign of an abnormality in the storage element based on the estimated state of the storage element ;
Stopping at least the supply of current during charging to the storage element identified based on the estimated state;
deriving resistance of the plurality of storage elements to which current has been stopped;
The estimation method includes discharging a de-energized storage element selected based on the derived resistance . Deriving estimated information including a change in thickness or a time rate of change in thickness of an electricity storage element having an exterior body and a negative electrode that deposits Li during charging and releases the Li into an electrolyte during discharging;
Estimating a state of the energy storage element, including the presence or absence of a sign of an abnormality in the energy storage element, based on the derived estimation information;
determining whether or not there is a sign of an abnormality in the storage element based on the estimated state of the storage element ;
Stopping at least the supply of current during charging to the storage element identified based on the determined sign of abnormality;
deriving resistance of the plurality of storage elements to which current has been stopped;
The estimation method includes discharging a de-energized storage element selected based on the derived resistance . Deriving estimated information including a change in thickness or a time rate of change in thickness of an electricity storage element having an exterior body and a negative electrode that deposits Li during charging and releases the Li into an electrolyte during discharging;
Estimating a state of the energy storage element, including the presence or absence of a sign of an abnormality in the energy storage element, based on the derived estimation information;
determining whether or not there is a sign of an abnormality in the storage element based on the estimated state of the storage element ;
Stopping at least the supply of current during charging to the storage element identified based on the estimated state;
deriving resistance of the plurality of storage elements to which current has been stopped;
Discharge the storage element that has been de-energized and selected based on the derived resistance.
A computer program that causes a computer to carry out processing. Deriving estimated information including a change in thickness or a time rate of change in thickness of an electricity storage element having an exterior body and a negative electrode that deposits Li during charging and releases the Li into an electrolyte during discharging;
Estimating a state of the energy storage element, including the presence or absence of a sign of an abnormality in the energy storage element, based on the derived estimation information;
determining whether or not there is a sign of an abnormality in the storage element based on the estimated state of the storage element ;
Stopping at least the supply of current during charging to the storage element identified based on the determined sign of abnormality;
deriving resistance of the plurality of storage elements to which current has been stopped;
Discharge the storage element that has been de-energized and selected based on the derived resistance.
A computer program that causes a computer to carry out processing.

Images