JP2013055842A - Cell system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a cell system capable of operating stably for a long period despite a possible certain degree of variation in the capacities of cells that constitute the cell system.SOLUTION: The present invention is applicable to a cell system comprising a plurality of cells connected in series. The number of cells in the plurality of cells is a number M pieces that are minimally required to operate the cell system plus a number mpiece(s) as a redundant cell(s). Also, the cell system has a cell-determination part that starts operating M pieces of cells among the plurality of cells in the order of larger capacity.

Description

  The present invention relates to a battery system using a secondary battery such as a lithium ion battery.

  In the future, the lithium-ion battery market is expected to expand not only to power supplies for mobile terminals such as notebook computers and mobile phones, which are the main markets in the future, but also to power sources for electric vehicles and hybrid vehicles. Yes.

  One electric vehicle uses lithium ion batteries for thousands of mobile phones. As described above, since lithium ion batteries are used in large quantities in the case of automotive applications, it is necessary to meet strict requirements regarding safety, reliability, and durability compared to mobile terminal applications. is there.

  On the other hand, the expansion of the market of storage batteries for leveling natural energy with large output fluctuations such as solar power generation and wind power generation, and stationary storage batteries for using cheap nighttime power is also expected. As storage batteries for power storage, sodium sulfur (NAS) batteries and lead storage batteries are currently used. In this power storage application, restrictions on installation space and weight are less than those of mobile terminals and electric vehicles. Thus, in power storage applications, the energy density / cost ratio is more important than simply high energy density.

  At present, lithium ion batteries are inferior to NAS batteries and lead-acid batteries in terms of energy density / cost ratio, but if electric vehicles become more widespread in the future, it is considered that the cost of lithium ion batteries will decrease due to mass production.

  In addition, lithium ion batteries lack qualification for automobile applications due to aging, but may still be reused because they still have sufficient performance for other applications. The cost of reusable batteries is lower than that of new batteries, so the market for reusable batteries may expand.

  Furthermore, since NAS batteries are handled as dangerous goods under the Fire Service Law, there are demerits such as requiring permission to install them and the operating temperature range being high (about 300 ° C.).

  Considering such a situation, in the future, it will be possible to find market competitiveness in lithium ion batteries for power storage.

  However, in general, a battery system for power storage is large-scale, and it is necessary to stably operate a battery pack composed of a large number of battery cells (hereinafter simply referred to as “cells”) over a long period of time. In addition, the lithium ion battery has a risk that the positive electrode becomes unstable due to overcharge, generates heat, and further ruptures and ignites.

  In addition, in the battery system for power storage, in particular, when lithium ion batteries (cells) are used in series, not only the protection circuit for avoiding overcharge and overdischarge, but also the electric capacity for each cell (hereinafter simply referred to as “capacity”). A cell balance circuit is also incorporated to deal with non-uniformity. This cell balance circuit takes measures such that at each time point, a cell with too little capacity is not discharged and a cell with too much capacity is not charged, thereby maintaining a balance between cells.

  In addition, in a battery system for power storage, a booster circuit may be used to obtain a necessary voltage while balancing with a cell balance circuit.

  In battery systems for power storage, long-term stable operation is considered to be particularly important in order to reduce the introduction cost and maintenance management cost. Therefore, in addition to incorporating a cell balance circuit, further stable operation is possible. The idea to do is required.

  FIG. 1 of Non-Patent Document 1 shows the relationship between the cell capacity and the number of charge / discharge cycles. In this figure, the graph in which the capacity decreases with respect to the number of charge / discharge cycles is classified into four regions (A) to (D). (A) is the first region of the charge / discharge cycle, and the capacity rapidly decreases. (B) is a region following (A), and the rate of decrease is moderate. (C) is an area following (B), and the rate of decrease is even slower than in (B). (D) is a region following (C), but here the capacity rapidly decreases again.

  FIG. 3 of Non-Patent Document 2 shows an example of measuring the relationship between the cell capacity and the number of charge cycles. In this figure, it is determined that the triangle and the rhombus are in the areas (A) to (C) of the above classification, and the square is in the areas (A) to (D).

  FIG. 8 of Non-Patent Document 3 shows the relationship between the cell capacity and the number of charge / discharge cycles when a B-MCF negative electrode and a graphite negative electrode are used. In this figure, the B-MCF negative electrode is in the region (A) to (C) and has not yet reached the region (D) until 500 cycles. On the other hand, the graphite negative electrode appears to have shifted to the region (D) around 200 cycles.

R. Spotniz, “Simulation of capacity fade in lithium-ion batteries”, J. Power Sources 113, 72, 2003 Hiroe Nakagawa, 3 others, “Development of high-performance lithium-ion secondary battery using phosphoric acid ester-containing flame retardant electrolyte containing alkyl fluoride group”, GS Yuasa Technical Report Vol. 6, p. 7, 2009 Norio Takami, "Ultra-thin aluminum laminate exterior lithium-ion battery", Toshiba review, 56 volumes, 10 pages, 2001 Kazuhiko Takeno, Tomomi Soda, “Capacity degradation characteristics of lithium-ion batteries for mobile terminals”, NTT Docomo Technical Journal, Vol. 13, p. 62

  In order to ensure stable operation of the battery system for a long period of time, even if the battery system contains cells that have deteriorated considerably, such as those in the second half of the (C) area and the (D) area, It is important to realize operation.

  However, in the above-described cell balance circuit, it is difficult to maintain the capacity required for the battery system when a large number of cells having deteriorated as described above are included.

  Therefore, an object of the present invention is to provide a battery system that enables stable operation for a long period of time even if there is a considerable capacity variation between cells constituting the battery system.

The battery system of the present invention includes:
A battery system comprising a plurality of cells connected in series,
The number of cells of the plurality of cells is a number obtained by adding m _R cells for redundant cells to the minimum M cells required for operation of the battery system.
It has a cell determination part which operates the said M cells in an order from a large capacity | capacitance among these cells.

  According to the present invention, even if there is a considerable capacity variation between cells constituting the battery system, it is possible to obtain an effect that a long-term stable operation is possible.

It is a circuit diagram of the battery system of this embodiment. FIG. 4 is a schematic diagram showing an example of the magnitude relationship between the capacities of two cells when capacity degradation due to storage of the cells is taken into consideration. It is a flowchart explaining the process sequence of the cell determination part H shown in FIG. It is a schematic diagram which shows the capacity | capacitance change with respect to the charging / discharging cycle number of an nth cell. It is a schematic diagram which shows the capacity | capacitance change with respect to the number of charging / discharging cycles of 15 cells when the working cell is determined by the comparative example 1 of this invention. It is a schematic diagram explaining the procedure which determines an operation cell by the comparative example 2 of this invention. It is a schematic diagram which shows the capacity | capacitance change with respect to the number of charging / discharging cycles of 15 cells when the working cell is determined by the comparative example 2 of this invention. It is a schematic diagram which shows the capacity | capacitance change with respect to the number of charging / discharging cycles of 15 cells when the working cell is determined by the Example of this invention. It is a schematic diagram which shows the capacity | capacitance change with respect to the number of charging / discharging cycles of a cell when the capacity | capacitance of a cell is estimated by the comparative example of this invention. It is a schematic diagram which shows the change of the parameter set ((alpha), (beta)) with respect to the number of charging / discharging cycles when the capacity | capacitance of a cell is estimated by the Example of this invention.

(1) Configuration of the present embodiment In the present embodiment, in order to realize long-term stable operation required for the battery system, the minimum number of cells required for the operation of the battery system (that is, the battery system is a predetermined number). Redundant cells are introduced beyond the minimum number of cells required to obtain a voltage. Then, an active cell and a standby cell are determined based on an algorithm that can be regarded as the longest life operation from all the cells including the redundant cell.

  As a result, long-term stable operation can be realized even if there is a considerable capacity variation between cells constituting the battery system.

Hereinafter, as the battery system of the present embodiment, a battery system having a configuration in which M + m _R cells (secondary batteries such as lithium ion batteries, etc., denoted as “CL” in the figure) are connected in series as shown in FIG. . Here, M is the minimum number of cells required for operation of the battery system, and m _R is the number of redundant cells.

  In each cell, a switch A is connected in series and a switch B is connected in parallel.

  According to Kirchhoff's law, when switch A is closed and switch B is opened, this cell is connected in series with an adjacent cell and becomes an active cell. On the other hand, when switch B is closed and switch A is opened, this cell is bypassed and becomes a standby cell. The algorithm for determining the operating cell will be described later.

One charge recording portion G is provided for M + m _R cells, and is connected in series to the cells. The charge recording unit G measures and records the current value that has passed through the cell at each time. Here, for example, by integrating the current values that have passed between _two times [t ₁ , t ₂ ], the amount of charge that has passed between these times can be found. Therefore, the charge recording unit G also records the amount of charge that has passed through the cell during each time. The charge amount recorded here is input to the cell determination unit H.

M + m _R voltage recording units V are provided corresponding to the M + m _R cells, respectively, and are connected in parallel to the corresponding cells. The voltage recording unit V measures and records the voltage value of the corresponding cell at each time. The voltage value recorded here is input to the cell determination unit H.

The cell determination unit H uses two pieces of information of the charge amount input from the charge recording unit G and the voltage value input from the voltage recording unit V, based on an algorithm described later, from among M + m _R cells. Determine the working cell. The switches A and B operate according to the determination of the cell determination unit H.

  The switch C connects the battery system to a load during discharging, and connects the battery system to an external power source during charging.

  Next, an algorithm for determining active cells will be described.

  First, consider cells that are inappropriate for operation among the cells constituting the battery system.

In a battery system in which cells are connected in series, the charge Q of each cell is equal. If some of the cells have a very small capacitance C, this cell will reach the voltage threshold V _th earlier than the other cells, and the remaining cells will remain at a voltage lower than V _th . For this reason, sufficient electric charge as a battery system cannot be stored.

Therefore, it is important not to operate a cell having a very small capacity C. Here, a cell having a capacity _Cth or less is regarded as inappropriate for operation and is not used. Such a cell is referred to as an unusable cell.

When the battery system is operated and the charge / discharge cycle proceeds, the number of unusable cells increases. When the number of unusable cells becomes m _R +1, the number of working cells becomes less than M and the battery system cannot be operated. This time is called the system stop time.

There can be various procedures for selecting M active cells from M + m _R cells at time t. For example, there is a procedure in which M operating cells are appropriately selected first, and if an unusable cell is included therein, they are sequentially replaced with redundant cells. Alternatively, there is a procedure of randomly selecting M operating cells from M + m _R cells. Alternatively, there is a procedure in which M + m _R cells are turned into operating cells in turn. As described above, among the various procedures, the procedure that can substantially delay the system stop time is to operate M cells in order from the largest in capacity among the M + m _R cells. .

  This procedure results in equalization of the remaining capacity of each cell, resulting in a substantially longest life operation. Although it is not mathematically proved that this procedure is the longest life operation, it is understood that there is no practical problem as described in (2-2) below.

Here, a problem in executing this procedure is that a capacity relationship between M + m _R cells at time t is required. In actual operation, the time for selecting an operating cell corresponds to a discrete number of charge / discharge cycles, and hence the time t and the cycle number t are regarded as the same.

The cell determination unit H receives the charge amount Q (j) (1 ≦ j ≦ t−1) in each cycle j from the charge recording unit G to the cycle t−1, and from the voltage recording unit V in each cycle j. The voltage V _i (j) of the cell i (1 ≦ i ≦ M + m _R ) is input.

Therefore, the cell determination unit H can obtain the capacity C _i (j) of the cell i in each cycle j, and the capacity C _i (t) of the cell i in the cycle t is used as the capacity history C _i before the cycle t. Estimated from (j). This estimation method will be described in (2-3) below.

  In general, as shown in FIG. 1 of Non-Patent Document 1 described above, the capacity of the cell gradually decreases with a convex function for a while from the start of the cycle, and then decreases more rapidly with a concave function. In the present embodiment, the following equation (1) having a general form is considered as a differential equation, and this equation (1) is adapted to follow the change in the capacity of the cell.

  Here, C is a capacity. Further, as f (t), a linear expression of t

  In the end, C is expressed by the following equation.

Here, C ₀ is an initial capacity when t = 0.

  In equation (2), α and β are constants that do not depend on t. However, as described in (2-3) below, it is difficult to predict the capacity C of the entire cycle with one set of α and β. . Therefore, α and β are set as parameters to be applied within a certain cycle interval, and the measured values in the subsequent cycles are sequentially predicted using α and β. If the deviation between the predicted measurement value and the actual measurement value exceeds the allowable range, α and β are determined again from the capacity history up to cycle t, and α and β are sequentially updated. That is, by changing α and β as the cycle progresses, the cell capacity is changed (by this, f (t) is not the first assumed linear expression, but this does not cause a problem).

Note that the capacity that needs to be estimated in the present embodiment is only the capacity value in the next cycle t, and it is appropriate to estimate it from C _i (j) in the last few cycles of the cycle t.

For the cycle interval [t ₁ , t ₂ ], equation (3) for the capacity C is as follows:

Here, C ₀ is the capacity at cycle t ₁ , and considering capacity C ′ = C / C ₀ where C is normalized by this value,

  It becomes.

If there are at least two measured values (t, C ′) in the range of (t ₁ <t ≦ t ₂ ), (α, β) will have (t−t ₁ ) and (t−t ₁ ) ² as variables. Can be determined by standard multiple regression analysis.

  The remaining problem with the handling so far is that the cell is not used and it is assumed that the cell will maintain capacity while waiting.

  That is, it is assumed that the capacity of the cell does not deteriorate when the cell is stored without being used. As a result, there is a possibility that the size relationship between the cells is erroneously determined. FIG. 2 schematically shows such a situation.

  Here, consider selecting a cell having a large capacity from the two cells cell1 and cell2 shown in FIG. When capacity degradation due to storage of cell2 is not considered, in cycles 5 and 6, the capacity of cell2 exceeds cell1. However, when capacity degradation due to storage of cell2 is taken into account, there is a possibility that A occurs when the capacity of cell2 exceeds cell1 only in cycle 6, and B occurs when the capacity of cell1 exceeds cell2 in all cycles 1-6.

  There are several possible solutions to this problem. For example, it is conceivable that the capacity decrease due to storage is determined from the capacity history of the cell, and the influence of capacity deterioration due to storage is estimated from the above equation (4). Alternatively, it is conceivable to devise a standby cell to avoid a charged state in which capacity deterioration is likely to proceed, such as full charge or complete discharge. However, as will be described in (2-2) below, unless the capacity variation between the first cells becomes very large or the number of redundant cells is not very large, a certain cell will continue to wait for a long time. I understand that there is no.

  In FIG. 3, the outline of the process sequence of the cell determination part H which is the principal part of a battery system is shown.

  The function of the cell determination unit H is to determine the operating cell in each cycle t.

  In step S1, the cell determination unit H first calculates the capacity of the active cell in the cycle (t-1), and then in step S2, estimates the capacity of each cell in the cycle t. In S3, an operating cell in cycle t is determined.

  In step S1, the capacity of the operating cell in the cycle (t-1) is determined based on the charge amount in the cycle (t-1) input from the charge recording unit G and the cycle (t-1) input from the voltage recording unit V. ) To calculate from the voltage of each cell. On the other hand, the capacity of the standby cell in the cycle (t-1) is the capacity in (t-2). Here, also in (t-2), if the corresponding cell is a standby cell, the capacity in (t-3) is used (hereinafter, the same operation is repeated in the case of a standby cell).

  In step S2, for the active cell in cycle (t-1), among the capacities of cycles 1 to (t-1), several capacity values near (t-1) are used, and the above ( 5) The values of α and β are determined by performing multiple regression analysis on the equation. And the capacity | capacitance of the working cell in the cycle t is estimated from the determined value of (alpha) and (beta), and (5) Formula. For the standby cell in cycle (t-1), the capacity in (t-1) is the capacity in t.

In step S3, the capacity estimated values in cycle t of M + m _R cells are arranged in ascending order, and the cells from the top to the Mth cell are determined as operating cells in cycle t.

That is, the battery system of the present embodiment is configured to operate M cells in order from the largest in capacity among M + m _R cells.
(2) Operation of the present embodiment (2-1) Operation of determining the magnitude relationship of the capacity between cells In the battery system of the present embodiment, in order to determine the operating cell, first, between the cells constituting the battery system Need information on the size of the capacity. To obtain this information, the following operations are performed.

First, an identification number is assigned to each cell, and M + m _R cells are divided into M groups. Here, when two groups are selected, the cells are divided so that cells belonging to both groups exist. For example, when M = 10 and m _R = 5, cells with identification numbers 1 to 10 are classified into group I, and cells with identification numbers 6 to 15 are classified into group II.

Next, charge / discharge of one cycle is performed on the group I. From the charge and voltage obtained there, the cell determining unit H calculates the capacity of each cell in the group I, and determines the magnitude relationship of the capacity in the group I. Next, the same operation is performed on the group II. If the capacity change of the cells (identification numbers 6 to 10) that are common in the two cycles of charging and discharging is ignored, it is possible to determine the magnitude relationship of the initial values of the capacity of the M + m _R cells.

  Information on the long-term tendency of the capacity change of the cell can be obtained from the parameter set (α, β) of equation (2). Usually, α immediately after the start of the cycle is a negative value and represents a rapid capacity decrease. After that, a cycle in which the capacity decrease eased continues for a while. This is represented by a positive value of β. As the cycle progresses from there, α increases while β decreases. When β decreases and changes from a positive value to a negative value, a region where the capacity rapidly deteriorates is entered. Thus, by monitoring the behavior of (α, β) with the progress of the cycle, the tendency of long-term cell capacity change can be understood.

  In a large-scale battery system, since it is considered that there are a plurality of battery packs including about 100 cells, it is possible to create a distribution function of α and β from the analysis results of many cells. By comparing this distribution with the values of α and β of each cell, it is possible to position the life of each cell in the cells constituting the battery system.

  Assuming that lithium-ion batteries used for automobiles and the like are reused, such lithium-ion batteries have already been used in units of several years, and it is considered that there is a cell that has considerably deteriorated. Therefore, there is a large variation in capacity between cells, and when reassembling an assembled battery from these cells, it is necessary to devise a technique that satisfies the set performance requirements even if the cells have deteriorated. This embodiment is suitable for such cell reuse.

Moreover, this embodiment is suitable for a battery system that requires stable operation over a long period of time.
(2-2) Operation for Determining Operating Cell As shown in FIG. 3 of Non-Patent Document 4, for example, the capacity of the cell decreases with respect to the number of charge / discharge cycles. This figure is an example of measuring the relationship between the capacity and the number of charge / discharge cycles for 20 cells.

From FIG. 3 of Non-Patent Document 4, in each cycle, the capacity C _{max of the} cell maintaining the capacity with the least capacity deterioration and the capacity C _min of the cell with the most capacity deterioration are read. . Since it is difficult to read the capacity change of the remaining cells from this figure, a uniform random number is generated between C _max and C _min in cycle t ₁ , and this uniform random number is converted to t _{1 of the} nth cell. _Let C _n (t ₁ ) at. That is, r _n a uniform random number interval [0, 1],

  It becomes.

Assuming that each cycle of t> t ₁ preserves the relative position of the capacity distribution at t ₁ ,

  Is established.

In the optimal procedure described in the following (2-2-3), it is considered that this assumption does not largely depend on the behavior of the cell capacity between Q _max and Q _min , and therefore there is no significant problem with this assumption. FIG. 4 schematically shows such a cell model setting.
(2-2-1) Comparative Example 1
For the model shown in FIG. 4, as a simple procedure for determining the operating cells, M operating cells are first selected, and those cells that become unavailable during system operation are sequentially selected as redundant cells. Consider the procedure to replace with. Here, M = 10, m _R = 5, t ₁ = 500, and C _th = 0.64. The result in this case is shown in FIG. Here, starting operation of the battery system at t ₁ = 500 corresponds to collecting cells used up to 500 cycles and reusing those cells to construct a battery system. Further, the lower arrow indicates a cycle in which the cell becomes unusable, and the upper arrow indicates a cycle in which use of the redundant cell is started.

In FIG. 5, the capacity of the M operating cells decreases with the number of charge / discharge cycles, and an unusable cell that becomes less than _Cth in 171 cycles appears. This unavailable cell is first replaced with a redundant cell. Since storage degradation is not considered here, the redundant cell keeps the capacity at the start of reuse until it is replaced. If there is no redundant cell, the system stop time is 171 cycles when the first cell becomes unusable, but here there are five redundant cells, so the system stops until 260 cycles when the sixth cell becomes unusable. Time is extended.

However, there is a large variation in the remaining capacity of the cells at the system stop time, and particularly, the replaced cells still have a large capacity.
(2-2-2) Comparative Example 2
Therefore, it is expected that the procedure of using all the cells as much as possible by the rotation can delay the system stop time as compared with the selection procedure of the active cell such as unusable-replacement as in the first comparative example.

Therefore, for the purpose of using the cells as evenly as possible, the following procedures (1) to (3) are considered (see FIG. 6).
(1) (M + m _R ) / m _R makes cells wait in turn every cycle (2) If the number of active cells is less than M in the above, select from (available) standby cells at random (3) Available If the number of cells becomes less than M, the system is stopped. FIG. 7 shows that these 15 cells in the case where M = 10, m _R = 5, t ₁ = 500, and C _th = 0.64 are set in this procedure. Indicates capacity change. (6) uniform random number r _n of formula, since the same as those used in Comparative Example 1, changes in the system stop time is dependent only on the difference of the steps. The lower arrow indicates a cycle in which the cell is disabled.

In FIG. 7, the first cell becomes unusable in 234 cycles. This is 1.37 times the 171 cycles of Comparative Example 1. On the other hand, since it is 287 cycles that the sixth cell becomes unusable, it is 1.10 times that of Comparative Example 1. In addition, there is a large variation in the capacity between cells at the system stop time, and there are some cells that have left a large capacity, so it can be seen that there is still a problem with load distribution in this procedure.
(2-2-3) Examples From these two comparative examples, it can be seen that it is important to eliminate cells having a large capacity in order to delay the system stop time. Therefore, it is considered appropriate to operate with priority from a cell having a large remaining capacity. Such a procedure is shown below.
(1) Sort (M + m _R ) cells for each cycle and use M cells with large capacity. (2) Stop the system when the number of operable cells is less than M. The capacity changes of these 15 cells when M = 10, m _R = 5, t ₁ = 500, and C _th = 0.64 are shown. Again, the uniform random number r _n of equation (6), since using the same as Comparative Example 1, changes in the system stop time is dependent only on the difference of the steps. Note that the inset in FIG. 8 is an enlarged view of a portion immediately after the start of system operation.

In FIG. 8, the capacity variation between cells rapidly decreases as the cycle progresses, and all the cells are disabled at 353 cycles almost simultaneously. Therefore, this procedure results in equalizing the remaining capacity of the cells. Since there is no surplus cell that leaves a capacity of C _th or more at the system stop time, this procedure can be regarded as a substantially optimal procedure that can operate the system for the longest time in the current model setting.
(2-3) Operation for Predicting Cell Capacity (2-3-1) Comparative Example In FIG. 9, white diamonds indicate the capacity measurement (graphite negative electrode) of the aluminum laminate exterior lithium ion battery of FIG. ) Is the measured value read from. Originally, α and β in the equation (2) do not depend on t. Therefore, it is considered to fit the entire measured value with one parameter set (α, β).

(Α, β) was calculated by performing multiple regression analysis using the equation (5) on the 20 measured values read. Here, t ₁ = 0 and t ₂ = 300 in the equation (5). The capacity reproduction value according to the equation (5) is indicated by a dotted line in FIG. This reproduced value does not deviate greatly from the white diamond, but it cannot be said that it sufficiently reflects the change in curvature observed in the measured value. Thus, it is difficult to predict the entire cycle with a single parameter set (α, β).
(2-3-2) Example Therefore, (α, β) is a parameter that is applicable in a certain cycle section, and the measured values in subsequent cycles are sequentially predicted using α and β. If the deviation between the predicted measurement value and the actual measurement value exceeds the allowable range, α and β are determined again from the capacity history up to cycle t, and α and β are sequentially updated. That is, by changing (α, β) with the progress of the cycle, the change in cell capacity can be followed (by this, f (t) deviates from the initial assumption of a linear expression, but this does not cause a problem. ).

  First, (α, β) is determined by multiple regression analysis from eight capacity measurement values in the cycle interval [0, 140], and the next measurement value (cycle t = 160) is predicted. At this time, if the difference between the predicted value and the actual capacity is within the allowable range (in this case, 1%), the next measured value (cycle t = 180) is also predicted with this (α, β). In this example, the predicted value and the measured value agree within 1% at t = 160, but they deviate by 1% or more at t = 180.

  Therefore, (α, β) is determined again by multiple regression analysis from the eight measured capacity values in the cycle interval [40, 180]. As a result, the capacity can be predicted with accuracy within 1% until the next cycle (t = 200 and t = 220). FIG. 10 shows changes in (α, β) with respect to the number of charge / discharge cycles at this time. In FIG. 10, the flat portion indicates that the parameter set (α, β) is not updated. Such operations are sequentially repeated to update (α, β). In this example, there is a case where the capacity cannot be predicted with an accuracy within 1% in successive cycles. However, this is not a problem because it is only necessary to predict the capacity of the next cycle.

As described above, in the present embodiment, M + m _R cells obtained by adding m _R cells corresponding to redundant cells to M cells necessary for the operation of the battery system are connected in series, and the largest capacity is sequentially installed. Activate M cells.

  As a result, the system can be operated for the longest time substantially.

  Furthermore, in the present embodiment, the cell capacity in the next cycle is predicted by a highly reliable calculation method using the capacity history in the previous charge / discharge cycle and multiple regression analysis, and the predicted capacity in descending order. Activate M cells.

  As a result, stable operation for a longer period of time is possible for large-scale battery systems such as stationary applications. For this reason, it is expected to contribute to reduction of introduction cost and maintenance management cost and resource saving.

  Although the present invention has been described with reference to the embodiments, the present invention is not limited to the above embodiments. Various changes that can be understood by those skilled in the art can be made to the configuration and details of the present invention within the scope of the present invention.

  For example, in the above embodiment, the capacity of the cell is estimated based on the capacity history in the previous charge / discharge cycle, and the M cells are operated in order from the largest estimated capacity. It is good also as operating M cell in an order from the thing with the largest capacity | capacitance in the last charging / discharging cycle.

  INDUSTRIAL APPLICABILITY The present invention can be used for construction of a battery system using a secondary battery that requires a long-term stable operation such as stationary use.

CL cell A, B, C switch G charge recording unit V voltage recording unit H cell determination unit

Claims (3)

A battery system comprising a plurality of cells connected in series,
The number of cells of the plurality of cells is a number obtained by adding m _R cells for redundant cells to the minimum M cells required for operation of the battery system.
A battery system comprising: a cell determining unit that operates M cells in order from the one having the largest capacity among the plurality of cells. The battery system according to claim 1,
The cell determination unit
Estimating the capacity of each of the plurality of cells based on the history of the capacity of each of the plurality of cells in a previous charge / discharge cycle, and operating M cells in order from the estimated capacity Battery system characterized. The battery system according to claim 1,
The cell determination unit
A battery system characterized in that M cells are operated in descending order of capacity in the most recent charge / discharge cycle among the plurality of cells.