JP6658289B2 - Battery recharge rate estimation system - Google Patents

Description

  The present invention relates to a system for estimating a state of charge of a secondary battery, and more particularly to a system for estimating a state of charge from a restrained load of a secondary battery.

  In recent years, hybrid vehicles using an engine and a motor as drive sources, and electric vehicles such as electric vehicles driven by a motor have been widely used. Such an electric vehicle is equipped with a chargeable / dischargeable secondary battery that supplies driving power to a driving motor or charges regenerative power. The charge and discharge power of the secondary battery is controlled by its charge rate (the ratio of the charge capacity (A × h) to the full charge capacity (A × h), hereinafter referred to as SOC). Therefore, in a power supply system using a secondary battery, it is required to accurately estimate the SOC of the secondary battery.

  As a method of estimating the SOC of the secondary battery, a volume change and a temperature of the electrode of the secondary battery are detected by a pressure sensor and a temperature sensor, and the SOC is estimated using a map defining a correlation between these and the SOC. A method has been proposed (see, for example, Patent Document 1).

JP-A-2006-12761

  By the way, a secondary battery is formed by stacking a plurality of secondary batteries with a flat resin member through which cooling air flows, sandwiching restraint plates at both ends thereof, and restraining the restraint plates integrally with restraint rods. In many cases, the battery module is mounted on an electric vehicle or the like as a laminated battery module. When the secondary battery repeats charge and discharge, the secondary battery repeats expansion and contraction. As a result, creep occurs in the resin member, the restraining rod, and the like, and the restraining load of the secondary battery decreases over time. In the method for estimating SOC described in Patent Document 1, there is a problem in that the estimation error of SOC increases over time because the reduction in the secular load due to creep of such members is not taken into account. Was.

  Therefore, an object of the present invention is to accurately estimate the SOC even when the restraint load of the secondary battery changes over time.

  The system for estimating a charging rate of a secondary battery according to the present invention includes: a secondary battery stacked via an insertion member; a constraining member for constraining the secondary battery and the insertion member in a stacking direction; A load sensor disposed between the batteries to detect a constrained load, a temperature sensor for detecting the temperature of the secondary battery, a constrained load detected by the load sensor, a temperature detected by the temperature sensor, and a constrained load. And a correction unit that estimates the charging rate of the secondary battery based on a correction coefficient that corrects the aging of the secondary battery. When charging the battery, the increment of the restraint load is calculated for each fixed charging capacity, and the restraint load increment ratio is calculated by dividing the present value of the restraint load increment by the previous value of the restraint load increment. Generate the measured curve of the constraint load increment ratio to the capacity A battery charge rate range in which the error between the reference curve of the restraint load increment ratio with respect to the charge rate prepared in advance and the actually measured curve is minimized, and the initial restraint load and the restraint load detected by the load sensor in the specified charge rate range are specified. The correction coefficient is calculated from the deviation from the above.

  The present invention can accurately estimate the SOC even when the restraint load of the secondary battery changes over time.

It is an explanatory view showing a system of a charge rate estimation system and a configuration of a battery module in an embodiment of the present invention. It is a side view of the battery cell which comprises a battery module. FIG. 3 is a sectional view of the battery cell shown in FIG. 2. 4 is a map showing a change in the electrode natural length with respect to the SOC. It is explanatory drawing which shows the physical model of one segment of a battery cell. 5 is a flowchart illustrating an operation of estimating the SOC from the electrode natural length. It is a figure showing change of restraint load F to SOC in an initial state and a creep state. FIG. 7 is a diagram showing a change in an increment (ΔF (n)) of a constraint load with respect to a charged capacity (A × h) in an initial state and a creep state. FIG. 7 is a diagram showing a change in a constraint load increment ratio (ΔF ratio) with respect to a charged capacity (A × h) in an initial state and a creep state. It is a flowchart which shows the operation | movement which acquires the measurement curve of the restraint load increment ratio ((DELTA) F ratio) with respect to a charging capacity (Axh). 11 is a flowchart showing an operation of calculating a correction coefficient α using the actually measured curve acquired in the operation of FIG. 10 and a reference curve of a constraint load increment ratio (ΔF ratio) with respect to SOC. It is explanatory drawing of the method of acquiring the increase ((DELTA) F (n)) of the restraint load with respect to a charging capacity (Axh). It is an actual measurement curve of the restraint load increment ratio (ΔF ratio) with respect to the charging capacity (A × h). Curve of constraint load increment ratio ΔF ratio (= ΔF (n) / ΔF (n−1)) to SOC and constraint load increment ratio ΔF ratio (= ΔF (n) / ΔF (n) to charge capacity (A × h) It is a figure which shows the relative relationship with the curve of -1)). FIG. 14 is an explanatory diagram of a fitting error calculation operation between the actual measurement curve shown in FIG. 13 and a reference curve of a constraint load increment ratio (ΔF ratio) with respect to SOC. FIG. 14 is an explanatory diagram of a fitting operation of the actual measurement curve shown in FIG. 13 and a reference curve of a constraint load increment ratio (ΔF ratio) with respect to SOC. FIG. 17 is a diagram illustrating a change in a fitting error with respect to a section obtained by the operations in FIGS. 15 and 16. FIG. 9 is a diagram in which a section in which a fitting error is minimized is superimposed on a curve indicating a change in a constraint load F with respect to SOC in an initial state and a creep state. It is explanatory drawing which shows the structure of the system of the charging rate estimation system in another embodiment of this invention, and a battery module. It is an explanatory view showing a physical model of one segment of a battery cell in other embodiments of the present invention. FIG. 9 is a diagram illustrating a change in a structural member shrinkage amount dX2 with respect to SOC in an initial state and a creep state. It is a figure which shows the change of the increment ((DELTA) dX2 (n)) of the contraction amount of a structural member with respect to charge capacity (Axh) in an initial state and a creep state. It is a figure which shows the change of the structural member shrinkage | increase amount ratio ((DELTA) dX2 ratio) with respect to charge capacity (Axh) in an initial state and a creep state. It is a flowchart which shows the operation | movement which acquires the actual measurement curve of the structural member shrinkage | increase amount ratio ((DELTA) dX2 ratio) with respect to charge capacity (Axh). 25 is a flowchart showing an operation of calculating a correction coefficient β using the actual measurement curve acquired in FIG. 24 and a reference curve of a structural member shrinkage amount increment ratio (ΔdX2 ratio) with respect to SOC. It is explanatory drawing of the method of acquiring the increase ((DELTA) dX2 (n)) of the contraction amount of a structural member with respect to charge capacity (Axh). 6 is an actual measurement curve of a structural member shrinkage amount increment ratio (ΔdX2 ratio) with respect to a charging capacity (A × h). FIG. 28 is an explanatory diagram of a fitting error calculation operation between the actual measurement curve shown in FIG. 27 and a reference curve of a structural member shrinkage amount increment ratio (ΔdX2 ratio) with respect to SOC. FIG. 28 is an explanatory diagram of a fitting operation of the actual measurement curve shown in FIG. 27 and a reference curve of a structural member shrinkage amount weight increment ratio (ΔdX2 ratio) with respect to SOC. FIG. 9 is a diagram in which a section showing a minimum fitting error is superimposed on a curve showing a change in the amount of contraction dX2 of the structural member with respect to the SOC in the initial state and the creep state. FIG. 4 is an explanatory diagram showing different usages of an SOC estimation method. FIG. 4 is an explanatory diagram illustrating a method of estimating SOC by combining different SOC estimation methods.

<Configuration of charging rate estimation system>
Hereinafter, embodiments of the present invention will be described with reference to the drawings. As shown in FIG. 1, a charging rate estimation system 100 according to the present embodiment includes a battery cell 10, a resin member 17 serving as an insertion member, restraint plates 11 and 12, restraint rods 13 to 16, a battery cell 10. And a temperature sensor 22 for detecting the temperature of the battery cell 10 and a calculation unit 30 for estimating the SOC of the battery cell 10.

  The battery cell 10 is a chargeable / dischargeable secondary battery such as a lithium ion battery or a nickel hydride battery. The resin member 17 is a flat resin molded member through which cooling air flows. As shown in FIG. 1, the battery cells 10 are stacked with a resin member 17 as an interposed member interposed therebetween, and the battery cells 10 are restrained by restraint plates 11 and 12 and restraint rods 13 to 16 arranged at both ends in the laminating direction. The battery 10 is mounted on an unillustrated electric vehicle as a battery module 50 which is a laminate in which the cells 10 and the resin member 17 are integrally restrained. Each battery cell 10 includes a positive terminal 111 and a negative terminal 112, and the positive terminal 111 and the negative terminal 112 of the adjacent battery cell 10 are connected by a bus bar (not shown).

  As shown in FIG. 1, the calculation unit 30 includes a CPU 31 that internally performs information processing and calculation processing, a memory 32 that stores various characteristic maps of the battery cells 10, an SOC estimation calculation program, and the like, a load sensor 21, This is a computer including a sensor interface 33 to which the sensor 22 is connected. The CPU 31, the memory 32, and the sensor interface 33 are connected by a data bus.

<Modeling of battery cell structure and segments>
As shown in FIG. 2 and FIG. 3A, the battery cell 10 is such that a wound body 120 that is a power generation element that performs charging and discharging is housed in a casing 110 made of metal such as aluminum. The casing 110 is in a closed state, and an electrolytic solution is injected therein. 2 and 3, the X axis and the Z axis are axes orthogonal to each other. In the present embodiment, the axis corresponding to the vertical direction is the Z axis. Note that an axis orthogonal to the X axis and the Z axis is defined as a Y axis.

  As shown in FIGS. 3A and 3B, the wound body 120 is formed by winding a belt-shaped body in which a positive electrode plate 121, a negative electrode plate 122, and a separator 123 are laminated in an elliptical shape. is there. As shown in FIG. 3B, the positive electrode plate 121 has a current collecting foil 121a and a positive electrode active material layer 121b formed on the surface of the current collecting foil 121a. The positive electrode active material layer 121b includes a positive electrode active material, a conductive material, a binder, and the like. The current collector foil 121a of the positive electrode plate 121 is connected to the positive electrode terminal 111 shown in FIG. In addition, the negative electrode plate 122 has a current collecting foil 122a and a negative electrode active material layer 122b formed on the surface of the current collecting foil 122a. The negative electrode active material layer 122b includes a negative electrode active material, a conductive material, a binder, and the like. The current collector foil 122a of the negative electrode plate 122 is connected to the negative electrode terminal 112 shown in FIG. As shown in FIG. 3B, in this embodiment, the positive electrode active material layer 121b, the separator 123, and the negative electrode active material layer 122b through which the electrolyte penetrates are referred to as electrodes 125. In addition, the length of the state in which the binding force is not applied to the electrode 125 is referred to as the electrode natural length.

  As shown in FIG. 4, the natural length of the electrode becomes longer or thicker as the SOC of the battery cell 10 becomes higher. Therefore, the SOC of the battery cell 10 can be estimated using the natural length of the electrode 125 of the battery cell. The map shown in FIG. 4 is stored in the memory 32 of the calculation unit 30.

  As described above with reference to FIG. 1, the battery cell 10 and the resin member 17 of the battery module 50 are integrally restrained by the restraint plates 11 and 12 and the restraint rods 13 to 16. As shown in (1), the length of one segment 40 composed of one battery cell 10 and resin members 17 on both sides is maintained at a constant length L. The physical model of the segment 40 will be described with reference to FIG. As shown in FIG. 5A, the segment 40 includes the battery cell 10 and the resin member 17. The battery cell 10 includes the electrode 125 (the positive electrode active material layer 121b, the separator 123, and the negative electrode active material layer 122b) whose thickness increases as the SOC increases due to the permeation of the electrolyte. On the other hand, the thickness of the resin member 17 and the casing 110 does not change even if the SOC increases. Therefore, as shown in FIG. 5B, the segment 40 has a positive electrode active material layer 121b whose thickness changes due to a change in SOC, an electrode 125 including the separator 123, and a negative electrode active material layer 122b, even if the SOC changes. The structural member 45 such as the resin member 17 and the casing 110 whose thickness does not change can be modeled as being laminated. On the other hand, as shown in FIG. 5C, both the electrode 125 and the structural member 45 are elastic bodies whose thickness changes when a load is applied. Therefore, the electrode 125 is a spring member having a spring constant k1, and the structural member 45 is a spring constant. k2 spring members, and these two spring members can be modeled as being connected in series. The spring constants k1 and k2 can be determined in advance by experiments or the like.

  As shown in FIG. 5C, when the SOC of the battery cell 10 increases, the thickness of the electrode 125 tends to increase by dX. On the other hand, the thickness of the structural member 45 does not change even when the SOC increases. Therefore, when the SOC of the battery cell 10 increases, the entire thickness of the segment 40 tends to be increased by the thickness dX of the electrode 125. However, since the thickness of the segment 40 is maintained at the thickness L by the constraint rods 13 to 16, the segment 40 receives the constraint load F that compresses the entire thickness by the thickness increase dX of the electrode 125. Become.

When the segment 40 is modeled as a series of a spring member having a spring constant k1 and a spring member having a spring constant k2, the equivalent spring constant ktotal is represented by the following equation (1).

ktotal = (k1 × k2) / (k1 + k2) (1)

Since the increase in the total thickness of the segment 40 is the increase dX in the thickness of the electrode 125 due to the increase in the SOC, the constraint load F is expressed by the following equation (2).

F = (k1 × k2) / (k1 + k2) × dX (2)

Therefore, if the load sensor 21 detects the constraint load F, the increase dX in the thickness of the electrode 125 due to the increase in the SOC can be obtained from the following equation (3).

dX = (k1 + k2) / (k1 × k2) × F (3)

<Basic operation of SOC estimation using map of electrode natural length with respect to SOC>
Next, a basic operation of estimating the SOC of the battery cell 10 using the map of the electrode natural length with respect to the SOC shown in FIG.

  As shown in step S101 in FIG. 6, the arithmetic unit 30 checks whether or not a large amount of heat has been generated in the battery cell 10 immediately before performing the processing from step S102 to step S105 shown in FIG. This is because, if the temperature rises due to the immediately preceding heat generation, the electrolytic solution thermally expands, and if the thickness changes transiently, the SOC estimation error increases, so in such a case, the SOC estimation is not performed. That's why.

  There are various methods for checking whether a large amount of heat is generated in the battery cell 10 immediately before. For example, when the square of the current Ib of the battery cell 10 exceeds a predetermined threshold value within a predetermined time immediately before, It may be determined that large heat generation has occurred in the battery cell 10 immediately before, or a temperature sensor is attached to each of the outer surface of the casing 110 of the battery cell 10 and the surface of the internal winding body 120, When the temperature difference detected by the two temperature sensors exceeds a predetermined threshold, it may be determined that the battery cell 10 has generated large heat immediately before.

  When YES is determined in step S101 of FIG. 6, arithmetic unit 30 determines that the estimation error of the SOC becomes large, and does not perform the SOC estimation processing. On the other hand, if NO is determined in step S101 of FIG. 6, it is determined that the SOC estimation accuracy can be ensured, and the process proceeds to step S102 of FIG.

  After proceeding to step S102 in FIG. 6, the calculation unit 30 calculates a thickness increase dX of the electrode 125 due to an increase in the SOC from an initial state where there is no constraint. As described above, the calculation unit 30 acquires the restraint load F by the load sensor 21 and calculates the thickness increase dX of the electrode 125 by Expression (3).

Next, the calculation unit 30 proceeds to step S103, and corrects the temperature change of the thickness dX of the electrode 125 by the following equation (4).

_{dX = dX-ΔT × k T} × X0 ········ (4)

In Expression (4), ΔT is a temperature change from the initial state, and is a temperature difference between the temperature detected by the temperature sensor 22 and the initial temperature set in the memory 32 in advance. _KT is a coefficient of linear expansion of the electrode 125. X0 is the thickness of the electrode 125 in the initial state without restriction.

Next, the calculation unit 30 proceeds to step S104, and adds the thickness increase dX of the electrode 125 corrected by the equation (4) to the thickness X0 of the electrode 125 in the initial state where there is no constraint, and calculates the equation (5). To calculate the natural length X of the electrode 125.

X = X0 + dX (5)

  Next, the calculation unit 30 proceeds to step S105, and estimates the SOC of the battery cell 10 using the map of the electrode natural length with respect to the SOC illustrated in FIG.

<Estimation of SOC of battery cell considering creep>
The operation of estimating the SOC of the battery cell 10 using the map of the electrode natural length with respect to the SOC has been described above. Next, the SOC of the battery cell 10 is determined in consideration of the creep of the resin member 17 and the constraint rods 13 to 16. A method for estimating will be described.

  FIG. 7 is a diagram illustrating a change in the constraint load F with respect to the SOC of the battery cell 10. The solid line in FIG. 7 shows the change in the restraint load F with respect to the SOC of the battery cell 10 in the initial state, and the broken line shows the SOC of the battery cell 10 in a state where creep has occurred in the resin member 17 and the restraint rods 13 to 16 due to aging. 3 shows a change in the constraint load F with respect to the following. When a constraint load F is applied to each battery cell 10 of the battery module 50 shown in FIG. 1, a compression load having the same magnitude as the constraint load F is applied to the resin member 17 sandwiched between the battery cells 10. I have. Further, a tensile load having a magnitude of １ ／ of the constraint load F is applied to each of the constraint rods 13 to 16. When this state continues for a long time, the compressive stress of the resin member 17 decreases due to the compressive load, and accordingly, the tensile stress of the restraining rods 13 to 16 also decreases. Such a phenomenon is creep.

  When creep occurs in the resin member 17 and the restraining rods 13 to 16 and the compressive stress of the resin member 17 decreases and the tensile stress of the restraining rods 13 to 16 decreases, the restraining load F of the battery cell 10 decreases. For this reason, as shown in FIG. 7, after creep occurs, the change in the constraint load F with respect to the change in the SOC becomes smaller than in the initial state.

When estimating the SOC of the battery cell 10 using the map of the electrode natural length with respect to the SOC, as described above, the thickness increase dX of the electrode 125 is calculated by Expression (3) (described again below). .

dX = (k1 + k2) / (k1 × k2) × F (3)

For this reason, when the restraint load F decreases due to creep, the increase dX in the thickness of the electrode 125 decreases even if the increase amount of the SOC is the same, the natural length X of the electrode 125 calculated by the equation (5) also decreases, and the battery cell 10 Is estimated to be smaller than the actual SOC.

  Therefore, a method of correcting the estimation error of the SOC and accurately estimating the SOC when the constraint load F of the battery cell 10 changes over time due to creep will be described below.

This method first corrects the equivalent spring constant ktotal (= (k1 × k2) / (k1 + k2)) of the segment 40 shown in Expression (1) using the following three principles (a) to (c). The correction coefficient α is calculated.
(A) As shown in FIGS. 8 and 9, the increment of the present constraint load F with respect to the increment of the previous constraint load F when the battery cell 10 is charged at a constant charge capacity ΔAh with respect to the increment ΔF (n−1). The constraint load increment ratio ΔF ratio (= ΔF (n) / ΔF (n−1)), which is the ratio of ΔF (n), which is the ratio between the initial state and the state after creep,
(B) The constraint load increment ratio ΔF ratio (= ΔF (n) / ΔF (n−1)) changes according to the change in the charge capacity (A × h), and becomes a curve as shown in FIG. 9;
(C) The constraint load increment ratio ΔF ratio (= ΔF (n) / ΔF (n−1)) changes according to the change in the SOC, and becomes like the reference curve shown in FIG.

When the correction coefficient α is used, an increase dX in the thickness of the electrode 125 due to an increase in the SOC can be obtained from the following equation (6).

dX = α × (k1 + k2) / (k1 × k2) × F (6)

<Curve acquisition operation of restraint load increment ratio ΔF ratio to charging capacity (A × h)>
Hereinafter, with reference to FIGS. 10 and 12 to 14, an operation of acquiring a curve of the constraint load increment ratio ΔF ratio (= ΔF (n) / ΔF (n−1)) with respect to the charging capacity (A × h) will be described. explain.

  As shown in step S201 in FIG. 10, the calculation unit 30 determines whether or not the battery cell 10 is being charged by an external power supply. This is because the calculation of the correction coefficient α requires a large change in the SOC as in the case of external charging. When the external charging is not being performed, the arithmetic unit 30 terminates the operation of the program without executing the processing of steps S202 to S210 shown in FIG.

  If the calculation unit 30 determines YES in step S201 in FIG. 10, the process proceeds to step S202 in FIG. 10 and sets 1 to a counter n to initialize. Then, the calculation unit 30 proceeds to step S203 in FIG. 10 and detects the constraint load F (1) at the point (1) shown in FIG. 12 by the load sensor 21 and proceeds to step S204.

  The arithmetic unit 30 detects the charging current Ib of the battery cell 10 with a current sensor (not shown) in step S204 of FIG. 10, and calculates the amount of charge of the battery cell 10 by integrating the detected charging current Ib. Then, the arithmetic unit 30 waits until the charge amount reaches a certain charge capacity ΔAh as shown in FIG. 12 in step S204. Then, when the charge amount of the battery cell 10 reaches the fixed charge capacity ΔAh, the calculation unit 30 proceeds to step S205, detects the constraint load F (2) at the next point (2) shown in FIG. Proceeding to S206, the constraint load increment ΔF (1) = F (2) −F (1) is calculated.

  In the case of n = 1, there is no restraint load increment ΔF (n−1) calculated previously, and the restraint load increment ratio ΔF ratio cannot be calculated. For this reason, the calculation unit 30 determines whether n> 1 in step S207 of FIG. 10, and when n> 1 is not satisfied, that is, when n = 1, does not proceed to step S208, and proceeds to step S211. n is incremented by 1 and n = 2, and the process returns to step S204. Then, when the charge amount of the battery cell 10 reaches the fixed charge capacity ΔAh, the process proceeds to step S205, where the constraint load F (3) at the next point (3) is detected, and the process proceeds to step S206, where the constraint load increment ΔF ( 2) = F (3) -F (2) is calculated.

Here, since n = 2, the calculation unit 30 determines YES in step S207 and proceeds to step S208 to calculate the restraint load increment ratio ΔF ratio (1) by the following equation (7).

ΔF ratio (1) = ΔF (2) / ΔF (1) (7)

  After calculating the restraint load increment ratio ΔF ratio, the calculation unit 30 proceeds to step S209 in FIG. 10, increments n by 1, proceeds to step S210, and determines whether n is a preset nend. Then, steps S204 to S209 are repeatedly executed until n becomes nend, and the load sensor 21 calculates the constraint loads F (4) to F (6) from points (4) to (6) shown in FIG. Detected, the constraint load increment ΔF (5) is calculated from the constraint load increment ΔF (3), and the constraint load increment ratio ΔF ratio (5) is calculated from the constraint load increment ratio ΔF ratio (2), and stored in the memory 32. . Then, when n = nend in step S210, the calculation unit 30 ends the execution of the program.

  The arithmetic unit 30 executes the above-described program to measure the restraint load increment ratio ΔF ratio (= ΔF (n) / ΔF (n−1)) with respect to the charge capacity (A × h) as shown in FIG. A curve can be obtained. As described above, this actually measured curve is the same curve both in the initial state and after creep.

<Calculation of correction coefficient α>
Next, an equivalent spring constant ktotal (= ((( The calculation of the correction coefficient α for correcting (k1 × k2) / (k1 + k2)) will be described.

  As shown in step S301 of FIG. 11, the calculation unit 30 calculates the constraint load increment ratio ΔF ratio (= ΔF (n) with respect to the charging capacity (A × h) as described with reference to FIGS. / ΔF (n−1)) is confirmed. If the acquisition of the actual measurement curve has not been completed, the calculation unit 30 determines that the calculation of the correction coefficient α is not ready, and ends the execution of the program. On the other hand, when the acquisition of the actual measurement curve has been completed, the calculation unit 30 determines that the calculation of the correction coefficient α is ready, and proceeds to step S302.

  In step S302, the arithmetic unit 30 prepares and stores in advance in the memory 32 the constraint load increment ratio ΔF (= ΔF (n) / ΔF (n−1)) of the battery cell 10 with respect to the SOC shown in FIG. Read out the reference curve.

<Procedure for Generating Reference Curve of Restrained Load Incremental Ratio ΔF Ratio (= ΔF (n) / ΔF (n−1)) of SOC of Battery Cell 10>
This reference curve is prepared in the following procedure. First, in the initial state, when the SOC and the full charge capacity (A × h) of the battery cell 10 at that time are known, as described with reference to FIGS. , The constraint load F is detected for each constant charge capacity ΔAh, the constraint load increment ΔF and the constraint load increment ratio ΔF ratio are calculated, and the charge capacity (A × h) of the battery cell 10 as shown in FIG. , A curve of the constraint load increment ratio ΔF ratio is generated. However, when generating the reference curve, the constraint load F is detected in a range of the charge capacity (A × h) wider than the curve shown in FIG. 13, and the constraint load increment ΔF and the constraint load increment ratio ΔF ratio are calculated. , A curve of the constraint load increment ratio ΔF ratio with respect to the charging capacity (A × h) is generated in advance. Since the SOC (%) and the full charge capacity (A × h) of the battery cell 10 at the time of starting charging are known, the charge starting curve of the constraint load increment ratio ΔF ratio with respect to the generated charging capacity (A × h) is obtained. At this time, the SOC (%) (= SOC0) and the ΔSOC (%) corresponding to the constant charge capacity ΔAh can be found. When the charge capacity (A × h) on the horizontal axis of the curve of the constraint load increment ratio ΔF ratio to the charge capacity (A × h) generated using this relationship is replaced with the SOC (%) of the battery cell 10, FIG. A reference curve of the restrained load increment ratio ΔF ratio (= ΔF (n) / ΔF (n−1)) with respect to the SOC of the battery cell 10 as shown can be generated. The generated reference curve is stored in the memory 32.

<Fitting of reference curve and measured curve>
When the battery cell 10 has not deteriorated and the full charge capacity (A × h) is the same as the initial state, the constant charge capacity of the measured curve shown in FIG. 13 is the same as in the case of generating the reference curve described above. ΔAh can be converted to ΔSOC (%) of the battery cell 10. Therefore, the calculation unit 30 generates an actual measurement curve obtained by converting ΔAh on the horizontal axis of the actual measurement curve shown in FIG. 13 into ΔSOC. Here, since the SOC at the time of starting the acquisition of the actual measurement curve is unknown, the calculation unit 30 adjusts the initial position of the actual measurement curve indicated by the broken line to the position of SOC1 shown in FIG. It is superimposed on the reference curve of the constraint load increment ratio ΔF ratio (= ΔF (n) / ΔF (n−1)).

Next, as shown in step S302 in FIG. 11, the calculation unit 30 calculates the fitting error between the measured curve shown by the broken line in FIG. 15 and the reference curve shown by the solid line, that is, in FIG. The fitting error of the section 1 shown is calculated.

Fitting error = Σ (ΔF ratio _(a n) _-ΔF ratio _{(b n)) ··· (8} )

In the formula (8), [Delta] F ratio _{(a n)} represents the value of [Delta] F ratio in _{a n} points on the measured curve, [Delta] F ratio _{(b n)} is the value of [Delta] F ratio at the point _{b n} on the reference curve Show.

  When the calculation of the fitting error is completed, the arithmetic unit 30 shifts the first position of the measured curve shown by the broken line to the position of SOC2 shown in FIG. 16 and superimposes it on the reference curve shown by the solid line as shown in FIG. . Then, the fitting error of the section 2 shown in FIG. 16 is calculated by the equation (8). Similarly, the calculation unit 30 shifts the initial position of the actually measured curve shown by the broken line to the position of SOC3 shown in FIG. Then, the fitting error of the section 3 shown in FIG. 16 is calculated by the equation (8).

  As described above, the calculation unit 30 calculates the fitting error in each section while shifting the start position of the actually measured curve indicated by the broken line to SOC1, SOC2, and SOC3. Then, the calculation unit 30 can generate a fitting error curve with the horizontal axis as a section, as shown in FIG. Then, the calculation unit 30 specifies a section where the fitting error is minimized, as shown in step S303 of FIG. In this case, as shown in FIGS. 16 and 17, since the fitting error is smallest in the section 2, the calculation unit 30 specifies that the actually measured curve corresponds to the section 2 of the reference curve.

Next, as illustrated in FIG. 18, the calculation unit 30 overlaps the specified section 2 with the curve of the constraint load F with respect to the SOC of the battery cell 10 described with reference to FIG. 7, and the process proceeds to step S304 in FIG. 11. The correction coefficient α is calculated as in the following equation (9) from the ratio of the initial constraint load F and the post-creep constraint load F in the specified section 2, that is, the specified SOC range.

Correction coefficient α = (initial restraint load F) / (restraint load F after creep) (9)

The calculation of the correction coefficient α may be the section 2 where the fitting error is minimum, that is, the average value of the entire SOC range where the fitting error is minimum, or the ratio in the section 2 or the median value of the SOC range. .

<Estimation calculation of SOC of battery cell using correction coefficient α>
After calculating the correction coefficient α, the calculation unit 30 obtains the increase dX in the thickness of the electrode 125 due to the increase in the SOC using the equation (6) described above in step S102 of FIG. 6 (described again below). .

dX = α × (k1 + k2) / (k1 × k2) × F (6)

Then, in the same manner as described above, in step S103 of FIG. 6, the temperature change of the thickness increase dX of the electrode 125 is corrected using Expression (4) (to be described again below),

_{dX = dX-ΔT × k T} × X0 ········· (4)

Proceeding to step S104 in FIG. 6, the natural length X of the electrode 125 is calculated by equation (5) (described below).

X = X0 + dX (5)

Then, in step S105 of FIG. 6, the calculation unit 30 estimates the SOC of the battery cell 10 using the map of the electrode natural length with respect to the SOC shown in FIG.

  As described above, the charging rate estimation system 100 according to the present embodiment is configured such that when the resin member 17 and the constraint rods 13 to 16 creep and the constraint load F of the battery cell 10 decreases over time, the constraint load F decreases. Since the SOC is estimated by correcting the SOC, the SOC of the battery cell 10 can be accurately estimated even when the resin member 17 and the constraint rods 13 to 16 creep and the constraint load F of the battery cell 10 decreases over time. can do.

<Estimation of SOC of Battery Cell Using Thickness Sensor>
In the embodiment described above, the operation of estimating the SOC of the battery cell 10 based on the constraint load F detected by the load sensor 21 has been described. However, the same applies to the case where the thickness sensor 23 is used instead of the load sensor 21. By the operation described above, the SOC can be estimated.

  The configuration of the charging rate estimation system 200 of the present embodiment will be described with reference to FIG. The charging rate estimation system 200 of the present embodiment has a thickness sensor 23 attached between the battery cells 10 instead of the load sensor 21 of the charging rate estimation system 100 described with reference to FIGS. The other configuration is the same as that of the charging rate estimation system 100.

  Next, a physical model of the segment 40 when the thickness sensor 23 is used will be described with reference to FIG. The basic physical model is the same as the physical model described with reference to FIG. As shown in FIGS. 20 (a) and 20 (b), in the segment 40 composed of one battery cell 10 and the resin members 17 on both sides, the electrode 125 and the SOC increase in thickness as the SOC increases. However, it can be modeled as being constituted by the structural member 45 whose thickness does not change. Since both the electrode 125 and the structural member 45 are elastic bodies whose thickness changes when a load is applied, as shown in FIG. 20C, the electrode 125 is a spring member having a spring constant k1, and the structural member 45 is a spring constant. k2 spring members, and these two spring members can be modeled as being connected in series. The spring constants k1 and k2 can be determined in advance by experiments or the like.

As shown in FIG. 20C, when the SOC of the battery cell 10 increases, the thickness of the electrode 125 tends to increase by dX. On the other hand, the thickness of the structural member 45 does not change even when the SOC increases. Therefore, when the SOC of the battery cell 10 increases, the entire thickness of the segment 40 tends to be increased by the thickness dX of the electrode 125. However, since the thickness of the segment 40 is maintained at the thickness L by the constraint rods 13 to 16, the segment 40 receives the constraint load F that compresses the entire thickness by the thickness increase dX of the electrode 125. Become. Then, the electrode 125 is contracted by the thickness dX1 by the constraint load F, and the structural member 45 is contracted by the thickness dX2.
Therefore,

dX = dX1 + dX2 (10)

As shown in FIG. 20A, since the thickness sensor 23 detects the interval between the casings 110 of the adjacent battery cells 10, the thickness sensor 23 detects the contraction amount dX2 of the structural member 45. Here, the relationship between the ratio between the amount of contraction dX1 of the electrode 125 and the amount of contraction dX2 of the structural member 45, and the relationship between the ratio between the spring constant k1 of the electrode 125 and the spring constant k2 of the structural member 45 is expressed by the following equation (11). become.

dX1: dX2 = k2: k1 (11)

From the expressions (10) and (11), the increase dX in the thickness of the electrode 125 due to the increase in the SOC is represented by the following expression (12).

dX = k2 / k1 × dX2 + dX2 (12)

Accordingly, when the thickness sensor 23 detects the contraction amount dX2 of the structural member 45, the increase dX in the thickness of the electrode 125 due to the increase in the SOC is obtained from the above equation (12).

<Estimation of SOC of Battery Cell Considering Creep Using Thickness Sensor>
Next, a method of detecting the amount of contraction dX2 of the structural member 45 by the thickness sensor 23 and estimating the SOC of the battery cell 10 in consideration of creep of the resin member 17, the constraint rods 13 to 16, and the like will be described.

  FIG. 21 is a diagram showing a change in a contraction amount dX2 of the structural member 45 (hereinafter, referred to as a structural member contraction amount dX2) with respect to the SOC of the battery cell 10. The solid line in FIG. 21 shows the change in the amount of contraction dX2 of the structural member with respect to the SOC of the battery cell 10 in the initial state, and the broken line shows the battery cell 10 in the state where creep has occurred in the resin member 17 and the restraining rods 13 to 16 due to aging. Of the structural member shrinkage dX2 with respect to the SOC of FIG. When a constraint load F is applied to each battery cell 10 of the battery module 50 shown in FIG. 1, a compression load having the same magnitude as the constraint load F is applied to the resin member 17 sandwiched between the battery cells 10. I have. Further, a tensile load having a magnitude of １ ／ of the constraint load F is applied to each of the constraint rods 13 to 16. If this state continues for a long time, the contraction amount dX2 of the structural member increases due to the compressive load, and conversely, the contraction amount dX1 of the electrode 125 decreases. Such a phenomenon is creep. Therefore, when creep occurs in the resin member 17, as shown in FIG. 21, the change in the structural member shrinkage dX2 with respect to the change in SOC becomes larger than in the initial state. Note that even when such creep occurs, the constraint load F decreases.

  When estimating the SOC of the battery cell 10 using the map of the electrode natural length with respect to the SOC, as described above, the thickness increase dX of the electrode 125 is calculated by the equation (12) described above. For this reason, when the amount of contraction dX2 of the structural member increases due to creep, the increase dX in the thickness of the electrode 125 increases even if the increase in the SOC is the same, and the natural length X of the electrode 125 calculated by the equation (5) also increases. This means that the SOC of the cell 10 is estimated to be larger than the actual value.

  Therefore, a method for correcting the estimation error of the SOC and accurately estimating the SOC when the contraction amount dX2 of the structural member detected by the thickness sensor 23 due to creep changes over time will be described below.

In this method, first, a correction coefficient β for correcting the increase dX in the thickness of the electrode 125 shown in Expression (12) is calculated using the following three principles (d) to (f).
(D) As shown in FIGS. 22 and 23, when the battery cell 10 is charged at a constant charge capacity ΔAh, the current structural member contraction relative to the increment ΔdX2 (n−1) of the previous structural member contraction amount dX2. The ratio of the structural member shrinkage amount increment ratio ΔdX2 (= ΔdX2 (n) / ΔdX2 (n-1)), which is the ratio of the amount increment ΔdX2 (n), is the same between the initial state and after creep;
(E) The structural member shrinkage amount increment ratio ΔdX2 ratio (= ΔdX2 (n) / ΔdX2 (n−1)) changes with the change in the charging capacity (A × h), resulting in a curve as shown in FIG.
(F) The structural member shrinkage amount increment ratio ΔdX2 ratio (= ΔdX2 (n) / ΔdX2 (n-1)) changes according to the change of the SOC, and becomes like a reference curve shown by a solid line in FIG.

When the correction coefficient β is used, an increase dX in the thickness of the electrode 125 due to an increase in the SOC is obtained from the following equation (13).

dX = β × [k2 / k1 × dX2 + dX2] (13)

<Curve Acquisition Operation of Increasing Ratio ΔdX2 Ratio of Structural Member Shrinkage to Charge Capacity (A × h)>
Hereinafter, referring to FIG. 24 and FIGS. 26 to 27, a curve of the ratio of the increase in the contraction amount of the structural member ΔdX2 (= ΔdX2 (n) / ΔdX2 (n−1)) to the charging capacity (A × h) is acquired. The operation will be described.

  In step S401 in FIG. 24, the calculation unit 30 determines whether the battery cell 10 is being charged by an external power supply. This is because the calculation of the correction coefficient β requires a large change in the SOC as in the case of external charging. When the external charging is not being performed, the arithmetic unit 30 terminates the operation of the program without executing the processing of steps S402 to S410 shown in FIG.

  When determining YES in step S401 in FIG. 24, the calculation unit 30 proceeds to step S402 in FIG. 24, sets 1 to a counter n, and performs initialization. Then, the calculation unit 30 proceeds to step S403 in FIG. 24, detects the structural member contraction amount dX2 (1) at the point (1) shown in FIG. 26 by the thickness sensor 23, and proceeds to step S404.

  The calculation unit 30 detects the charging current Ib of the battery cell 10 with a current sensor (not shown) in step S404 of FIG. 24, and calculates the amount of charge of the battery cell 10 by integrating the detected charging current Ib. Then, the arithmetic unit 30 waits in step S404 until the charge amount reaches a certain charge capacity ΔAh as shown in FIG. Then, when the charge amount of the battery cell 10 reaches a certain charge capacity ΔAh, the operation unit 30 proceeds to step S405, and detects the structural member contraction amount dX2 (2) at the next point (2) shown in FIG. In step S406, the structural member shrinkage increment ΔdX2 (1) = dX2 (2) −dX2 (1) is calculated.

  When n = 1, there is no structural member shrinkage amount increment ΔdX2 (n−1) calculated previously, and the structural member shrinkage amount increase ratio ΔdX2 ratio cannot be calculated. Therefore, the arithmetic unit 30 determines whether n> 1 in step S407 of FIG. 24. If n> 1 is not satisfied, that is, if n = 1, the processing unit 30 does not proceed to step S408 and proceeds to step S411. n is incremented by 1 and n = 2, and the process returns to step S404. When the charge amount of the battery cell 10 reaches a certain charge capacity ΔAh, the process proceeds to step S405, where the structural member shrinkage dX2 (3) at the next point (3) is detected, and the process proceeds to step S406, where the structural member shrinkage is performed. Calculate the quantity increment ΔdX2 (2) = dX2 (3) −dX2 (2).

Here, since n = 2, the arithmetic unit 30 determines YES in step S407 and proceeds to step S408 to calculate the structural member shrinkage amount increment ratio ΔdX2 ratio (1) by the following equation (14). .

ΔdX2 ratio (1) = ΔdX2 (2) / ΔdX2 (1) (14)

  After calculating the structural member shrinkage amount increment ratio ΔdX2 ratio, the calculation unit 30 proceeds to step S409 in FIG. 24, increments n by 1, proceeds to step S410, and determines whether n is a preset nend. Steps S404 to S409 are repeated until n becomes nend, and the thickness sensor 23 changes the structural member contraction amount dX2 (4) from point (4) to point (6) shown in FIG. (6) is detected, the structural member contraction amount increment ΔdX2 (3) is converted to the structural member contraction amount increment ΔdX2 (5), and the structural member contraction amount increment ratio ΔdX2 ratio (2) is converted to the structural member contraction amount increment ratio ΔdX2 ratio (5). Is calculated and stored in the memory 32. Then, when n = nend in step S410, the calculation unit 30 ends the execution of the program.

  By executing the above-described program, the arithmetic unit 30 executes the above-described program to thereby increase the ratio of the structural member shrinkage amount increment ratio ΔdX2 to the charge capacity (A × h) (= ΔdX2 (n) / ΔdX2 (n−1)). Can be obtained. As described above, this actually measured curve is the same curve both in the initial state and after creep.

<Calculation of correction coefficient β>
Next, the calculation of the correction coefficient β using the actually measured curve of the ratio ΔdX2 (= ΔdX2 (n) / ΔdX2 (n−1)) of the structural member shrinkage amount increment ratio to the acquired charging capacity (A × h) will be described.

  As shown in step S501 of FIG. 25, the calculation unit 30 calculates the ratio ΔdX2 (= ΔdX2 (r) of the structural member shrinkage amount increment ratio to the charging capacity (A × h) as described with reference to FIGS. (n) / ΔdX2 (n-1)) is confirmed. If the acquisition of the actual measurement curve has not been completed, the calculation unit 30 determines that the calculation of the correction coefficient β is not ready, and ends the execution of the program. On the other hand, if the acquisition of the actual measurement curve has been completed, the calculation unit 30 determines that the calculation of the correction coefficient β is ready, and proceeds to step S502.

  In step S502, the arithmetic unit 30 prepares and stores in the memory 32 in advance the solid member contraction amount increment ratio ΔdX2 ratio (= ΔdX2 (n) / ΔdX2 ( The reference curve of n-1)) is read.

<Preparation of reference curve>
This reference curve is prepared in the following procedure as described above. First, in the initial state, when the SOC and the full charge capacity (A × h) of the battery cell 10 at that time are known, the battery cell 10 is charged by the external power supply, and the structural member shrinks at a constant charge capacity ΔAh. The amount dX2 is detected, and the structural member contraction amount increment ΔdX2 and the structural member contraction amount increment ratio ΔdX2 ratio are calculated, and a curve of the structural member contraction amount increment ratio ΔdX2 ratio with respect to the charge capacity (A × h) of the battery cell 10 is generated. Then, the charging capacity (A × h) on the horizontal axis is replaced with SOC (%), and the ratio of increase of the structural member shrinkage amount to the SOC of the battery cell 10 as shown by the solid line in FIG. 28, ΔdX2 ratio (= ΔdX2 (n) / A reference curve of [Delta] dX2 (n-1) is generated. The generated reference curve is stored in the memory 32.

<Fitting of reference curve and measured curve>
When the battery cell 10 has not deteriorated and the full charge capacity (A × h) is the same as the initial state, the constant value in the external charging shown in FIG. 27 is the same as in the case of generating the reference curve described above. The charge capacity ΔAh can be converted to ΔSOC (%) of the battery cell 10. Therefore, the calculation unit 30 generates an actual measurement curve obtained by converting ΔAh on the horizontal axis of the actual measurement curve shown in FIG. 27 into ΔSOC. Here, since the SOC at the time of starting the acquisition of the actual measurement curve is unknown, the calculation unit 30 adjusts the initial position of the actual measurement curve shown by the broken line to the position of SOC1 shown in FIG. , On the reference curve of the ratio of the increase in the contraction amount ΔdX2 of the structural member to the SOC (= ΔdX2 (n) / ΔdX2 (n−1)).

Next, as shown in step S502 of FIG. 25, the calculation unit 30 calculates the fitting error between the actually measured curve and the reference curve shown by the broken line in FIG. 28, that is, the section 1 shown in FIG. Calculate the fitting error of.

Fitting error = Σ (ΔdX2 ratio _(a n) _-ΔdX2 ratio _{(b n)) ·· (15} )

In the formula (15), ΔdX2 ratio _{(a n)} represents the value of DerutadX2 ratio at the point _{a n} on the measured curve, DerutadX2 ratio _{(b n)} is the value of DerutadX2 ratio at the point _{b n} on the reference curve Show.

  When the calculation of the fitting error is completed, the arithmetic unit 30 shifts the initial position of the measured curve shown by the broken line to the position of SOC2 shown in FIG. 29 and superimposes it on the reference curve shown by the solid line as shown in FIG. . Then, the fitting error of the section 2 shown in FIG. 29 is calculated by the equation (15). Similarly, the calculation unit 30 shifts the first position of the actually measured curve shown by the broken line to the position of SOC3 shown in FIG. Then, the fitting error of the section 3 shown in FIG. 29 is calculated by the equation (15).

  As described above, the calculation unit 30 calculates the fitting error in each section while shifting the start position of the actually measured curve indicated by the broken line to SOC1, SOC2, and SOC3. Then, as in the case of the charging rate estimation system 100 described above, the calculation unit 30 can generate a curve of the fitting error with the horizontal axis as a section, as shown in FIG. Then, as shown in step S503 of FIG. 25, the calculation unit 30 specifies a section in which the fitting error is minimized. In this case, as shown in FIG. 29 and FIG. 17, since the fitting error is the smallest in the section 2, the calculation unit 30 specifies that the actually measured curve corresponds to the section 2 of the reference curve.

Next, as shown in FIG. 30, the calculation unit 30 superimposes the specified section 2 on the curve of the structural member contraction amount dX2 with respect to the SOC of the battery cell 10 described with reference to FIG. 21, and proceeds to step S504 in FIG. , The correction coefficient β is calculated from the ratio between the initial structural member shrinkage dX2 of the specified SOC range and the structural member shrinkage dX2 after creep as shown in the following equation (9). .

Correction coefficient β = (initial structural member shrinkage dX2) / (creep structural member shrinkage dX2)
... (16)

The calculation of the correction coefficient β may be performed in the section 2 where the fitting error is minimized, that is, the average value of the entire SOC range in which the fitting error is minimized, or may be calculated in the section 2 or a ratio of the median value of the SOC range. .

<Estimation calculation of SOC of battery cell using correction coefficient β>
After calculating the correction coefficient β, the calculation unit 30 obtains the increase dX in the thickness of the electrode 125 due to the increase in the SOC using the equation (13) described above in step S102 of FIG. 6 (to be described again below). .

dX = β × [k2 / k1 × dX2 + dX2] (13)

Then, in the same manner as described above, in step S103 of FIG. 6, the temperature change of the thickness increase dX of the electrode 125 is corrected using Expression (4) (to be described again below),

_{dX = dX-ΔT × k T} × X0 ········· (4)

Proceeding to step S104 in FIG. 6, the natural length X of the electrode 125 is calculated by equation (5) (described below).

X = X0 + dX (5)

Then, in step S105 of FIG. 6, the calculation unit 30 estimates the SOC of the battery cell 10 using the map of the electrode natural length with respect to the SOC shown in FIG.

  As described above, when the structural member 45 creeps and the structural member shrinkage dX2 of the battery cell 10 increases over time, the charging rate estimation system 200 of the present embodiment increases the structural member shrinkage dX2 by Therefore, the SOC of the battery cell 10 can be accurately estimated even when the structural member 45 creeps and the structural member shrinkage dX2 of the battery cell 10 increases over time. .

  As described above, the operation of estimating the SOC of the battery cell 10 based on the constrained load F detected by the load sensor 21 in the charging rate estimation system 100 and the amount of contraction dX2 of the structural member detected by the thickness sensor 23 in the charging rate estimation system 200 determine the battery cell 10 Has been described, the SOC is estimated from the OCV in the SOC range where the change in the OCV is large with respect to the change in the SOC, and as shown in FIG. The SOC may be estimated from the restrained load F or the contraction amount dX2 of the structural member in the SOC range that is not used. Further, as shown in FIG. 32, the SOC estimation reflection rate based on the restraint load F or the contraction amount dX2 of the structural member may be changed depending on the SOC range. As shown in FIG. 32, in the SOC range where the change in OCV is large with respect to the change in SOC, the SOC estimation reflection rate based on the constraint load F or the contraction amount dX2 of the structural member is set to zero, and the SOC estimation based on OCV is mainly performed. On the other hand, in the SOC range where the OCV hardly changes, the SOC estimation based on the constrained load F or the contraction amount dX2 of the structural member is set to 1 assuming that the SOC estimation reflection ratio based on the constrained load F or the contraction amount dX2 of the structural member is 1 and the SOC estimation is mainly performed. Then, the SOC estimation reflection rate by the constraint load F or the structural member shrinkage dX2 is changed from zero to 1 to increase the SOC, and the SOC estimation reflection rate by the constraint load F or the structural member shrinkage dX2 is increased. It may be.

  Reference Signs List 10 battery cell, 11, 12 restraint plate, 13 to 16 restraint rod, 17 resin member, 21 load sensor, 22 temperature sensor, 23 thickness sensor, 30 calculation unit, 31 CPU, 32 memory, 33 sensor interface, 34 data bus, 40 segment, 45 structural member, 50 battery module, 100, 200 charge rate estimation system, 110 casing, 111 positive terminal, 112 negative terminal, 120 wound body, 121 positive plate, 121a, 122a current collector foil, 121b positive active material Layer, 122 negative electrode plate, 122b negative electrode active material layer, 123 separator, 125 electrode.

Claims (1)

A secondary battery stacked via an insertion member,
A restraining member for restraining the secondary battery and the insertion member in a stacking direction,
A load sensor disposed between the secondary batteries to detect a restraint load;
A temperature sensor for detecting the temperature of the secondary battery,
A calculation unit configured to estimate a charging rate of the secondary battery based on the restraint load detected by the load sensor, the temperature detected by the temperature sensor, and a correction coefficient for correcting a secular change of the restraint load. A system for estimating a charge rate of a secondary battery,
The arithmetic unit includes:
When performing the charging of the secondary battery, calculate the increase of the restraint load for each fixed charge capacity, calculate the restraint load increment ratio by dividing the current value of the restraint load increment by the previous value of the restraint load increment. Calculate and generate an actual measurement curve of the constraint load increment ratio with respect to the charging capacity,
A battery charge rate range in which an error between the reference curve of the restraint load increment ratio with respect to the charge rate prepared in advance and the actually measured curve is minimized, and the initial restraint load and the restraint load detected by the load sensor in the specified charge rate range are specified. And a charge rate estimating system for the secondary battery that calculates the correction coefficient from the deviation from

Images