JP4682509B2 - Battery open voltage calculation method and charge amount calculation method - Google Patents

Description

The present invention relates to a battery internal resistance calculation method, an open-circuit voltage calculation method, and a charge amount calculation method.

  As one method of calculating the amount of charge (SOC) of a secondary battery for driving a motor (hereinafter referred to as a battery) in a hybrid vehicle or the like, a current value I, a terminal voltage V, an integrated value, and a current value I at the time of charging / discharging are calculated. There is a method of calculating based on the Ah integrated value obtained by integration and the battery capacity. In such a charge amount calculation method, a charge amount calculation error is likely to occur due to accumulation of current / voltage detection errors. Therefore, a method has been proposed in which, when the calculation error reaches a predetermined value, the charge amount calculation is corrected using the correction charge amount calculated by a method other than the integration (see, for example, Patent Document 1). ).

  There are three methods for calculating the correction charge amount. The first is to calculate the correction charge amount based on the terminal voltage of the battery when there is no load. Second, the open circuit voltage is calculated by power calculation based on the battery terminal voltage and the current value at the time of charging / discharging, and the correction charge amount is calculated based on the open circuit voltage. Thirdly, an open circuit voltage is calculated based on the battery terminal voltage, current value, and internal resistance value during charging / discharging, and a correction charge amount is calculated based on the open circuit voltage. When traveling in a hybrid vehicle in which charging and discharging are switched in a short time, error correction using the third correction charging amount is important because there are few no-load conditions and power calculation timing.

JP 2000-150003 A

  However, it has been found that in a high-power system that extracts a large current from a battery in a short time, the internal resistance tends to increase as the discharge current increases rapidly. That is, a region having no linearity in IV characteristics appears on the high current region side. Such deviation from linearity is likely to occur when conditions such as a sudden increase in current value or a low SOC overlap. For this reason, there is a problem in that the open circuit voltage cannot be accurately calculated because the IV characteristics are not linear as the current reaches the high current region side, and as a result, an accurate charge amount cannot be obtained.

The present invention relates to a battery open-circuit voltage calculation method for calculating an open-circuit voltage based on a terminal voltage and a discharge current at the time of battery discharge and a virtual internal resistance corresponding to the current region of the discharge current , or a battery for calculating an internal resistance. This is applied to the internal resistance calculation method . The battery discharge current area is classified into at least two current areas according to the magnitude of the current value, and current / voltage sampling data synchronously detected in a low current area having a low current value among a plurality of current areas, and The linear regression calculation based on the current / voltage sampling data synchronously detected in the current region including the discharge current is performed, and calculation is performed from the slope of the IV characteristic line obtained by the linear regression calculation.
Moreover, in the charge amount calculation method by this invention, it calculates with the open circuit voltage-charge amount correlation showing the correlation with the open circuit voltage of a battery, and the charge amount, and the open circuit voltage calculation method in any one of Claims 1-3. The charge amount of the battery is calculated based on the open circuit voltage.

  According to the present invention, since the open circuit voltage is calculated based on the virtual internal resistance corresponding to the current region, the open circuit voltage can be accurately calculated even when the IV characteristic is not linear, The amount of charge calculated using the open circuit voltage can also improve accuracy.

Hereinafter, the best mode for carrying out the present invention will be described with reference to the drawings.
-First embodiment-
FIG. 1 is a block diagram showing a power drive system used in an electric vehicle (EV) and a hybrid electric vehicle (EHV). Reference numeral 1 denotes a travel battery. When the motor is driven, DC power output from the battery 1 is converted into AC power by the inverter 2 and supplied to the travel motor 3. In the regenerative control, the running energy of the vehicle is reversely converted into electric energy via the motor 3 and the inverter 2, and the battery 1 is charged and the vehicle is subjected to regenerative braking.

  The voltage sensor 5 detects the voltage (total voltage) V across the battery 1, and the current sensor 6 detects the current I flowing through the battery 1. A temperature sensor 7 detects the temperature T of the battery 1. The controller 4 calculates a battery state of the battery 1 based on the voltage value V and the current value I detected by the voltage sensor 5 and the current sensor 6, and performs output control and regeneration of the inverter 2 according to the calculated battery state. Control and so on.

Next, a charge amount calculation method performed by the controller 4 in FIG. 1 will be described. FIG. 2 is a conceptual diagram showing a comparison between the calculated charge amount RSOC and the true charge amount SOC. The charge amount RSOC is calculated by the following equation (1) based on a current integrated value (hereinafter referred to as an Ah integrated value) obtained by integrating the current value I during charging / discharging. In the formula (1), Ah is an Ah integrated value, and Ah0 is a battery capacity of the battery 1. For example, the rated capacity of the battery 1 is used as the battery capacity Ah0.
RSOC (%) = [1-Ah / Ah0] × 100 (1)

  In the charge amount RSOC shown in Equation (1), if there is a detection error in the current value, the error is accumulated in the Ah integrated value. Therefore, as shown in FIG. 2, even if the charge amount RSOC at the start of the calculation is equal to the true charge amount SOC, the error with respect to the true charge amount SOC increases as the calculation is performed. By the way, in the lithium ion battery etc. which are used as the battery 1, it is known that there is a certain correlation between the open circuit voltage and the charge amount SOC regardless of the battery deterioration and the battery temperature, and the reproducibility is very good. Therefore, if the correlation between the open circuit voltage E and the charge amount SOC as shown by the curve L0 in FIG. 3 is prepared in advance as a table, the current charge amount SOC can be obtained from the calculated open circuit voltage E3.

  As described above, when the SOC calculation deviation due to accumulation of detection errors becomes large, the charge amount obtained by applying the open circuit voltage E3 to the correlation of FIG. 3 is used as the correction charge amount (CAPSOC), and the charge amount The RSOC may be reset. In the present embodiment, the calculation accuracy of the charge amount RSOC is improved by increasing the calculation accuracy of the open circuit voltage E3 used for calculating the correction charge amount (CAPSOC).

  Next, a charge amount calculation method using the open circuit voltage E3 will be described. Below, the calculation method of the virtual internal resistance deterioration coefficient used when calculating the open circuit voltage E3 will be described first, and then the calculation of the open circuit voltage E3 and the calculation of the charge amount (CAPSOC) will be described.

<< Calculation Method of Virtual Internal Resistance Degradation Coefficients γ1 to γ5 >>
First, the procedure for calculating the virtual internal resistance deterioration coefficients γ1 to γ5 based on the sampling data will be described with reference to the flowchart shown in FIG. In step S1, the voltage value 5 and the current value I during discharging are sampled by the voltage sensor 5 and the current sensor 6 in synchronization. This sampling process is repeated until it is determined in step S3 that the discharge power amount has reached a predetermined value. In step S2, the detected sampling data is classified into a plurality of current regions and stocked in a storage unit (not shown) of the controller 4.

The storage unit of the controller 4 is provided with seven storage units CL1 to CL3 and CH1 to CH7 according to the current range of the discharge current value. As shown in FIG. 5, data (current value I, voltage value V) in which the current value I is included in the current range 0 <I ≦ I ₁ is stocked in the storage unit CL1, and the current range I ₁ is stored in the storage unit CL2. Data of <I ≦ I ₂ is stored, and data of current region I ₂ <I ≦ I ₃ is stocked in the storage unit CL3. Further, in each of the storage units CH1 to CH4, data of current region I ₃ <I ≦ I ₄ , data of current region I ₄ <I ≦ I ₅ , data of current region I ₅ <I ≦ I ₆ , Data of current region I ₆ <I ≦ I ₇ is stocked.

In the present embodiment, the current range 0 <I ≦ I ₃ of the data stocked in the storage units CL1 to CL3 is set as the low current region, and the current range I ₃ <I ≦ I of the stock data stored in the storage units CH1 to CH7. the I ₇ to set up a high-current region. In the example shown in FIG. 5, the number of stocks that can be stored in each of the storage units CL1 to CL3 and CH1 to CH7 is 3, respectively. For example, three sets of sampling data are stocked in the storage unit CL2, and if the fourth sampling data (i10, v10) is detected, the oldest data (i5, v5) is deleted. Stock the latest data (i10, v10).

  In this way, the storage units CL1 to CL3 and CH1 to CH7 are not stocked with more than 3 sampling data, thus preventing a large amount of sampling data from being biased to a specific current region. it can. Therefore, it is not necessary to secure an enormous memory capacity in the controller 4, and the calculation accuracy of the IV characteristic line regression calculation based on the sampling data is increased.

  In step S3, it is determined whether or not the discharge power amount has reached a predetermined amount after starting sampling. If it is determined that the predetermined amount has not been reached, the process returns to step S1 and sampling is continued. On the other hand, if it is determined in step S3 that the discharge power amount has reached a predetermined amount, the process proceeds to step S4. In step S4, it is determined whether or not the data stocked in the storage units CL1 to CL3 has a sufficient number of data for the regression calculation in the low current region in step S5 and the data distribution is not biased.

  As an example of the determination condition in step S4, the total number of data stored in the storage units CL1 to CL3 is sufficient for regression calculation, and the data is stored in all the storage units CL1 to CL3. As a condition. As a condition for determining that the data distribution varies, instead of the above condition that data is stocked in all the storage units CL1 to CL3, the storage units CL1 and CL3 on both sides of the storage unit CL2 are interposed. You may use the condition that data is stocked. By making the determination under such conditions, it is possible to prevent regression calculation using only sampling data within a narrow current range, and it is possible to improve the calculation accuracy of IV characteristics.

If YES is determined in step S4, the process proceeds to step S5. If NO is determined, that is, if the number of data stock is small or the data distribution is biased, the process proceeds to step S7. In step S5, a linear regression calculation of IV characteristics is performed based on the sampling data of the low current region stocked in the storage units CL1 to CL3 to obtain a regression line (IV characteristic line) L1 shown in FIG. The IV characteristic straight line L1 is expressed as shown in Expression (2), and the magnitude of the slope corresponds to the internal resistance R1, and the V intercept corresponds to the open circuit voltage E0.
V = E0−I × R1 (2)

The IV characteristic straight line L1 is calculated based on the actually measured voltage value and current value, and the internal resistance R1, which is the slope thereof, reflects the state (temperature, deterioration) of the battery 1. Internal resistance R1 calculated in step S5 has been calculated based only on the sampling data of the low current range 0 <I ≦ I _3, represents the internal resistance in the low current range 0 <I ≦ I ₃ . Hereinafter, such an internal resistance R1 is referred to as a virtual internal resistance R1.

  FIG. 7 shows the relationship between the current region and the virtual internal resistance initial value. The virtual internal resistance R1 described above reflects the state (temperature, deterioration) of the battery 1, but the virtual resistance initial values Rint1 to Rint5 shown in FIG. 7 are reference temperatures in the battery 1 in the initial state without deterioration. The virtual internal resistance at the time of (for example, normal temperature) is represented. These virtual internal resistance initial values Rint1 to Rint5 are stored in the storage unit of the controller 4 in advance.

In the example shown in FIG. 7, virtual internal resistance initial values Rint1 to Rint5 are set for every five current regions. The setting of the current range is as follows: 0 <I ≦ I ₃ , 0 <I ≦ I ₄ , 0 <I ≦ I ₅ , 0 <I ≦ I ₆ , 0 <I ≦ I with respect to the current values I _{1 to} I ₇ in FIG. ₇ of 5. The current value is larger in the lower current region in FIG. 6, and the initial value of the virtual internal resistance is larger in the lower region. The storage unit of the controller 4 also stores a correlation between a current region as shown in FIG. 8 and a virtual internal resistance deterioration coefficient γ described later. The set values of the virtual internal resistance deterioration coefficients are γ1 to γ5 for each current region similar to the case of FIG.

In step S6, using the virtual internal resistance R1 calculated in step S5, the virtual internal resistance initial value Rint1 in FIG. 6 and the temperature deterioration coefficient α, the virtual internal resistance deterioration coefficient γ1 (new) is calculated by the following equation (3). calculate. The temperature deterioration coefficient α is a parameter representing the internal resistance change due to the temperature change, has a different value depending on the battery temperature, and is stored in advance as a table in the storage unit of the controller 4.
γ1 (new) = Rint1 / (α × R1) (3)

Further, the calculated virtual internal resistance deterioration coefficient γ1 (new) and the virtual internal resistance deterioration coefficient γ1 of the same current region already stored in the storage unit are weighted and averaged as shown in the following equation (4) to obtain the virtual internal resistance A resistance deterioration coefficient γ1 is calculated. Then, γ1 in FIG. 8 is updated by replacing it with the virtual internal resistance deterioration coefficient γ1 calculated by the equation (4). In Equation (4), the current virtual internal resistance deterioration coefficient already stored in the storage unit is expressed as γ1 (old). x is a weighted average coefficient.
γ1 = {(x−1) × γ1 (old) + γ1 (new)} / x (4)

  In step S7, when the regression calculation is performed using the data stored in the storage unit CL1 and each of the storage units CH1 to CH4, the storage unit CL1 and each of the storage units CH1 to CH4 have a sufficient number of data for the regression calculation. It is determined whether or not it has been done. In step S8, which will be described later, the regression calculation is performed using the sampling data of the storage unit CL1 and any one of the storage units CH1 to CH4. However, by performing the determination process of step S7, The calculation accuracy can be improved by assuming that there is data not only in the flow areas (CH1 to CH4) but also in the low current area (CL1) close to no load.

If YES is determined in step S7, the process proceeds to step S8. If NO is determined, that is, if the number of data stock is not sufficient, the process proceeds to step S10. In step S8, using the sampling data stored in the storage unit CL1 and the storage unit CH1, that is, IV characteristics by sampling data of the low current region 0 <I ≦ I ₁ and the high current region I ₃ <I ≦ I ₄ The IV regression line L2 shown in FIG. 6 is obtained. The IV characteristic straight line L2 is expressed as Equation (5), and the magnitude of the inclination corresponds to the virtual internal resistance R2.
V = E ₀ −I × R 2 (5)

  Similarly, an IV characteristic line L3 is calculated from the sampling data stocked in the storage part CL1 and the storage part CH2 by linear regression, and the IV characteristic straight line L4 is obtained from the sampling data stocked in the storage part CL1 and the storage part CH3. Are calculated from the sampling data stocked in the storage part CL1 and the storage part CH4, respectively. And virtual internal resistance R3-R5 in each electric current area | region is calculated | required from the inclination of each IV characteristic straight line L3-L5. FIG. 6 shows the IV characteristic straight lines L3 to L5.

  In step S9, based on the virtual internal resistances R2 to R5 calculated in step S8, virtual internal resistance deterioration coefficients γ2 to γ5 are calculated, and virtual internal resistance deterioration coefficients γ2 to γ5 of FIG. 8 are calculated. Replace with deterioration coefficients γ2 to γ5, respectively. Since the calculation method is the same for any of the virtual internal resistances R2 to R5, only the case of the virtual internal resistance deterioration coefficient γ2 will be described below.

First, from the calculated virtual internal resistance R2, the virtual internal resistance initial value Rint2 in the same current region of FIG. 7 and the temperature deterioration coefficient α, the latest virtual internal resistance deterioration coefficient γ2 (new) is obtained by the following equation (6). calculate.
γ2 (new) = Rint2 / (α × R2) (6)

Next, the calculated virtual internal resistance degradation coefficient γ2 (new) and the virtual internal resistance degradation coefficient γ2 in the same current region already stored in the storage unit are weighted and averaged as in the following equation (7), and the virtual An internal resistance deterioration coefficient γ2 is calculated. In equation (7), γ2 (old) represents the stored virtual internal resistance deterioration coefficient γ2 in FIG. X is a weighted average coefficient. Then, γ2 in FIG. 8 is replaced with the virtual internal resistance deterioration coefficient γ2 calculated by Expression (7).
γ2 = {(x−1) × γ2 (old) + γ2 (new)} / x (7)

  Next, in step S10, all the sampling data stocked in the storage units CL1 to CL3 and CH1 to CH4 are erased, and then the process returns to step S1. As described above, by performing the series of processes shown in FIG. 4, the latest virtual internal resistances R1 to R5 and virtual internal resistance deterioration coefficients γ1 to γ5 in each current region are calculated, and the values of γ1 to γ5 in FIG. Is updated.

  The calculation and update of the virtual internal resistance deterioration coefficients γ1 to γ5 described above may be performed all the time, or may be performed at predetermined time intervals. In this way, virtual internal resistance coefficients γ1 to γ5 used for open circuit voltage calculation are calculated in advance. In the open circuit voltage calculation to be described later, a table of detected voltage value V and current value I, virtual internal resistance coefficients γ1 to γ5 and virtual part resistance initial values Rint1 to Rint5 for each current region (see FIGS. 7 and 8). ), That is, the open circuit voltage is calculated based on the virtual internal resistance Rj.

《Charging amount calculation method》
Next, calculation of the open circuit voltage E3 and the charge amount will be described with reference to the flowchart of FIG. In step S101, the voltage value V and the current value I detected by the voltage sensor 5 and the current sensor 6 are detected, and the voltage value V and the current value I being discharged are sampled synchronously. In step S102, the open circuit voltage E3 is calculated by the following equation (8) using the voltage value V and the current value I sampled in step S101.
E3 = V + I × R (8)

The internal resistance R in the equation (8) is an internal resistance after the temperature correction and the deterioration correction, and in this embodiment, the virtual internal resistance Rj (j = 1, 2, 3, 4, 5) shall be used. That is, the virtual internal resistance Rj is expressed as the internal resistance initial value Rint corrected by the temperature deterioration coefficient α and the internal resistance deterioration coefficient γ. In this case, in which current region of FIG. 6 the current value I is included. Is corrected using the virtual internal resistance initial value Rintj and the virtual internal resistance deterioration coefficient γj in the current region. Therefore, the open circuit voltage E3 is expressed as the following equation (9).
E3 = V + I × Rintj / (α × γj) (9)

Specifically, when the current value I is included in the current region 0 <I ≦ I ₃ , Rint1 and γ1 are used as the virtual internal resistance initial value and the virtual internal resistance deterioration coefficient, and the current value I is the current region I. _{3 <if} included in the I ≦ I ₄ is, Rint2 and γ2 is used as a virtual internal resistance initial and virtual internal resistance deterioration coefficient. Similarly, when the current value I is included in the current regions I ₄ <I ≦ I ₅ , I ₅ <I ≦ I ₆ , and I ₆ <I ≦ I ₇ , the virtual internal resistance initial value and the virtual internal resistance degradation, respectively. A combination of (Rint3, γ3), (Rint4, γ4), (Rint5, γ5) is used as a coefficient.

FIG. 10 shows a case where the data (Ia, Va) detected in step S101 is included in the current region I ₅ <I ≦ I ₆ . As the internal resistance initial value Rintj and the internal resistance deterioration coefficient γj in equation (9), Rint4 and γ4 in FIGS. 7 and 8 are used, and the open circuit voltage E3 is calculated by the following equation (10).
E3 = Va + Ia × Rint4 / (α × γ4) (10)
In the subsequent step S103, the charge amount (CAPSOC) corresponding to the open circuit voltage E3 calculated in step S102 is calculated using the open circuit voltage-charge amount characteristic shown in FIG.

  Conventionally, as shown in FIG. 11, regardless of the magnitude of the current value Ia used for the calculation of the open circuit voltage E3, the internal resistance R is calculated from the IV characteristic straight line calculated by linear regression calculation of sampling data in the entire current range. The open-circuit voltage E3 was calculated by the equation (8) using the obtained internal resistance R. However, as shown in FIG. 11, the actual IV characteristic has a tendency that the linearity is lost in the high current region and the internal resistance is increased. Therefore, when linear regression calculation is performed from sampling data in the entire current range, an error tends to occur in the direction in which the open circuit voltage E3 (calculated value) becomes larger than the true value due to the influence of the data in the high current range, and the charge amount due to the error A decrease in calculation accuracy has been a problem.

  On the other hand, in the present embodiment, the virtual internal resistance initial value Rintj and the virtual internal resistance deterioration coefficient γj based on the sampling data for each current region are calculated, and the current value I used for the calculation of the open circuit voltage E3 is set in the current region. The corresponding virtual internal resistance initial value Rintj and virtual internal resistance deterioration coefficient γj are used. Therefore, as shown in FIG. 10, since the open circuit voltage E3 is calculated based on the IV characteristic which is a regression line using data in a low current region close to no load and data in an actual high current region, the true value is obtained. It is possible to calculate an open circuit voltage E3 (calculated value) that is very close to. That is, the open-circuit voltage calculation with high accuracy reflecting the internal resistance that changes in accordance with the current range can be performed, and the calculation accuracy of the charge amount can be improved.

-Second Embodiment-
FIG. 12 is a diagram for explaining virtual internal resistance calculation processing in the second embodiment, and shows a flowchart in place of FIG. 4 described above. The processing from step S201 to step S203 in FIG. 12 is the same as the processing from step S1 to step S3 in FIG. That is, the voltage value V and the current value I being discharged are sampled synchronously in step S201, and sampling data is stocked as shown in FIG. 5 in step S202. In step S203, it is determined whether or not the discharge of the predetermined power has been completed since the start of sampling. If it is determined that the discharge has not been completed, the process returns to step S201 to continue sampling.

  In step S204, it is determined whether or not the dampling data stocked in the storage units CL1 to CL3 and CH1 to CH4 is stocked to such an extent that linear regression calculation in step S205 described later can be performed. In the linear regression calculation in step S205, linear regression calculation is performed using sampling data of a region obtained by combining three consecutive current regions. Therefore, in the determination in step S204, whether or not data is stocked is determined as a determination condition whether data is stocked in three consecutive current regions or two regions sandwiching one region. As a result, linear regression calculation using sampling data in a narrow current range can be prevented, and calculation accuracy can be improved.

If YES is determined in the step S204, the process proceeds to a step S205. If NO is determined, the process proceeds to a step S207 to delete the stocked sampling data. In step S205, it performs a linear regression calculation based on the sampling data of the current range 0 _<I ≦ I _3, and calculates a virtual internal resistance R11 to calculate the IV characteristic line L11. Similarly, linear regression calculation is performed using each data of current regions I ₁ <I ≦ I ₄ , I ₂ <I ≦ I ₅ , I ₃ <I ≦ I ₆ , I ₄ <I ≦ I ₇ , and IV Characteristic straight lines L12, L13, L14, L15 and virtual internal resistances R12, R13, R14, R15 are calculated (see FIG. 13).

In step S206, the virtual internal resistances R11 to R15 adjacent to each other are compared with each other in the virtual internal resistances R15 to R15 to find the place where the change is most significant. When the most changed virtual internal resistance is Rb, it can be determined that there is an IV characteristic shift in the current region when the virtual internal resistance Rb is calculated. In the example shown in FIG. 14, there is a change in the current region I ₂ <I ≦ I ₅ of the virtual internal resistance R13, and the current threshold values in the change region are Ia and Ib.

As a method for obtaining the current threshold values Ia and Ib, for example, a current region boundary value close to the intersection of the IV characteristic line L13 and the IV characteristic line L12 is defined as the current threshold value Ia, and the IV characteristic line L13 and the IV characteristic line are obtained. A current region boundary value close to the intersection with L14 is defined as a current threshold value Ib. In FIG. 4, the _{Ia = I 3, Ib = I} 5. The virtual internal resistance in the current region Ia <I ≦ Ib is Rb, the virtual internal resistance in the current region 0 <I ≦ Ia is Ra, and the virtual internal resistance in the current region I> Ib is Rc. In FIG. 14, Ra = R11, Rb = R13, and Rc = R15.

  In step S207, all the sampling data stocked in the storage units CL1 to CL3 and CH1 to CH4 are deleted, and the process returns to step S201. The open circuit voltage E3 is calculated using the virtual internal resistances Ra, Rb, Rc and the current threshold values Ia, Ib thus calculated, and the charge amount (CAPSOC) is obtained from the correlation between the open circuit voltage and the charge amount SOC.

  Here, the calculation method of the open circuit voltage E3 is demonstrated with reference to FIGS. First, as in the case of the first embodiment, the voltage value V and the current value I during discharge are sampled in synchronization. And when calculating the open circuit voltage E3 using this data, a calculation method is changed according to the magnitude | size of the electric current value I. FIG.

<< If I≤Ia >>
As shown in FIG. 15, when the current value I is I ≦ Ia, the open circuit voltage E3 is calculated by the following equation (11) using the virtual internal resistance Ra in the current region 0 <I ≦ Ia.
E3 = V + I × Ra (11)

<< If Ia <I≤Ib >>
As shown in FIG. 16, when the current value I is Ia <I ≦ Ib, the IV characteristic line L13 is expressed by the following equation (12), and the point P1 (I, V) is on the IV characteristic line L13. It is in. E ′ is the V intercept of the straight line L13. In the IV characteristic line L13, the voltage value V ′ at the current value Ia is expressed by the following equation (13).
V = E′−I × Rb (12)
V ′ = E′−Ia × Rb
= V + (I-Ia) * Rb (13)

On the other hand, the IV characteristic straight line L11 is expressed by the following formula (14), and the open circuit voltage E3 is calculated by the following formula (15) when the values (Ia, Ea) of the point P3 on the straight line L11 are used.
V = E3-I × Ra (14)
E3 = Ea + Ia × Ra (15)

Since the point P2 (Ia, V ′) on the IV characteristic line L13 and the point P3 (Ia, Ea) on the IV characteristic line L11 are very close to each other, the value of Ea is expressed by the following equation (16). Is equal to V ′. As a result, the open circuit voltage E3 can be calculated by the following equation (17).
Ea = V '
= V + (I-Ia) * Rb (16)
E3 = V + (I-Ia) * Rb + Ia * Ra (17)

<< If Ib <I >>
In FIG. 17, the IV characteristic straight line L15 is represented by the following equation (18), and therefore the voltage value V ′ at the point P4 (Ib, V ′) is represented by the following equation (19). E ′ is the V intercept of the straight line L15.
V = E′−I × Rc (18)
V ′ = E′−Ib × Rc
= V + (I-Ib) * Rc (19)

Since the point P5 (Ib, Eb) on the IV characteristic line L13 is very close to the point P4, it is assumed here that the voltage value Eb is equal to the voltage value E ′, and the following equation (20) holds. At this time, the voltage value Ea of the point P3 (Ia, Ea) is obtained in the same manner as in FIG. 16, and can be calculated by the following equation (21).
Eb = V + (I−Ib) × Rc (20)
Ea = Eb + (I-Ia) * Rb (21)

Therefore, it can be seen from the equations (20) and (21) that the open circuit voltage E3 is calculated by the following equation (22). The charge amount (CAPSOC) is calculated from the correlation between the open circuit voltage E3 thus obtained and the open circuit voltage-charge amount SOC shown in FIG.
E3 = Ea + Ia × Ra
= Eb + (I-Ia) * Rb + Ia * Ra
= V + (I-Ib) * Rc + (I-Ia) * Rb + Ia * Ra (22)

  In the correspondence between the embodiment described above and the elements of the claims, the current thresholds Ia and Ib are the boundary current values of the local area, the virtual internal resistance Rb is the first virtual internal resistance, and the virtual internal resistance. Ra constitutes a second virtual internal resistance, and virtual internal resistance Rc constitutes a third virtual internal resistance. In addition, the present invention is not limited to the above embodiment as long as the characteristics of the present invention are not impaired.

It is a block diagram which shows the electric power drive system used for an electric vehicle (EV) and a hybrid electric vehicle (EHV). It is a conceptual diagram which shows charge amount RSOC and true charge amount SOC. It is a figure which shows the correlation of the open circuit voltage E and charge amount SOC. It is a flowchart for demonstrating the calculation method of virtual internal resistance degradation coefficient (gamma) 1- (gamma) 5. It is a figure explaining storage parts CL1-CL3, CH1-CH4. It is a figure which shows IV characteristic straight lines L1-L5. It is a figure which shows the correlation with an electric current area | region and virtual internal resistance initial value. It is a figure which shows the correlation with an electric current area | region and virtual internal resistance degradation coefficient (gamma). It is a flowchart which shows the calculation procedure of the open circuit voltage E3 and charge amount. It is a figure explaining the characteristic of the open circuit voltage E3 of 1st Embodiment. It is a figure explaining the open circuit voltage E3 by the conventional open circuit voltage calculating method. It is a flowchart explaining the virtual internal resistance calculation process in 2nd Embodiment. It is a figure which shows IV characteristic straight lines L1-L5 in 2nd Embodiment. It is a figure explaining the current threshold value Ia and Ib of a shift area. It is a figure explaining the calculation method of the open circuit voltage E3 in case the electric current value I is I <= Ia. It is a figure explaining the calculation method of the open circuit voltage E3 in case the electric current value I is Ia <I <= Ib. It is a figure explaining the calculation method of the open circuit voltage E3 in case the electric current value I is Ib <I.

Explanation of symbols

1 Battery 2 Inverter 3 Motor 4 Controller 5 Voltage Sensor 6 Current Sensor 7 Temperature Sensor

Claims (10)

A first step of classifying the discharge current region of the battery into at least two current regions according to the magnitude of the current value;
A second step of performing linear regression on the IV characteristic line of the low current region based on current / voltage sampling data synchronously detected in a low current region lower than a predetermined current value among the plurality of current regions;
The high current region based on the current / voltage sampling data synchronously detected in the high current region higher than the predetermined current value in the plurality of current regions and the current / voltage sampling data synchronously detected in the low current region A third step of linear regression calculation of the IV characteristic line of
A fourth step of calculating a virtual internal resistance in the low current region from the slope of the IV characteristic line in the low current region;
A fifth step of calculating a virtual internal resistance of the high current region from the slope of the IV characteristic straight line of the high current region,
A battery internal resistance calculation method, wherein a virtual internal resistance in a current region corresponding to a discharge current at the time of battery discharge is used as an internal resistance at the time of battery discharge. In the battery internal resistance calculation method according to claim 1,
The fourth step includes
Correcting the virtual internal resistance of the low current region using a temperature degradation coefficient based on the temperature of the battery,
The fifth step includes
A method for calculating an internal resistance of a battery, comprising a step of correcting a virtual internal resistance in the high current region using a temperature degradation coefficient based on a temperature of the battery. In the battery internal resistance calculation method according to claim 1 or 2,
The fourth step includes
Correcting the virtual internal resistance of the low current region using an internal resistance deterioration coefficient based on the deterioration of the battery,
The fifth step includes
A method for calculating an internal resistance of a battery, comprising a step of correcting a virtual internal resistance in the high current region using an internal resistance deterioration coefficient based on the deterioration of the battery. A first step of classifying the discharge current region of the battery into a plurality of current regions according to the magnitude of the current value;
A second step of performing linear regression on an IV characteristic line based on current / voltage sampling data detected synchronously in the current region group for each current region group consisting of a predetermined number of the current regions;
Based on the slopes of the plurality of IV characteristic lines calculated for each current region group, the first virtual internal resistance of the shift region where the IV characteristics of the battery change greatly, the second current in the lower current region than the shift region. A third step of calculating a virtual internal resistance and a third virtual internal resistance in a higher current region than the transformation region,

A battery internal resistance calculation method, wherein a virtual internal resistance in a current region corresponding to a discharge current at the time of battery discharge is used as an internal resistance at the time of battery discharge. In the battery open voltage calculation method for calculating the open voltage based on the terminal voltage and discharge current at the time of battery discharge, and the virtual internal resistance corresponding to the current region of the discharge current,
A first step of classifying the discharge current region of the battery into at least two current regions according to the magnitude of the current value;
A second step of performing linear regression on the IV characteristic line of the low current region based on current / voltage sampling data synchronously detected in a low current region lower than a predetermined current value among the plurality of current regions;
The high current region based on the current / voltage sampling data synchronously detected in the high current region higher than the predetermined current value in the plurality of current regions and the current / voltage sampling data synchronously detected in the low current region A third step of linear regression calculation of the IV characteristic line of
A fourth step of calculating a virtual internal resistance in the low current region from the slope of the IV characteristic line in the low current region;
A fifth step of calculating a virtual internal resistance of the high current region from the slope of the IV characteristic line of the high current region;
And a sixth step of calculating an open-circuit voltage of the battery from the terminal voltage and discharge current during battery discharge and the virtual internal resistance in a current region corresponding to the discharge current during battery discharge. Battery open voltage calculation method. In the battery open circuit voltage calculation method according to claim 5,
Correcting the virtual internal resistance of the low current region and the virtual internal resistance of the high current region,
In the sixth step, the open circuit voltage of the battery is calculated using a virtual internal resistance corrected in a current region corresponding to the discharge current at the time of battery discharge.
Open-circuit voltage calculation method of battery-. The battery open circuit voltage calculation method according to claim 6,
The sixth step includes
A method for calculating an open-circuit voltage of a battery, wherein the virtual internal resistance in the low current region and the virtual internal resistance in the high current region are corrected using a temperature degradation coefficient based on the temperature of the battery. The battery open circuit voltage calculation method according to claim 6,
The sixth step includes
A method for calculating an open-circuit voltage of a battery, wherein the virtual internal resistance in the low current region and the virtual internal resistance in the high current region are corrected using an internal resistance deterioration coefficient based on the deterioration of the battery. In the battery open voltage calculation method for calculating the open voltage based on the terminal voltage and discharge current at the time of battery discharge, and the virtual internal resistance corresponding to the current region of the discharge current,
A first step of classifying the discharge current region of the battery into a plurality of current regions according to the magnitude of the current value;
A second step of performing linear regression on an IV characteristic line based on current / voltage sampling data detected synchronously in the current region group for each current region group consisting of a predetermined number of the current regions;
Based on the slopes of a plurality of IV characteristic lines calculated for each of the current area groups, the boundary current value of the transformation area where the IV characteristics of the battery change greatly, the first virtual internal resistance of the transformation area, the transformation A third step of calculating a second virtual internal resistance in a lower current region than the region and a third virtual internal resistance in a higher current region than the transformation region;
A fourth step of calculating an IV characteristic linear function for each current region group using the boundary current value, the first virtual internal resistance, the second virtual internal resistance, and the third virtual internal resistance;
A battery open-circuit voltage calculation method comprising: a fifth step of calculating a battery open-circuit voltage from the IV characteristic linear function in a current region corresponding to a terminal voltage and a discharge current during battery discharge. The charge amount of the battery is calculated based on the open-circuit voltage-charge amount correlation representing the correlation between the open-circuit voltage of the battery and the charge amount, and the open-circuit voltage calculated by the open-circuit voltage calculation method according to claim 5. A method for calculating a charge amount of a battery, characterized in that:

Images