JP2015206758A - Device for estimating parameter of equivalent circuit of secondary battery for vehicle - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a device for estimating a parameter of an equivalent circuit model of a secondary battery that is designed to be able to precisely estimate the parameter of the equivalent circuit model while a vehicle is actually running.SOLUTION: An arithmetic unit 8 calculates the parameter of an RC series-parallel circuit in an equivalent circuit model after calculating Warburg impedance Zw. When calculating Warburg impedance Zw, the arithmetic unit 8 calculates the impedance Zw0 of a predetermined frequency f0, calculates the impedance Zw1 of frequencies f1 gradually low from the frequency f0, and, assuming that the impedance Zw1 of a first frequency f1 where the sum of a difference in the real part (Re(Zw1)-Re(Zw0)) and a difference in the imaginary part (Im(Zw1)-Im(Zw0)) of these is greater than or equal to a prescribed value Ze is a curved portion on a Nyquist diagram, calculates diffusion resistance Rand time constant τby using the impedance Zw1 of this frequency f1 (S11-S19).

Description

  The present invention relates to a parameter estimation device for an equivalent circuit of a vehicular secondary battery.

  A secondary battery is mounted on a vehicle such as a hybrid vehicle (HV) or an electric vehicle (EV). The state of charge (SOC) of the vehicle battery is used to calculate the travelable distance and the like, and sequentially estimating the state of charge SOC of the vehicle battery greatly reduces the capacity of the battery pack. It is important to contribute.

  As a method for estimating the state of charge of the vehicular secondary battery, there is a method of estimating the state of charge according to the integrated current value flowing through the vehicular secondary battery. This estimation method is a method of estimating the state of charge by integrating the current values flowing through the current sensor. When this estimation method is applied, errors of the current sensor accumulate and the estimation accuracy is poor.

Further, a regression line is calculated from the detected current and the detected voltage, and an open circuit voltage (OCV, V _OCV ) is calculated from the intersection of the regression line and current 0, and a predetermined ideal open circuit is determined. There is a method for estimating the state of charge SOC according to the calculated open circuit voltage V _OCV based on the relationship between the voltage V _OCV and the state of charge SOC. When this estimation method is applied, it is clear that the state of charge SOC shifts according to the influence of the electrochemical polarization of the secondary battery. Here, the polarization is a phenomenon mainly caused by the bias of the ion concentration distribution in the battery electrolyte, and if the ion concentration distribution is biased, it deviates from the estimated impedance value when no polarization is generated. The calculated value of the open circuit voltage V _OCV is also shifted.

In order to solve this problem, a method for estimating the state of charge SOC using an equivalent circuit model has been studied. Conventionally, in this estimation method, a vehicle secondary battery is represented by an electric circuit model, and parameters of each impedance circuit are estimated using a difference equation or the like according to the energization current and voltage of the vehicle secondary battery, The open circuit voltage V _OCV is calculated by calculating the voltage applied to the impedance circuit and subtracting this calculated voltage from the voltage of the secondary battery.

  If the diffusion resistance caused by the influence of polarization is Zw, the diffusion resistance Zw can theoretically be expressed using an infinite number of RC parallel circuits. However, for example, the time required for the in-vehicle device to calculate the diffusion resistance Zw is limited, and it is desirable to limit the practical use to a finite number of RC parallel circuits. In order to improve the calculation accuracy of the diffusion resistance parameter, it is preferable to increase the number of connections of the RC parallel circuit. However, increasing the number of connections increases the amount of calculation.

With respect to such a problem, in the technique described in Non-Patent Document 1, when the Warburg impedance Zw is expressed by the following equation (1), it is approximated by using a Cowell-type equivalent circuit, and the equation (2) Thus, by replacing with the continued fraction expansion formula, the unknown variables are only the diffusion resistance R _d for the direct current and the time constant τ _d of the diffusion process, so that a model of an infinite order can be handled practically.

E.Kuhn, C.Forgez, P.Lagonotte, and G.Friedrich: Modeling Ni-mH battery using Cauer and Foster structures, Journal of Power Sources, Vol.158, No.2, SI, pp.1490-1497 (2006 )

  In the technique described in Non-Patent Document 1, the nonlinear least square method is used to estimate the parameter. However, if the parameter is estimated using the nonlinear least square method, the calculation becomes complicated, and the equivalent circuit model is obtained during actual running of the vehicle. This method is not appropriate as a method for estimating the parameters in a short time.

  An object of the present invention is to provide a parameter estimation device for an equivalent circuit model of a secondary battery for a vehicle that can calculate parameters of the equivalent circuit model in a short time during actual traveling of the vehicle.

  According to the first aspect of the present invention, since the computing unit estimates the parameters of the RC series-parallel circuit in the equivalent circuit model after calculating the Warburg impedance, it can calculate the impedance of each circuit separately. Thus, even if each impedance is calculated in an actual vehicle environment, the parameters of the equivalent circuit model can be calculated in as short a time as possible.

1 is an electrical configuration diagram schematically illustrating an example of a system block configuration of an apparatus for estimating an equivalent circuit parameter of a vehicle secondary battery in the first embodiment. A diagram schematically showing an example of an equivalent circuit model of a secondary battery Nyquist diagram schematically showing examples of impedance characteristics of secondary batteries A flowchart schematically showing a method for estimating the state of charge Flowchart (part 1) schematically showing an example of Warburg impedance estimation processing Illustration of Warburg impedance estimation process Flowchart schematically showing an example of Warburg impedance estimation processing (part 2) Transient response characteristics showing step response of voltage value Characteristic diagram schematically showing open circuit voltage vs. state of charge Characteristic diagram representing fundamental impedance and actual measurement result on Nyquist diagram in the second embodiment The figure which shows schematically the equivalent circuit model example of a secondary battery in consideration of C component element of open circuit voltage Flowchart (part 1) schematically showing an example of Warburg impedance estimation processing Flowchart schematically showing an example of Warburg impedance estimation processing (part 2)

  Hereinafter, some embodiments of the present invention will be described with reference to the drawings. Elements having the same configuration or similar configuration, the same function or similar function between the embodiments are denoted by the same or similar reference numerals, and description thereof is omitted as necessary.

(First embodiment)
1 to 9 show a first embodiment. FIG. 1 is a schematic block diagram showing an example of the electrical configuration of an equivalent circuit parameter estimating apparatus 1 for a vehicle secondary battery. A load 2 by a main motor is mounted in the vehicle. The load 2 is supplied with power from a cell group of the secondary battery 3. The secondary battery 3 is a lithium ion battery or the like. The voltage value measuring unit 4 is provided for measuring the voltage value of the cell of the secondary battery 3, and the temperature measuring unit 5 is provided for measuring the temperature of the secondary battery 3. A current value measuring unit 7 is provided for measuring the current flowing through the load 2. Further, another load (not shown) is connected to the secondary battery 3. This other load represents a load other than the load 2 by the main motor, and examples thereof include a DCDC converter (switching regulator), an air conditioner compressor, and an inverter.

  In addition, a current value measuring unit 7 is provided for measuring a current flowing through the load 2 (another load as required). The calculation unit 8 is configured using, for example, a microcomputer and functions as an estimation unit. The calculation unit 8 includes, for example, an A / D conversion unit 8 a and a memory 8 b in a microcomputer, voltage information based on the voltage value v (t) measured by the voltage value measurement unit 4, and a temperature measured by the temperature measurement unit 5. Temperature information based on the value T (t) and current information based on the current value i (t) measured by the current value measuring unit 6 are acquired through the A / D conversion unit 8a, and the measured voltage data, measured temperature data, The measured current data is stored in the memory 8b. The calculation unit 8 calculates the parameter values of the equivalent circuit of the secondary battery 3 using at least some or all of these values.

  A method for estimating the state of charge SOC of the secondary battery in the above-described configuration will be described. FIG. 2 shows an equivalent circuit model, and FIG. 3 schematically shows a Nyquist diagram of the impedance of the secondary battery.

The secondary battery 3 can be simulated by an equivalent circuit model 9 shown in FIG. In FIG. 2, V _OCV (OCV: Open Circuit Voltage) indicates an open circuit voltage, and the open circuit voltage V _OCV indicates a potential difference between the electrodes in an electrochemical equilibrium state. Further, since the secondary battery 3 is electrochemically influenced by polarization, the Warburg impedance Zw can be defined as shown in FIG.

  As shown in FIG. 2, when the secondary battery 3 is represented by an equivalent circuit model 9, the equivalent circuit model 9 can be divided into a pure voltage element 10 of the secondary battery 3 and other internal impedance elements. Strictly speaking, there are inductor components and the like, but they are not shown because they are negligible, for example.

  The internal impedance element can be divided into an impedance of an RC series / parallel circuit (corresponding to an RC circuit) 11 indicating a linear element of the secondary battery 3 and a Warburg impedance Zw indicating an influence of polarization, for example, as shown in FIG. In addition, these impedances can be expressed by connecting them in series, for example. For example, among the internal impedance elements shown in FIG. 2, the RC series-parallel circuit 11 is expressed in a form in which a resistor R0 that is a zero-order component and two RC parallel circuits 12 having different time constants are connected in series. Can do.

  A method for estimating the state of charge SOC of the secondary battery 3 in the above configuration will be described with reference to FIG. The flowchart shown in FIG. 4 schematically shows the contents of processing executed by the calculation unit 8 for each cycle T0, and is processing performed in an environment where the estimation device 1 is mounted in the vehicle. Here, the period T0 indicates a sampling interval of the current value i (t) by the current value measuring unit 7, the voltage value v (t) by the voltage value measuring unit 4, and the temperature value T (t) by the temperature measuring unit.

  As shown in FIG. 4, the calculation unit 8 measures the current value i (t) measured by the current value measurement unit 7, the voltage value v (t) measured by the voltage value measurement unit 4, and the temperature measurement unit 5. The obtained temperature value T (t) is acquired and stored in the internal memory 8b (S1). Then, the calculation unit 8 calculates and estimates the Warburg impedance Zw based on the measurement data of the current value i (t), the voltage value v (t), and the temperature value T (t) (S2). In step S2 of FIG. 4, the calculation unit 8 can calculate the Warburg impedance Zw by dividing the measured voltage value v (t) by the current value i (t) within a predetermined frequency region. Details of this processing will be described later.

  Then, the calculation unit 8 calculates, estimates (identifies), and updates parameters of the equivalent circuit model 9 (for example, resistance values of the resistors R0 and R1, and capacitance values of the capacitors C1 and C2) (S3). As this method, a voltage applied to the Warburg impedance Zw is calculated, a residual voltage is calculated by subtracting the calculated voltage from the voltage value v (t), and a difference equation or the like is calculated together with the current value i (t). A method is used in which the remaining impedance is calculated and estimated.

Specifically, in step S3, the calculation unit 8 calculates a voltage value Vw applied to the Warburg impedance Zw according to the Warburg impedance Zw and the current value i (t), and stores the voltage value Vw in the memory 8b. The arithmetic unit 8 subtracts the voltage value Vw applied from the voltage value v (t) in the Warburg impedance Zw, thereby, the total voltage value of the voltage value _{V RC} according to the open circuit voltage _V OCV + RC serial-parallel circuit 11 only The calculation is performed to remove the influence of the voltage value Vw applied to the Warburg impedance Zw, and this value is stored in the memory 8b.

Then, the arithmetic unit 8, a voltage value _{V RC} according to the open circuit voltage _V OCV + RC serial-parallel circuit 11, in response to a current value i (t), the impedance of the RC serial-parallel circuit 11 by using a difference equation like to calculate the Z _RC. Then, the calculation unit 8 calculates the voltage value V _RC applied to the RC series / parallel circuit 11 according to the impedance Z _RC and the current value i (t) of the RC series / parallel circuit 11, and stores the voltage value V _RC in the memory 8b.

And the calculating part 8 calculates the open circuit voltage _VOCV of the secondary battery 3 according to these calculation contents (S4). At this time, the arithmetic unit 8, calculated in step S2, S3, estimated Warburg impedance Zw to such voltage value Vw, reads a voltage value _{V RC} according to the impedance of the RC serial-parallel circuit 11 from the memory 8b, the voltage value v The open circuit voltage V _OCV is calculated by subtracting these voltage values Vw and V _RC from (t).

Then, the calculation unit 8 has an open circuit voltage V _OCV calculated in step S4, a map of the open circuit voltage V _{OCV −charged} state (SOC) calculated in advance and stored in the memory 8b (see FIG. 9), and Then, the state of charge SOC of the secondary battery 3 is estimated according to the information of the temperature value T (t) by the temperature measuring unit 5 (S5). With this flow, the state of charge SOC of the secondary battery 3 can be estimated.

  FIG. 5 is a flowchart showing an example of processing contents for estimating (identifying) the Warburg impedance Zw in step S2 of FIG. In order to estimate the Warburg impedance Zw, it is preferable to obtain a plurality of impedances in a predetermined frequency region along an impedance curve on the Nyquist diagram using frequency analysis. The reason is as follows. In the following formula (1),

ｔａｎｈ√（τ_ｄ×ｓ）は√（τ_ｄ×ｓ）が大きいとき、ｔａｎｈ√（τ_ｄ×ｓ）≒１となる。したがって、ｓが大きい（すなわち周波数が大きい）ときにはｔａｎｈ√（τ_ｄ×ｓ）≒１となる。ｔａｎｈ√（τ_ｄ×ｓ）≒１と近似可能な周波数領域では、Ｚ_ｗ＝Ｒ_ｄ／√（τ_ｄ×ｓ）となる。ｔａｎｈ≒１となる周波数領域Ｆ０の範囲では１／√ｓに比例するため、図３のナイキスト線図上では−４５度の直線に近似できる。図３では縦軸を負としているため横軸より上に４５度の傾きの直線となることに留意する。

tanh√ (τ _d × s) becomes tanh√ (τ _d × s) ≈1 when √ (τ _d × s) is large. Therefore, when s is large (that is, the frequency is large), tanh√ (τ _d × s) ≈1. In a frequency region that can be approximated as tanh√ (τ _d × s) ≈1, Z _w = R _d / √ (τ _d × s). Since it is proportional to 1 / √s in the range of the frequency region F0 where tanh≈1, it can be approximated to a straight line of −45 degrees on the Nyquist diagram of FIG. Note that in FIG. 3, since the vertical axis is negative, the straight line is inclined at 45 degrees above the horizontal axis.

For example, when the time constant τ _d is about several hundreds [s], the frequency corresponding to the straight line portion of the Warburg impedance Zw is several to several tens [mHz]. In the lower frequency range, the tanh term becomes semicircular due to the gradually decreasing effect.

(1) Since unknowns _{R d,} tau _d is two, these unknowns _{R d,} may make more than one equation to determine the tau _d. Then, the unknown numbers R _d and τ _d can be obtained individually. In the frequency region that is a straight line on the Nyquist diagram, tanh≈1, and real part = imaginary part. Therefore, only one equation can be obtained for two variables. Therefore, it is desirable to establish an equation including a frequency region that becomes a curve on the Nyquist diagram. Note that the lower the frequency region, the less likely the frequency components desired to be obtained in the current and voltage waveforms are. For this reason, it is difficult to obtain impedance by frequency analysis processing. Therefore, except for the low frequency region where the desired frequency component is hard to be included, the frequency is gradually lowered from the predetermined frequency, and the curve is plotted on the Nyquist diagram on the condition that the size of the real part and the imaginary part differ by a predetermined amount or more. It is good to judge it as a part.

  In order to realize this process, the calculation unit 8 executes the process shown in FIG. 5 in the in-vehicle mounting environment. First, the calculation unit 8 sets a predetermined frequency f0 as a reference (S11). The predetermined frequency f0 is a frequency obtained by detecting the impedance of the secondary battery 3 from an experiment or simulation in advance, and is predetermined as a frequency that allows the setting of the Warburg impedance Zw. The calculation unit 8 sets this frequency f0 by reading it from the memory 8b in the calculation unit 8. A plurality of candidates may be stored in the memory 8b as the value of the predetermined frequency f0, and the calculation unit 8 may appropriately select the value of the predetermined frequency f0.

  The calculation unit 8 determines whether or not the measured current value i (t) and voltage value v (t) include the component of the frequency f0 (S12), and when the frequency f0 is not included, (S12: NO), it is determined that the frequency analysis process cannot be performed, and the process is terminated. Conversely, when determining that the frequency f0 is included, the calculation unit 8 calculates the impedance Zw0 of the frequency f0 using a frequency analysis method such as wavelet transform processing (S13). Then, the calculation unit 8 subtracts the difference frequency Δf from the frequency f0 to calculate the frequency f1 (S14). This difference frequency Δf is also a frequency value determined in advance from experiments or simulations and stored in the memory 8b. There may be a plurality of candidates for the difference frequency Δf, and the calculation unit 8 may appropriately select the value of the difference frequency Δf.

  In addition, the calculation unit 8 determines whether or not the component of the frequency f1 is included in the measured current value i (t) and voltage value v (t) (S15). It is determined that analysis cannot be performed, and the process is terminated. Conversely, when it is determined that the component of the frequency f1 is included (S15: YES), the calculation unit 8 calculates the impedance Zw1 of the frequency f1 using a frequency analysis method such as wavelet transform processing (S16). ).

  Then, the calculation unit 8 calculates the difference between the real part Re (Zw1) of the impedance Zw1 and the real part Re (Zw0) of the impedance Zw0, and the imaginary part Im (Zw1) of the impedance Zw1 and the imaginary part Im ( Difference from Zw0) is calculated. Then, the calculation unit 8 determines whether or not these sums are equal to or greater than a predetermined value Ze (for example, several tens of μΩ) that is a threshold value (S17). The arithmetic unit 8 determines that the slope of the impedance on the Nyquist diagram is approximately −45 ° between the frequencies f0 and f1 when the condition of step S17 is not satisfied, and the difference frequency Δf again from the frequency f1. Is subtracted (S18), and the processes of steps S15 to S18 are repeated until the condition of step S17 is satisfied.

Computing unit 8, when it is condition is satisfied in step S17 (S17: YES) the real part Re (Zw1) impedance Zw1 of and an imaginary part adopts Im (Zw1), the diffusion resistance _{R d} and the time constant τ _d is calculated (S19). At this time, since there are two unknowns, the diffusion resistance R _d and the time constant τ _d , the equation (1) is divided into two parts by dividing the real part and the imaginary part, and the real part of the impedance Zw1 described above If the two calculated values of Re (Zw1) and imaginary part Im (Zw1) are put into an organized expression, the diffusion resistance R _d and the time constant τ _d can be calculated.

FIG. 6 partially shows a portion related to the Warburg impedance Zw in the Nyquist diagram shown in FIG. As shown in FIG. 6, in the frequency region F0 in which the slope is kept almost constant when the real part and the imaginary part are plotted on the Nyquist diagram, the calculation unit 8 repeats the processing of steps S15 to S18 to calculate the frequency. The first impedance that is gradually lowered and the condition that the plot points of the real part and the imaginary part do not satisfy the inclination becomes almost constant is adopted as Zw1 (S17: NO), and the value of the real part Re (Zw1) of this impedance Zw1 The diffusion resistance R _d and the time constant τ _d are calculated according to the values of the imaginary part Im (Zw1) (S19). Thereby, since the real part Re (Zw1) and the imaginary part Im (Zw1) become two with respect to the unknown 2, the diffusion resistance R _d and the time constant τ _d can be calculated. Then, the arithmetic unit 8, using the diffusion resistance _{R d} and the time constant tau _d, can be calculated Warburg impedance Zw (S20). By calculating along this flow, the Warburg impedance Zw can be calculated.

  FIG. 7 is a flowchart schematically showing the flow of another method for the calculation unit 8 to estimate (identify) the Warburg impedance Zw. The method shown in FIG. 7 is a method that is particularly suitable in the extremely low frequency region, and is a method that can also be applied in the extremely low frequency region that cannot be calculated using the method shown in FIG. Further, the calculation unit 8 may collate by performing both the method shown in FIG. 5 and the method shown in FIG. 7 and judge the validity according to the collation result.

In the method shown in FIG. 7, the calculation unit 8 relaxes the number of data of the current value i (t) and the voltage value v (t) measured by the current value measurement unit 7 and the voltage value measurement unit 4 and stored in the memory 8b. It is determined whether data necessary for obtaining the characteristics is obtained (S21). Here, the relaxation characteristic indicates a characteristic that settles in a certain range (steady state) after the voltage value v (t) changes transiently when the current value i (t) changes rapidly. In this step S21, for example, it determines whether the operation unit 8 has acquired predetermined data number required for calculating the error as a predetermined below DC resistance component _{R0 + R1 + R2 + R d} .

  For example, when the vehicle is traveling normally, a constant current flows through the driving load 2 and the like. However, when the vehicle suddenly decelerates, stops suddenly, the ignition key switch is turned off and the engine stops, the current flowing through the driving load 2 decreases rapidly, and the degree of decrease in the voltage value v (t) increases. Thereafter, the voltage value v (t) settles in a certain level range. On the equivalent circuit model 9, the charge accumulated in the capacitors C1 and C2 and the charge accumulated in the capacitance component of the Warburg impedance Zw are rapidly discharged, and then settled in a certain level range.

  Conversely, for example, when the vehicle is normally stopped, almost no current flows through the driving load 2 or the like. However, when the vehicle engine is started, the current flowing through the driving load 2 rapidly increases, the degree of increase in the voltage value v (t) increases, and then the voltage value v (t) is within a certain level range. Will settle down. On the equivalent circuit model 9, charges are rapidly accumulated in the capacitance components of the capacitors C1 and C2 and the Warburg impedance Zw, and the amount of accumulated charges settles within a substantially constant level range.

FIG. 8 shows the step response of the equivalent circuit model 9 at the time of rapid discharge. As shown in FIG. 8, for example, when the current value i (t) changes from the constant current I0 to almost zero in steps, it takes time for the voltage value v (t) to settle to a constant level range in a steady state. Immediately after the current value i (t) changes stepwise, the voltage value v (t) changes gradually according to the influence of the capacitors C1 and C2 and the capacitance components of the Warburg impedance Zw, but if the time elapses, the DC value Only the ingredients affect it. Thus, (1) tanh√ when the s → 0 in formula _{(τ d × s) ≒ √} (τ d × s) for the next Zw ≒ _{R d,} when enough time passes, this Zw ≒ _{R d} The voltage value v (t) changes by a voltage obtained by multiplying the combined resistance value obtained by adding the resistors R0 + R1 + R2 and the constant current I0. That is, the voltage change in this case is _{(R0 + R1 + R2 + R} d) × I0.

Therefore, acquisition in order to calculate the DC resistance component R0 + R1 + R2 + R _d described above, after the current value i (t) is changed stepwise, a predetermined time or more the (for example, several hundreds of seconds or more) lapse voltage value v (t) It is desirable to do. Therefore, the arithmetic unit 8 determines in step S21, whether the number of data acquired every predetermined period T0 is a predetermined number of data than is needed to calculate these DC resistance component _{R0 + R1 + R2 + R d} . This determination condition is satisfied when the above-described operation situation occurs in which the discharge characteristics (or charge characteristics) of the capacitors C1 and C2 and the capacitance components of the Warburg impedance Zw are sufficiently settled within a predetermined level range.

Computing unit 8, if the number of acquired data has reached the predetermined number of data required for calculating the DC resistance component _{R0 + R1 + R2 + R d} (S21: NO), determines that the frequency analysis impossible at present, acquires data number predetermined data The processing is repeated until the number reaches, but if the required number of data is reached, it is determined that the frequency analysis is possible.

  Then, the calculation unit 8 first sets a reference frequency f0 (S22). The predetermined frequency f0 is a frequency obtained by measuring the impedance of the secondary battery 3 in advance, and is set in advance to a predetermined frequency within a frequency region in which the Warburg impedance Zw is settable. The calculation unit 8 determines this frequency f0 by reading it from the memory 8b.

  The computing unit 8 determines whether or not the component of the frequency f0 is included in the measured current value i (t) and voltage value v (t) (S23), and if the frequency f0 is not included, (S23: NO), it is determined that the frequency analysis cannot be performed, and the process is terminated. Conversely, if the calculation unit 8 determines that the frequency f0 is included (S23: YES), it determines that the frequency analysis is possible. Therefore, the calculation unit 8 calculates the impedance Zw0 of the frequency f0 using a frequency analysis method such as wavelet transform processing (S24).

  Assuming that the Warburg impedance Zw is on a straight line with an inclination of approximately −45 degrees on the Nyquist diagram, “real part Re (Zw0) −imaginary part Im (Zw0)” is the origin of the Warburg impedance Zw. Become. Therefore, the calculation unit 8 calculates the real part Re (Zw0) −the imaginary part Im (Zw0), whereby the origin resistance component of the Warburg impedance Zw can be calculated (S25). Here, considering the equivalent circuit model 9, the origin resistance component of the Warburg impedance Zw is R0 + R1 + R2.

Next, the arithmetic unit 8, in accordance with the obtained data so far, and calculates the voltage measured voltage change until the regulated current value i synthesis DC resistance component by dividing by (t) R0 + R1 + R2 + R d (S26). As shown in FIG. 8, it calculates the combined resistance component R0 + R1 + R2 + _{R d} from the discharge characteristic of the capacitance component of the capacitors C1, C2 and Warburg impedance Zw (or charging characteristics).

Then, the arithmetic unit 8, from the calculated combined resistance component R0 + R1 + R2 + _{R d} by subtracting the origin resistance component R0 + R1 + R2 of the Warburg impedance Zw, calculates the diffusion resistance _{R d} (S27). And the calculating part 8 substitutes this calculated diffused resistance _Rd to (1) Formula. Here, it should be noted that since the frequency f0 is a straight line portion on the Nyquist diagram, it can be replaced with tanh√ (τ _d × s) ≈1. And the calculating part 8 calculates time constant (tau) _d using the impedance Zw0 calculated | required by the frequency analysis method (S28). Then, the calculation unit 8 calculates the Warburg impedance Zw using the calculated diffusion resistance R _d and time constant τ _d (S29).

<Summary>
According to the present embodiment, the calculation unit 8 calculates the parameters of the RC series-parallel circuit 11 in the equivalent circuit models 9 and 19 after calculating the component of the Warburg impedance Zw (S2) (S4). The impedance of each circuit can be calculated separately, and the parameter (impedance) of the equivalent circuit model 9 can be calculated in as short a time as possible even if the parameter is calculated in an actual vehicle environment.

The Warburg impedance Zw can be divided into a straight line portion and a curved portion (generally a semicircular portion) having an inclination ≈−45 degrees on the Nyquist diagram, and the Warburg impedance Zw is expressed by the above-described equation (1). At this time, there are two unknown variables: a diffusion resistance R _d and a time constant τ _d . When the impedance of a predetermined frequency is obtained using the frequency analysis method, the real part size and the imaginary part size, that is, two pieces of information can be obtained from this information. However, since the real part and the imaginary part are in a proportional relationship in the straight line part, only one unknown is substantially obtained. Here, if the impedance on the curved portion of the Nyquist diagram can be obtained, two equations can be established for two unknowns, and an unknown variable can be obtained.

Therefore, in the present embodiment, the calculation unit 8 calculates the impedance Zw1 of the frequency f1 on the curved portion of the Nyquist diagram, and uses the real part Re (Zw1) and the imaginary part Im (Zw1) of the impedance Zw1. In the equation (1), two variables of the diffusion resistance R _d and the time constant τ _d are calculated. If a parameter is calculated from the difference equation using a least square method or the like using the conventional technique, it may fall into a local optimum solution. However, according to this embodiment, the Warburg impedance Zw is calculated without falling into the local optimum solution. it can.

In addition, the calculation unit 8 calculates an impedance Zw0 of a predetermined frequency f0, calculates an impedance Zw1 of a frequency f1 that gradually decreases from the frequency f0, and calculates a difference between these real parts (Re (Zw1) −Re ( Zw0)) and the difference between the imaginary part (Im (Zw1) −Im (Zw0)) is considered to be the impedance Zw1 of the first frequency f1 at which the sum is equal to or greater than the predetermined value Ze as a curve portion on the Nyquist diagram. The diffusion resistance R _d and the time constant τ _d are calculated using the impedance Zw1.

As a result, the “curve portion on the Nyquist diagram” can be determined in the highest frequency range. For example, when the impedance Zw1 is calculated in the low frequency region, a large amount of data with a small fluctuation component of the voltage value v (t) and the current value i (t) is required, and is extremely low when voltage fluctuation and current fluctuation increase in an actual vehicle environment. It may be difficult to obtain the impedance Zw1 in the frequency domain. In the present embodiment, since the so can calculate the impedance Zw1 in as much as possible the high frequency range, it is possible to easily calculate the impedance Zw1 using frequency analysis techniques, the calculation of the thus diffused resistor R _d and the time constant tau _d Accuracy can be increased.

In addition, since the calculation unit 8 calculates the diffusion resistance R _d and the time constant τ _d only by the real part Re (Zw1) and the imaginary part Im (Zw1) of the impedance Zw1 of the frequency f1, The diffusion resistance R _d and the time constant τ _d can be calculated by the minimum calculation process of two equations, and the calculation efficiency can be improved.

Further, in the method shown in FIG. 7, the calculation unit 8 calculates the diffusion resistance R _d in the equation (1) based on the relaxation characteristics of the measured voltage data by the voltage value measurement unit 4, and calculates the calculated diffusion resistance R _d and Nyquist The time constant τ _d is calculated using the impedance Zw0 of the predetermined frequency f0 that is a straight line portion on the diagram, and the Warburg impedance is calculated using the calculated diffusion resistance R _d and the time constant τ _d . In particular, it is suitable for estimating the state of charge SOC of the secondary battery 3 in an actual vehicle environment.

At this time, the arithmetic unit 8, the combined resistance component R0 + R1 + R2 + _{R d} is calculated by using the relaxation characteristics of the voltage, and calculates the diffusion resistance _{R d} by subtracting the origin resistance component R0 + R1 + R2. Even with this method, the Warburg impedance Zw can be calculated with as little calculation as possible and without falling into a local optimal solution. In particular, it is suitable for estimating the state of charge SOC of the secondary battery 3 in an actual vehicle environment. Thereby, the state of charge SOC of the secondary battery 3 can be accurately estimated.

  For example, the Warburg impedance Zw calculated by the method shown in FIG. 7 may be collated with the Warburg impedance Zw calculated by the method shown in FIG. 5, or an average value of these may be output. For example, in the method shown in FIG. 5, when it is determined NO in step S15 and calculation is impossible, for example, the Warburg impedance Zw may be calculated using the method shown in FIG. That is, the Warburg impedance Zw may be calculated by combining the method shown in FIG. 5 and the method shown in FIG.

(Second Embodiment)
10 to 13 show a second embodiment. In the second embodiment, a mode in which the C component element of the open circuit voltage V _OCV is taken into consideration is shown.
In consideration of the portion of the equivalent circuit model 9 excluding the open circuit voltage V _OCV , the fundamental impedance characteristic on the Nyquist diagram is as shown by the broken line characteristic FA in FIG. However, when the inventors substantially repeat the measurement, it is obtained as shown by the solid line characteristic FB in FIG.

As shown in the state of charge state SOC-open circuit voltage V _{OCV in} FIG. 9, the open circuit voltage V _OCV is not a constant voltage, but changes according to a change in the state of charge SOC of the secondary battery 3. It turns out. Considering this influence, it is estimated that the characteristic FB is obtained as shown by the solid line in FIG. 10 that the capacitor circuit is substantially joined in series. That is, the solid line characteristic FB in FIG. 10 is considered that the C component of the open circuit voltage V _OCV is superimposed on the impedance of the broken line characteristic FA in FIG.

As shown in FIG. 9, the usable area RA is allocated in advance as the state of charge SOC of the secondary battery 3. In the usable area RA shown in FIG. 9, the dependence of the state of charge SOC on the open circuit voltage V _OCV is approximately proportional and can be linearly approximated. Therefore, the dependence of the open circuit voltage V _{OCV on} the state of charge SOC can be assumed to be a capacity component due to the capacitor, and the form of an equivalent circuit model that equivalently represents the secondary battery 3 by removing this capacity component. The circuit equivalence of can be improved. Thereby, the estimation accuracy of the parameters of these equivalent circuit models can be increased, and as a result, the estimation accuracy of the state of charge SOC of the secondary battery 3 can be improved.

FIG. 11 shows a new equivalent circuit model 19 replaced with the equivalent circuit model 9 of FIG. 2 in consideration of this influence. As shown in FIG. 11, the equivalent circuit 20 of the open circuit voltage V _OCV can be divided into a pure voltage element 20a and a C component element 20b. Assuming that the voltage applied to the C component element 20b is a constant voltage + V _C , the capacitance value of the C component element 20b of the open circuit voltage V _OCV is C = Q / V from the portion of the SOC-V _OCV characteristic of FIG. It can be calculated in advance according to the relational expression.

Here, the charge Q [C] is obtained by assuming that the discharge capacity [Ah] of the secondary battery 3 is in the charged state SOC = 100%, and (linear variation possible area) (SOC change amount) × (discharge capacity) × 3600. It can be calculated. Then, the capacitance value of the C component element 20b can be calculated by dividing by the amount of change in the open circuit voltage V _OCV . The reason why 3600 is applied is because unit conversion [Ah] → [A · s] = [C]. The calculation unit 8 subtracts the impedance of the C component element 20b calculated in this way from the impedance characteristic FB calculated from the voltage value v (t) and the current value i (t), so that the equivalent circuit model 19 The parameter estimation accuracy can be increased.

  For example, an example applied to the method shown in FIG. 5 of the first embodiment is schematically shown in FIG. In the description with reference to FIG. 12, only the parts different from FIG. 5 will be described. The calculation unit 8 reads the capacitance value Ca of the C component element 20b obtained in advance as described above from the memory 8b (S10a). Then, the calculation unit 8 calculates the impedance Zw0 of the frequency f0 in step S13, and calculates the impedance Zw1 of the frequency f1 in step S16. The calculation unit 8 subtracts the capacitive impedance component 1 / (j · ω0 · Ca) of the C component element 20b from the impedance Zw0 calculated in step S13, and calculates the C component element from the impedance Zw1 calculated in step S16. The capacitive impedance component 1 / (j · ω1 · Ca) of 20b is subtracted (S16a). Here, ω0 is 2 × π × f0, and ω1 is 2 × π × f1.

Then, the estimation accuracy of the diffusion resistance R _d and the time constant τ _d calculated by the calculation unit 8 in step S19 can be increased, and the estimation accuracy of the Warburg impedance Zw calculated in step S20 can be increased. Thereby, the parameter estimation accuracy of the equivalent circuit model 19 can be increased.

Further, for example, FIG. 13 schematically shows an example applied to the method shown in FIG. 7 of the first embodiment. In the description shown in FIG. 13, only portions different from those in FIG. 7 will be described. The calculation unit 8 reads the capacitance value Ca of the C component element 20b obtained in advance as described above from the memory 8b (S20a). The computing unit 8 calculates the impedance Zw0 of the frequency f0 in step S24, and subtracts the capacitive impedance component 1 / (j · ω0 · Ca) of the C component element 20b from the impedance Zw0 calculated in step S24. (S24a). Then, the estimation accuracy of the diffusion resistance R _d and the time constant τ _d calculated by the calculation unit 8 in steps S28 and S29 can be increased, and the estimation accuracy of the Warburg impedance Zw can be increased. Thereby, the parameter estimation accuracy of the equivalent circuit model 19 can be increased.

<Summary>
According to the present embodiment, the equivalent circuit model 19 is set based on the C component element 20b calculated according to the open circuit voltage V _{OCV -charge} state SOC characteristic. Therefore, the calculation unit 8 subtracts the capacitive impedance component of the C component element 20b from the calculated impedances Zw0, Zw1 (FIG. 12), and Zw0 (FIG. 13), respectively, thereby estimating the estimation accuracy of the Warburg impedance Zw. And the parameter estimation accuracy of the equivalent circuit model 19 can be improved.

(Other embodiments)
The present invention is not limited to the above-described embodiment. For example, the following modifications or expansions are possible.
The RC circuit is configured using the RC series parallel circuit 11 and the RC parallel circuit 12, but is not limited to this circuit form. For example, an equivalent circuit model to which other capacitor components or inductor components are added may be used.

  In addition, the code | symbol with the parenthesis attached | subjected to the claim attaches | subjects the code | symbol corresponding to the component of this-application specification, and gives an example of the component. Therefore, the invention according to the present application is not limited to the elements indicated by the reference numerals attached to the constituent elements of the claims, and can be variously expanded in terms of the claims or the equivalents thereof.

  In the drawings, 1 is a parameter estimation device for an equivalent circuit of a secondary battery, 3 is a secondary battery, 4 is a voltage value measurement unit, 7 is a current value measurement unit, 8 is a calculation unit (estimation unit, first to fourth calculations). Means), 9, 19 are equivalent circuit models, 11 is an RC series / parallel circuit (RC circuit), 12 is an RC parallel circuit (RC circuit), and Zw is Warburg impedance.

Claims (7)

A current value measuring unit (7) for measuring a current flowing in a secondary battery (3) mounted on a vehicle such as a hybrid vehicle or an electric vehicle;
A voltage value measuring unit (4) for measuring a voltage applied to the secondary battery (3);
The equivalent circuit model (9, 19) is obtained by using a circuit in which the secondary battery (3) is connected in series with the RC circuit (11, 12) and the Warburg impedance (Zw) generated in accordance with the influence of the polarization of the secondary battery. In the replaced state, the estimation unit (8) that estimates the parameters of the equivalent circuit model (9, 19) of the secondary battery according to the measurement current data of the current value measurement unit and the measurement voltage data of the voltage value measurement unit ) And
The estimation unit (8)
First calculation means (8) for calculating the Warburg impedance (Zw) in the equivalent circuit model (9, 19);
And second calculation means (8) for calculating parameters of the RC circuit (11, 12) in the equivalent circuit model (9) after the Warburg impedance is calculated by the first calculation means. A parameter estimation device for an equivalent circuit of a vehicle secondary battery. The equivalent circuit parameter estimation apparatus for a vehicle secondary battery according to claim 1,
The estimation unit (8)
Third calculation means (8) for calculating the impedance (Zw1) of the frequency of the curved portion of the Nyquist diagram on the impedance characteristics of the secondary battery (3);
Using the real part (Re (Zw1)) and imaginary part (Im (Zw1)) of the impedance (Zw1) calculated by the third calculation means, the diffusion resistance (R _d ) in the following formula (1) and Fourth calculating means (8) for calculating two variables of time constant (τ _d ),
The first calculation means (8) calculates the Warburg impedance using the diffusion resistance (R _d ) and the time constant (τ _d ) calculated by the fourth calculation means. A parameter estimation device for an equivalent circuit of a secondary battery.

In the vehicle secondary battery equivalent circuit parameter estimation device according to claim 2,
The third calculating means (8) of the estimating unit calculates an impedance (Zw0) of a predetermined frequency, calculates an impedance (Zw1) of a gradually lower frequency from the frequency, and calculates a difference between these real parts ( Re (Zw1) -Re (Zw0)) and the difference between the imaginary part (Im (Zw1) -Im (Zw0)) is the Nyquist line at the first frequency impedance (Zw1) at which the sum is equal to or greater than a predetermined value (Ze). A parameter estimation device for an equivalent circuit of a vehicular secondary battery, characterized in that the impedance (Zw1) at this frequency is taken as a calculation result, which is regarded as the impedance of the curved portion in the figure. In the parameter estimation apparatus of the equivalent circuit of the secondary battery for vehicles of Claim 2 or 3,
The third calculation means (8) of the estimation unit (8) includes a real part (Re (Zw1)) and an imaginary part (Im (Zw1)) of the impedance (Zw1) of the frequency of the curve part on the Nyquist diagram. The diffusion resistance (R _d ) and the time constant (τ _d ) are calculated only by means of the parameter estimation device for an equivalent circuit of a vehicular secondary battery. The equivalent circuit parameter estimation apparatus for a vehicle secondary battery according to claim 1,
The first calculation means (8) calculates a diffusion resistance (R _d ) in the following equation (1) based on the relaxation characteristics of the measured voltage data by the voltage value measurement unit (4), and the calculated diffusion resistance A time constant (τ _d ) is calculated using (R _d ) and an impedance (Zw 0) of a predetermined frequency (f 0) that is a linear portion on the Nyquist diagram, and the calculated diffusion resistance (R _d ) and the A parameter estimation device for an equivalent circuit of a vehicular secondary battery, wherein the Warburg impedance is calculated using a time constant (τ _d ).

In the vehicle secondary battery equivalent circuit parameter estimation apparatus according to claim 5,
The first calculating means (8) calculates a combined resistance component (R0 + R1 + R2 + R _d ) based on the relaxation characteristics of the measured voltage data by the voltage value measuring unit (4), and an origin resistance component (R0 + R1 + R2) of the Warburg impedance (Zw). A parameter estimation device for an equivalent circuit of a vehicular secondary battery, wherein the diffusion resistance (R _d ) is calculated by subtracting. The parameter estimation device for an equivalent circuit of the vehicle secondary battery according to any one of claims 1 to 6,
The equivalent circuit model (19) is set based on a C component element (20b) calculated according to a predetermined open circuit voltage (V _OCV ) _-charge state (SOC) characteristic,
The estimation unit (8) subtracts the capacitive impedance component of the C component element (20b) from the calculated impedance (Zw1 or Zw0), and estimates the parameters of the equivalent circuit of the vehicular secondary battery. apparatus.

Images