Battery state estimation device and method of estimating battery state.

(54) Title: BATTERY STATE ESTIMATION DEVICE AND BATTERY STATE ESTIMATION METHOD.

(54) Title: BATTERY STATE ESTIMATION DEVICE AND METHOD OF ESTIMATING BATTERY STATE.

(57) Summary

A battery status estimating device and a method for estimating a battery status are provided where a parameter of a rechargeable battery is identified with high precision. The current and terminal voltage of the rechargeable battery are detected, the terminal voltage of a rechargeable battery is estimated based on a given battery model using the sensed current and measured value of the terminal voltage and the parameter of the rechargeable battery is identified from so that the difference between the measured value of the terminal voltage and the estimated value of the terminal voltage converges to zero. When identifying the parameter of the rechargeable battery, filtering the measured value of the terminal voltage and the estimated value of the terminal voltage were used, respectively applied to it using a slow-pass filter having the common high-frequency breakdown characteristics and the value measured terminal voltage and the estimated value of terminal voltage at which the filter was applied.

(57) Abstract

Provided are a battery State estimation device and a method of estimating a battery State where a parameter of a rechargeable battery is identified with high accuracy. The current and the terminal voltage of the rechargeable battery are detected, the terminal voltage of a rechargeable battery is estimated based on a given battery model using the detected current and the measured value of the terminal voltage, and the parameter of the rechargeable battery is identified such that the difference between the measured value of the terminal voltage and the estimated value of the terminal voltage converges to zero. When identifying the parameter of the rechargeable battery, filtering of the measured value of the terminal voltage and the estimated value of the terminal voltage is applied respectively thereon using a low pass filter having the common high frequency breaking characteristics, and the measured value of the terminal voltage and the estimated valué of the terminal voltage to which the filtering was applied are used.

BATTERY STATUS ESTIMATION DEVICE AND METHOD OF

BATTERY STATUS ESTIMATE

TECHNICAL FIELD

The present invention relates to a battery status estimation device and a battery status estimation method.

PREVIOUS TECHNIQUE

Japanese Patent Unexamined Publication No. 2003-185719 describes a secondary battery control device as discussed below. That is, the secondary battery monitoring device defines a predetermined battery model and a measured current value and a measured terminal voltage value of the secondary battery are converted to a state quantity using a variable state filter that is based on a battery model. Using the state amount, the secondary battery monitoring device estimates the secondary battery terminal voltage based on the battery model. The secondary battery monitoring device then identifies a secondary battery parameter such that a difference between the measured voltage value and the estimated terminal voltage based on the battery model is changed to zero.

SUMMARY OF THE INVENTION

In the Unexamined Japanese Patent Publication

No. 2003-185719, however, a portion of the measured current value and the measured terminal voltage value of the secondary battery is used to estimate the terminal voltage without undergoing filter treatment by a variable-state filter. Herein, the measured current value and the measured terminal voltage value of the secondary battery are those measured by an ammeter or a voltmeter, therefore observation noise is commonly included. Therefore, Japanese Patent Unexamined Publication No. 2003-185719 has such a problem that an influence by observation noise causes insufficient pressure to identify the parameter of the secondary battery.

<td>Is</td><td>a</td><td>objective</td><td>of</td><td>the present invention provide</td><td>a</td>

<td>device</td><td>of</td><td>estimate</td><td>of</td><td>battery status and a method</td><td>of</td>

<td>estimate</td><td>of</td><td>state</td><td>of</td><td>battery they are capable</td><td>of</td>

<td>identify</td><td>the</td><td>parameter</td><td>of</td><td>the secondary battery with</td><td>a</td>

high pressure.

To solve the above problem, one aspect of the present invention includes a battery status estimation device and a battery status estimation method, wherein a current from a secondary battery and a terminal voltage from the secondary battery are detected, then using the measured value of current and the measured value of terminal voltage detected in this way, The terminal voltage of the secondary battery whose terminal voltage is based on a predetermined battery model is estimated, and then a secondary battery parameter is identified such that a difference between the measured terminal voltage value and the estimated voltage value becomes to zero. In the identification of the secondary battery parameter, the measured value of terminal voltage and the estimated value of terminal voltage are subjected to a filter treatment by a slow-pass filter having a common high-frequency switch characteristic and the measured value terminal voltage and the estimated terminal voltage value that is subjected to the filter treatment is used.

ADVANTAGEOUS EFFECTS OF THE INVENTION

In accordance with the aspect of the present invention, to identify the secondary battery parameter, the measured terminal voltage value and the estimated terminal voltage value are subjected to a filter treatment by a slow-pass filter having a characteristic of Common high frequency interruption and terminal voltage measured value and estimated terminal voltage value undergoing filter treatment are used. By this operation, the influence of the observation noise included in the measured current value and the measured terminal voltage value can be effectively removed, as a result, making it possible to identify a parameter of the secondary battery with high precision.

BRIEF DESCRIPTION OF THE DRAWINGS

Figure 1 is a diagram showing the structure of a secondary battery control system according to a first embodiment.

Figure 2 is a functional block diagram of an electronic control unit 30 according to the first embodiment.

Figure 3 is a diagram showing an equivalent circuit model showing a battery model of the secondary battery.

Figure 4 is a block diagram of an adaptive identification system according to the first embodiment.

Figure 5 is a drawing showing the frequency band characteristics of a measured current value I (k), a measured value of voltage V (k), and an estimated value of voltage V^to(k).

Figure 6 is a relative drawing showing a relationship between a cutoff frequency of a Gipf slow pass filter and an influence that may be caused by an identification of a battery parameter that lacks the information necessary to obtain a battery characteristic and by an observation noise.

Figure 7 is a block diagram of an adaptive identification system that is an equivalent transformation of Figure 4.

Figure 8 is a drawing showing an example of an open circuit voltage in relation to the characteristic state of charge of the secondary battery.

Figure 9 is a flow graph showing estimation processes for the battery parameter and state of charge, according to the first modality.

Figure 10 is a drawing showing simulation results of the charge state estimation process, according to the first modality.

Figure 11 is a block diagram of an adaptive identification system according to a second embodiment,

Figure 12 is a drawing showing simulation results of the charge state estimation processes, according to the second modality.

Figure 13 is an example of a block diagram of an adaptive identification system, according to a third embodiment.

Figure 14 is another example of the block diagram of the adaptive identification system, according to the third embodiment.

DESCRIPTION OF MODALITIES

Hereinafter, the embodiments of the present invention will be exhibited based on the drawings.

First modality

A control system shown in Figure 1 is an example where a secondary battery control device according to the present invention is applied to a system that directs a load such as motor by a secondary battery or charges the secondary battery. by an electrical power caused by an engine regeneration or by an electrical power generated by an alternator with a motor as a power source.

The secondary battery 10 has such a structure that a plurality of unit cells are connected in series. As the unit cell included in secondary battery 10, for example, a serial lithium secondary battery such as secondary lithium ion battery originates. As a load 20, for example, a motor and the like originate.

A current sensor 40 is a sensor for detecting a charge-discharge current flowing through the secondary battery 10. A signal detected by a current sensor 40 is sent to an electronic control unit 30. A voltage sensor 50 it is a sensor for detecting a terminal voltage of the secondary battery 10. A signal detected by the voltage sensor 50 is sent out of the electronic control unit 30.

The electronic control unit 30 is a control unit for controlling the secondary battery 10 and a microcomputer that includes a CPU for the operation of a program, a ROM and a RAM that memorizes the results of the program or operation, an electronic circuit and the like.

As shown in Figure 2, the electronic control unit 30 includes a current detector 301, a voltage detector 302, a battery parameter estimator 303, an open circuit voltage estimator 304, and a SOC estimator 305. The battery parameter estimator 303 includes a slow-pass filter operator 3031, a state variable filter operator 3032, and an adaptive identification operator 3033.

The current detector 301 obtains a signal from ammeter 40 in a predetermined period and then, based on the signal from ammeter 40, detects the charge-discharge current flowing through secondary battery 10, thereby obtaining a measured value current I (k). Current detector 301 gives the measured current value

I (k) thus measured to the battery parameter estimator 303.

The voltage detector 302 obtains a signal from a voltmeter 50 in a predetermined period and then, based on the signal from the voltmeter 50, detects the terminal voltage of the secondary battery 10, thereby obtaining a measured value of voltage V (k ). The voltage detector 302 gives the measured value of voltage V (k), to the battery parameter estimator 303.

The battery parameter estimator 303 defines a battery model of the secondary battery 10 and then of the measured current value I (k) detected by the current detector 301 and the measured voltage value V (k) detected by the detector voltage 302, collectively estimates a battery parameter φ^Α(k) of the battery model of the secondary battery 10 through an adaptive digital filter operation.

Hereby, the ^TO added to a right shoulder of φ of φ (k), top right V of V₀(k) and top right S of SOC (k). As shown by the following expression (1), however, these are respectively equivalent to φ^TO(k), V₀^TO(k) and SOC^TO(k). Henceforth, the same goes for V^to(k).

Expression 1 ff (£) = f¿ (Á)

SOC \ k) = SOC (k) (1)

Hereafter, a method of estimating the battery parameter Φ^Α(k) of the secondary battery 10 by the battery parameter estimator 303 will be displayed.

First, the battery model used in accordance with this modality will be displayed. Figure 3 is an equivalent circuit model of the battery model of the secondary battery 10 and the equivalent circuit model shown in Figure 3 is given by the following expression (2).

Expression 2

K (r) ₌ . _{/ (í) + and (()}

C_x · S +1 (2)

Herein, a model input is a current () [A] (a positive value denotes charge while a negative value denotes discharge), a model output denotes a terminal voltage V [V], Ri [Ω] denotes a pure resistance, Ci [F] denotes an electric double layer capacity and V₀[V] denotes an open circuit voltage. In expression (2), s denotes a differential operator. The battery model according to this modality is a reduction model (first - first order) that does not specifically separate a positive electrode from a negative electrode, however, it can accurately and positively show the charge-discharge characteristic of a current battery. As established earlier in accordance with this modality, the structure will be displayed with an order of the battery model established for primary (first order) as an example.

Then Ri's expression, R₂, Ci for the following expression (3) allows the previous expression (2) to be given for the following expression (4).

Expression 3

<img file="MX2012002807A_D0001.tif" />

(3)

Expression 4

<img file="MX2012002807A_D0002.tif" />

(4)

According to this modality, from the battery model shown in the previous expression (4), the battery parameter estimator 303 estimates the battery parameter φ^Λ (k) of the battery model shown in Figure 3 through the adaptive digital filter. Hereinafter, an explanation will be given regarding the method for estimating the battery parameter φ^Λ(k) by the estimator of the battery parameter 303.

First, it is conceived that an open circuit voltage V₀(t) is obtained by integration, from an initial state, a current U (t) multiplied by a variable parameter h. In this case, the open circuit voltage V₀(t) can be given by the following expression (5).

Expression 5

<img file="MX2012002807A_D0003.tif" />

(5)

Then, replacing the previous expression (5) in the previous expression (4) gives the following expression (6) to be led to the next expression 87) after a disposition.

<img file="MX2012002807A_D0004.tif" />

Expression 6 ^ • (Γ₂·5 + 1) ·/«+-·/(0

Expression 7 m =

KT₂-s² + (K + T_}-h) -s + h

T \ s² + s

I (í) (7)

The previous expression (2) and the previous expression (7) correspond respectively to the following expression (8) and the following expression (9) and correspond to those where the order of each of A (s) and B (s) the primary (first order) is established in the following expression (8) and following expression (9).

Expression 8

K (f) =

B (s)

A (s) • 7 «+ F„ (f) (8)

Expression 9

V (t) = s Bis) + H (í) -hs A (s)

ΛΟ (9)

Here A (s) and B (s) are each a polynomial function of s, where A (s) and B (s) have the same order.

Then, introducing a known constant ki (i = l, 2 ,,, n) to the previous expression (7) can give the following expression (10) and expression (11).

Expression 10 f „(t) = fn (t) =

V (t) l (t) (10)

Expression 11

<img file="MX2012002807A_D0005.tif" />

(11)

In the previous expression (11), Ii and boi are each a parameter that includes an unknown parameter (Τχ, T₂, K, h). fu are each a conversion state quantity obtained by subjecting the values ​​I (k), V (k) measured respectively by ammeter 40 and voltmeter 50 to filter treatment by a variable state filter. Since the previous expression (11) is a product sum of these, the previous expression (11) coincides with the following expression 812) which is a standard form of the adaptive digital filter.

Expression 12 γ (ί) = φ^τω (12)

In expression (12) above, φ^τ- [I, b_Oi] „Ω =

Then, based on an algorithm shown in the following expression (13), an identification of the battery parameter or^TO(k) of the battery model is implemented from a conversion state quantity ω (k) according to an adaptive adjustment rule so that there is a difference between an estimated value of voltage V^to (k which is an estimated terminal voltage value of the secondary battery 10 estimated by the previous battery model and the measured voltage value (V (k) which is a current measured value detected by the voltmeter 50 and obtained by the Voltage 302 can change to 0. In this case, in accordance with this mode, both limits plot the augmentation method where a logical defect of an adaptive digital filter by the improved least squares approach that can be used. Herein, the logical flaw means that once the estimated value converges, an accurate estimate cannot be achieved again even when the parameter changes.

Expression 13 tW = fak -1) - · F (fr -1) · ζψ) · e (k) e (k) = V (k) -V (k)

V (k) = £^T(k) -¿ (kl) (B)

The previous expression (13) includes sequential expressions to adaptively calculate the battery parameter φ^Λ(k), and (k) and r (kl) are each an adaptive increase of these and (k) is a scalar increase (error increase) while T (kl) is an inline increase (signal increase). Then, when a quantity of state ς (k) at a point of time k is obtained by the previous expression (13), it is possible to calculate e (k) which is a difference between the estimated voltage value V^to(k) which is the estimated terminal voltage value of the secondary battery 10 estimated from the battery model and the estimated voltage value 10 from the battery model and the measured voltage value V (k) detected by the voltmeter 50 and obtained by voltage detector 302. By converting this e (k) to zero you can sequentially calculate the battery parameter &lt;£^TO(k).

Herein, in accordance with this embodiment, as shown in Figure 2, the battery parameter estimator 303 has the slow-pass filter operator 3031, the state variable filter operator 3032 and the adaptive identification operator 3033 Then, in accordance with this mode, the slow pass filter operator 3031, the state variable filter operator 3032 and the adaptive identification operator 3033 calculate the battery parameter φ^Λ(k) by the method that will be shown later. Hereinafter, the method for calculating the battery parameter is^TO(k) in accordance with this embodiment will be displayed with reference to a block diagram of an adaptive identification system shown in Figure 4.

That is, according to this modality, to calculate the battery parameter φ^Λ(k) According to the above method, the slow-pass filter operator 3031 implements the filter treatment by a Gipf slow-pass filter, the variable-state filter operator 3032 obtains the conversion state amount or (k) , (conversion of state quantities roi (k), GD2 (k), ro₃(k), ra₄(k), ®₅(k)), using the state variable filter, as shown in

Figure 4. That is, according to this modality, in the previous expressions (10) and (11), such as I (t) and V (t), those subjected to the filter treatment using the slow-pass filter are used. Gi_pf and is implemented by the slow-pass filter operator 3031 to remove the observation noise itself. Then, based on the conversion state quantity m (k), the variable state filter operator 3032 calculates the estimated value of voltage V^to(k) which is the estimated terminal voltage value based on the battery model.

Then, as shown in Figure 4, using the conversion state quantity m (k), obtained by the state variable filter operator 3032 as well as the measured current value I (k) and the measured voltage value V (k) from which the observation noise was removed by the slow-pass filter Gi_Pf, the adaptive identification operator 3033 implements an identification of the battery parameter φ (k) (φ_η&gt; φ₂Α φ₃&gt; φ₄&gt; φ₅) of the battery model. That is, according to this modality, in the previous expression (13), as V (k) and V^to (k), those subjected to the filter treatment using the slow-pass filter Gi are used_Pfy is implemented by the 3031 slow-pass filter operator to remove the observation noise.

In this way, the measured current value I (k) and measured voltage value V (k) are subjected to the filter treatment G_ipf, in order to remove the observation noise. For this operation, in the identification of the battery parameter φ^Λ(k) converging to zero the difference e (k) between the estimated value of voltage V^to(k) and the measured value of voltage V (k), an influence by observation noise can be effectively removed. As a result, the estimation precision of the battery parameter φ^Λ(k) can be improved.

Herein, the Gi slow pass filter_pf used in accordance with that embodiment is not specifically limited, however, that given by the following expression 814) may be elevated.

Expression 14

<img file="MX2012002807A_D0006.tif" />

<img file="MX2012002807A_D0007.tif" />

<img file="MX2012002807A_D0008.tif" />

(14)

Herein, as shown in Figure 5, the measured current value 1 (k) and the measured voltage value V (k) each include a frequency band that is necessary to obtain the battery characteristic, for example , a battery characteristic band (0 Hz to fi

Hz) shown in Figure 5 and a frequency band based on observation noise, for example an observation noise band (fi Hz to f₂ Hz) in the example shown in Figure 5. The frequency band based on the observation noise is considered as a higher frequency side than the frequency band necessary to obtain the battery characteristic. Furthermore, Figure 5 shows the characteristic frequency band of the measured value of current I (k) and measured value of voltage V (k). Herein, the characteristic battery frequency band can be measured, for example, by the cole-cole graph and the like. Furthermore, the frequency band based on the observation noise can be measured by FFT (fast Fourier transform) and the like.

Then, as shown in Figure 6, an influence by the lack of information necessary to obtain the battery characteristic and an influence by observation noise are in a marketing relationship between them, with respect to identification performance. of the battery parameter in relation to a cutoff frequency of the slow-pass filter Gi_pf. Therefore, as the cutoff frequency of the Gipf slow pass filter used by the slow pass filter operator 3031 it is preferably greater than or equal to the cutoff frequency of a state variable filter used by the variable filter operator. 3032, more preferably equal to the cutoff frequency of the state variable filter. Because of this, it is possible to extract the frequency band necessary to obtain the battery characteristic while reducing the observation noise enough. As a result, the adaptive identification operator 3033 can improve the identification pressure in identifying the battery parameter or^TO(k) of the battery model. Furthermore, this can be applied in the same way to the method of carrying out the filter treatment by the slow-pass filter Gi_Pf at the estimated voltage value V^to(k) and measured value of voltage V (k) which are each an equivalent transformation of Figure 4 (refer to Figure 7).

Furthermore, according to this embodiment, it is preferable that the slow-pass filter G_lpf having the same characteristic can be used for the filter treatment of the measured value of current I (k), measured value of voltage (V (K) and estimated value of voltage V^TO(k). This can eliminate a phase change, thus increasing the identification accuracy of the drum parameter φ^Λ(k) of the battery model.

Then along with the amount of conversion status ro^TO(k), I read attery parameter or^TO(k) of the secondary battery 10 thus calculated is sent out of the open circuit voltage estimator 304 of the battery parameter estimator 303, as shown in FIG. 2.

The open circuit voltage estimator 304 estimates the open circuit voltage of the secondary battery 10 based on the battery parameter &lt;j&gt;^TO(k) and the conversion state amount ro (k) that is calculated by the battery parameter estimator 303, to thereby calculate the intended value of open circuit voltage V₀^TO(k). Hereinafter, the method for calculating the estimated value of open circuit voltage Vo^TO(k) will be exhibited.

That is, according to this modality, substituting in expression (4) the battery parameter Ó^TO(k) calculated by the previous expression (13) and the conversion state quantity ra (k) calculated by the previous expression (10) calculates the estimated value of open circuit voltage V₀^TO(k).

Herein, the battery parameter or^TO(k) corresponds to the parameters Ii, boi that includes the unknown parameters (Ti, R2, K, h), as previously shown. Therefore, substituting in expression 84) the battery parameter or^TO(k) and the amount of conversion state ro (k) that are calculated by the battery parameter estimator 303 can calculate the estimated value of open circuit voltage V0^TO(k). The open circuit voltage estimator

304 sends the estimated value of the calculated open circuit voltage V out₀^TO(k) to the SOC 305 estimator.

From the estimated value of open voltage V₀^TO(k) calculated by the open circuit voltage estimator 304, the SOC estimator 305 calculates the estimated value state of charge SOC^TO(k) based on a predetermined open circuit voltage in relation to the charge characteristic state of the secondary battery 10. In addition, Figure 8 shows an example of the open circuit voltage in relation to the charge characteristic state of the secondary battery. secondary battery 10. According to this embodiment, the open circuit voltage in relation to the state of charge characteristic of the secondary battery 10 is memorized in advance in the RAM provided in the electronic control unit 30. The open circuit voltage relative to the state of charge characteristic of the secondary battery 10 can be calculated by obtaining the relationship between the open circuit voltage and the state of charge through the experiments and the like implemented for the secondary battery. 10.

Then, the battery parameter estimation processes φ^Λ(k) and the status of the estimated SOC load value^TO(k) according to this modality will be displayed with reference to the flow graph shown in Figure 9. Herein, the processes shown in Figure 9 will be implemented in a constant period, for example, every 100 msec. In the following explanation, I (k) denotes a current value of the current implementation period, that is, the present measured value, while I (kl) denotes a current value of the previous implementation period (a time before), is that is, an initial measured value (a time before). Values ​​other than the current value will be denoted itself. Furthermore, the processes that will be exhibited later are implemented by this electronic control unit 30.

First, in step SI, current detector 301 and voltage detector 302 respectively obtain the measured current value I (k) and measured voltage value V (k) respectively. The measured current value I (k) is sent out of the battery parameter estimator 303.

In step S2, the low-pass filter operator 3031 of the battery parameter estimator 303 subjects the measured current value I (k) and measured voltage value V (k) to the filter treatment using the slow-pass filter Gi_pf, in order to remove the observation noise. Then, according to the above expressions (10) and (11), the variable state filter operator 3032 of the battery parameter estimator 303 submits the measured value of current I (k) and the measured value of voltage V (k ) (with the observation noise removed) to the filter treatment using the state variable filter, to calculate the conversion state amount ü (k).

In step S3, using the conversion state amount o (k) calculated in step S2 and the measured voltage value V (k) with the observation noise removed, the adaptive identification operator 3033 of the battery parameter estimator 303 implements battery parameter identification φ^Α(k) of the battery model according to the previous expression (139. Herein, the estimated value of voltage V^to(k) using to implement the battery parameter identification φ '(k) is one that underwent the filter treatment that is implemented by the 3031 slow-pass filter operator and that uses the Gi slow-pass filter_PF.

In step S4, according to the above expressions (4), (10) and (13), the open circuit voltage estimator 304 calculates the estimated value of open circuit voltage V₀^TO(k) based on battery parameter φ^Α(^ and the conversion state quantity o (k) that is calculated by the battery parameter estimator 303. Then the estimated value of the calculated open circuit voltage V₀^TO(k) is shipped outside the SOC 305 estimator.

In step S5, using the estimated value of open circuit voltage V₀^TO(k) calculated by the open circuit voltage estimator 304, the SOC estimator 305 calculates the estimated value state of charge SOC^TO(k) based on the predetermined open circuit voltage relative to the state of charge characteristic of the secondary battery.

According to this modality, as it was shown before, the battery parameter φ^Α (k) of the battery model of the secondary battery 10 and the state of estimated value of charge SOC^TO(k) of the secondary battery 10 are estimated.

Figure 10 shows results to verify the effects of this modality by a simulation using the battery model. In Figure 10 shown from the top are a profile showing the change in the measured value of current I (k), a profile showing the change in the measured value of voltage V ((k), a profile showing the change in the difference e (k) = V (k) - V^to (k) and a profile showing the change in the estimated value of the SOC charge status. Then, of these, with respect to the difference e (k) and the estimated value of the state of charge SOC, the simulated ones using the measured value of current I (k) and the measured value of voltage V (k) that are subjected to the Filter treatment by the Gipf slow-pass filter are denoted by a solid line, while simulated using the measured current value I (k) and the measured voltage value V (k) that is not subjected to filter treatment by the Gi slow pass filter_Pf are denoted by a dotted line. In addition, regarding the estimated value of the state of charge

SOC, an actual value is denoted by a string line in addition to the previously estimated values.

As shown in Figure 10, when the filter treatment was not implemented by the Gipf slow pass filter, the difference e (k) = V (k) -V^to (k) causes a large carryover, as a result, the estimated value of the SOC state of charge deviates from the actual value. On the other hand, when the measured current value I (k) and the measured voltage value V (k) were subjected to the filter treatment by the Gipf slow-pass filter, the difference e (k) = V (k) - V^to (k) converged to zero. It can be verified that this makes it possible to estimate the battery parameter preferably, as a result, making it possible to estimate the SOC state of charge with high precision.

According to this modality, the estimated value of voltage V^to (k) which is based on the battery model of the secondary battery 10 is calculated using the measured current value I (k) and the measured voltage value V (k). To remove the influence of the measured noise included in the measured current value I (k) and the measured voltage value V (k), the measured current value I (k) and the measured voltage value V (k) are subjected to filter treatment by Gi slow-pass filter_PF. Then, the measured current value I (k) and the measured voltage value V (k) that undergo the filter treatment are used to estimate the battery parameter φ^Α (k) so that the difference e (k) between the measured value of voltage V (k) and the estimated value of voltage V^to (k) converge to zero. By this operation, according to this modality, the influence of the measured noise included in the measured value of current I (k) and the measured value of voltage V (k) can be effectively removed, thus making it possible to easily converge the difference e (k ) between the measured value of voltage V (k) and the estimated value of voltage V^to(k) to zero. Through this, the identification precision of the battery parameter φ^Α(^ can be improved. Also, according to this modality, the identification capacity of the battery parameter φ^Α(k) with high pressure can increase the estimation precision of each estimated value of open circuit voltage V₀^TO(k) and the state of estimated charge

SOC^TO(k). Furthermore, this also applies to the method for carrying out the filter treatment by the Gipf slow-pass filter at the estimated value of voltage V^to (k) and the measured voltage value V (k) that each are an equivalent transformation of Figure 4 (refer to Figure 7).

Furthermore, according to this mode, the cut-off frequency of the pass filter Gi_Pf becomes greater than or equal to the cutoff frequency of the state variable filter or becomes the same as the cutoff frequency of the state variable filter. By this, the observation noise can be selectively reduced without attenuating the information necessary to obtain the battery characteristic, specifically, the observation noise is selected from i) the information necessary to obtain the battery characteristic included in the measured value of current I (k) and the measured value of voltage V (k) and ii) the observation noise. Therefore, the identification precision of the battery parameter D^TO (k) can be further improved. Especially, the cut-off frequency of the slow-pass filter Gi is made_pf equals the cutoff frequency of the state variable filter.

By this, the observation noise can be suppressed to a minimum without attenuating the information necessary to obtain the battery characteristic.

Second Modality

Next, an explanation will be made about a second embodiment of the present invention.

According to the second mode, such as the Gi slow pass filter_pf used for filter treatment by the slow weight filter operator 3031 and the state variable filter used for filter treatment by the state variable filter operator 3032, those that do not include a differentiator (differential operator) are used ). Other structures are substantially the same as those according to the first embodiment.

Referring to Figure 11, the structure of an adaptive identification system according to the second embodiment will be displayed. According to the second mode, such as the Gi slow pass filter_Pf, one that does not have the differentiator, that is, the differential operator is used. Also, the state variable filters used according to the first modality, one that has the differentiator, i.e. s / (s²+ ki.s + k₂) is subjected to an application of a partial fractional break, as shown by the following expression (15). Therefore, the state variable filter that the differentiator has is returned to a form that the differentiator does not have, that is, the differential operator s. Therefore, the adaptive identification system according to the second embodiment has the structure similar to that shown in Figure 11.

Expression 15 s² + k_} -s + k₂ s _ AB (s + k_to) (s-yk_F) (s + k_to) (s + k_and) (15)

According to the second modality, the following effects can be taken in addition to the effects of the first modality.

That is, according to the second modality, such as the Gi slow-pass filter_pf used for filter treatment by the slow-pass filter operator 3031 and the state variable filter used for filter treatment by the state variable filter operator 3032, those not including the differentiator (differential operator s) are use. Because of this, even a short data length can perform precise operation, further reducing the influence of observation noise. Therefore, it is possible to further improve the estimation precision of the battery model battery parameter and the estimation precision of the SOC charge state.

Especially, when a high-functionality CPU having a FPU function cannot be used as a CPU of the electronic control unit 30 due to the reduction of cost and power consumption, it is necessary to implement an operation with integer type variables . Then when the Gi low pass filter_pf o state variable filter each having a differential characteristic is calculated by the variables of integer type, it is necessary to set the resolution thickly because the dynamic range of the variable is very small. Therefore, in such a case, using the Gi slow-pass filter_Pfo state variable filter each having the differentiator can influence the identification performance of the battery parameter φ^Λ(k) even by a small observation noise or they may cause an error to the estimated value of each of the battery parameter and the SOC state of charge. Contrary to this, according to the second mode, such as the Gi slow-pass filter_Pf and the state variable filter, those not included in the differentiator (differential operator s) are used. Thus, even though the high-functionality CPU that has the FPU casting cannot be used, the influence of observation noise is unlikely to be caused by the performance of battery parameter identification φ^Λ(1 &lt;), thus making it possible to improve the estimation precision of the battery parameter and the estimation precision of the SOC state of charge.

Figure 12 shows results to verify the effects of the second modality by a simulation using the battery model. In Figure 12, shown from the top are a profile showing the change of the measured value of current I (k), a profile showing the change of the measured value of voltage V (k) and a profile showing the change of the estimated value of the SOC charge status. Then of these, with respect to the estimated value of the SOC state of charge, a solid line denotes a case where filters that do not include the differentiator (differential operator s) are used as the slow-pass filter Gi_pf and the state variable filter while a dotted line denotes a case where filters that include the differentiator (differential operator s) are used as the slow-pass filter Gi_Pf and the state variable filter. In addition, with respect to the estimated value of the SOC state of charge, a real value is denoted by a chain line in addition to the values ​​estimated before. In addition, Figure 12 shows the simulation results obtained where the operations related to the treatment by the slow-pass filter Gi_pf and the state variable filter are of the integer type and the resolution is largely set.

As shown in Figure 12, when the filter that has the differentiator (differential operator s) is used as the slow-pass filter Gi_Pfy state variable filter, the estimated value of the SOC state of charge released in the middle course of testing. Contrary to this, when the filter does not have the differentiator (differential operator s9 is used as the Gi slow pass filter_Pf and the variable state filter, the battery parameter can preferably be estimated. As a result, it can be verified that the state of the SOC load can be estimated with high precision.

Third Modality

Next, an explanation will be given about a third embodiment of the present invention.

According to the third mode, such as the Gi slow pass filter_Pf used for filter treatment by the slow pass filter operator 3031 of the battery parameter estimator 303 and as the state variable filter used for filter treatment by the state variable filter operator 3032 of the battery estimator battery parameter 303, the primary filters will be used respectively. Other structures are substantially the same as those according to the first modality.

That is, according to the third modality, as shown by the following expression (16), a partial fractional break applies as / (s² + kx-s + k2) and 1 / (s² + kxs' k₂) that each is a variable state filter used according to the first modality.

Expression 16

AB —----- 1-s + k_x-s + k₂ (s + k_to) (s + kp)

CD

------- 1-5 + k_x-s + k₂ (s + k_to) (s + k_b) (16)

Then, in the previous expression (16), with k_to= kc, yk_b= k_p the common roots between the state variable filter and the Gi slow pass filter_Pf are joined and placed together to form a structure where the primary slow-pass filters are connected in parallel. Therefore, the structure of the adaptive identification system can be one shown in Figure 13. Furthermore, with k₃ = 2. V (k₂), designing the Gi slow pass filter_Pf to become a multiple root makes a structure where the first slow pass filters are connected in series. Therefore, the structure of the adaptive identification system can be one shown in Figure 14. In addition, in Figures 13 and 14 qi (k), ς₂ (k), ς₃ (k), q₄(k), ς₅ (k), are conversion status quantities while θι, θ₂, θ_3ζ θ₄ and θ₅ are battery parameters of the battery model.

According to the third modality, the following effects can be taken in addition to the effects of the first modality.

That is, according to the third embodiment, the slow-pass filter operator 3031 and the state variable filter operator 3032 implements the filter treatment using the primary slow-pass filter as the slow-pass filter Gi_Pf and the state variable filter. Therefore, the number of operations required for the filter treatment can be reduced. For example, in the adaptive identification system shown in Figure 4, the number of multiplications is eight and the number of additions and subtractions is four, while the adaptive identification system shown in Figures 13 and 14, the number of multiplications is four and the number of additions-subtractions is two. With this, the operating load of the battery parameter identification can be reduced.

Furthermore, in the above embodiments, the current detector 301 corresponds to a current detection means of the present invention, the voltage detector 302 corresponds to the voltage detector means of the present invention, the slow-pass filter operator 3031 of the battery parameter estimator 303 corresponds to a slow-pass filter operating means of the present invention, the state variable filter operator 3032 of the battery parameter estimator 303 corresponds to a terminal voltage estimating means of the present invention, the adaptive identification operator 30333 of the battery parameter estimator 303 corresponds to a means of identifying the present invention, the open circuit voltage estimator 304 corresponds to an open circuit voltage estimating means of the present invention and the SOC 3 05 estimator corresponds to a state of charge estimating means of the present invention.

As discussed above, although the embodiments of the present invention have been explained, these embodiments are set forth to make the present invention easy and understandable, and therefore are not set forth to limit the present invention. Therefore, each item described in the

<td>modalities includes all</td><td>the</td><td>changes</td><td>of</td><td>design</td><td>or</td>

<td>equivalents included</td><td>in</td><td>the reaching</td><td colspan="2">technician of</td><td>the</td>

<td>present invention.</td><td></td><td></td><td></td><td></td><td></td>

<td>INDUSTRIAL APPLICATION</td><td></td><td></td><td></td><td></td><td></td>

<td>By device</td><td>of</td><td>estimate</td><td>of</td><td>state</td><td>of</td>

<td colspan="2">battery and estimation method</td><td>of State</td><td>of</td><td>drums</td><td>of</td>

<td colspan="2">according to the present invention,</td><td colspan="2">the influence</td><td>of noise</td><td>of</td>

<td colspan="5">measurement included in the measured value of current I (k)</td><td>and</td>

Measured value of voltage V (k) can be effectively removed, thus making it easy to converge the difference e (k) between the measured value of voltage V (k) and the estimated value of voltage V^to (k) to zero. Therefore, the identification precision of the battery parameter φ^Α(k) can be improved. Also, the ability to identify the battery parameter φ^Λ(k) with high precision can increase the estimation precision of each estimated value open circuit voltage V₀^TO(k) and the estimated SOC load value^TO(k). Therefore, the battery status estimation device and battery status estimation method according to the present invention has an industrial application.