JP2012159414A - Battery status estimation device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a battery status estimation device capable of enhancing calculation accuracy of the internal state of a secondary battery including a nonlinear region thereof.SOLUTION: A battery status estimation device includes: current detection means for detecting a current of a secondary battery as a current measuring value; voltage detection means for detecting a terminal voltage of the secondary battery as a voltage measurement value; terminal voltage estimation means that defines a battery model of the secondary battery, and calculates a conversion state quantity by a state quantity conversion of the current measuring value and the voltage measurement value by using a state variable filter based on the battery model in order to estimate a terminal voltage of the secondary battery based on the battery model as a voltage estimation value, from the conversion state quantity; identification means for identifying a parameter of the secondary battery so that a difference between the voltage measurement value and the voltage estimation value converges to zero; and nonlinear region detection means for detecting a presence ratio of a nonlinear region in characteristics between current and voltage in the secondary battery. The terminal voltage estimation means sets a cutoff frequency of the state variable filter according to the presence ratio.

Description

  The present invention relates to a battery state estimation device for estimating a state inside a secondary battery.

  As a control device for a secondary battery, a predetermined battery model is defined, and the measured values of the secondary battery current and terminal voltage are converted into state quantities using a state variable filter based on the battery model. A control device that estimates the terminal voltage of the secondary battery based on the model and identifies the parameters of the secondary battery so that the difference between the voltage measurement value and the terminal voltage estimated based on the battery model converges to zero. It is known (see Patent Document 1).

JP 2003-185719 A

  However, the conventional control device is a device that estimates the internal state of the secondary battery using the input detection value based on the linear region of the current-voltage characteristic of the secondary battery. Therefore, when the internal state of the secondary battery including the nonlinear region is estimated by the conventional control device, the conventional control device is affected by the nonlinear region and has a problem that the calculation accuracy is lowered.

  The problem to be solved by the present invention is to provide a battery state estimation device that can improve the calculation accuracy of the internal state of a secondary battery including a nonlinear region.

  The present invention detects a current and a terminal voltage of a secondary battery, estimates a terminal voltage of a secondary battery based on a predetermined battery model using a measured value of the detected current and voltage, and a state variable filter. In the battery state estimation device that identifies the parameters of the secondary battery so that the difference between the measured value of the terminal value and the estimated value of the terminal voltage converges to zero, according to the existence ratio of the non-linear region of the secondary battery, The above problem is solved by setting the cutoff frequency of the state variable filter.

  According to the present invention, in the state variable filter, the conversion state quantity is calculated and the parameter is identified using the signal that becomes the linear region by attenuating the signal in the nonlinear region. Can be increased.

It is a figure which shows the structure of the battery state estimation apparatus which concerns on embodiment of invention. It is a block diagram of the electronic control unit of FIG. It is a figure which shows the equivalent circuit model which shows the battery model of the secondary battery of FIG. 2 is a graph showing current-voltage characteristics of the secondary battery of FIG. 1. It is a flowchart which shows the control procedure of the battery state estimation apparatus of FIG. It is a graph which shows the current-voltage characteristic (battery temperature-25 degree) of a secondary battery. It is a graph which shows the current-voltage characteristic (battery temperature-0 degree | times) of a secondary battery. It is a graph which shows the current-voltage characteristic (battery temperature-30 degree | times) of a secondary battery. It is a graph which shows the current-voltage characteristic (before deterioration) of a secondary battery. It is a graph which shows the current-voltage characteristic (after deterioration) of a secondary battery. It is a block diagram of the electronic control unit 30 of the battery state estimation apparatus which concerns on other embodiment of invention. It is a figure which shows the simulation result of the estimation process of the charging rate in Example 1. FIG. It is a figure which shows the simulation result of the estimation process of the charging rate in Example 1. FIG. It is a figure which shows the simulation result of the estimation process of the charging rate in Example 2. FIG. It is a figure which shows the simulation result of the estimation process of the charging rate in Example 2. FIG. It is a figure which shows the simulation result of the estimation process of the charging rate in Example 2. FIG. It is a figure which shows the simulation result of the estimation process of the charging rate in Example 2. FIG. It is a figure which shows the simulation result of the estimation process of the charging rate in Example 2. FIG.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings.

<< First Embodiment >>
FIG. 1 is a diagram illustrating a configuration of a battery state estimation device for a secondary battery according to the present embodiment. The battery state estimation apparatus shown in FIG. 1 is used in a system for driving a load such as a motor with a secondary battery, or for charging a secondary battery with electric power generated by motor regeneration or electric power generated by an alternator using an engine as a power source. It is the example which applied the control apparatus of the secondary battery which concerns on invention.

  The secondary battery 10 is formed by connecting a plurality of unit batteries in series. Examples of the unit battery constituting the secondary battery 10 include a lithium secondary battery such as a lithium ion secondary battery. An example of the load 20 is a motor.

  The current sensor 40 is a sensor that detects a charging / discharging current flowing through the secondary battery 10, and a signal detected by the current sensor 40 is sent to the electronic control unit 30. The voltage sensor 50 is a sensor that detects the terminal voltage of the secondary battery 10, and a signal detected by the voltage sensor 50 is sent to the electronic control unit 30.

  The temperature sensor 60 is a sensor that detects the battery temperature of the secondary battery 10, and a signal detected by the temperature sensor 60 is sent to the electronic control unit 30. The deterioration sensor 70 is a sensor that detects the degree of deterioration of the secondary battery 10, and a signal detected by the deterioration sensor 70 is sent to the electronic control unit 30. Note that an estimation unit that estimates the degree of deterioration may be used instead of the deterioration sensor.

  The electronic control unit 30 is a control unit for controlling the secondary battery 10 and includes a CPU that calculates a program, a microcomputer that includes a ROM and RAM that stores programs and calculation results, an electronic circuit, and the like. .

  FIG. 2 shows a functional block diagram of the electronic control unit 30. As shown in FIG. 2, the electronic control unit 30 includes a current detector 301, a voltage detector 302, a state variable filter calculator 303, an adaptive identifier 304, a state observer 305, an SOC converter 306, and a nonlinear region detector 308. Is provided.

  The current detection unit 301 acquires a signal from the ammeter 40 at a predetermined cycle, and detects a charge / discharge current flowing through the secondary battery 10 based on the signal from the ammeter 40, thereby measuring a current measurement value I (k). To get. The current detection unit 301 sends the acquired current measurement value I (k) to the state variable filter calculation unit 303.

  The voltage detection unit 302 acquires a signal from the voltmeter 50 at a predetermined cycle, and acquires a voltage measurement value V (k) by detecting a terminal voltage of the secondary battery 10 based on the signal from the voltmeter 50. To do. The voltage detection unit 302 sends the acquired current measurement value V (k) to the state variable filter calculation unit 303.

  The state variable filter calculation unit 303 defines a battery model of the secondary battery 10, and the current measurement value I (k) detected by the current detection unit 301 and the voltage measurement value V (k) detected by the voltage detection unit 302. Then, the state quantity is converted by adaptive digital filter calculation to calculate the converted state quantity ω (k).

  The adaptive identifier 304 collectively estimates the battery parameter φ ^ (k) of the battery model of the secondary battery 10 from the calculation result of the state variable filter calculation unit 303, and if necessary, the battery parameter φ ^ (k ).

Here, “^” attached to the right shoulder in φ ^ (k) indicates that the value is an estimated value. In FIG. 2, “^”, which is an estimated value, is set directly above “φ” of φ (k), directly above “V” of V ₀ (k), and “S” of SOC (k). However, as shown in the following equation (1), this is synonymous with φ ^ (k), V ₀ ^ (k), and SOC ^ (k). The same applies to V ^ (k) below.

以下、状態変数フィルタ演算部３０３による変換状態量ω（ｋ）の演算方法及び、適応同定器３０４による電池パラメータφ＾（ｋ）の推定方法について説明する。

Hereinafter, the calculation method of the conversion state quantity ω (k) by the state variable filter calculation unit 303 and the estimation method of the battery parameter φ ^ (k) by the adaptive identifier 304 will be described.

  First, the “battery model” used in the present embodiment will be described. FIG. 3 is an equivalent circuit model showing a battery model of the secondary battery 10. And the battery model of the secondary battery 10 is defined like Formula (1).

ここで、Ａ（ｓ）、Ｂ（ｓ）の次数を１次とした例で状態変数フィルタの導入を簡単に説明する。図３に示す等価回路モデルは、式（２）の形で表すと、下記の式（３）で表される。

Here, the introduction of the state variable filter will be briefly described with an example in which the orders of A (s) and B (s) are the first order. The equivalent circuit model shown in FIG. 3 is expressed by the following expression (3) when expressed in the form of expression (2).

ここで、モデル入力は電流Ｉ［Ａ］（正値は充電、負値は放電）、モデル出力は端子電圧Ｖ［Ｖ］であり、Ｒ_１〔Ω］は電荷移動抵抗、Ｒ_２［Ω］は純抵抗、Ｃ_１［Ｆ］は電気二重層容量、Ｖ_０［Ｖ］は開路電圧である。また、上記式（２）中、ｓは微分オペレータである。なお、本実施形態に係る電池モデルは、正極、負極を特に分離していないリダクションモデル（１次）であるが、実際の電池の充放電特性を比較的正確に示すことが可能である。このように本実施形態においては、電池モデルの次数を１次にした構成を例として説明する。

Here, the model input is current I [A] (positive value is charging, negative value is discharging), model output is terminal voltage V [V], R ₁ [Ω] is charge transfer resistance, R ₂ [Ω] Is a pure resistance, C ₁ [F] is an electric double layer capacitance, and V ₀ [V] is an open circuit voltage. In the above formula (2), s is a differential operator. Note that the battery model according to the present embodiment is a reduction model (primary) in which the positive electrode and the negative electrode are not particularly separated, but it is possible to show the actual charge / discharge characteristics of the battery relatively accurately. As described above, in the present embodiment, a configuration in which the order of the battery model is first will be described as an example.

_When the representative of R _1, R 2, _{C 1} as shown in Equation (4), the equation (3), and thus represented by the following formula (5).

そして、本実施形態においては、上記式（５）に示される電池モデルから、状態変数フィルタ演算部３０３によって、変換状態量ω（ｋ）を演算し、適応同定器３０４は電池パラメータφ＾（ｋ）の推定を行なう。

In this embodiment, the state variable filter calculation unit 303 calculates the conversion state quantity ω (k) from the battery model shown in the above formula (5), and the adaptive identifier 304 uses the battery parameter φ ^ (k ) Is estimated.

First, assuming that the open circuit voltage V ₀ (t) is obtained by integrating a current I (t) multiplied by a variable parameter h from an initial state, the open circuit voltage V ₀ (t) is expressed by the following formula (6 ).

そして、上記式（５）に、上記式（６）を代入すると、下記式（７）となり、これを整理すると下記式（８）となる。

When the above formula (6) is substituted into the above formula (5), the following formula (7) is obtained, and when this is arranged, the following formula (8) is obtained.

そして、式（８）より、状態空間方程式を適応同定器３０４の設計に便利な非最小実現へ変換する。

Then, from Equation (8), the state space equation is converted into a non-minimum realization that is convenient for the design of the adaptive identifier 304.

  Here, the state variable filter calculation unit 303 of this example sets the cutoff frequency of the state variable filter according to the existence ratio of the non-linear region of the current-voltage characteristic of the secondary battery 10. In the state variable filter calculation unit 303, a cutoff frequency (λa) and a cutoff frequency (λb) are set in advance. λa is a higher frequency than λb. The existence ratio of the non-linear region is specified based on a non-linear region existence ratio detection unit 308 described later. Then, the state variable filter calculation unit 303 of this example performs a state variable filter with a cutoff frequency (λa) and a state variable filter with a cutoff frequency (λb) according to the existence ratio of the nonlinear region of the secondary battery 10. The following calculation is performed properly.

The state variable filter calculation unit 303 includes state variable filters represented by the following expressions (9) and (10), and expressions (11) and (12). Expressions (9) to (12) are obtained by introducing known constants k _i (i = 1, 2,..., N) into Expression (8).

式（９）及び式（１０）は、カットオフ周波数がλａである場合の状態変数フィルタを示し、式（１１）及び式（１２）は、カットオフ周波数がλｂである場合の状態変数フィルタを示す。

Equations (9) and (10) show the state variable filter when the cutoff frequency is λa, and Equations (11) and (12) show the state variable filter when the cutoff frequency is λb. Show.

In the above formulas (10) and (12), I _{i_1} , b _{0i_1} , I _{i_2} , b _{0i_2} are parameters including unknown parameters (T ₁ , T ₂ , K, h), and f _{Vi — 1} , f _{Ii_1} , f _{Vi_2} , and f _{Ii_2} are conversion state quantities obtained by filtering I (k) and V (k), which are values measurable by the ammeter 40 and the voltmeter 50, with a state variable filter. And since the said Formula (10) and Formula (12) are these product-sum formulas, it corresponds with following formula (13) which is a standard form of an adaptive digital filter.

上記式（１３）中、φ_１ ^Ｔ＝［Ｉ_ｉ＿１，ｂ_０ｉ＿１］、ω_１＝［ｆ_Ｖｉ＿１，ｆ_Ｉｉ＿１］ であり、φ_２ ^Ｔ＝［Ｉ_ｉ＿２，ｂ_０ｉ＿２］、ω_２＝［ｆ_Ｖｉ＿２，ｆ_Ｉｉ＿２］ である。

In the above formula ^{_{(13), φ 1 T =}} [I i_1, b 0i_1], ω 1 = [f Vi_1, f Ii_1] _{^{a, φ 2 T = [I i_2}} , b 0i_2], ω 2 = [f Vi_2 , F _{Ii — 2} ].

Then, from the conversion state quantity ω ₁ (k) or ω ₂ (k), a voltage estimated value V ^ (k) that is an estimated value of the terminal voltage of the secondary battery 10 estimated from the battery model described above, The following equation (14) is obtained by the adaptive adjustment law so that the difference between the actual measured value detected by the voltmeter 50 and the voltage measured value V (k) obtained by the voltage detector 302 converges to zero. Based on the algorithm shown, the adaptive identifier 304 identifies the battery parameter φ ^ (k) of the battery model. At this time, in this embodiment, the logical disadvantage of a simple “adaptive digital filter based on the least square method” (once the estimated value converges, accurate estimation cannot be performed again even if the parameter changes thereafter). It is possible to use a “both limit trace gain method” that improves the above.

上記式（１４）は、電池パラメータφ＾（ｋ）を適応的に求める逐次式であり、γ（ｋ）、Γ（ｋ−１）は、共に適応ゲインであり、これらのうち、γ（ｋ）はスカラゲイン（誤差ゲイン）であり、Γ（ｋ−１）は行列ゲイン（信号ゲイン）である。そして、上記式（１３）により、ｋ時点における状態量ζ（ｋ）が得られた際には、電池モデルから推定される二次電池１０の端子電圧の推定値である電圧推定値Ｖ＾（ｋ）と、電圧計５０で検出され電圧検出部３０２により取得された電圧計測値Ｖ（ｋ）との差分であるｅ（ｋ）を求めることができ、このｅ（ｋ）をゼロに収束させることにより、電池パラメータφ＾（ｋ）を逐次的に算出することができる。ここで、λａの状態変数フィルタにより算出された電池パラメータを、φ_１＾（ｋ）、λｂの状態変数フィルタにより算出された電池パラメータを、φ_２＾（ｋ）とする。

The above equation (14) is a sequential equation for adaptively obtaining the battery parameter φ ^ (k), and γ (k) and Γ (k−1) are both adaptive gains, and among these, γ (k ) Is a scalar gain (error gain), and Γ (k−1) is a matrix gain (signal gain). Then, when the state quantity ζ (k) at time point k is obtained by the above equation (13), the estimated voltage value V ^ () that is the estimated value of the terminal voltage of the secondary battery 10 estimated from the battery model. k) and a difference e (k) between the voltage measurement value V (k) detected by the voltmeter 50 and acquired by the voltage detector 302 can be obtained, and this e (k) is converged to zero. Accordingly, the battery parameter φ ^ (k) can be calculated sequentially. Here, the battery parameter calculated by the state variable filter of λa is φ ₁ ^ (k), and the battery parameter calculated by the state variable filter of λb is φ ₂ ^ (k).

Cell parameters _{φ 1 ^ (k), φ} 2 ^ (k) , as described above, the parameter _I i_1 containing unknown parameters _{(T 1, T 2, K} , h), b 0i_1, I i_2, the _{b 0I_2} Each corresponds. Therefore, the state observer 305 uses the battery parameter φ ^ (k) and the converted state quantity ω (k) calculated by the state variable filter calculation unit 303, and substitutes them into the above equation (4). An open circuit voltage estimated value V ₀ ^ (k) can be obtained. The open circuit voltage estimated value V ₀ ^ (k) obtained in this way is sent to the SOC converter 306 by the state observer 305.

The SOC conversion unit 306 converts and calculates the SOC that is the state of charge of the secondary battery 10 based on the open circuit voltage estimated value V ₀ ^ (k). Since there is a predetermined relationship between the open circuit voltage and the SOC, for example, the SOC conversion unit 306 stores a table indicating the relationship between the open circuit voltage and the SOC in advance, and calculates the SOC with reference to the table. Thereby, the internal state of the secondary battery 10 is estimated.

  Next, the non-linear region existence ratio detection unit 308 will be described with reference to FIG. FIG. 4 is a graph showing the current-voltage characteristics of the secondary battery 10.

For example, the non-linear region existence ratio detection unit 308 calculates the open-circuit voltage V _OCV of the secondary battery 10 when the estimation device of this example is activated and the secondary battery 10 is in a no-load state. Alternatively, the non-linear region presence ratio detection unit 308 calculates V _OCV from the SOC calculated by integrating the measurement values of the current detection unit 301. When integrating the measurement values, initialization may be performed periodically so that integration errors are not accumulated. Then, the non-linear region existence ratio detection unit 308 plots the sampling data and the open-circuit voltage V _OCV on the current-voltage characteristics using the sampling data of the predetermined period of the current detection unit 301 and the voltage detection unit 302. Then, an area having a high plot density is extracted. The existence ratio of the non-linear region is determined depending on whether or not the extracted area and the current-voltage characteristics fall within a predetermined region.

Here, the predetermined area is specified by a predetermined width centered on the IV reference straight line (P) with the calculated open-circuit voltage V _OCV as an intercept, as indicated by the hatched portion in FIG. It is an area to be done. For example, the internal resistance in the initial state of the secondary battery 10 is used for the inclination of the IV reference straight line (P).

  For example, when the secondary battery 10 is used in a low-middle current region, the current-voltage characteristics are kept linear, so that the sampling data is within the shaded portion (FIG. 4). Area a). On the other hand, when the secondary battery 10 is used in the current region in the low to middle region (area b in FIG. 4), the current-voltage characteristic has nonlinearity, and the sampling data is the shaded portion. Deviate from.

  When the area extracted by the sampling data is within the shaded area (area a in FIG. 4), the non-linear area existence ratio detection unit 308 maintains the linearity, and the non-linear area existence ratio. Is determined to be lower than a predetermined ratio. On the other hand, when the area extracted by the sampling data is outside the shaded area (area b in FIG. 4), the secondary battery 10 does not maintain linearity, and the existence ratio of the non-linear region is higher than the predetermined ratio. Judged to be high. Then, the nonlinear region existence ratio detection unit 308 transmits a signal based on the determination result to the state variable filter calculation unit 303, and the state variable filter calculation unit 303 determines that the existence ratio is small based on the existence ratio. When the cutoff frequency is λa and the existence ratio is large, the cutoff frequency is λb.

  Next, the control procedure of the battery state estimation apparatus of this example will be described with reference to FIG. FIG. 5 is a flowchart showing a control procedure of the battery state estimation device of this example.

  First, in step S <b> 1, the current measurement value I (k) and the voltage measurement value V (k) are acquired by the current detection unit 301 and the voltage detection unit 302 and sent to the state variable filter calculation unit 303. The

In step S2, the non-linear region existence ratio detection unit 308 detects the non-linear region existence ratio of the battery 12 based on the open circuit voltage V _OCV and the measured value in step S1. Next, in step S3, the state variable filter calculation unit 303 sets the cutoff frequency of the state variable filter to λa or λb according to the existence ratio in step S2.

  In step S4, the current variable value I (k) and the voltage measurement value V (k) are calculated by the state variable filter calculation unit 303 according to the above formulas (9) and (10) or (11) and (12). Then, a filter process using a state variable filter based on the set cutoff frequency is performed, and a conversion state quantity ω (k) is calculated.

  In step S5, using the conversion state quantity ω (k) and the voltage measurement value V (k) calculated in step S4, the adaptive identifier 304 performs battery parameter φ ^ of the battery model according to the above equation (14). Identification of (k) is performed.

In step S6, based on the battery parameter φ ^ (k) and the converted state quantity ω (k) calculated by the state observer 305 by the state variable filter calculation unit 303 and the adaptive identifier 304, the open circuit voltage estimated value V Calculation of ₀ ^ (k) is performed.

In step S7, SOC conversion unit 306 estimates the charging rate based on the predetermined open circuit voltage-SOC characteristic of secondary battery 10 using open circuit voltage estimated value V ₀ ^ (k) in step S5. The value SOC ^ (k) is calculated.

  In the present embodiment, the battery parameter φ ^ (k) and the estimated charge rate SOC ^ (k) of the battery model of the secondary battery 10 are estimated as described above.

  As described above, this example calculates the conversion state quantity using the state variable filter based on the battery model of the secondary battery 10 using the current measurement value I (k) and the voltage measurement value V (k), and the voltage The estimated value V ^ (k) is calculated, and the battery parameter φ ^ (k) of the battery model is set such that the difference e (k) between the measured voltage value V (k) and the estimated voltage value V ^ (k) converges to zero. ) Is set, the cutoff frequency of the state variant filter is set according to the existence ratio of the non-linear region of the secondary battery 10. As a result, the current measurement value I (k) and the voltage measurement value V (k) in the non-linear region are attenuated, and the conversion state quantity is calculated using the measurement value in the linear region efficiently, and the parameter φ ^ (K) can be estimated.

  The present invention also detects the existence ratio of the nonlinear region based on the measured current value I (k) and the measured voltage value V (k), and if the existence ratio is larger than the predetermined existence ratio, the state variant filter is cut. The off frequency is set to a low cutoff frequency (λb). Thereby, the current-voltage characteristics of the secondary battery 10 change over time depending on the environment and use conditions. Therefore, the secondary battery 10 is grasped from the current measurement value I (k) and the voltage measurement value V (k), and the secondary battery 10 is detected by detecting the existence ratio of the non-linear region of the secondary battery 10. When 10 nonlinear regions increase, the cutoff frequency is lowered to improve the identification performance. Thereby, calculation accuracy can be improved.

  In the present example, the cut-off frequency may be set by detecting the existence ratio of the non-linear region of the secondary battery 10 based on the battery temperature of the secondary battery 10. 6A to 6C show an example of the characteristics (IV characteristics) of the voltage between terminals with respect to the current at each temperature, FIG. 6A shows the IV characteristics when the battery temperature is 25 degrees, and FIG. 6B shows the IV characteristics when the battery temperature is 0 degrees. FIG. 6 c shows the IV characteristics when the battery temperature is −30 degrees. As shown in FIGS. 6a to 6c, the IV characteristics correlate with the battery temperature, and the non-linear region existence ratio increases as the battery temperature decreases. 6a to 6c are determined in advance by the characteristics of the battery used for the secondary battery 10. Therefore, in this example, the non-linear region existence ratio detection unit 308 previously records a correlation table between the battery temperature and the non-linear region existence ratio, and specifies the existence ratio based on the temperature detected by the temperature sensor 60. Then, the state variable filter calculation unit 303 determines the cutoff frequency when the non-linear area of the secondary battery 10 increases according to the non-linear area existence ratio specified by the non-linear area existence ratio detection unit 308. To improve identification performance. Thereby, calculation accuracy can be improved.

  In the present example, the cut-off frequency may be set by detecting the existence ratio of the non-linear region of the secondary battery 10 based on the degree of deterioration of the secondary battery 10. 7A and 7B show an example of the characteristics (IV characteristics) of the voltage between the terminals with respect to the current during deterioration, FIG. 7A shows the IV characteristics of the initial battery, and FIG. 7B shows the IV characteristics of the battery after deterioration. As shown in FIGS. 7a and 7b, the IV characteristic has a correlation with the degree of deterioration, and the existence ratio of the non-linear region increases as it deteriorates. The characteristics shown in FIGS. 7 a and 7 b are determined in advance by the characteristics of the battery used for the secondary battery 10. Therefore, in this example, the non-linear region existence ratio detection unit 308 records a correlation table between the deterioration degree and the non-linear area existence ratio in advance, and specifies the existence ratio based on the detected temperature of the deterioration degree sensor 70. Then, the state variable filter calculation unit 303 determines the cutoff frequency when the non-linear area of the secondary battery 10 increases according to the non-linear area existence ratio specified by the non-linear area existence ratio detection unit 308. To improve identification performance. Thereby, calculation accuracy can be improved.

  In this example, the current current measurement value is smaller than the predetermined current value when the non-linear region exists in proportion to the current and voltage measurement values, battery temperature, or battery deterioration level. The cutoff frequency may be set to a high cutoff frequency (λa). Even when the existence ratio of the non-linear region of the secondary battery 10 increases, the linearity of the secondary battery 10 is maintained in the region where the absolute value of the current measurement value is low (small and medium current region). On the other hand, in the region having a high absolute value (high current region), the linearity of the secondary battery 10 is not maintained. Therefore, the state variable filter calculation unit 303 is a case where the existence ratio of the non-linear region is determined to be larger than a predetermined rate, and the small / medium current region is used based on the measurement value of the current detection unit 301. In this case, the cutoff frequency is set to a high cutoff frequency (λa). Thereby, this example can improve calculation accuracy.

  Note that if the magnitude of the current is frequently switched and the cut-off frequency is switched each time, the parameter convergence may be deteriorated. Therefore, for example, the cutoff frequency may be changed based on an average current within a certain time.

  The current sensor 40 of this example corresponds to the “current detection means” of the present invention, the voltage sensor 50 as the “voltage detection means”, the state variable filter calculation unit 303 as the “terminal voltage estimation means”, and the adaptive identifier. Reference numeral 304 corresponds to “identification means”, the state observer 305 corresponds to “open circuit voltage calculation means”, and the non-linear area existence ratio detection unit 308 corresponds to “non-linear area existence ratio detection means”.

<< Second Embodiment >>
FIG. 8 is a block diagram of the electronic control unit 30 of the battery state estimation device according to another embodiment of the invention. In this example, a part of the electronic control unit 30 is different from the first embodiment described above. Since the other configuration is the same as that of the first embodiment described above, the description thereof is incorporated.

  As shown in FIG. 8, the electronic control unit 30 includes a parameter conversion unit 307, and the state variable filter calculation unit 303 includes a first calculation unit 3031 and a second calculation unit 3032. The state variable filter calculation unit 303 calculates the converted state quantity ω (k) using a plurality of state variable filters. The first calculation unit 3031 performs calculation using a state variable filter (first state variable filter) having a cutoff frequency (λa) represented by Expression (9) and Expression (10), and the second calculation unit 3032. Is calculated using a state variable filter (second state variable filter) having a cut-off frequency (λb) represented by Expression (11) and Expression (12).

The first calculation unit 3031 calculates the converted state quantity ω ₁ (k) using the first state variable filter based on the battery model, and the second calculation unit 3032 uses the second state variable filter based on the battery model. Is used to calculate the conversion state quantity ω ₂ (k). The adaptive identifier 304 is a voltage estimation value V ^ (k) from the conversion state quantity ω ₁ (k) and an actual measurement value detected by the voltmeter 50 and acquired by the voltage detection unit 302. Based on the algorithm shown in the following equation (14), the adaptive identifier 304 uses the battery parameter φ ₁ ^ of the battery model by the adaptive adjustment rule so that the difference from the voltage measurement value V (k) converges to zero. (K) is identified.

The parameter conversion unit 307 refers to a preset table or mathematical expression, and uses λa and λb to change the parameter φ ₁ ^ (k) identified by the adaptive identifier 304 to the parameter φ ₂ ^ (k). Convert. Then, the state observer 305 uses the parameter φ ₂ ^ (k) converted by the parameter conversion unit 307 and the converted state quantity ω ₂ (k) calculated by the second calculation unit 3032, so that the open circuit voltage estimated value is obtained. V ₀ ^ (k) is calculated.

  As described above, this example includes the first calculation unit 3031 that calculates using the state variable filter of λa and the second calculation unit 3032 that calculates using the state variable filter of λb. In this example, the adaptive identifier 304 calculates based on the calculation result of the first calculation unit 3031, and the second calculation unit 3032 calculates based on the calculation result of the second calculation unit 3032. When the results calculated using the state variable filter having a common cutoff frequency are output to the adaptive identifier 304 and the state observer 305, respectively, the influence of the frequency characteristics of the state variable filter on the parameter identification performance; The effect on the state quantity estimation performance becomes a trade-off, and there is a possibility that the calculation accuracy cannot be fully exhibited. In this example, different cut-off frequencies are set for each state variable filter so that the input in the nonlinearity region can be attenuated in the state observer 305 without wasting the input frequency band necessary for identification in the adaptive identifier 304. Therefore, the open circuit voltage can be accurately estimated without being affected by the non-linearity of the secondary battery 10.

  In this example, at least one of the cutoff frequency of the first state variable filter or the cutoff frequency of the second state variable filter may be set according to the presence ratio detected by the nonlinear region presence ratio detection unit 308. Good. In this example, at least one of the cutoff frequency of the first state variable filter or the cutoff frequency of the second state variable filter may be set to a low frequency according to the existence ratio. When the ratio of the non-linear region of the secondary battery 10 is small, it is not always necessary to perform the calculation using the filter of λb. Therefore, in this example, the cutoff frequency is increased depending on the state of the secondary battery 10. By adjusting the height, the calculation accuracy can be increased.

  The first calculation unit 3031 of this example corresponds to the “first calculation unit” of the present invention, and the second calculation unit 3032 corresponds to the “second calculation unit” of the present invention.

Example 1
Since the effect of the battery state estimation device according to the first embodiment of the present invention was verified by simulation using a battery model, description will be made with reference to FIGS. 9 to 11, from the top, a profile showing a change in the current measurement value I (k), a profile showing a change in the voltage measurement value V (k), a profile showing a change in the identification parameter (T1), and an SOC error (%). And a profile showing a change in the estimated value of the SOC.

  FIG. 9 shows a simulation result when the current-voltage characteristic is in a linear region, and FIG. 10 shows a simulation result when the current-voltage characteristic is in a non-linear region. FIG. 11 is an enlarged view of a dotted line portion of FIG. 9 to 11, the alternate long and short dash line indicates the true value obtained during the simulation, the broken line indicates the case where the cutoff frequency is λ1, and the solid line indicates the case where the cutoff frequency is λ2. The cutoff frequency shown here has a relationship of λ1> λ2.

  As shown in FIG. 9, in the linear region of the current-voltage characteristics, the convergence performance of the identification parameter (T1) is better and the SOC estimation error is smaller at λ1 than at the cutoff frequency λ2. Therefore, in the linear region, the calculation accuracy is improved by setting the cutoff frequency to λ1.

  As shown in FIGS. 10 and 11, when the IV characteristic reaches the non-linear region, the SOC estimation difference is large at the cut-off frequency λ1 set in the linear region, and the identification parameter (T1) is affected by the non-linearity. ) Sometimes vibrates (see X in FIG. 11). At the cutoff frequency λ2, the vibration of the identification parameter was suppressed. Therefore, in the nonlinear region, the calculation accuracy is improved by setting the cutoff frequency to λ2.

Example 2
Similar to Example 1, the effect of the battery state estimation device according to Embodiment 2 of the present invention was verified, and will be described with reference to FIGS. 12 to 15, from the top, a profile showing a change in the current measurement value I (k), a profile showing a change in the voltage measurement value V (k), a profile showing a change in the identification parameter (T1), and an SOC error (%). ) Indicating the change in the estimated value of the SOC.

  12 and 13 show simulation results when the current-voltage characteristic is in the linear region, and FIGS. 14 and 15 show simulation results when the current-voltage characteristic is in the non-linear region. A one-dot chain line is a true value obtained at the time of simulation. The broken lines in FIGS. 12 and 14 indicate the case where the cutoff frequency λ1 of the state variable filter of the first calculation unit 3031 and the second calculation unit 3032 is set. A solid line indicates a case where the state variable filter cutoff frequency λ1 of the first calculation unit 3031 is set and the state variable filter cutoff frequency λ3 of the second calculation unit 3032 is set. The broken lines in FIGS. 13 and 15 indicate the case where the cutoff frequency λ2 of the state variable filter of the first calculation unit 3031 and the second calculation unit 3032 is set. A solid line indicates a case where the cutoff frequency λ2 of the state variable filter of the first calculation unit 3031 and the cutoff frequency λ3 of the state variable filter of the second calculation unit 3032 are used. The cut-off frequency shown here has a relationship of λ1> λ2> λ3.

  12 and 13, when the current-voltage characteristic is in the linear region, the convergence performance of the identification parameter (T1) is better at λ1 (FIG. 12) than the cutoff frequency λ2 (FIG. 13), and the SOC estimation error is increased. Became smaller. Therefore, in the linear region, the calculation accuracy is improved by setting the cutoff frequency to λ1.

  14 and 15, when the current-voltage characteristic reaches the nonlinear region, the cutoff frequency λ1 (FIG. 14) set in the linear region has a large SOC estimation difference and is identified by the influence of nonlinearity. There was a case where the parameter (T1) became oscillatory (see the Y portion in FIG. 14). At the cutoff frequency λ2 (FIG. 15), the vibration of the identification parameter was suppressed. Therefore, in the non-linear region, the calculation accuracy is improved by setting the cutoff frequency to λ2.

DESCRIPTION OF SYMBOLS 10 ... Secondary battery 20 ... Load 30 ... Electronic control unit 301 ... Current detection part 302 ... Voltage detection part 303 ... State variable filter calculating part 3031 ... 1st calculating part 3032 ... 2nd filter calculating part 304 ... Adaptive identifier 305 ... State observer 306 ... SOC conversion unit 307 ... parameter conversion unit 308 ... nonlinear region existence ratio detection unit 40 ... current sensor 50 ... voltage sensor 60 ... temperature sensor 70 ... degradation degree sensor

Claims (11)

Current detection means for detecting the current of the secondary battery as a current measurement value;
Voltage detection means for detecting the terminal voltage of the secondary battery as a voltage measurement value;
A battery model of the secondary battery is defined, and the current measurement value and the voltage measurement value are converted into a state quantity using a state variable filter based on the battery model to calculate a conversion state quantity, and the conversion state quantity From the terminal voltage estimation means for estimating the terminal voltage of the secondary battery based on the battery model as a voltage estimated value,
Identification means for identifying parameters of the secondary battery so that a difference between the voltage measurement value and the voltage estimation value converges to zero;
A non-linear region detecting means for detecting the existence ratio of the non-linear region among the current-voltage and characteristics of the secondary battery,
The terminal voltage estimation means sets a cutoff frequency of the state variable filter according to the existence ratio, and the battery state estimation device. 2. The battery state estimation device according to claim 1, wherein the terminal voltage estimation unit sets the cutoff frequency lower than a predetermined cutoff frequency when the presence ratio is larger than a predetermined existence ratio. The terminal voltage estimation means includes
If the abundance ratio is less than a predetermined abundance ratio, the cutoff frequency is set to a first cutoff frequency,
3. The battery state according to claim 1, wherein when the existence ratio is larger than a predetermined existence ratio, the cutoff frequency is set to a second cutoff frequency lower than the first cutoff frequency. Estimating device. The terminal voltage estimation means uses a first state variable filter to calculate a conversion state quantity and a second state variable filter different from the first state variable filter to convert the state The battery state estimation apparatus according to claim 1, further comprising a second calculation unit that calculates the amount. Using the state conversion amount and the parameter, further comprising an open circuit voltage calculating means for calculating an open circuit voltage of the secondary battery,
The identification means identifies the parameter based on the conversion state quantity calculated by the first calculation unit,
5. The battery state estimation device according to claim 4, wherein the open circuit voltage calculation means calculates the open circuit voltage based on a conversion state quantity calculated by the second calculation unit. 5. The terminal voltage estimating means sets the cut-off frequency of at least one of the first state variable filter and the second state variable filter according to the existence ratio. Or the battery state estimation apparatus of 5. The terminal voltage estimation means is configured to cut off at least one of a cutoff frequency of the first state variable filter and a cutoff frequency of the second state variable filter when the existence ratio is larger than a predetermined existence ratio. The battery state estimation device according to any one of claims 4 to 6, wherein the frequency is set lower than a predetermined cutoff frequency. The nonlinear area detecting means detects the existence ratio of the nonlinear area based on the current measurement value and the voltage measurement value,
The said terminal voltage estimation means makes the said cut-off frequency lower than a predetermined cut-off frequency, when the said presence ratio is larger than a predetermined presence ratio, The Claim 1 characterized by the above-mentioned. The battery state estimation device described. The non-linear area detecting means detects the existence ratio of the non-linear area based on the temperature of the secondary battery,
The said terminal voltage estimation means makes the said cut-off frequency lower than a predetermined cut-off frequency, when the said presence ratio is larger than a predetermined presence ratio, The Claim 1 characterized by the above-mentioned. The battery state estimation device described. The non-linear area detecting means detects the existence ratio of the non-linear area based on the degree of deterioration of the secondary battery,
The said terminal voltage estimation means makes the said cut-off frequency lower than a predetermined cut-off frequency, when the said presence ratio is larger than a predetermined presence ratio, The Claim 1 characterized by the above-mentioned. The battery state estimation device described. 11. The terminal voltage estimation unit, wherein the cutoff frequency is higher than the predetermined cutoff frequency when the measured current value is smaller than a predetermined current value. The battery state estimation device according to item.

Images