JP2008268161A - Method and device for estimating battery residual capacity and battery power supply system - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a method for estimating battery residual capacity, reducing a calculation load, and extremely highly accurately estimating residual capacity of a battery based on stable open circuit voltage, by highly accurately estimating the stable open circuit voltage. <P>SOLUTION: When finishing acquisition of a voltage measurement value (Step S2), optimization is progressed renewing each value of coefficients in the following iterating calculations (Step S4) with an initial value of the coefficients set at Step S3 as a start point. Once an optimum value of each coefficient in the approximation is determined at the Step S4, the stable open circuit voltage is calculated by the optimized reciprocal function using the optimal value at the Step S5. And then based thereon, the battery residual capacity is calculated by the predetermined conversion method (Step S6). <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a technical field such as a battery remaining capacity estimation method that estimates a remaining capacity of a battery that supplies power to a load based on a stable open circuit voltage.

  In recent years, there has been an increasing demand for accurate estimation of the charging rate or remaining capacity of an installed battery power source due to portability in electronic devices, hybridization in automobiles, and idling stop. As a method for estimating the charging rate or the remaining capacity, a method of estimating based on the stable open circuit voltage (OCV) of the battery is the simplest and effective as an implementation means.

  However, in the actual usage environment of the power supply system, since charging / discharging current frequently flows from the battery, polarization always occurs inside the battery. For this reason, even when charging / discharging is not performed, the polarization voltage is superimposed on the battery voltage, and it is hardly expected that the battery voltage matches the OCV voltage.

  For example, Patent Documents 1 to 3 disclose a method for estimating a stable open circuit voltage in consideration of the influence of such polarization. In Patent Literatures 1 to 3, voltage behavior in which the battery charge / discharge polarization is relaxed with time and the battery voltage converges to a stable open circuit voltage is expressed as a function of time. A technique for predicting a stable open circuit voltage after a long time has been proposed by learning this time function using a result of measuring a battery voltage for a relatively short time.

で表現している。上記関数を用いて特性関数Ｆ(ｔ)＝ｔ×ｄＶ(ｔ)／ｄｔを定義し、これを基にＦ(ｔ)を最大とするｔ＝ｔmax（＝ａ／ｂ）を求め、さらに安定開回路電圧を係数Ｃ＝２Ｖ（ｔmax）−Ｖ（ｔ0）として算出している。 As an example, in Patent Document 1, a battery voltage change V (t) when charging / discharging is not performed is represented by a linear inverse function of time t.

It is expressed with. Using the above function, the characteristic function F (t) = t × dV (t) / dt is defined, and based on this, t = tmax (= a / b) that maximizes F (t) is obtained, and further stable The open circuit voltage is calculated as a coefficient C = 2V (tmax) -V (t0).

として表現し、最小二乗演算により係数α及びＯＣＶの最適値を決定し、係数であるＯＣＶの最適値を安定開回路電圧とする方法が提案されている。 Moreover, in patent document 2, the voltage change V (t) of the battery when charging / discharging is not performed,

And the optimum value of the coefficient α and the OCV is determined by the least square operation, and the optimum value of the OCV as the coefficient is used as a stable open circuit voltage.

を用いて表現し、係数Ａｉ、Ｂｉ、Ｖ０を最小二乗演算により最適化して決定する方法が提案されている。
特開平７−９８３６７号公報 特開２００２−２３４４０８号公報 特開２００５−４３３３９号公報 Further, in Patent Document 3, a battery voltage change V (t) when charging / discharging is not performed is expressed by an exponential function of a fourth or higher order,

And a method of optimizing and determining the coefficients Ai, Bi, and V0 by least square calculation.
JP-A-7-98367 JP 2002-234408 A JP 2005-43339 A

  However, the above conventional techniques have the following problems. In Patent Document 1, the voltage change of the battery when charging / discharging is not performed is expressed by using (Equation 1), but the polarization relaxation behavior of the battery voltage is only the inverse function of (Equation 1). It is known that it is not a simple behavior that can be expressed, and Patent Document 1 has a problem that it is not possible to estimate the remaining battery capacity by predicting the stable open circuit voltage with sufficient accuracy.

  In Patent Document 1, a characteristic function is defined and a stable open circuit voltage is calculated based on the characteristic function. However, in such a method, a voltage change is calculated using an optimum coefficient calculated by a least square method or the like. Compared to the estimation method, it is difficult to predict a stable open circuit voltage with high accuracy.

  In Patent Document 2, the battery voltage change when charging / discharging is not performed is expressed using (Equation 2), and the coefficient α and the value of OCV when the power number D is approximately −0.5 are optimal. Although the value is determined, there is a problem that even the function thus determined is quite different from the actual polarization relaxation behavior of the battery voltage.

  On the other hand, in Patent Document 3, the voltage change of the battery when charging / discharging is not performed is expressed by (Equation 3) using an exponential function of the fourth order or higher, so that the accuracy is extremely high. It is possible to predict a stable open circuit voltage. However, (Equation 3) has a problem that the calculation load becomes extremely high because an exponential function that is complicated and has a high calculation load is frequently used. Still further, the exponential function fluctuates greatly as the value of the time variable changes, and sometimes an astronomical value is shown in the calculation process. This is because, for example, when using an arithmetic element with a limited capacity such as a controller mounted on an automobile, the numerical value exceeds the range that the element can handle, and it is difficult to perform stable calculation. there were.

  Therefore, the present invention has been made to solve these problems, and by making it possible to estimate the stable open circuit voltage with high accuracy, the remaining capacity of the battery can be determined with extremely high accuracy based on the stable open circuit voltage. An object of the present invention is to provide a battery remaining capacity estimation method that can be estimated and has a light calculation load.

で近似し、前記ｎ次の逆数関数の係数の値を、前記電圧値取得開始から所定時間内に前記バッテリの電圧値を前記係数の個数以上取得し、取得された前記電圧値を基に、最小二乗法又はカルマンフィルタ演算又はニューラルネットワークによって決定し、前記決定された係数の値を用いて前記ｎ次の逆数関数から前記安定開回路電圧を算出し、前記算出された安定開回路電圧に基づいて前記残存容量を推定する ことを特徴とする。 A first aspect of the battery remaining capacity estimating method of the present invention is a battery remaining capacity estimating method for estimating a remaining capacity of a battery that supplies power to a load based on a stable open circuit voltage, wherein t is a voltage of the battery. The elapsed time from the start of value acquisition, where n is an integer of 2 or more, F (t) is an arbitrary function of the time t, and C is a constant, the charge / discharge polarization generated in the battery is gradually eliminated and the A function in which an open circuit voltage change stabilized by a stable open circuit voltage is a function whose denominator is an nth order polynomial of the time t (hereinafter referred to as an nth order inverse function).

The coefficient value of the n-th order reciprocal function is obtained within a predetermined time from the start of obtaining the voltage value, and the battery voltage value is obtained by the number of the coefficients or more, and based on the obtained voltage value, Determined by a least square method, a Kalman filter operation or a neural network, and calculates the stable open circuit voltage from the n-order reciprocal function using the determined coefficient value, and based on the calculated stable open circuit voltage The remaining capacity is estimated.

とする ことを特徴とする。 In another aspect of the battery remaining capacity estimation method of the present invention, the n-th order reciprocal function is

It is characterized by.

  According to another aspect of the battery remaining capacity estimation method of the present invention, when the time from the charge / discharge stop to the start of acquiring the voltage value of the battery is equal to or less than a first reference time, the voltage change during open circuit is changed to the n-th order. When the time until the start of voltage value acquisition exceeds the first reference time and is less than or equal to the second reference time is approximated by an inverse function, the voltage change during open circuit is expressed as the (n-1) th order inverse function. In the following, each time the time until the voltage value acquisition starts exceeds a predetermined reference time, the order of the reciprocal function is sequentially reduced by 1 until it reaches 1 and approximated.

で近似し、前記ｎ次の逆数関数の係数の値を、前記電圧値取得開始から所定時間内に前記バッテリの電圧を前記電圧センサで測定して取得した前記係数の個数以上の電圧値を基に、最小二乗法又はカルマンフィルタ演算又はニューラルネットワークによって決定し、前記決定された係数の値を用いて前記ｎ次の逆数関数から前記安定開回路電圧を算出し、前記算出された安定開回路電圧に基づいて前記バッテリの残存容量を推定している ことを特徴とする。 A first aspect of a battery remaining capacity estimating device according to the present invention is a battery remaining capacity estimating apparatus that estimates a remaining capacity of a battery that supplies power to a load based on a stable open circuit voltage, and measures the voltage of the battery. And a control unit that executes and controls calculation for estimating the remaining capacity, wherein the control unit is t is an elapsed time from the start of acquiring the voltage value of the battery, and n is an integer of 2 or more. , F (t) is an arbitrary function of time t, and C is a constant, the charge-discharge polarization generated in the battery is gradually eliminated and the change in open circuit voltage is stabilized at the stable open circuit voltage. An nth-order reciprocal function whose denominator is an nth-order polynomial of the time t

And the coefficient value of the n-th order reciprocal function is based on a voltage value equal to or greater than the number of coefficients obtained by measuring the voltage of the battery with the voltage sensor within a predetermined time from the start of voltage value acquisition. The stable open circuit voltage is calculated from the n-th reciprocal function using the value of the determined coefficient, determined by a least square method, a Kalman filter operation, or a neural network, and the calculated stable open circuit voltage Based on this, the remaining capacity of the battery is estimated.

を用いていることを特徴とする。 In another aspect of the battery remaining capacity estimation device of the present invention, the control unit is configured as the nth-order reciprocal function.

It is characterized by using.

  In another aspect of the battery remaining capacity estimation device according to the present invention, the control unit may change the open circuit voltage change when the time from the charge / discharge stop until the battery voltage value acquisition start is equal to or shorter than a first reference time. Approximating with the nth-order reciprocal function, and when the time until the voltage value acquisition start exceeds the first reference time and is equal to or less than the second reference time, the voltage change during open circuit is (n−1) th order. In the following, approximation is performed by decreasing the order of the inverse function one by one until the time until the voltage value acquisition start exceeds a predetermined reference time until the degree reaches 1. And

  According to a first aspect of the battery power supply system of the present invention, the battery and any one of the battery remaining capacity estimation devices described above are provided.

  According to the present invention, since the time characteristic of the open circuit voltage of the battery is approximated by a reciprocal function of the second or higher order, the convergence value of the open circuit voltage is obtained within a short time, and the charging rate of the battery is reduced. Thus, it is possible to provide a charging rate estimation method and the like that can be accurately and stably estimated.

  Configurations of a battery remaining capacity estimation method, a battery remaining capacity estimation apparatus, and a battery power supply system according to a preferred embodiment of the present invention will be described in detail with reference to the drawings. In addition, about each structural part which has the same function, the same code | symbol is attached | subjected and shown for simplification of illustration and description.

  FIG. 2 is a block diagram showing a schematic configuration of the battery remaining capacity estimation device and the battery power supply system according to the embodiment of the present invention. The battery power supply system 100 of this embodiment includes a battery 110, a charging circuit 120, and a battery remaining capacity estimation device 200 of this embodiment, and a load 10 is connected to the battery 110.

  Further, the remaining battery capacity estimation device 200 is configured by a control unit 210, a storage unit 220, and a voltage sensor 230. The controller 210 controls the execution of the calculation for estimating the remaining capacity using the embodiment of the battery remaining capacity estimating method of the present invention, and further controls the operation of the battery power supply system 100 as a whole. It is also possible to do. In FIG. 2, the control unit 210 controls the charging circuit 120 to charge the battery 110.

  The voltage sensor 230 measures the voltage between the terminals of the battery 110 and sends the detected voltage measurement value to the control unit 210. The storage unit 220 stores various parameters used for estimation calculation of the remaining capacity processed by the control unit 210, voltage measurement values detected by the voltage sensor 230, and the like.

  When the battery power supply system 100 of the present embodiment is applied to a vehicle power supply system, a vehicle lead-acid battery is used as the battery 110, and power is supplied to a load 10 such as a motor mounted on the vehicle. . In a lead storage battery for a vehicle, an alternator is provided as the charging circuit 120.

  Next, the battery remaining capacity estimation method of the present embodiment for estimating the remaining capacity of the battery 110 in the battery power supply system 100 of the present embodiment will be described. As described above, the remaining capacity of the battery 110 has a strong correlation with the stable open circuit voltage of the battery 110, and a correlation equation that uniquely calculates the remaining capacity from the stable open circuit voltage is set in advance. Is possible. By using such a correlation equation, it is possible to estimate the remaining capacity of the battery 110 with high accuracy by estimating the stable open circuit voltage of the battery 110 with high accuracy.

  On the other hand, in the actual operation state of the battery 110, since charging / discharging from the battery 110 is frequently repeated, polarization is generated in the battery 110 for most of the period. Polarization voltage is superimposed. The influence of such polarization gradually decreases when the charging / discharging of the battery 110 is stopped, but it takes an extremely long time such as several tens of hours to several days until the influence becomes sufficiently small. Therefore, it is extremely difficult to measure the open circuit voltage at the time when the polarization is eliminated, and this cannot be expected in actual operation.

  Therefore, in the battery remaining capacity estimation method of the present embodiment, a function that can accurately approximate the time variation of the open circuit voltage after stopping charging and discharging is used, and this is learned using the voltage measurement value detected by the voltage sensor 230. Thus, the open circuit voltage when the battery 110 is stabilized can be estimated with high accuracy.

ここで、時間ｔをバッテリ１１０の電圧値取得開始からの経過時間、Ｖ（ｔ）をバッテリ１１０の開回路電圧、ｎを２以上の整数、Ｆ（ｔ）を時間ｔの任意の関数、Ｃを定数としている。 The battery remaining capacity estimation method of this embodiment will be described in detail below. In the present embodiment, as a function that approximates the time change of the open circuit voltage after stopping charging and discharging with high accuracy, a function such as the following equation in which the denominator is an nth order polynomial of time t (hereinafter, an nth order reciprocal function Is used).

Here, the time t is the elapsed time from the start of acquiring the voltage value of the battery 110, V (t) is the open circuit voltage of the battery 110, n is an integer of 2 or more, F (t) is an arbitrary function of the time t, C Is a constant.

  In the present embodiment, the open circuit voltage V (t) is expressed by an inverse function as shown in (Equation 4), and the order n of the inverse function is at least 2 or more, so that the time variation of the open circuit voltage of the battery 110 is changed. Can be approximated with high accuracy. In (Equation 4), the time change of the polarization voltage after charge / discharge stop is approximated using the reciprocal function of the first term on the right side, and F (t) of the second term can be arbitrarily set as necessary. I am doing so.

In the following, as another embodiment, a battery remaining capacity estimation method when using the reciprocal function of the following equation that does not use F (t) of the second term will be described.

In the battery remaining capacity estimating method of the present embodiment executed in the battery remaining capacity estimating apparatus 200, after stopping charging / discharging from the battery 110, the open circuit voltage of the battery 110 is measured using the voltage sensor 230, and this voltage is detected. The coefficients A _i , B _i , and C in (Equation 5) are determined based on the measured values.

ここで、Ｖｍ（ｋ）は電圧センサ２３０で取得した電圧測定値を表わしており、ｔ_ｋは充放電停止後の電圧測定点の時間を表わしている。電圧測定点は、全部でＫ点あるものとしている。 As a method for determining the coefficients A _i , B _i , and C in (Expression 5), a least square method, a Kalman filter operation, or a neural network can be used. If for example, using the least square method, the coefficient A _i, B i, define the sum of squared deviations function as following equation to determine the optimum value and _C, the coefficient A _i far is the _smallest, B _i , C is obtained.

Here, Vm (k) represents the measured voltage value acquired by the voltage sensor 230, t _k denotes the time of the voltage measurement point after stopping charging and discharging. There are a total of K voltage measurement points.

As a method for obtaining the coefficients A _i , B _i , and C that minimize the deviation sum of squares M shown in the above (formula 6), M is sequentially set from predetermined initial values such as Gauss-Newton method or Levenberg-Marquardt method. A sequential calculation method may be used in which the coefficients A _i , B _i , and C are updated to optimum values so as to decrease.

は、

と書き替えられる。 Taking the quadratic reciprocal function as an example, an example in which the optimal coefficient solution of this function is obtained by the Gauss-Newton method is shown below. Quadratic reciprocal function

Is

Will be rewritten.

と書き替えられる。 Furthermore, when the open circuit voltage V (t) is a discrete value sampled at ΔT intervals,

Will be rewritten.

とし、 次に、最小二乗法の適用に際し（式６）の各係数α１〜α５に関する偏微分項を下記（式１１）によって各サンプリングタイミング毎に求める。

The measured battery voltage measured at ΔT intervals is Vm (n) (n = 1 to Ns), and the difference from the value calculated from the above formula is

Next, when applying the least square method, partial differential terms relating to the coefficients α1 to α5 of (Equation 6) are obtained at each sampling timing by the following (Equation 11).

And based on the partial differential term regarding each coefficient (alpha) 1- (alpha) 5 for every obtained sampling timing, the Hessian matrix B (5 * 5 square symmetry matrix which fits the simultaneous equations of the least squares method by the following (Formula 12). i, j) is calculated.

とすると、係数α１〜α５の逐次計算における補正変分ｄｄ１〜ｄｄ５は下記（式１４）によって

と算出される。 Further, dR1 to dR5 represented by the following (formula 13) are calculated based on the partial differential terms relating to the coefficients α1 to α5 at the respective sampling timings similarly obtained.

Then, the correction variations dd1 to dd5 in the sequential calculation of the coefficients α1 to α5 are expressed by the following (formula 14).

Is calculated.

The coefficients α1 to α5 are updated in accordance with the following (Equation 15) until the above dd1 to dd5 become sufficiently small, and (Equation 9) to (Equation 15) are repeatedly calculated.

  By the least square method as described above, the open circuit voltage V (t) represented by the reciprocal function of (Equation 5) is approximated to the voltage measurement value Vm (k) with high accuracy, and the open circuit voltage V It becomes possible to estimate the change of the circuit voltage with high accuracy. In the battery remaining capacity estimation method according to the present embodiment, since no exponential function is used in the above calculation process, the processing load for optimizing the coefficient of the approximate expression of the open circuit voltage is greatly reduced.

図３では、太い実線３１０が測定値を示しており、破線３２０が開回路電圧の（式１６）を示している。同図に示すとおり、上記の２次の逆数関数で近似された開回路電圧の式３２０は、測定値３１０と極めてよい一致を示している。また、開回路電圧式３２０と測定値３１０との一致度を示すＲ^２は、０．９９５２５と１に近い値を示している。 An example in which the open circuit voltage is approximated using the reciprocal function shown in (Equation 5) is shown in FIG. The approximate expression shown in the figure uses the following quadratic reciprocal function.

In FIG. 3, the thick solid line 310 indicates the measured value, and the broken line 320 indicates the open circuit voltage (Equation 16). As shown in the figure, the open circuit voltage equation 320 approximated by the above-described quadratic reciprocal function shows a very good agreement with the measured value 310. R ² indicating the degree of coincidence between the open circuit voltage equation 320 and the measured value 310 is 0.99525 and a value close to 1.

As other methods for determining the coefficients A _i , B _i , and C in (Equation 5), there are a Kalman filter operation and a neural network. In the Kalman filter operation, the coefficients A _i , B _i , and C or a mathematical expression including these coefficients are in a state. A vector X _k is set, and a Jacobian matrix for determining a time-series first-order prediction of the state vector and an observation equation for calculating an observed value V (t _k ) from the state vector are determined. As a result, starting from the initial values of predetermined coefficients A _i , B _i , C and the first actual measurement value V ₁ (t ₁ ) of K = 1, the actual measurement value Vm (t _k ) and the calculated value V (t _k ) so that the error is minimized, the Kalman gain calculation, the sequential update of the state vector X _k , and the one-point prediction of the state vector are repeated in accordance with the Kalman filter algorithm, so Thus, the coefficients A _i , B _i and C can be optimized.

When a neural network is used, an appropriate network such as a multi-stage perceptron is selected, and various examples of measured values of voltages A _i , B _i , and C optimized in advance based on actual measured values of voltages are selected. Educate the network by using appropriate methods such as back-propagation using these as teacher signals, and give the measured values of the voltages to the network as inputs to optimize the coefficients A _i , B _i , and C A value can be obtained as an output.

  The battery remaining capacity estimation method executed in the battery remaining capacity estimation apparatus 200 of this embodiment using the least square method will be described in more detail with reference to the flowchart shown in FIG. First, when it is determined in step S1 that charging / discharging from the battery 110 is stopped, the voltage (open circuit voltage) of the battery 110 is measured using the voltage sensor 230 in step S2. In this measurement, voltage measurement values equal to or greater than the number of coefficients included in the reciprocal function of (Equation 4) or (Equation 5) approximating the open circuit voltage are obtained.

  When the acquisition of the voltage measurement value is completed, the initial value of the coefficient of the reciprocal function optimized by the least square method is set in step S3. In step S4, the optimization is advanced while updating the value of each coefficient in the subsequent iterative calculation starting from the initial value of the coefficient set in step S3. When the optimum value of each coefficient of the approximate expression is determined in step S4, in step S5, a stable open circuit voltage is calculated with an optimized inverse function using the coefficient.

  In step S6, the remaining battery capacity is calculated by a predetermined conversion method based on the stable open circuit voltage calculated in step S5. For the conversion from the stable open circuit voltage to the remaining battery capacity, a conversion formula created in advance and stored in the storage unit 220 can be used.

Another embodiment of the battery remaining capacity estimation method of the present invention will be described below. Also in this embodiment, the reciprocal function of (Equation 4) or (Equation 5) is used as a function that approximates the open circuit voltage of the battery 110. Hereinafter, as an example, the order in the reciprocal function of (Equation 5) is used. This will be described using an example in which n is 4. At this time, the open circuit voltage V (t) of the battery 110 is expressed by the following equation.

  The change in the open circuit voltage of the battery 110 after charging / discharging stops, for example, as shown in FIG. 3, shows a rapid change immediately after stopping charging / discharging, and becomes gradual as time passes. The change in the open circuit voltage as shown in FIG. 3 can be approximated with high accuracy by using (Equation 17). In particular, in order to approximate a rapid change immediately after stopping charging / discharging with high accuracy, it is preferable to use a high-order reciprocal function such as (Equation 17).

  However, the change in the open circuit voltage is gentle with the passage of time, and in order to approximate the gentle change in the open circuit voltage after the lapse of a predetermined time after stopping charging / discharging, it is not always a high-order inverse. There is no need to use a function. Therefore, in the present embodiment, the order n of the reciprocal function used for approximating the open circuit voltage is reduced according to the length of time from the stop of charging / discharging to the start of measurement of the open circuit voltage of the battery 110.

First, assuming that the time from when the charging / discharging of the battery 110 is stopped until the voltage measurement of the battery 110 is started by the voltage sensor 230 in order to obtain the voltage measurement value Vm (k), the time tx is the first time. When the reference time has not been reached, (Equation 17) is used as an approximate expression of the open circuit voltage. When the time tx exceeds the first reference time, the open circuit voltage is approximated. As an equation, the following equation obtained by reducing the order of the reciprocal function of (Equation 17) by one is used.

Further, when the time tx exceeds the second reference time, the following expression obtained by reducing the order of the reciprocal function of (Expression 18) by 1 is used as an approximate expression of the open circuit voltage.

Further, when the time tx exceeds the third reference time, the following expression obtained by reducing the order of the reciprocal function of (Expression 19) by 1 is used as an approximate expression of the open circuit voltage.

  Since the order n in (Equation 20) has reached 1, the reciprocal function in which the order n is further reduced is not used.

  As described above, by reducing the order n of the reciprocal function that approximates the open circuit voltage according to the time from the charge / discharge stop of the battery 110 to the start of voltage measurement, the calculation time required for the optimal approximation of the reciprocal function can be shortened. It becomes possible.

  The method for estimating the remaining battery capacity in the present embodiment will be described in more detail with reference to the flowchart shown in FIG. First, when it is determined in step S11 that charging / discharging from the battery 110 is stopped, an elapsed time since charging / discharging is measured in step S12.

  In step S13, it is determined whether or not the elapsed time from the charge / discharge stop has reached a predetermined time. If it is determined that the predetermined time has been reached, the voltage of the battery 110 (open circuit voltage) is determined using the voltage sensor 230 in step S14. Measure. In this measurement, voltage measurement values equal to or greater than the number of coefficients included in the reciprocal function of (Equation 4) or (Equation 5) approximating the open circuit voltage are obtained.

  When the acquisition of the voltage measurement value is completed, the order of the reciprocal function of (Equation 4) or (Equation 5) that approximates the open circuit voltage is determined in step S15 according to the length of the elapsed time until the voltage measurement of the battery 110 starts. . In step S16, the initial value of the coefficient of the inverse function optimized by the least square method is set. In step S17, the optimization is advanced while updating the value of each coefficient in subsequent iterations, starting from the initial value of the coefficient set in step S16. When the optimum value of each coefficient of the approximate expression is determined in step S17, a stable open circuit voltage is calculated with an optimized inverse function using this in step S18.

  In step S19, the remaining battery capacity is calculated by a predetermined conversion method based on the stable open circuit voltage calculated in step S17. For the conversion from the stable open circuit voltage to the remaining battery capacity, a conversion formula created in advance and stored in the storage unit 220 can be used.

  Even in the battery remaining capacity estimation method of the present embodiment, since no exponential function is used in the above calculation process, the processing load for optimizing the coefficient of the approximate expression of the open circuit voltage is greatly reduced.

  As described above, in the battery remaining capacity estimation method according to the present invention, the time characteristic of the open circuit voltage of the battery is approximated by a reciprocal function of the second or higher order, so that the convergence value of the open circuit voltage is obtained within a short time. Thus, it is possible to accurately and stably estimate the charging rate of the battery with a small calculation load. In addition, in the method using the exponential function, the value fluctuates greatly according to the change in the value of the time variable, and it may be difficult to perform stable calculation. In the battery remaining capacity estimation method, it is possible to perform stable calculation.

  In this embodiment, the open circuit voltage is approximated only by the reciprocal function, but other functions such as an exponential function may be combined as appropriate. At this time, the order of the reciprocal function and the order of the exponential function are appropriately selected in consideration of the calculation load and the required accuracy. In this case, a combination of a first-order reciprocal function and a third-order or lower exponential function may be used.

  In addition, the description in this Embodiment shows an example of the battery remaining capacity estimation method, battery remaining capacity estimation apparatus, and battery power supply system which concern on this invention, and is not limited to this. The detailed configuration and detailed operation of the battery remaining capacity estimation method and the like in the present embodiment can be changed as appropriate without departing from the spirit of the present invention.

It is a flowchart which shows the flow of a process of the battery remaining capacity estimation method which concerns on the 1st Embodiment of this invention. 1 is a block diagram showing a schematic configuration of a remaining battery capacity estimation device and a battery power supply system according to a first embodiment of the present invention. It is a figure which compares the measured value of an open circuit voltage, and the approximate expression of a reciprocal function. It is a flowchart which shows the flow of a process of the battery remaining capacity estimation method which concerns on the 2nd Embodiment of this invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 Load 100 Battery power supply system 110 Battery 120 Charging circuit 200 Battery remaining capacity estimation apparatus 210 Control part 220 Storage part 230 Voltage sensor

Claims (7)

A battery remaining capacity estimation method for estimating a remaining capacity of a battery that supplies power to a load based on a stable open circuit voltage,
When t is an elapsed time from the start of acquiring the voltage value of the battery, n is an integer of 2 or more, F (t) is an arbitrary function of the time t, and C is a constant, the charge / discharge polarization generated in the battery is A function in which an open circuit voltage change that is gradually resolved and stabilized at the stable open circuit voltage is a function whose denominator is an nth order polynomial of the time t (hereinafter referred to as an nth order reciprocal function).

And approximate with
The value of the coefficient of the nth-order reciprocal function is acquired within the predetermined time from the start of the acquisition of the voltage value, the battery voltage value is equal to or greater than the number of the coefficient, and based on the acquired voltage value, the least square method or Determined by Kalman filter operation or neural network,
The stable open circuit voltage is calculated from the n-th reciprocal function using the determined coefficient value,
The remaining battery capacity is estimated based on the calculated stable open circuit voltage. The remaining battery capacity estimation method, comprising: The reciprocal function of order n

The battery remaining capacity estimation method according to claim 1, wherein: When the time from the charge / discharge stop to the start of acquiring the voltage value of the battery is equal to or shorter than the first reference time, the open circuit voltage change is approximated by the nth-order inverse function,
When the time until the voltage value acquisition start exceeds the first reference time and is equal to or less than the second reference time, the open circuit voltage change is approximated by the (n−1) -order reciprocal function,
3. The method according to claim 1, further comprising: approximating the reciprocal function by decreasing the order of the reciprocal function one by one until the voltage value acquisition start time exceeds a predetermined reference time. The battery remaining capacity estimation method as described. A battery remaining capacity estimation device that estimates a remaining capacity of a battery that supplies power to a load based on a stable open circuit voltage,
A voltage sensor for measuring the voltage of the battery;
A control unit that executes and controls a calculation for estimating the remaining capacity,
The control unit generates the battery when t is an elapsed time from the start of acquiring the voltage value of the battery, n is an integer of 2 or more, F (t) is an arbitrary function of the time t, and C is a constant. The n-th order reciprocal function whose denominator is the n-th order polynomial of the time t is the voltage change at the time of open circuit in which the charged / discharge polarization is gradually eliminated and stabilized at the stable open circuit voltage.

And the coefficient value of the n-th order reciprocal function is based on a voltage value equal to or greater than the number of coefficients obtained by measuring the voltage of the battery with the voltage sensor within a predetermined time from the start of voltage value acquisition. The stable open circuit voltage is calculated from the n-th reciprocal function using the value of the determined coefficient, determined by a least square method, a Kalman filter operation, or a neural network, and the calculated stable open circuit voltage The remaining battery capacity is estimated based on the remaining battery capacity estimation apparatus. The control unit is configured as the nth-order inverse function.

The battery remaining capacity estimation device according to claim 4, wherein: The controller obtains the voltage value by approximating the open circuit voltage change by the nth-order reciprocal function when the time from the charge / discharge stop to the battery voltage value acquisition start is equal to or shorter than a first reference time. When the time until the start exceeds the first reference time and is less than or equal to the second reference time, the open circuit voltage change is approximated by the (n-1) th order reciprocal function, and then the voltage value acquisition is performed. 6. The remaining battery according to claim 4, wherein when the time until the start exceeds a predetermined reference time, the order of the reciprocal function is approximated by decreasing by 1 until the degree reaches 1. Capacity estimation device. A battery power supply system comprising: the battery; and the remaining battery capacity estimation device according to any one of claims 4 to 6.

Images