JPH1132442A - Method and apparatus for estimating residual capacity of storage battery - Google Patents

Abstract

PROBLEM TO BE SOLVED: To accurately estimate the residual capacity of a storage battery, even when charging and discharging are repeated in a condition in which there is no standard of the residual capacity of the battery in a hybrid electric vehicle or the like. SOLUTION: To digitally process the charging and discharging voltage and current of a load 1 such as a motor and a storage battery, the voltage and current are converted into digital signals with an A/D converter 5 and an A/D converter 6 respectively and subsequently converted into a compound spectrum with a frequency converter (FFT) 7 for voltage and another frequency converter (FFT) 8 for current. From these complex spectra of the voltage and current so obtained with the storage battery in use, impedance is calculated with an impedance calculating part 9 to obtain a radius (rj), a characteristic amount of impedance, from the storage battery in operation. This radius (rj) is compared with the radius (rj) which has been obtained beforehand and stored in a battery residual capacity calculating part 10, so that the residual capacity of the storage battery is estimated from the interrelation.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for estimating the remaining capacity of a storage battery by measuring the voltage, current and battery temperature of a storage battery used in an electric vehicle and the like at the time of charging and discharging, and determining the remaining capacity of the storage battery. It relates to a system for estimating.

[0002]

2. Description of the Related Art From the end of this century to the beginning of the 21st century, demand for electric vehicles starts, and it is predicted that electric vehicles will become popular in response to the tight supply of oil.
Since an electric vehicle travels with electric energy stored in a storage battery, the remaining capacity indication of the storage battery is very important in knowing the possible travel distance and in controlling the charge / discharge amount of the storage battery.

(Configuration of First Conventional Example) An example of the above-described conventional storage battery remaining capacity display device will be described with reference to the drawings. FIG. 11 shows a conventional storage battery remaining amount display device. 110 is a drive motor, 112 is a storage battery, 114 is a meter, 116 is a switch device, 11
8 is a motor start switch, 120 is a switch actuating means, 122 is an amplifier, 124 is a detector, and 126 is a logic circuit. The storage battery remaining capacity display device of the first conventional example is provided with a meter 114 for indicating the remaining capacity of the storage battery 112, a switch device 116 for connecting the meter to the storage battery, and means 120 for operating the switch device. .

(Operation of First Conventional Example) Regarding the storage battery remaining capacity display device of the first conventional example configured as described above,
The operation will be described below. The switch device operating means 120 detects the magnitude of the current supplied from the storage battery 112 to the load 110 with the detector 124 and switches the meter 114 only when the magnitude of the current is within a predetermined range. Otherwise, the switch device is operated to disconnect the meter 114 from the storage battery 112 at other times. In this storage battery remaining capacity display device, the remaining battery capacity can be displayed on a meter with the same accuracy as when a measured current is applied to a standard load and the measured terminal voltage of the storage battery is displayed.
305514). Also, instead of using a DC method, a method of superposing an AC signal on the charge / discharge current waveform of the storage battery and estimating the state of charge of the storage battery from the obtained milliohm-order impedance value has been studied (Kracimir Nenov, Stefan -Lipka, Electrochemical Society Proceedings, Vol. 9
2-5, p. 374).

[0005]

However, the above configuration has the following problems. For example, when estimating the remaining capacity of the storage battery using charge / discharge characteristics using a direct current, in an electric vehicle, when driving on an uphill and downhill or on a flat ground, the driving load of the motor differs, and Because of the hysteresis, the remaining capacity of the storage battery was not uniquely determined, and it was difficult to estimate the remaining battery capacity only from the voltage characteristics of the storage battery. Also,
The capacity of the storage battery of a hybrid vehicle is as small as about 1/10 of the capacity of the storage battery of an electric vehicle, and a more accurate estimation method is required. However, it has been difficult to satisfy this by a method using direct current. When the AC signal is superimposed on the charge / discharge current, it is considered that the remaining battery capacity can be estimated more accurately than when DC is used.However, accurate impedance measurement is difficult due to the influence of inverter noise and ignition noise. Was. In a hybrid electric vehicle, the remaining capacity 1 which is a reference of the remaining capacity of the battery is 1
Since the vehicle runs in an intermediate state without full charge corresponding to 00% or zero charge corresponding to 0% of the remaining capacity, and charging and discharging are repeated during running, there is no absolute reference for the remaining battery capacity. Therefore, unlike the case of an electric vehicle, it is basically difficult to estimate the remaining battery capacity from the voltage characteristics of the storage battery during traveling. In addition, it is difficult to formulate a model such as battery reaction resistance, temperature correction and hysteresis, and it is difficult to estimate the remaining capacity of the storage battery in real time.

The present invention has been made in view of the above-described problems, and is intended to accurately estimate the remaining battery capacity of a storage battery in real time even when charging and discharging are repeated without a reference for the remaining battery capacity, such as in a hybrid electric vehicle. An object of the present invention is to provide a storage battery remaining capacity estimation method and a system thereof.

[0007]

In order to solve the above problems, a method for estimating a remaining capacity of a storage battery according to the present invention measures a voltage and a current signal of a storage battery connected to a load during charging and discharging, and measures the measured voltage. Each current signal is mapped to a frequency domain, and a characteristic amount of an impedance is calculated from a voltage and a current signal in a predetermined frequency region.Based on a relational expression between the characteristic amount of the impedance obtained in advance and the remaining capacity of the storage battery, This is a method for estimating the remaining battery capacity.

According to the present invention, with the above-described configuration, the remaining battery capacity of a storage battery used in a hybrid electric vehicle or the like is calculated during charging / discharging operation, particularly during running, and the impedance of the storage battery is measured beforehand. By comparing with the impedance of the remaining battery capacity, the remaining battery capacity is estimated. When a load such as data during operation does not operate at the characteristic frequency, a signal of a predetermined frequency characteristic of the remaining capacity of the storage battery is generated by using a signal generator (a charge / discharge current reference waveform generation device). By superimposing on the charging / discharging current, removing an AC signal having a frequency other than a predetermined frequency, calculating an impedance based only on the superimposed frequency, and comparing the calculated impedance with the impedance of the remaining battery capacity of the storage battery, which is measured in advance. Will be estimated. In addition, it is provided with knowledge data obtained by preliminarily learning various cases based on empirical data accumulated through actual operation by using various charge / discharge current voltage state information and load state information during traveling or stopping. Each state information is processed, and the remaining capacity of the storage battery is estimated based on knowledge data learned in advance.

[0009]

DESCRIPTION OF THE PREFERRED EMBODIMENTS A method and system for estimating a remaining battery capacity according to an embodiment of the present invention will be described below.

According to the present invention, (a) a voltage and a current signal at the time of charging and discharging of a storage battery connected to a load are measured, and (b) the measured voltage and current signals are respectively mapped as complex spectra in a frequency domain. (C) calculating an impedance and an impedance characteristic amount from a complex spectrum of a voltage and a current in a predetermined frequency region; and (d) determining the impedance characteristic based on a relational expression between the impedance characteristic amount obtained in advance and the remaining capacity of the storage battery. It consists of a method of estimating the remaining battery capacity from the amount.

The impedance of the storage battery is calculated in a wide frequency range in advance, and the imaginary part Im (Z (w)) and the real part Re are calculated.
The relationship between the radius of one or more circles drawn from (Z (w)) and the remaining battery capacity is examined. Thus, the radius of the circle serving as the feature value is obtained, and the frequency of the AC signal that provides the feature value is obtained. Here, the selected AC signal is 1
Or two or more are desirable.

The measured voltage and current signals are A / D
After the conversion, the voltage and current are respectively converted into complex spectra by a frequency converter (FFT). The complex impedance is calculated from the obtained voltage and current values at each frequency. Further, by displaying the impedance on a complex plane, the real part Re (Z (w)) and the imaginary part I
Obtain one or more circles drawn by m (Z (w)). From this, the radius of the circle, which is the feature amount, is calculated.

(Configuration of First Embodiment) FIG. 1 is a block diagram of a storage battery remaining capacity estimation system according to an embodiment of the present invention. In FIG. 1, 1
Is a load, 2 is a battery, 5 is a voltage A / D converter, 6 is a current A / D converter, and 7 is a voltage frequency converter (FFT: Fa).
stFourier Transform), 8 is a current frequency converter (F
FT), 9 is a division calculator, 10 is an impedance calculator, 11 is a remaining battery charge calculator, 12 is a battery thermometer, 13
Denotes a battery remaining capacity temperature correction unit.

The operation of the storage battery remaining capacity estimating system configured as described above will be described below with reference to FIGS. 1, 2 and 3.

(Operation of the First Embodiment) First, FIG. 2 shows the measurement result of the impedance Z (w) of the storage battery. The complex spectra Sv (w) and Si (w) of the voltage V and the current I when the storage battery is used are obtained, and the ratio Sv
(W) / Si (w) is calculated as the impedance Z (w) of the storage battery, and the ordinate represents the imaginary part Im (Z
(W)), and the horizontal axis is shown on the complex plane as a real part Re (Z (w)). The impedance is calculated in advance in a wide frequency range, the relationship between the remaining capacity of the storage battery and the radius of the drawn circle is examined, and the frequency range in which the characteristic amount is given is obtained. Next, the real part Re of the angular frequencies W0, W1,..., Wi,.
A radius ri, which is a feature quantity, is obtained in advance from the size of a circle drawn from (Z (wi)) and the imaginary part Im (Z (wi)).

FIG. 3 shows the measurement results of the relationship between the radius ri obtained in FIG. 2 and the remaining capacity of the storage battery. The relationship between the radius ri and the remaining battery charge is stored in the remaining battery charge calculator 11.

Next, a battery remaining capacity estimation system according to an embodiment of the present invention will be described with reference to the block diagram of FIG. In FIG. 1, in a hybrid electric vehicle or the like, for example, charging or discharging is repeated between a load 1 such as a motor and a storage battery 2 while the vehicle is running or stopped, and the voltage and current are digitally processed based on the measured voltage and current. To make it possible, the digital signal is converted by the A / D converter 5 and the A / D converter 6. Further, the current and voltage are converted into a complex spectrum by a voltage frequency converter (FFT: Fast Fourier Transform) 7 and a current frequency converter (FFT) 8. From the obtained complex spectra Sv (w) and Si (w) of the voltage V and the current I when the storage battery is used, the ratio Sv
(W) / Si (w) is the impedance Z (w) of the storage battery
, Wj,..., Wn, and the real part Re (Z) of the predetermined angular frequencies W0, W1,.
(Wj)) and the magnitude of the imaginary part Im (Z (wj)) are obtained from the storage battery that is repeatedly charged and discharged. Then, a radius rj in the angular frequency domain for providing the feature amount is obtained.
The radius rj is compared with the radius ri obtained in advance and stored in the remaining battery capacity calculator 11, and the remaining battery capacity is calculated from the mutual relationship. The calculation of the remaining battery capacity is performed by using the radius rj
Is calculated using the least squares method or the spline function of the difference between the radius and the radius ri, and the remaining capacity is estimated.

Since the remaining capacity of the storage battery changes depending on the temperature of the storage battery, the battery temperature is measured by the 12 battery thermometers, and the remaining battery capacity estimated by the remaining battery temperature correction unit 13 is corrected and displayed on the display unit. . Also, in a storage battery such as a nickel-metal hydride battery, there is hysteresis even with the same remaining battery capacity, so that the measured value of the remaining battery capacity differs between charging and discharging. In this case, the complex spectrums Sv (w) and Si (w) of the voltage V and the current I when the storage battery is used are determined from those at the time of charging and those at the time of discharging, respectively, and the spectrum ratio Sv (w) / Si ( w), the spectral ratio Sv at the time of discharge
(W) / Si (w) is the impedance Zj when charging the storage battery.
(W), calculated by the impedance calculator 9 as the discharge impedance Zh (w) and given angular frequencies W0, W
1, ---, Wj,-, real part Re (Zj (wj), Z of Wn
h (wj)) and the imaginary part Im (Zj (wj), Zh (w
The magnitude of j)) is obtained from the storage battery during the operation of repeating charge and discharge. Hereinafter, by the same method as described above, the remaining capacity estimation value at the time of charging and the remaining capacity estimation value at the time of discharging can be obtained, and the remaining capacity of the storage battery can be obtained comprehensively.

As described above, according to the present embodiment, the complex spectrum of the voltage V and the current I when the battery is repeatedly charged and discharged is obtained, the impedance of the battery is calculated, and the real part Re () at a predetermined angular frequency wj is calculated. From the size of a circle drawn from Z (wj)) and the imaginary part Im (Z (wj)), a radius rj as a feature amount is obtained. By comparing the radius rj with the radius ri obtained in advance and stored in the remaining battery capacity calculation unit 11, and estimating the remaining battery capacity, it is possible to reduce the estimation accuracy of an operating electric vehicle or the like that has conventionally had a low estimation accuracy. The remaining capacity of the storage battery can be estimated with high accuracy.

Hereinafter, a second embodiment of the present invention will be described with reference to the drawings.

(Configuration of Second Embodiment) FIG. 4 is a block diagram of a storage battery remaining capacity estimation system according to a second embodiment of the present invention. In FIG. 4, 1 is a load, 2 is a battery, 3 is a noise eliminator for voltage,
4 is a current noise removing device, 5 is a voltage A / D converter,
6 is a current A / D converter, 7 is a voltage frequency converter (F
FT), 8 is a current frequency converter (FFT), 9 is a division operation section, 10 is an impedance calculation section, 11 is a remaining battery capacity calculation section, 12 is a battery thermometer, 13 is a remaining battery temperature correction section, 14 Denotes an engine unit, 15 denotes a charger, and 16 denotes a signal generator.

The operation of the storage battery remaining capacity estimating system configured as described above will be described below with reference to FIGS. 4, 5 and 6.

(Operation of the Second Embodiment) First, as in the first embodiment, FIG. 5 shows the impedance Z of the storage battery.
(W) is calculated, and the vertical axis is the imaginary part Im (Z
(W)), and the horizontal axis represents the real part Re (Z (w)).
If the charging / discharging voltage and current include a large amount of inverter noise or ignition noise, AC signals having a frequency other than a predetermined frequency are removed by the noise removing devices (3, 4). Real part Re of predetermined angular frequencies W0, W1, ---, Wi, ---, Wn
A characteristic radius ri is obtained in advance from the size of a circle drawn from (Z (wi)) and the imaginary part Im (Z (wi)). However, a load such as a motor during operation does not always operate at the characteristic angular frequency, and as shown in FIG.
In the frequency ranges of w0, w1, w2, and w3, data may be insufficient, and it may not be possible to estimate the remaining capacity of the storage battery with high accuracy. In the second embodiment, a signal corresponding to a predetermined frequency giving a characteristic amount of the remaining capacity of the storage battery is superimposed on the charging / discharging current during operation, and the impedance Z (w) of the frequency corresponding to the characteristic of the remaining capacity of the storage battery is measured. And the remaining battery capacity is always estimated with high accuracy.

Next, a battery remaining capacity estimating system according to a second embodiment of the present invention will be described with reference to the block diagram of FIG. 4, in a hybrid electric vehicle or the like, for example, charging / discharging is repeated by a load 1, such as a motor, a storage battery 2, and a charger 14. The voltage and current of the storage battery are passed through a noise elimination device to remove an AC signal having a frequency other than a predetermined frequency, and are converted by an A / D converter 5 and an A / D converter 6. The frequency converter (FFT) 8 obtains the complex spectrum of the current and the voltage, and obtains the spectrum Sv (w), Si of the voltage V and the current I.
(W), the ratio Sv (w) / Si (w) is defined as the impedance Z (w) of the storage battery,
Calculate with 0. Predetermined angular frequencies W0, W1, ---, wj,
---, the real part Re (Z (wj)) and the imaginary part Im (Z
The magnitude of (wj)) is obtained from the storage battery during the operation of repeating charge and discharge. However, with a load such as a running motor,
In some cases, the charge / discharge current does not include signals in the frequency range of w0, w1, w2, and w3. At that time, it is detected that there is no signal in the frequency domain of w0, w1, w2, and w3, which is obtained in advance and stored in the remaining battery capacity calculation unit 10, and the signal generator 15
To superimpose the charging / discharging current at each characteristic frequency wj on the operating battery charging / discharging current. FIG. 6 shows a sine wave superimposed on the charge / discharge current waveform.
In this way, the real part Re (Z (wj)) and the imaginary part Im of the impedance at each characteristic frequency wj
(Z (wj)) is obtained. Then, similarly to the first embodiment, a radius rj which is a feature amount is obtained. This radius rj
Is compared with the radius ri obtained in advance and stored in the remaining battery capacity calculation unit 11, the remaining battery capacity is calculated from the mutual relationship, and the remaining capacity is estimated.

As described above, according to the present embodiment, the complex spectrum of the voltage V and the current I at the time of using the storage battery which repeats charging and discharging is obtained after removing the AC signal other than the predetermined frequency, and the impedance of the storage battery is calculated. Then, the real part Re (Z (wj)) and the imaginary part Im (Z) at a predetermined angular frequency wj
From the size of the circle drawn from (wj)), a radius rj, which is a feature quantity, is obtained. By comparing the radius rj with the radius ri obtained in advance and stored in the remaining battery capacity calculation unit 11, and estimating the remaining battery capacity, an operating hybrid electric vehicle or the like that has conventionally had a low estimation accuracy can be obtained. Can be estimated with high accuracy.

Hereinafter, a third embodiment of the present invention will be described with reference to the drawings.

(Configuration of Third Embodiment) FIG. 7 is a block diagram showing a storage battery remaining capacity estimation system according to a third embodiment of the present invention. In FIG. 7, 1 is a load, 2 is a battery, 15 is a charger, 16 is a signal generator, 17 is a battery remaining capacity estimating unit, 18 is input means, 19
Denotes an arithmetic unit, 20 denotes an output unit, 12 denotes a battery thermometer, and 21 denotes a battery remaining capacity display unit.

The operation of the storage battery remaining capacity estimating system configured as described above will be described below with reference to FIGS. 7, 8 and 9. (Operation of Third Embodiment) In the third embodiment, a case where a neural network is used for the estimating unit will be described. First, similarly to the second embodiment, information necessary for an operation for estimation (e.g., an automobile, if necessary) is obtained from the load 1, the battery 2, the charger 15, the signal generator 16, the battery thermometer 12, and the like. Information on accelerator opening, climbing, descending, etc.)
Input means 1 of the remaining battery capacity estimating unit 17
Enter 8 This state information is represented by a vector. This vector amount becomes input information of the arithmetic unit 19 such as a neural network. The arithmetic unit 19 such as a neural network assumes a multi-layer structure model as shown in FIG. 8, for example. The input signal propagates from the input layer to the output layer via the intermediate layer. Now, assuming that the input information accelerates while climbing a hill and the remaining battery capacity is less than, for example, 40%, it is assumed that the generator is fully opened to supply a sufficient amount of current to the storage battery 2 and the load 1. In the learning, the value of the synaptic connection is changed by the back propagation (BP: back propagation) method so that the error between the desired output (teacher data) and the output that has just come out becomes zero. In the case of BP, the value of the synaptic connection is changed in the reverse direction from the output layer to the input layer. This process is repeated until a desired output is obtained for all the input information, for example, the battery temperature, the accelerator opening, and the charge / discharge state. For all possible input information, learning ends when a desired output is obtained such that the error between the teacher data and the NN (Neural Network) output is equal to or less than an allowable value.

FIG. 9 shows the relationship between the remaining battery capacity of the storage battery and the open circuit voltage. As shown in the figure, the difference between the open voltage and the remaining battery capacity is almost 80% to 20% of the remaining battery capacity, and it is difficult to estimate the remaining battery capacity with sufficient accuracy by using simple state information as input information. . As in the second embodiment, for example, the impedance Z of the storage battery shown in FIG.
The measurement result of (w) is calculated, and predetermined angular frequencies W0 and W
The radius ri which is a feature quantity is obtained in advance from the size of a circle drawn from the real part Re (Z (wi)) and the imaginary part Im (Z (wi)) of 1, ---, Wi, ---, Wn. . Then, a signal corresponding to a predetermined angular frequency that gives the characteristic amount of the storage battery remaining capacity is superimposed on the charging / discharging current during operation, and the impedance Z (w) of the frequency corresponding to the characteristic of the storage battery remaining capacity can be measured. Let it be a part of the input information vector amount of the neural network operation.

Next, a storage battery remaining capacity estimation system according to a third embodiment of the present invention will be described with reference to the block diagram of FIG. 7, in a hybrid electric vehicle or the like, for example, charge / discharge is repeated by a load 1, such as a motor, a storage battery 2, and a charger 15. From the load 1, the battery 2, the charger 15, the signal generator 16, the battery thermometer 12, and the like, information necessary for the calculation for the estimation (information such as the accelerator opening degree of the vehicle, climbing, descending, etc., as necessary) ) Is input to the input means 18 of the remaining battery capacity estimating unit 17 as state information. Also, as in the second embodiment,
The voltage and the current of the storage battery are passed through a noise eliminator to remove an AC signal having a frequency other than a predetermined frequency, and are converted by an A / D converter provided in the input means 18. Then, the complex spectrum of the current and the voltage is converted by the frequency converter. From the complex spectra Sv (w) and Si (w) of the voltage V and the current I, and the ratio Sv (w)
/ Si (w) is obtained by the input means 18 as the impedance Z (w) of the storage battery. Predetermined angular frequencies W0, W1,-
The magnitudes of the real part Re (Z (wj)) and the imaginary part Im (Z (wj)) of-, Wj, ..., Wn are obtained from the storage battery that is repeatedly charged and discharged. However, in the case of a load such as an operating motor, the charge / discharge current may not have a signal in the frequency range of w0, w1, w2, and w3. At that time, the charge / discharge current at each frequency wj that gives the characteristic amount is superimposed on the active battery charge / discharge current. Thus, each frequency w
The real part Re (Z (wj)) and the imaginary part Im (Z (wj)) of the impedance at j are obtained. And the second
In the same manner as in the embodiment, the radius rj, which is the feature amount, is obtained.
This radius rj is used as a part of the input information vector amount for the neural network operation. Desired outputs are obtained for all possible input information. The output means 20 outputs the obtained remaining battery capacity.

As described above, according to the present embodiment, the complex spectrum of the voltage V and the current I when the battery is repeatedly charged and discharged is obtained, the impedance of the battery is calculated, and the real part Re in a predetermined angular frequency range wj is calculated. From the size of a circle drawn from (Z (wj)) and the imaginary part Im (Z (wj)), a radius rj which is a feature amount is obtained. This radius rj is used as a part of the input information vector amount for the neural network operation. This is repeated until desired output can be obtained for all input information, for example, storage battery temperature, accelerator opening, and charging / discharging states. Learning is completed when desired output can be obtained for all conceivable input information. By comparing the calculated remaining radius with the radius ri previously obtained and stored in the remaining battery capacity calculator 11, the remaining battery capacity of an operating hybrid electric vehicle or the like, which has conventionally been estimated with low accuracy, can be calculated. It can be estimated with high accuracy.

In the first embodiment, the size of the circle drawn from the real part Re (Z (wj)) and the imaginary part Im (Z (wj)) of the impedance at a predetermined angular frequency wj giving the characteristic amount. From the above, the radius rj is obtained, and the obtained value is compared with the radius ri previously obtained and stored in the battery remaining capacity calculation unit to estimate the storage battery remaining capacity. It may be an absolute value or an absolute value of a plurality of complex impedances. In addition, the center coordinate and absolute value of one or a plurality of complex impedance circles in which the feature quantity draws a circle, the center coordinate of one or more complex impedance circles in which one or more complex impedances draw a circle, one or more complex impedances in which a circle is drawn And the absolute value of the complex impedance of the circle.

In the second embodiment, the characteristic value of the impedance of the storage battery during running obtained by superimposing the charge / discharge current reference waveform of the sine wave and the charge / discharge current waveform is compared to estimate the remaining capacity of the storage battery. However, the sinusoidal charge / discharge current reference waveform may be a rectangular wave or a sawtooth wave. Further, in the second embodiment, the storage battery remaining capacity is estimated by comparing the characteristic amount of the impedance during running of the storage battery obtained by superimposing the charge / discharge current reference waveform of the sine wave and the charge / discharge current waveform. Alternatively, the charge / discharge current waveform and the sinusoidal charge / discharge current reference waveform may be superimposed during running and stopping. Further, the charge / discharge current reference waveform of the sine wave may be a sine wave of a plurality of different frequencies, and the sine waves of each frequency may be sequentially superimposed on the charge / discharge current waveform.

In the third embodiment, the remaining capacity of the storage battery is estimated by using the radius rj, which is a feature quantity, as a part of the input information vector quantity of the neural network operation. The unit may use fuzzy inference or multiple regression calculation.

[0035]

As described above, the present invention relates to a method and a system for estimating the remaining capacity of a storage battery connected to a device comprising a storage battery and a load connected to the storage battery. By comparing the characteristic value of the impedance of the storage battery according to each temperature, each charging / discharging current, and the traveling state with the characteristic amount of the impedance of the storage battery obtained during traveling, and estimating the remaining capacity of the storage battery, the estimation accuracy has been conventionally known. It is possible to highly accurately estimate the remaining capacity of the storage battery that has no absolute reference, such as a hybrid electric vehicle that repeatedly charges and discharges while traveling. In addition, since there is hysteresis in the charge / discharge characteristics, the remaining capacity of the storage battery cannot be uniquely determined. The remaining capacity can be estimated with high accuracy.

FIG. 10 shows a measured value (A) when a current is applied to a conventional standard load to display the terminal voltage of the storage battery, and the remaining battery capacity is estimated using neural network calculations. (B) is shown. With the conventional measurement value (A), the estimation accuracy is low, and it is difficult to estimate the remaining battery capacity of an operating hybrid electric vehicle or the like. Can be estimated with high accuracy.

[Brief description of the drawings]

FIG. 1 is a block diagram of a storage battery remaining capacity estimation system according to a first embodiment of the present invention.

FIG. 2 shows an impedance measurement result of the storage battery remaining capacity estimation system for explaining the operation in the first embodiment.

FIG. 3 shows a measurement result of a remaining capacity of a storage battery and an impedance which is a feature amount in the first embodiment.

FIG. 4 is a block diagram of a storage battery remaining capacity estimation system according to a second embodiment of the present invention.

FIG. 5 shows a result of impedance measurement of a storage battery remaining capacity and a feature quantity in the second embodiment.

FIG. 6 shows a charge / discharge current waveform on which a sine wave is superimposed.

FIG. 7 shows a block diagram of a storage battery remaining capacity estimation system according to a third embodiment of the present invention.

FIG. 8 shows a neural network operation unit having a multilayer structure (three-layer structure including an input layer, an intermediate layer, and an output layer).

FIG. 9 is a relationship diagram between the remaining battery capacity of the storage battery and the open circuit voltage.

FIG. 10 shows a measured value (A) when a current is applied to a conventional standard load to display the terminal voltage of a storage battery and a measured value when the remaining battery capacity is estimated by using a neural network operation. (B) is shown.

FIG. 11 shows a storage battery remaining capacity display device of a first conventional example.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 Load 2 Battery 3 Voltage noise eliminator 4 Current noise eliminator 5 Voltage A / D converter 6 Current A / D converter 7 Voltage frequency converter (FFT: Fast Fourier Transformer)
form) 8 Current frequency converter (FFT) 9 Division calculator 10 Impedance calculator 11 Battery remaining capacity calculator 12 Battery thermometer 13 Battery remaining capacity temperature correction unit 14 Engine unit 15 Charger 16 Signal generator 17 Battery remaining capacity Estimation unit 18 Input unit 19 Operation unit such as neural network 20 Output unit 21 Battery remaining capacity display unit

 ──────────────────────────────────────────────────続 き Continuing from the front page (72) Inventor Hirotaka Ishihara 1006 Kadoma Kadoma, Osaka Prefecture Matsushita Electric Industrial Co., Ltd.

Claims (11)

[Claims] (A) measuring voltage and current signals during charging and discharging of a storage battery connected to a load; (b) mapping the measured voltage and current signals as complex spectra in the frequency domain, respectively; And (d) calculating impedance and impedance characteristic amounts from complex spectra of voltage and current in a predetermined frequency region; and (d) determining impedance characteristics based on a relational expression between the previously determined impedance characteristic amounts and the remaining capacity of the storage battery. A method for estimating a remaining capacity of an operating storage battery, comprising estimating a remaining capacity of the storage battery from the amount. 2. The method according to claim 1, wherein the characteristic quantity of the impedance is a radius of one or more circles drawn in a predetermined frequency range when the impedance is plotted on a complex plane using frequency as a parameter. The remaining battery capacity estimation method described in the above. 3. 1 kHz for providing a characteristic amount of impedance
An AC signal having one or more different frequencies equal to or less than z,
2. The method according to claim 1, wherein the method is superimposed on the charge / discharge current of the storage battery. 4. The method for estimating a remaining capacity of a storage battery according to claim 1, wherein an AC signal other than providing an impedance characteristic amount is removed by a noise removing device. 5. A relational expression between a characteristic amount of impedance corresponding to each temperature, each charge / discharge current, and a running state and a remaining capacity of each storage battery corresponding to each remaining capacity of the storage battery, and a charging time obtained during the running. 2. The remaining capacity estimation method according to claim 1, wherein the remaining capacity of the storage battery is estimated using a neural network, fuzzy logic, or multiple regression calculation from the characteristic amount of the impedance of the storage battery at the time of discharging. 6. A storage battery and a load connected to the storage battery, a voltage measuring device and a current measuring device, a device for frequency-converting the voltage and the current, respectively, and a storage battery remaining capacity estimating means. Voltage measuring means at both ends, means for connecting the current measuring means flowing to the storage battery, means for mapping the voltage and current measured by the voltage measuring means and the current measuring means to the frequency domain, respectively, from the mapped voltage and current A storage battery remaining capacity estimation system comprising means for calculating impedance and means for estimating remaining battery capacity from the impedance. 7. A means for removing a frequency component other than providing a characteristic amount of impedance by using a noise removing device for a voltage and a current measured by voltage measuring means and a current measuring means connected to both ends of a storage battery, and charging and discharging. 7. The storage battery according to claim 6, further comprising means for mapping in a frequency domain according to the charge / discharge current reference waveform generated by the current reference waveform generation device, and calculating impedance from the mapped voltage and current. Estimation system. 8. The system according to claim 7, wherein the noise removing device is a band-pass filter. 9. The storage battery remaining capacity estimation system according to claim 7, wherein an A / D converter and a frequency converter (FFT) are used when frequency-converting the voltage and the current, respectively. 10. Detecting the absence of a signal in a frequency domain that provides an impedance feature, and superimposing a signal corresponding to the frequency that provides the feature on a charge / discharge current using a charge / discharge current reference waveform generator. The storage battery remaining capacity estimation system according to claim 7, characterized in that: 11. An input means for inputting charge / discharge current / voltage information and load state information to the storage battery remaining capacity estimating means,
It has knowledge data in which various cases for estimating the remaining battery capacity based on the experience data accumulated through actual operation are learned in advance by an algorithm, and based on this knowledge data, each information provided from the input unit is obtained. 7. An estimating unit for performing processing to estimate the remaining battery capacity, and output means for providing instruction information to the generator according to the output value estimated by the remaining battery capacity estimating unit. A storage battery remaining capacity estimation system according to the above.