WO2000031557A1

WO2000031557A1 - Method and apparatus for determining characteristic parameters of a charge storage device

Info

Publication number: WO2000031557A1
Application number: PCT/KR1999/000345
Authority: WO
Inventors: Chul Oh Yoon; Yevgen Barsukov; Jong Hyun Kim
Original assignee: Korea Kumho Petrochemical Co., Ltd.
Priority date: 1998-11-19
Filing date: 1999-06-28
Publication date: 2000-06-02
Also published as: ATE303601T1; EP1153311B1; DE69927058D1; JP2000156248A; EP1153311A1; TW444128B; KR20000033041A; DE69927058T2; US6160382A; KR100395516B1; JP3190313B2; AU4532099A

Abstract

Provided with a method and an apparatus of determining characteristic parameters of a charge storage device based on wide frequency range of impedance measurement and a non-linear equivalent circuit model by which the parameters of the non-linear equivalent circuit model indicative of the characteristics of various charge storage devices such as primary battery, secondary battery, capacitor, supercapacitor and fuel cell, the method comprising the steps of: 1) measuring voltage and current characteristics in process of charging/discharging of charge storage device by applying a voltage/current at a predetermined discharge rate; 2) measuring impedance spectra at a predetermined range of frequency by measuring the current and voltage from both terminals of the charge storage device or from an electrical load directly connected to the charge storage device at a plurality of states of within the entire charge/discharge interval; and 3) obtaining the parameters of the non-linear equivalent circuit of the charge storage device from the charge or discharge characteristics measured in step (1) and the characteristic impedance spectrum in the predetermined range of frequency measured in step (2).

Description

METHOD AND APPARATUS FOR DETERMINING CHARACTERISTIC PARAMETERS OF A CHARGE STORAGE DEVICE

Technical Field

The present invention relates to a method and apparatus for determining

characteristic parameters of a charge storage device based on wide frequency range of

impedance measurement and a non-linear equivalent circuit model by which the

parameters of the non-linear equivalent circuit model indicative of the characteristics of

various charge storage devices such as primary battery, secondary battery, capacitor,

supercapacitor and fuel cell are determined.

Background Art

In general, charge storage devices of all types such as primary battery,

secondary battery capacitor, supercapacitor and fuel cell are manufactured in

accordance with various product standards based on their material, size and type.

Rechargeable secondary batteries are classified further into lead-acid, nickel-

cadmium, nickel-metalhydride. lithium ion and lithium polymer battery according to

their chemistry and material.

These secondary batteries are different in various aspects such as average

discharge voltage, discharge voltage profile, internal resistance, limiting discharge

current, temperature characteristics, charge cut-off voltage and the like.

In order to select a battery appropriate to the requirements of particular

application among various battery products, many kinds of factors are to be considered

such as battery capacity, size, price of the product or the like including characteristics of the battery described above. As a reference for the choice of the battery for suitable

application it is required to use a method for quantitatively representing the

performance of a battery.

For example, a battery used to drive a DC motor must have the drive voltage

greater than the minimum operational voltage of the DC motor. A voltage drop caused

by the internal resistance of the battery during operation has a very significant effect

upon the operating time of the DC motor depending upon the size of the drive current.

For a battery used as a portable power source for GSM or CDMA type cellular

phone, precise analysis is needed for the characteristics of battery in association with

the discharge of periodic high current pulses. A rechargeable battery for electric vehicle

requiring an instantaneous high power also has to be analyzed precisely particularly for

the transient current response.

As a method to analyze the operation characteristics of battery precisely, it is

possible to examine and analyze the internal parameters related to the mechanism of a

battery as well as general characteristics such as battery capacity, average discharge

voltage, discharge voltage profile, internal resistance, temperature characteristic and

charge cut-off voltage.

These internal parameters of a battery can be determined through measurement

of impedance characteristic of the battery at a wide range of frequency.

The present inventors have applied for a patent on an effective method for

measuring and predicting battery capacity with the results of this study (Refer to.

Korean patent application Nos. 22540 (1998) and 24134 (1998)).

The most effective method for selecting a battery for the appropriate purpose is

to represent an application device including a specified battery as an equivalent circuit and then simulate it in a numerical manner.

This method can evaluate the efficiency and stability of the device by a virtual

combination of a selected battery and an electric or electronic device as an electronic

circuit using a computer.

Such a numerical simulation of electric and electronic devices has been

developed as a general computer program such as SPICE at California Berkely

University in America and widely used in many research institutes and manufacturing

companies.

A conventional equivalent circuit of battery is however oversimplified in

construction and has been utilized for the simulation of devices rather than the battery

itself.

The conventional equivalent circuit of battery is made up of a constant voltage

source and serial resistors without consideration upon the variation of discharge voltage

over time. Thus the model works effectively only for a short period of discharge time

under a DC current.

A method using the Peukert parameter disclosed in U.S. Patent No. 4,952,862 is

not appropriate to precisely describe the discharge characteristics of a battery since it

employs a phenomenological model which simply approximates the discharge profile

linearly at the initial stage of discharge and exponentially at the late stage of discharge.

A method for simulating a thermal battery as described in U.S. Patent No.

5.428.560. which co\ ers discharge voltage and internal resistance of the battery, is valid

in a case of a long time DC discharge but cannot describe sufficiently the characteristics

of the batten' under transient conditions of discharge. The non-linear circuit model of a battery suggested in the present invention is

designed to calculate the voltage response of a battery with respect to a predetermined

electrical load such as DC, AC, pulse or impedance, which implements a precise and

effective simulation of the operational characteristics of electrical circuitry including a

battery.

Disclosure of Invention

An objective of the present invention is to provide a method and apparatus for

characterizing internal parameters of a charge storage device based on wide frequency

range of impedance measurement and a non-linear equivalent circuit model, which can

be used for quality control of battery products and testing of batteries by analyzing or

predicting the operational properties of charge storage devices such as primary battery,

secondary battery, capacitor, supercapacitor and fuel cell or condenser using computer,

or can be used to determine the parameters of the non-linear equivalent circuit model of

the charge storage device in order to design an electric/electronic device and batteries

appropriate for this device through digital simulation of an electric circuit.

To accomplish the objective of the present invention, provided is a method and

apparatus for characterizing internal parameters of a charge storage device based on

wide frequency range of impedance measurement and a non-linear equivalent circuit

model, which is to experimentally characterize internal parameters of the charge storage

device such as primary battery, secondary battery capacitor, supercapacitor and fuel cell,

the method including the steps of: ( 1 ) measuring voltage and current characteristics by-

applying a voltage/current having a predetermined discharge rate to the capacitor to charge/discharge the capacitor; (2) measuring a characteristic impedance spectrum in a

predetermined range of frequency from the current and voltage applied to both

terminals of the capacitor or a directly connected impedance load in a plurality of

charge states within the entire charge/discharge interval; and (3) characterizing the

parameters of the non-linear equivalent circuit model of the charge storage devices

from the voltage characteristic measured in step (1) and the impedance spectra in the

predetermined range of frequency measured in step (2). The present invention suggests

the numerical model and the equivalent circuit of a battery for accurately prediction and

description of the operational performance of the battery under a well-defined battery

charge/discharge conditions within 100 percent depth of discharge, such as DC

discharge of at least C/0.5 discharge rate, AC discharge in the MHz to 20 kHz range of

frequency, current density corresponding to at least C/0.5 rate, pulse discharge in the

same frequency range, or discharge for a fixed or variable impedance load. The present

invention implements methods of experimental measurement, numerical analysis and

nonlinear equivalent circuit, and an apparatus for the measurement and analysis.

Brief Description of Drawings

FIG. 1 is an equivalent circuit diagram based on the transmission line model of

a battery:

FIG. 2 is a graph in which an impedance spectrum measured in a fully charged

lithium ion battery (manufactured by Sony) is compared to an impedance spectrum

derived from parameters obtained by fitting to a static equivalent circuit model:

FIG. 3 is an dynamic model of equivalent circuit considering nonlinear behavior of model parameters at each state of charge of a battery;

FIG. 4 is a block diagram showing the construction of an apparatus for

characterizing parameters of a charge storage device in the present invention;

FIG. 5 is a signal flow chart showing a parameterization method of the present

invention;

FIG. 6A is a graph showing experimental and simulation results on discharge

characteristics at various discharge rates for a lithium ion battery manufactured by

Sony;

FIG. 6B is a graph showing experimental and simulation results on discharge

characteristics at various discharge rates for a lithium ion battery manufactured by Matsushita;

FIG. 6C is a graph showing experimental and simulation results on discharge

Sanyo;

FIG. 6D is a graph showing experimental and simulation results on discharge

characteristics at various discharge rates for a nickel hydrogen cell manufactured by

Hitachi;

FIG. 7 is a graph showing experimental and prediction results on transient

discharge characteristics upon on pulse currents applied to a lithium ion battery

manufactured by Sony:

FIG. 8 is an equivalent circuit diagram used to predict operation of DC motor

powered by a battery:

FIG. 9A is a graph showing experimental and prediction results on rotation speed of a DC motor connected in series with a battery;

FIG. 9B is a graph showing experimental and prediction results on the discharge

voltage of a battery connected in series with a DC motor; and

FIG. 10 is a graph showing experimental and prediction results on the discharge

characteristics of a battery pack comprising four serial and parallel lithium ion cells

manufactured by Matsushita.

Best Mode for Carrying out the Invention

Hereinafter, a method and apparatus for characterizing internal properties of a

charge storage device based on wide frequency range of impedance measurement and a

non-linear equivalent circuit model according to the present invention will be described

in detail with reference to the accompanying drawings.

The equivalent circuit model of a battery proposed in the present invention is a

circuit model in which the electrochemical mechanism associated with active layers of

anode and cathode constituting the battery is described with a transmission line model

as shown in FIG. 1.

FIG. 1 is an equivalent circuit diagram for a transmission line model which is

made up of macroscopic resistance p of electrode active layers, impedance Z_s of thin

electrode

ers cut infinitely in the vertical direction to a current collector, and serial

resistance R_^eι relevant to the interconnecting resistance through electrolyte, current

collectors and wires.

In FIG. 1. impedance Z_s of electrode layers is expressed by an equivalent circuit

made up of solid state diffusion coefficient D inside the electrode-forming particles. charge storage capacitance C_s corresponding to the chemical reaction associated with

energy storage, double layer capacitance C_d] and charge transfer resistance R_ct.

Such a circuit model is proven to be effective in describing the discharge

mechanism of active materials such as lithium intercalation compounds.

In a case where the cathode and anode of the battery can be described by the

same model, the equivalent circuit model having the electrodes in series can be of the

same form.

The equivalent circuit of a single battery comprising cathode and anode can be

constructed in consideration of the impedance of electrolyte as serial resistance R_s.

The validity of static equivalent circuit model comprising linear elements such

as resistor and capacitor at given state of charge of battery can be verified as shown in

FIG. 2 by measuring the impedance spectrum of the battery and then by fitting the

spectrum to the equivalent circuit by using complex nonlinear least square method.

From that result, the characteristic factors are determined.

This static equivalent circuit model is valid only when the change of battery-

state is small. If the discharge time is long or the discharge current is high, however,

consideration must be taken upon the nonlinear variation of parameters on the state of

charge of the battery.

The impedance spectrum for several state of charge of the battery is measured

and anah zed within the entire discharge interval of the battery, so that the parameters

for the static equivalent circuit are determined as a function of the state of charge.

On the other hand, charge storage capacitance C_s normally has a large value and

is difficult to measure within an appropriate period of time through the impedance method.

Accordingly, a discharge voltage curve is obtained for the entire discharge

interval of the battery and the charge storage capacitance C is determined from the

slope of discharge voltage profile.

The non-linear equivalent circuit model designed in consideration of the

nonlinear variation of parameters depending upon the state of charge of the battery, as

shown in FIG. 3. comprises the finite number of ladder-type two-terminal networks in

place of the transmission line, and nonlinear resistors and nonlinear capacitors whose

values depends upon voltage influenced by charge storage capacitance C_s.

A method of numerical simulation of operation voltage or current of a battery

provided by the present invention can be achieved by directly solving the

corresponding system of non-linear first order differential equations or by using the

finite-difference method used in an electronic simulator such as SPICE.

The desired precision of the solution can be obtained by control of the number

of ladder-type two-terminal networks representing the transmission line.

The results of simulation based on non-linear equivalent circuit model of a

battery correspond to the experimental results of measurement for the discharge

characteristics of a battery with accuracy, which will be later described in detail with

reference to the embodiments.

A method of obtaining parameters in the non-linear circuit model of a batten

proposed in the present invention will now be described.

A battery for test is fully charged by following predefined procedure, and the

impedance spectrum is measured in a certain range of frequency after the open circuit voltage of the battery reaches its equilibrium.

Provided that the total number of impedance measurements at different states of

charge is N. the battery is discharged by the amount of charge Q/(N-1) at a constant

current I where Q is the discharge capacity of the battery. After the open circuit voltage

of the battery reaches the equilibrium again, the next measurement of impedance

spectrum is performed. Such a procedure of discharge and impedance measurement is

repeated until the battery reaches the predefined end of discharge.

In this procedure, the discharge rate of the battery can be selected within a range

that the dE/dq of discharge voltage E according to the change in the amount of

discharge q of the battery is not changing as discharge rate as further decreased.

Normally C/10 rate can be applied.

One of the parameters, charge storage capacitance C_s is calculated by

multiplying dE/dt of discharge voltage E over time t by current I.

The range of frequency in measurement of impedance must be wide enough to

provide information concerning time constants associated with the electrochemical

reaction of the battery.

The high frequency region is usually limited to about 60 kHz due to the linearity

of an electrochemical system, but actually for a battery it is limited to 10 kHz

corresponding to a pulse of 0.1 msec interval for a battery.

The low frequency limit is determined from the specification of a measurement

instrument and the time required for measurement. The lowest frequency is important

to describe accurately the change of parameters associated with the reaction

characteristic of the battery for the entire discharge interval. The lower frequency is usually near 1 mHz.

A frequency response analyzer (FRA) has been typically used for a method of

measuring an impedance spectrum in which the frequency is sequentially scanned to

obtain the impedance from input and output signals.

The FRA. however, has a disadvantage that too much time is required for

measurement because the frequency scan is performed one by one from the minimum

frequency to the maximum.

Another method for measuring impedance spectra is a Fourier transform method

in which a perturbation current signal is applied to both terminals of a battery under

galvanostatic condition and a voltage response signal recorded in time domain is

Fourier-transformed to frequency domain to calculate the impedance impedance

spectrum.

The perturbation current signal used as an input signal can be obtained by

superposition of multiple sine waves at selected frequencies.

If the minimum frequency of measurement is f_mιn. for example, the perturbation

current signal has an odd multiple of the minimum frequency f_mιn such as 3f_mm. 5f_mm.

7f_mιn or the like. This Fourier transform method using non-overlapping multiple

frequencies is different from another Fourier transform method using a pulse (U.S.

Patent No. 5.633.801).

The principle of the measurement is described in detail by G. S. Popkirov and R.

N. Schindler. Rev. Sci. Instrum.. 63. 5366 (1992).

The maximum frequency measurable through the Fourier transform method

using non-o\ erlapping multiple frequencies is limited depending upon the sampling time of a signal detector. The time required for measurement corresponds to twice of

the inverse of the lowest measurement frequency, because of waiting time necessary for

excluding transient effects.

Compared to the FRA, the Fourier transform method is advantageous in that the

time required to measure the impedance is reduced by half or more.

Linearity of measured system is preserved if amplitudes of selected frequencies

are at least 3 orders of magnitude larger than that of additional frequencies obtained

during Fourier-transform impedance measurement.

The linearity of a measured system at given current applied to the battery in case

of Fourier transform impedance measurement can be checked by comparing the

amplitude of complex voltages for a selected frequencies and additional frequencies.

Another method for measuring impedance spectra is to use a simple current

waveform instead of multiply superposed perturbation current signals.

For example, a constant current I(t) = I₀ is applied to both terminals of a battery

and a linear or non-linear least square regression is performed based on the Laplace

transform of response voltage signal and input current measured in time domain to

calculate the impedance spectrum.

If the sampling time is t_samp and the time required for measurement is t_mcs. the

valid frequency range is l/2t_mes to l/2t_samp.

Assuming that the sampling speed is 1kHz and the time for measurement is 500

sec. for example, the frequency range is 1 mFIz to 500 Hz.

This demonstrates that the time required for measurement is shorter than that in

the Fourier transform method. 2000 sec. because no waiting time is necessary to remove transient effects.

If I(s) is the Laplace transformed function of I(t). the impedance function in a

Laplace space is Z(s) = V(s)/I(s).

The Laplace transformed impedance spectrometer is advantageous in that a

desired spectrum can be obtained by numerical analysis faster than FRA or Fourier

transform method.

The parameters except for charge storage capacitance Cs are determined by

fitting of the impedance spectrum measured at different state of charge of a battery to

the static model. The algorithm used for fitting can be made of complex non-linear least

square fitting described in J. R. Macdonald, J Schoonman and A. P. Lehnen, Solid State

Ionics V5. 137(1981), or non-linear fitting using a neural network or any other suitable

fitting methods.

The initial conditions used in the fitting may be arbitrarily chosen and. if the

result of the previous fitting is used as initial conditions for the successive fitting, the

performance and quality of the fit can be improved.

FIG. 4 is a block diagram showing the construction of a battery

parameterization system apparatus according to the present invention.

In the figure, reference numeral 10 denotes control means of determining

parameters of charge storage device 20 with a computer or the like and simulating

various operational characteristics of an electronic circuit including battery.

The control means 10 has an I/O interface unit 1 1 for interfacing the charge

storage device 20. a central processing unit(CPU) 13 for controlling the measurement

of voltage, current and impedance for the charge storage device 20. a voltage/current measuring means 15 for measuring the voltage and current characteristics of the charge

storage device 20 under the control of the central processing unit 13. a characteristic

impedance measuring means 17 for measuring the impedance of the charge storage

device 20 under the control of the central processing unit 13. and a parameterization

means 19 for obtaining the parameters of the non-linear equivalent circuit model of the

charge storage device 20 with the voltage and current characteristics of the charge

storage device 20 measured by the voltage/current characteristic measuring means 15

and the characteristic impedance of the charge storage device 20 measured by the

characteristic impedance measuring means 17.

Reference numeral 30 denotes a digital-to-analog (D/A) converter for

converting voltage and current data output from the control means 10 via the interface

unit 11 into analog signals, reference numeral 40 an amplifier for amplifying the output

signal of the digital-to-analog converter 30. reference numeral 50 a galvanostat for

feeding the output signal of the amplifier 40 into the capacitor 20 to output the voltage

and current of the capacitor 20.

The galvanostat 50 has an inversion amplifier 51 for inversion amplifying the

output signal from the amplifier 40 and applying it to the charge storage device 20.

buffers 53 and 55 for generating voltages at both terminals of the charge storage device

20. and a current-detecting resistance 57 connected in series to the charge storage

device 20 for detecting the current of the charge storage device 20.

Reference numerals 60 and 70 denote first and second filters/amplifiers for

filtering and amplify ing the voltage of the charge storage device 20 output from the

buffers 53 and 55 of the galvanostat 50 and the voltage based on the current of the charge storage device 20 detected by the current-detecting resistor 57. respectively.

Reference numeral 80 represents an analog-to-digital (A/D) converter for converting

the output signals of the first and second filters/amplifiers 60 and 70 into digital signals

and applying them to the I/O interface unit 1 1 of the control means 10.

Reference numeral 90 denotes a battery circuit simulator for simulating different

operational characteristics according to the values of the parameters obtained by the

parameterization unit 19 of the control means 10 for a specified battery circuit

including the same battery as the charge storage device 20.

If the voltage and current characteristics are measured first in the battery

parameterization system of the present invention with such a construction, the control

means 10 has the voltage/current characteristic measuring means 15 outputting voltage

and current digital data of a predetermined level via the I/O interface unit 1 1 under the

control of the central processing unit 13.

The digital voltage and current data are converted into analog signals at the

digital-to-analog converter 30. boosted through the amplifier 40. inversion amplified

via the inversion amplifier 51 of the galvanostat 50 and then applied to the charge

storage device 20. Thus voltage and current based on the digital output data from the

voltage characteristic measuring means 15 are produced in the charge storage device

20.

After a predetermined period of time under conditions described above, the

control means 10 applies the voltage and current of the charge storage device 20 to

gah anostat 50. first and second filters/amplifiers 60 and 70. and analog-to-digital

converter 80.

lδ The voltages at both terminals of the charge storage device 20 are fed into the

first filter/amplifier 60 via buffers 53 and 54 of the galvanostat 50. while the current of

the charge storage device 20 flows through the current-detecting resistor 57 in such a

manner that a voltage drop occurs in the current-detecting resistor 57. The voltages at

both terminals of the resistor 57 based on the current flowing through the current-

detecting resistor 57 are input to the second filter/amplifier 70.

Then, the first and second filters/amplifiers 60 and 70 filter the voltages at both

terminals of the capacitor 20 and the detected voltage based on the current of the charge

storage device 20 to remove noises and for amplification, respectively. The output

signals of the first and second filters/amplifiers 60 and 70 are converted into digital data

at the analog-to-digital converter 80 and applied to the I/O interface unit 1 1. The

voltage/current characteristic measuring means 15 is receptive to the digital data from

the voltages at both terminals of the charge storage device 20 and the detected voltage

based on the current of the charge storage device 20, which voltages are applied to the

I/O interface unit 11 , and then measures the voltage and current characteristics of the

charge storage device 20.

The control means 10 can use a separate voltage/current generator (not shown)

without generating voltage and current digital data to charge the charge storage device

20 directly, so that the voltage/current generator outputs predetermined voltage and

current, charging the charge storage device 20 under the control of the control means

10.

The characteristic impedance measuring means 17 measures the characteristic

impedance of the charge storage device 20 under the control of the central processing unit 13.

The measurement of the characteristic impedance includes FRA. Fourier

transform and Laplace transform methods, one or more of which may be used solely or

in combination in measurement of the characteristic impedance.

In the FRA method using the sine waves of predetermined frequency, digital

data at a predetermined frequency is output and applied to the charge storage device 20

via digital-to-analog converter 30. amplifier 40 and galvanostat 50 sequentially to

charge the charge storage device 20. The voltage and current of the charge storage

device 20 are input and stored in the control means 10 through galvanostat 50. first and

second filters/amplifiers 60 and 70 and analog-to-digital converter 80.

After the completion of such a measurement for one predetermined frequency,

the digital data of another frequency are output in sequence, and subsequently, the

voltage and current of the charge storage device 20 are measured and stored.

In the Fourier transform method, the voltage or current digital data in the

specified form is stored beforehand. The stored digital data is applied to the charge

storage device 20 through digital-to-analog converter 30, amplifier 40 and galvanostat

50 in sequence, or signals based on the digital data are generated from a separate

voltage/current generator and applied to the charge storage device 20.

The voltage and current of the capacitor 20 based on the applied signals are

input and stored in the control means 10 via galvanostat 50. first and second

filters/amplifiers 60 and 70 and analog-to-digital converter 80.

When the measurement of the characteristic impedance of the charge storage

device 20 is completed, Fourier transform and/or Laplace transform, the central processing unit 13 converts the stored result of measurement into a complex impedance

value as a function of frequency.

After a predetermined number of the repeated measurements for the voltage

characteristic and the characteristic impedance, numerical differentiation of discharge

curve and complex non-linear least square fitting are performed in parameterization

unit 19 and the characteristic impedance values of charge storage device in order to

obtain the parameters of the non-linear equivalent circuit.

The obtained parameters are used to simulate various operational properties of

the charge storage device in the charge and discharge process at arbitrary load by using

an electronic circuit simulator 90.

Simulation can be performed for the operational characteristics of a circuit

including battery, e.g.. pulse, AC. impedance and transient characteristics as well as

standard characteristics of battery, e.g.. constant voltage or constant current

charge/discharge characteristics.

The control means 10 is. for example, made up of a computer system or

hardware which is capable of obtaining parameters of the charge storage device 20 and

simulating various operational characteristics of electronic circuit including a charge

storage device by using the parameters.

FIG. 5 is a signal flow chart showing a method for obtaining parameters of the

charge storage device 20 with control means 10 that is made up of a computer system.

As illustrated in FIG. 5. when the power is ON, the hardware makes a diagnosis

of itself and becomes initialized, in step S10. Control variables are input, in step SI 2.

The control variables to be input include sampling rate, measurement mode,

1? voltage, current, memory size, gain, and the like.

A measurement mode is activated in step SI 4.

If voltage/current measurement is chosen in step SI 4, a control variable

corresponding to voltage/current measurement is set among the input control variables

(SI 6) and the voltage/current being output to charge the charge storage device 20 (SI 8).

The voltage/current of the charge storage device 20 are input and stored, in step S22.

A decision is made upon as to whether or not the measurement is completed, in

step S24. If the decision indicates that the measurement is uncompleted, the flow of the

process goes back to step S14 in which the measurement mode is selected and

activated.

If frequency scan is chosen in step SI 4, a control variable corresponding to the

frequency scan is set (S26) and a signal of one frequency among a plurality of

predetermined frequencies being output to charge the charge storage device 20 (S28).

The voltage and current of the charge storage device 20 are then input, in step S30.

Use is made of amplitude/phase shift or Fourier transform method to calculate

the value of the input voltage/current, in step S32, and impedance Z(f) is determined

from the calculated voltage/current value, in step S34.

In step S36. a charge storage device is made upon as to whether or not the

measurement of predetermined entire frequencies is completed. If the decision indicates

that the measurement is not completed, the flow of the process returns to step S28 in

which a signal of the next predetermined frequency is output and the procedures for

measuring the voltage/current and then performing a calculation on the impedance Z(f)

are repeated. When the decision made in step S36 indicates that the measurement of

predetermined entire frequencies has been completed, the impedance Z(f) determined is

stored, in step S38.

If Fourier transform is chosen in step SI 4, a control variable corresponding to

the Fourier transform is set (S40) and a perturbation signal of multiply selected

frequencies being output to perturb the charge storage device 20 (S42). The voltage and

current of the charge storage device 20 are input, in step S44.

In step S46, Fourier transform is performed to calculate V(t) and I(t). The power

spectrum of V(t) and I(t) is then determined (S48), and impedance Z(f) being calculated

(S50) and stored (S52).

If Laplace transform is chosen in step SI 4, a control variable corresponding to

the Laplace transform is set (S54) and voltage and current data in a specified form

being output to charge the charge storage device 20 (S57). The voltage and current of

the charge storage device 20 are input, in step S58.

In step S60, linear regression is performed to calculate V(t) and 1(f). After

determination of impedance Z(f) in step S62, the error function of impedance Z(f) is

calculated (S64) and stored (S66).

When the parameters are measured by frequency scan. Fourier transform and/or

Laplace transform method after measurement of voltage/current, the parameterization is

performed (S68) and print out the result as a file or graph.

The present invention method and apparatus for battery parameterization system

will now be described with reference to preferred embodiments.

Embodiment 1 An apparatus for successively measuring constant current discharge of a battery

and impedance spectrum by Fourier transform method in a repeated manner is

constructed as follows.

The apparatus is designed to perform a repetitive procedure for applying a

predetermined current (for example, C/10) to a battery by using a galvanostat for a

defined period of time, reading out the voltage value measured on terminals of the

battery, and measuring the impedance spectrum of the battery in that state by using a

Fourier transform impedance spectrometer.

The impedance spectrometer is constructed in the way to apply an input current

signal having overlapped sine waves of frequencies which are a multiple of the

minimum frequency to the battery via 16-bit digital-to-analog converter and galvanostat,

transfer the measured digital current and voltage signals to a computer with a two-

channel 16-bit analog converter, and perform a calculation on complex impedance by

using a digital discrete Fourier transform algorithm.

In the impedance spectrum and the voltage curve that are obtained through a

predetermined number (for example, ten) of the serial experiments as described above

and parameters of static model (FIG. 1) are obtained for each spectrum, system of the

non-linear ordinary differential equation corresponding to the dynamic circuit model

shown in FIG. 3 is solved to simulate performance of battery under different load

conditions.

At this stage. N_d - 8 and N, = 4 in FIG. 3. The Livermore stiff ordinary

differential equation solver is used to solve the system of ordinary differential

equations. This calculation takes about 10 min. by using an personal computer with

pentium processor.

The predescribed method is applied to all four secondary batteries, three types

of lithium ion batteries having the regulated capacity of 1300 mAh (manufactured by

Sony, Sanyo Electric Co., Matsushita Electric Co.) and one type of nickel hydrogen

accumulator having the regulated capacity of 1300 mAh (manufactured by Hitachi), in

which the Fourier transform impedance spectrometer as described in Embodiment 1 is

used to obtain 10 impedance spectra measured at each state of charge decreased by 130

mAh until the batteries are completely discharged.

With the voltage curve and the impedance spectra in each state of charge, the

differentiation and complex nonlinear least square fitting is performed as described in

Embodiment 1 to calculate the parameters for the respective states of charge.

Table 1 shows the set of parameters for each battery in the full charge state.

Table 1

Embodiment 2

The discharge curve of each battery is obtained at discharge rates of 10 h, 5 h, 2

h and 1 h with the characteristic factors of four batteries calculated in Embodiment 1

and the battery circuit model shown in FIG. 3 through the prescribed numerical

simulation.

A constant current corresponding to each discharge rate is applied to each

battery and the voltage at this moment is read out and output so as to obtain the

discharge characteristics of battery at each discharge rate experimentally. Thus obtained

discharge characteristics are compared to the results of simulation, as shown in FIGs.

6A-6D.

As shown in FIGs. 6A-6D, the results of simulation are consistent with those of

experiment in all cases.

Embodiment 3

Another advantage of the present invention lies in that the transient discharge

characteristics of a battery under application of current pulses can be predicted.

A discharge current is applied to a fully charged lithium ion battery having a

regulated capacity of 1300 mAh (manufactured by Sony) at a discharge rate of 2 hrs

and placed in an open circuit for 900 sec to restore the voltage.

Charge and discharge pulse currents with duration of 400 sec are applied to the

battery at a discharge rate of 10 h again. With the above-specified current being applied virtually to the battery cell

model of Sony lithium ion battery obtained in Embodiment 1. the system of ordinary

differential equations is solved through a numerical analysis for the change of voltage.

During measurement, voltage and current are read out and compared to the

results of simulation, which is shown in FIG. 7.

As shown in FIG. 7, the results of experiment exactly correspond to those

predicted.

Embodiment 4

The present invention makes it possible to predict the change of voltage and

various electrical characteristics of an electrical load when a battery is connected to any

type of electrical load.

A dc motor is connected to a fully charged lithium ion battery having a

regulated capacity of 1300 mAh (manufactured by Sony), and the rotation speed and

the operational voltage of the motor are measured.

To measure the rotation speed of the motor, the front end of the DC motor is

connected to an optical chopper and the frequency of a light signal is measured.

The internal parameters of the DC motor used in the experiment are determined

by establishing an equivalent circuit consisting of capacitor and inductor, as illustrated

in FIG. 8. and measuring the impedance characteristic.

The change of voltage is simulated as a function of time from the battery circuit

model of Sony batten obtained in Embodiment 1 and the equivalent circuit of the DC

motor. The rotation speed of the motor is predicted from the change of voltage and

compared to the experimental values, which is shown in FIGs. 9A and 9B. As shown in FIGs. 9A and 9B. the results of experiment are exactly in accord

with those predicted.

Embodiment 5

Using the battery circuit model of a single battery, the discharging

characteristics of a battery pack consisting of a plurality of parallel or serial batteries

are predicted and compared to the results of experiment.

Four lithium ion batteries having the same nominal capacity of 1300 mAh

(manufactured by Matsushita) are used. The lithium ion batteries are divided into two

groups, where each of the group has two lithium ion batteries connected with each other.

The two groups are connected in parallel with each other to constitute a battery pack.

While the battery pack is subjected to a discharge at a discharge rate of 10 h. the

voltage is measured.

On the other hand, the discharge voltage of the battery pack is simulated based

on the circuit model of the Matsushita lithium ion battery obtained in Embodiment 1.

The predicted values are shown in FIG. 10.

As shown in FIG. 10. the results of simulation exactly correspond to those of

experiment. Although the present invention has been described with reference to the

preferred embodiments of specified cases and methods for the purpose of detailed

description, the invention is not limited to these preferred embodiments and includes

within the scope of the appended claims all equivalent methods.

For example, another measurement method may be applied instead of the

prescribed Fourier or Laplace transform in the measurement of the impedance spectrum,

independent measurement can be performed in order to obtain a discharge curve through experimental means. It is also possible to determine parameters by direct fit of

the time domain response signal of voltage and current to be measured in the

experiment for impedance for determining the parameters to another algorithm.

Another type of charge storage device such as primary cell or capacitor may be

analyzed based on the same circuit model.

Other various methods for numerical analysis can be adapted in the process for

obtaining the numerical solution for the operational characteristics of a battery

equivalent circuit. Especially, in the process using the transmission line model, the

system of differential equation can be directly solved or another numerical method

describing the transmission line model can be introduced instead of using a ladder-type

four-terminal network.

Industrial Applicability

Use of the numerical model of a battery proposed in the present invention

enables a precise diagnosis of a specified battery and may be used for the purpose of

quality control in the manufacture of battery or examination of battery conditions in

application devices.

Furthermore, with the battery numerical model of the present invention, a

simulation can be performed on the voltage response characteristics of the battery for

an electrical load such as DC. AC. pulse or impedance. A combination of the numerical

model with a common electric circuit digital simulator like SPICE enables a simulation

of the operational performance of the electrical circuit.

The electrical circuit simulation method using based on numerical model of a battery according to the present invention is applicable to the design of various

electric/electronic devices using batteries such as computing devices, communication

apparatus, power tools for driving a motor with batteries, electric vehicles, and batteries

appropriate to these devices.

Claims

WHAT IS CLAIMED IS :

1. A method of determining characteristic parameters of a charge storage device

based on w ide frequency range of impedance measurement and a non-linear equivalent

circuit model, which is to obtain the parameters of the charge storage device such as

primary battery, secondary battery, capacitor, supercapacitor and fuel cell, the method

comprising the steps of:

(1) measuring voltage and current characteristics in process of charging

/discharging of charge storage device by applying a voltage/current at a predetermined

discharge rate;

(2) measuring impedance spectra at a predetermined range of frequency by

measuring the current and voltage from both terminals of the charge storage device or

from an electrical load directly connected to the charge storage device at a plurality of

states of within the entire charge/discharge interval; and

(3) obtaining the parameters of the non-linear equivalent circuit of the charge

storage device from the charge or discharge characteristics measured in step (1) and the

characteristic impedance spectrum in the predetermined range of frequency measured in

step (2).

2. The method as claimed in claim 1. wherein the step (1 ) of measuring the voltage

and current characteristics comprises measuring voltage at both terminals of the charge

storage

ice when an input current having a predetermined discharge rate is applied

to the charge storage device, calculating an instantaneous change of the discharge

voltage and storing data at each state of charge.

3. The method as claimed in claim 1 or 2, wherein in step (1). the input current

having the predetermined discharge rate is sufficiently low enough to maintain steady-

state for the instantaneous change (dE/dq) of the voltage (E) with respect to a discharge

capacity (q) when it is applied to the charge storage device, the input current being

expressed in mA less than a value obtained by dividing a nominal capacity expressed in

mAh by one or more firs.

4. The method as claimed in claim 1 , wherein in step (2), the characteristic

impedance spectrum is measured by a frequency scan method in which a series of sine

waves having selected frequencies in a predetermined range of frequency are applied to

the charge storage device as an input signal and current and voltage signals measured at

both terminals of the charge storage device are analyzed.

5. The method as claimed in claim 1, wherein in step (2), the characteristic

impedance spectrum is measured by a Fourier transform method in which an input

current or voltage composed by superposing multiple waves at selected non-

overlapping frequencies in a predetermined range of frequency is applied to both

terminals of the charge storage device and thus obtained current and voltage are Fourier

transformed to obtain the characteristic impedance spectrum.

6. The method as claimed in claim 1. wherein in step (2). the characteristic

impedance spectrum is measured by a Laplace transform method in which voltage and current measured by applying a current or voltage having a specified waveform in a

predetermined range of frequency to the both terminals of the charge storage device or

connecting an electrical load to the charge storage device is Laplace transformed

through a numerical analysis to obtain the impedance spectrum.

7. The method as claimed in claim 1. wherein in step (2). the characteristic

impedance spectrum is measured by at least two methods selected from a frequency

scan method in which a series of sine waves having selected frequencies in a

predetermined range of frequency are applied to the charge storage device as an input

signal and current and voltage signals measured at both terminals of the charge storage

device are analyzed, a Fourier transform method in which an input current or voltage

composed by superposing multiple waves at selected non-overlapping frequencies is

applied to both terminals of the charge storage device and thus obtained current and

voltage are Fourier transformed, and a Laplace transform method in which voltage and

current measured by applying a current or voltage having a specified waveform in a

connecting an electrical load to the charge storage device is Laplace transformed by a

numerical analysis.

8. The method as claimed in any one of claims 1 and 4 to 7. wherein the

predetermined range of frequency is selected to have the ratio of the maximum

frequency to the minimum frequency greater than 100.

9. The method as claimed in claim 1. wherein the non-linear equivalent circuit of

the charge storage device in step (3) includes at least two elements selected from

selected from resistor, capacitor, inductor, transmission line or a finite number of

ladder-type two-terminals in place of the transmission lines.

10. The method as claimed in claim 1. wherein the non-linear equivalent circuit of

constant current source, constant voltage source, non-linear resistor, non-linear

capacitor, non-linear inductor, non-linear transmission line or a finite number of ladder-

type non-linear two-terminals in place of the transmission lines.

1 1. The method as claimed in claim 9 or 10. the ladder type two-terminals have a

predetermined number of two-terminal circuits in the range of 20 μsec to 2.000 sec of

time constant connected in parallel.

12. The method as claimed in claim 1. wherein the parameters in step (3) are

components constituting the non-linear equivalent circuit of the charge storage device

and include at least two elements selected from serial resistance, macroscopic

distributed resistance, charge transfer resistance, double layer capacitance, charge

storage capacitance, pseudocapacitance. diffusion coefficient, open circuit voltage and

instantaneous change rate of voltage.

13. The method as claimed in claim 1 or 12. wherein the parameters in step (3) are determined from a complex impedance spectra by a non-linear least square fitting.

14. The method as claimed in claim 1 or 12, wherein the parameters in step (3) are

determined from a slope of voltage characteristics vs. discharge capacity.

15. A method of determining characteristic parameters of a charge storage device

based on wide frequency range of impedance measurement and a non-linear equivalent

primary battery, secondary battery, capacitor, supercapacitor and fuel cell in a certain

state of charge, the method comprising the steps of:

(1) measuring a impedance spectrum in a predetermined range of frequency

from voltage and current applied to both terminals of the charge storage device in the

corresponding state of charge; and

(2) determining parameters of the non-linear equivalent circuit of the charge

storage device in the corresponding state of charge from the measured impedance

spectrum in the predetermined range of frequency.

16. The method as claimed in claim 15, wherein in step (1). the impedance

spectrum is measured by a frequency scan method in which a series of sine waves

having selected frequencies in a predetermined range of frequency are applied to the

charge storage device as an input signal and current and voltage signals measured at

both terminals of the charge storage device are analyzed.

17. The method as claimed in claim 15. wherein in step (1), the impedance

spectrum is measured by a Fourier transform method in which an input current or

voltage composed by superposing multiple waves at selected non-overlapping

frequencies in a predetermined range of frequency is applied to both terminals of the

charge storage device and thus obtained current and voltage are Fourier transformed to

measure the impedance spectrum.

18. The method as claimed in claim 15. wherein in step (1), the impedance

spectrum is measured by a Laplace transform method in which voltage and current

measured by applying a current or voltage having a specified waveform in a

numerical analysis to obtain the impedance spectrum.

19. The method as claimed in claim 15. wherein in step (1). the impedance

spectrum is measured by at least two methods selected from a frequency scan method in

which a series of sine waves having selected frequencies in a predetermined range of

frequency are applied to the charge storage device as an input signal and current and

voltage signals measured at both terminals of the charge storage device are analyzed, a

Fourier transform method an input current or voltage composed by superposing

multiple waves at selected non-overlapping frequencies in a predetermined range of

frequency is applied to both terminals of the charge storage device and thus obtained

current and voltage are Fourier transformed, and a Laplace transform method in which voltage and current measured by applying a current or voltage having a specified

waveform in a predetermined range of frequency to the both terminals of the charge

storage device or connecting an electrical load to the charge storage device is Laplace

transformed by a numerical analysis.

20. The method as claimed in any one of claims 15 and 16 to 19, wherein the

predetermined range of frequency is selected to have the ratio of the maximum

frequency to the minimum frequency greater than 100.

21. The method as claimed in claim 15, wherein the non-linear equivalent circuit of

the charge storage device in step (2) includes at least two elements selected from

ladder-type two-terminals in place of the transmission lines.

22. The method as claimed in claim 15, wherein the non-linear equivalent circuit of

type non-linear two-terminals in place of the transmission lines.

23. The method as claimed in claim 21 or 22. the ladder type two-terminals have a

time constant connected in parallel.

24. The method as claimed in claim 15. wherein the parameters in step (2) are

and include at least two elements selected from serial resistance, macroscopic

instantaneous change rate of voltage.

25. The method as claimed in claim 15 or 24, wherein the parameters in step (2) are

determined from a complex impedance spectra by a non-linear least square fitting.

26. The method as claimed in claim 15 or 24, wherein the parameters in step (2) are

determined from a slope of voltage characteristics vs. discharge capacity..

27. The method as claimed in claim 15. wherein the parameters of the non-linear

equivalent circuit of the charge storage

ice are obtained, and an open circuit voltage

of the charge storage device in a predetermined state of charge is also measured at the

same time.

28. An apparatus for determining parameters of a charge storage device wide

frequency range of impedance measurement and a non-linear equivalent circuit model

comprising:

a control means for determining parameters of the charge storage device and simulating various operational characteristics of a circuit including a charge storage

device using the determined parameters;

a digital-to-analog converter for converting a predetermined voltage and

current data output from the control means to an analog signal;

a potentiostat and/or galvanostat for applying the output signal of the digital-to-

analog converter to the charge storage device to output the voltage and current of the

capacitor;

an analog-to-digital converter for converting a voltage based on the voltage and

current of the charge storage device output from the potentiostat and/or galvanostat and

feeding them into the control means; and

a circuit simulator for simulating various operational characteristics of a charge

storage device based on the parameters obtained from the control means for a

predetermined equivalent circuit including the charge storage device.

29. The apparatus as claimed in claim 28. wherein the control means comprises:

an I/O interface unit for interfacing the capacitor and a predetermined data:

a central processing unit for controlling an operation for measuring voltage,

current and impedance spectra of the charge storage device;

means for measuring the voltage/current characteristic of the charge storage

de\ ice under the control of the central processing unit:

means for measuring the impedance spectra of the charge storage device under

the control of the central processing unit; and

means for determining the non-linear parameters by differentiating the measured values from the voltage/current measuring means and the impedance spectra

as a function of charge capacity and performing a non-linear least square fitting from

the differentiated values.

30. The apparatus as claimed in claim 29, wherein the impedance spectra measuring

means measures the impedance spectra by detecting voltage and current of the charge

storage device while it outputs digital data of the voltage and current based on a

plurality of predetermined frequencies sequentially, and performing a calculation on the

detected voltage and current of the charge storage device.

31. The apparatus as claimed in claim 30, wherein the detected voltage and current

of the charge storage device are calculated by amplitude/phase shift or Fourier

transform method.

32. The apparatus as claimed in claim 29. wherein the impedance spectra measuring

means measures the impedance spectra by measuring voltage and current while it

applies superposed multiple wave signals selected at predetermined non-overlapping

frequencies to the charge storage device, performing the Fourier transform of the

measured voltage and current, and performing a calculation on a power spectrum.

33. The apparatus as claimed in claim 29. wherein the impedance spectra measuring

means measures the impedance spectra by measuring voltage and current while it

outputs voltage and current data having a specified waveform and applies them to the charge storage device, performing a linear regression of the measured voltage and

current to calculate the impedance spectra, and performing a calculation on an error

function of the impedance.

34. The apparatus as claimed in claim 28. further comprising a separate

voltage/current generator for generating a voltage/current based on the digital data

output from the control means and applying it to the charge storage device.

35. The apparatus as claimed in claim 28. wherein the galvanostat comprises:

an inversion amplifier for inversion amplifying the output signal of the digital-

to-analog converter and applying it to the (-) terminal of the charge storage device;

a buffer for outputting voltages at both terminals of the charge storage device;

and

a current-detecting resistance connected in series with the charge storage device

for detecting the current of the charge storage device.

36. The apparatus as claimed in claim 28, further comprising first and second

filters/amplifiers between the galvanostat and the analog-to-digital converter for

filtering the voltage and current of the charge storage device detected by the galvanostat

to remove noises and amplifying the voltage and current.