CN116224109A - Method for rapidly measuring capacity of lithium ion battery - Google Patents
Method for rapidly measuring capacity of lithium ion battery Download PDFInfo
- Publication number
- CN116224109A CN116224109A CN202310224928.7A CN202310224928A CN116224109A CN 116224109 A CN116224109 A CN 116224109A CN 202310224928 A CN202310224928 A CN 202310224928A CN 116224109 A CN116224109 A CN 116224109A
- Authority
- CN
- China
- Prior art keywords
- battery
- impedance
- capacity
- lithium ion
- model
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/36—Arrangements for testing, measuring or monitoring the electrical condition of accumulators or electric batteries, e.g. capacity or state of charge [SoC]
- G01R31/385—Arrangements for measuring battery or accumulator variables
- G01R31/387—Determining ampere-hour charge capacity or SoC
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/36—Arrangements for testing, measuring or monitoring the electrical condition of accumulators or electric batteries, e.g. capacity or state of charge [SoC]
- G01R31/389—Measuring internal impedance, internal conductance or related variables
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
- Y02E60/00—Enabling technologies; Technologies with a potential or indirect contribution to GHG emissions mitigation
- Y02E60/10—Energy storage using batteries
Abstract
The invention relates to a method for rapidly measuring the capacity of a lithium ion battery, which comprises the following steps: 1) Obtaining the model of a battery; 2) Using impedance measurement equipment to measure and obtain the impedance spectrum of the battery in a set frequency range, and simultaneously measuring the current temperature and open-circuit voltage of the battery; 3) Performing data verification of the battery impedance spectrum by using a linear Kramers-Kronig verification algorithm; 4) Obtaining a first impedance characteristic of a battery impedance spectrum through a battery equivalent circuit model parameter identification algorithm; 5) And inputting the first impedance characteristic and the current temperature of the battery as well as the open-circuit voltage into a trained multiple regression model to obtain the battery capacity. Compared with the prior art, the invention has the advantages of high measurement accuracy, high speed and the like.
Description
Technical Field
The invention relates to the field of testing of power batteries of new energy automobiles, in particular to a method for rapidly measuring the capacity of a lithium ion battery.
Background
The capacity of a lithium ion battery is an important technical index of a lithium ion battery. After the lithium ion batteries are connected in series to form a group, the battery cells with the lowest content in the battery pack can limit the capacity of the whole battery pack due to the short-plate effect of the capacity, so that the charging and discharging capacity and the whole service life of the whole battery pack are affected. Therefore, the sorting of the consistency of the cell capacities before the grouping of lithium ion cells is an important measure for improving the performance and the overall life of the battery pack, and the sorting of the consistency of the capacities depends on the capacity measurement of the lithium ion cells. In addition, in the decommissioning power battery recycling scene, the residual capacity of the decommissioning power battery is an important index for measuring the residual value of the decommissioning power battery, so that the battery capacity measurement is also significant in the recycling price and secondary utilization scene evaluation of the decommissioning power battery.
The existing lithium ion battery capacity measurement method is mainly based on a charge-discharge experimental method. The accuracy of the measurement method is relatively high, but the charging and discharging experiment on the lithium ion battery monomer usually needs to consume 4-8 hours, small waste is caused to electric energy, and the cost is very high for large-scale engineering application, so that the current capacity measurement can only realize spot check, and the capacity measurement requirements of capacity consistency sorting and retired power battery residual value evaluation cannot be met. In addition, the existing method does not fully consider that the relation between the electrochemical impedance spectrum and the capacity is influenced by the temperature and the state of charge of the battery, and the accuracy of capacity measurement is not high.
Disclosure of Invention
The invention aims to overcome the defects of the prior art, and provides a rapid measuring method for the capacity of a lithium ion battery, which has high measuring accuracy and high speed.
The aim of the invention can be achieved by the following technical scheme:
a method for rapidly measuring the capacity of a lithium ion battery comprises the following steps:
1) Connecting a lithium ion battery with impedance measurement equipment, and obtaining the model of the battery;
2) Using impedance measurement equipment to measure and obtain the impedance spectrum of the battery in a set frequency range, and simultaneously measuring the current temperature and open-circuit voltage of the battery;
3) Performing data verification on the battery impedance spectrum by using a linear Kramers-Kronig verification algorithm, if the deviation between the battery impedance spectrum obtained by fitting and the battery impedance spectrum obtained by measurement falls within a first threshold range, determining that the data is legal, executing 4), otherwise, adjusting the connection between the battery and impedance measurement equipment, returning to 2), and carrying out impedance measurement again;
4) Performing equivalent circuit fitting on the battery impedance spectrum subjected to the verification of 3) through a battery equivalent circuit model parameter identification algorithm, and extracting ohmic impedance corresponding to the real axis intersection point of the Nyquist diagram and inflection point impedance corresponding to a battery diffusion process in the battery impedance spectrum, so as to obtain a first impedance characteristic of the battery impedance spectrum;
5) And inputting the first impedance characteristic, the current temperature of the battery and the open-circuit voltage into a trained multiple regression model, and obtaining the battery capacity, wherein the multiple regression model corresponds to the model of the battery.
Further, the method comprises the steps of,
the multiple regression model is trained based on a Gaussian process regression model, and the training process specifically comprises the following steps:
acquiring second impedance characteristics of lithium ion batteries with known capacities of specific models at different temperatures and different open-circuit voltages respectively, wherein the second impedance characteristics, the temperatures and the open-circuit voltages form a training set;
taking the second impedance characteristic, the temperature and the open-circuit voltage in the training set as first independent variables, taking the capacity as first dependent variables, and establishing a relational expression between the capacity and the impedance characteristic, the temperature and the open-circuit voltage;
and (3) rewriting the relation into a random variable form, determining posterior distribution of the independent variable coefficient vector based on the random variable form, determining a Bayesian regression model, introducing a Gaussian function in a radial basis function into the Bayesian regression model as a kernel function, and determining a multiple regression model.
Further, the independent variable coefficient vector obeys gaussian distribution, and the expression is:
wherein, sigma p Is the covariance matrix of the independent variable coefficient vector.
Further, the posterior distribution of the independent variable coefficient vector is gaussian distribution, expressed as:
wherein X is the whole training set of independent variables, Y is the whole training set of independent variables, ω is the independent variable coefficient vector,
to treat the dependent variable Y as the variance of gaussian noise with respect to the independent variable X, a is the inverse of the independent variable coefficient vector posterior distribution covariance matrix,
the expression of the inverse a of the independent variable coefficient vector posterior distribution covariance matrix is:
wherein, sigma p Is the covariance matrix of the independent variable coefficient vector.
Further, the expression of the battery capacity is:
wherein y is * For battery capacity, ω is the independent coefficient vector obtained by Gaussian process regression training, x * Is the first impedance characteristic and the current temperature and open circuit voltage of the battery.
Further, the step of performing data verification of the battery impedance spectrum by using a linear Kramers-Kronig verification algorithm comprises the following specific steps:
and (3) carrying out linear fitting on the battery impedance spectrum based on the multi-element series equivalent RC circuit, wherein the linear fitting process is to continuously increase the number M of series RC pairs from 1, the fitting result gradually goes from under fitting to over fitting along with the increase of the number M, and when the proportion of negative resistance in the multi-element series equivalent RC circuit to the total series RC pair exceeds a second threshold value, calculating whether the deviation between the battery impedance spectrum at the moment and the measured battery impedance spectrum falls within a first threshold value range or not, if yes, executing the step (4), otherwise, returning to the step (2).
Further, the expression of the linear fit is:
wherein Z (omega) is the alternating current impedance of the battery when the angular frequency is equal to omega, M is the number of RC pairs connected in series, R Ohm R is ohmic resistance component in AC impedance of battery k Resistance value of the kth RC pair, τ k Is the time constant of the kth RC pair, i.e. the product of the resistance and capacitance.
Further, the first threshold is 10%.
Further, a battery equivalent circuit model parameter identification algorithm is realized based on a random abrupt differential evolution algorithm and an improved Rankine circuit model, wherein the Rankine circuit model consists of an inductor, an ohmic resistor, two resistor-constant phase elements and a Warburg element which are connected in series;
the specific process of performing equivalent circuit fitting by adopting a battery equivalent circuit model parameter identification algorithm is as follows:
and randomly initializing a plurality of parameter vectors in a parameter search space formed by parameters of circuit elements of the Rankine circuit model by adopting a random abrupt change differential evolution algorithm, randomly moving the parameter vectors in the search space by a random combination or random abrupt change mode in each iteration, calculating the impedance of the Rankine circuit model by utilizing the iterated parameters, carrying out residual calculation with the impedance of the measured battery impedance spectrum, finding out model parameters which can lead the impedance of the Rankine circuit model to be closest to the impedance of the measured battery impedance spectrum in the whole parameter search space within a certain iteration limit, bringing the model parameters into the Rankine circuit model to obtain a battery equivalent circuit model, and calculating the ohmic impedance of the battery equivalent circuit model and inflection point impedance corresponding to a battery diffusion process.
Further, the specific process of measuring the battery impedance spectrum of the battery in the set frequency range is as follows:
the impedance measurement equipment generates a voltage or current disturbance signal through a DAC excitation generation circuit, wherein the voltage or current disturbance signal is a sine wave or a multi-sine-wave superposition wave, and the voltage or current disturbance signal is used as an excitation signal;
the method comprises the steps that through a synchronous sampling circuit, voltage and current data of a battery are collected simultaneously, and response of the lithium ion battery under a given excitation signal is obtained;
after voltage and current data are obtained, impedance corresponding to excitation signals with different frequencies is calculated in a set frequency range through Fourier transformation, and a battery impedance spectrum is obtained, wherein the set frequency range is 0.01 Hz-5 kHz.
Compared with the prior art, the invention has the following beneficial effects:
(1) According to the invention, the influence of the battery temperature and the state of charge on the relation between the electrochemical impedance spectrum and the capacity is considered, and the mathematical relation is established by a regression method, so that the capacity estimation can be realized based on the electrochemical impedance spectrum under any battery temperature and any state of charge, the control requirement on the battery state is reduced, and the application flexibility is improved.
(2) The invention establishes the relation between the electrochemical impedance spectrum of the lithium ion battery and the battery capacity, can rapidly calculate the battery capacity by utilizing the characteristic information of the electrochemical impedance spectrum obtained by measurement, and controls the capacity measurement time of the lithium ion battery at the minute level.
Drawings
FIG. 1 is a flow chart of the present invention;
FIG. 2 is an impedance feature extracted from the electrochemical impedance spectrum of a cell in accordance with the present invention;
FIG. 3 is a result of fitting the battery equivalent circuit by the battery equivalent circuit model parameter identification method of the present invention;
FIG. 4 is a graph of the relationship between the capacity of a lithium ion battery and the electrochemical impedance spectrum at the same temperature and open circuit voltage in accordance with the present invention;
FIG. 5 is a graph of the relationship between the capacity and electrochemical impedance characteristics of a lithium-ion battery corresponding to FIG. 4 in accordance with the present invention;
FIG. 6 is a graph of electrochemical impedance characteristics versus temperature for a lithium ion battery at the same capacity and open circuit voltage in accordance with the present invention;
fig. 7 is a graph of electrochemical impedance characteristics versus open circuit voltage for a lithium ion battery at the same capacity and temperature in accordance with the present invention.
Detailed Description
The invention will now be described in detail with reference to the drawings and specific examples. The present embodiment is implemented on the premise of the technical scheme of the present invention, and a detailed implementation manner and a specific operation process are given, but the protection scope of the present invention is not limited to the following examples.
The invention provides a method for rapidly measuring the capacity of a lithium ion battery, and a flow chart of the method is shown in figure 1.
The method comprises the following steps:
1) Connecting a lithium ion battery with impedance measurement equipment, and obtaining the model of the battery;
2) Using impedance measurement equipment to measure and obtain the impedance spectrum of the battery in a set frequency range, and simultaneously measuring the current temperature and open-circuit voltage of the battery;
3) Performing data verification on the battery impedance spectrum by using a linear Kramers-Kronig verification algorithm, if the deviation between the battery impedance spectrum obtained by fitting and the battery impedance spectrum obtained by measurement falls within a first threshold range, determining that the data is legal, executing 4), otherwise, adjusting the connection between the battery and impedance measurement equipment, returning to 2), and carrying out impedance measurement again;
4) Performing equivalent circuit fitting on the battery impedance spectrum subjected to the verification of 3) through a battery equivalent circuit model parameter identification algorithm, and extracting ohmic impedance corresponding to the real axis intersection point of the Nyquist diagram and inflection point impedance corresponding to a battery diffusion process in the battery impedance spectrum, so as to obtain a first impedance characteristic of the battery impedance spectrum;
5) And inputting the first impedance characteristic and the current temperature of the battery as well as the open-circuit voltage into a trained multiple regression model to obtain the battery capacity. The multiple regression model corresponds to the model of the battery.
1) And 2) the impedance measuring device can generate voltage or current disturbance signals, such as sine waves, multi-sine overlapping waves and the like, as excitation signals of the battery pack to be measured through an excitation generating circuit such as a DAC (digital-to-analog converter); and simultaneously acquiring voltage and current data of the battery through the synchronous sampling circuit to obtain the response of the battery pack to be tested under a given excitation signal. After the voltage and current data are obtained, the impedance corresponding to the excitation signals with different frequencies can be calculated in the frequency interval (0.01 Hz-5 kHz) through Fourier transformation, so that the electrochemical impedance spectrum of the battery is obtained.
3) After the electrochemical impedance spectrum is obtained, in order to eliminate the interference caused by the quality of the electrical connection between the battery and the equipment, the quality of the electrochemical impedance spectrum obtained by measurement needs to be checked. In the study of the Kramers and the Kronig in the linear time-invariant system, a corresponding relation exists between a real part and an imaginary part of a transfer function of the linear time-invariant system at each frequency, and the relation can be summarized as a Kramers-Kronig relation, and is described by the following formula:
however, in practical applications, since the measurement conditions are limited, the measurement instrument cannot measure the battery impedance at the frequency of 0 and at the frequency of positive infinity, and therefore the continuous Kramers-Kronig relationship cannot be directly applied to impedance quality check. For this reason, it is a relatively classical practice to linearly fit the measured battery impedance spectrum by means of a finite element series RC circuit, as follows
The linear Kramers-Kronig verification algorithm has the following problems: if the number of series RC pairs is too small, this can result in a battery impedance spectrum under-fit; too large a number of series RC pairs can result in a overfitting of the battery impedance spectrum. The former may cause the algorithm to misunderstand that a legal impedance spectrum that is consistent with the linear time invariant property is illegal, and the latter may cause the algorithm to misunderstand that an illegal impedance spectrum that is not consistent with the linear time invariant property is legal. To solve this problem, the number M of series RC pairs can be increased continuously by starting from 1. As M increases, the fit results gradually go from under-fit to over-fit, and the index evaluating the degree of over-fit can be measured by the specific gravity of the negative resistance in the fit circuit to the total series RC pair. When this specific gravity just exceeds a certain empirical threshold, it is shown that the fitting effect is almost as good as global optimum. Then, the error analysis is performed by using the fitted impedance spectrum and the measured impedance spectrum, and if the error between the fitted impedance spectrum and the measured impedance spectrum is within a certain threshold, for example, the Kramers-Kronig verification is performed by using the real part of the measured impedance spectrum, and the root mean square error between the imaginary part of the fitted result and the imaginary part of the measured impedance spectrum is less than 10%, the impedance spectrum data is considered legal and can be used according to the fact that the measured impedance spectrum satisfies the linearity.
After obtaining the broadband impedance spectrum of the battery, the coordinates of the intersection point of the electrochemical impedance curve and the nyquist real axis and the coordinates of the inflection point reflecting the diffusion process in the electrochemical impedance curve can be found by manually analyzing the impedance spectrum of the battery, as shown in fig. 2.
4) In the method, through a battery equivalent circuit model parameter identification algorithm, global optimal fitting parameter searching is automatically carried out through input of battery impedance data. Taking a battery equivalent circuit model parameter identification algorithm based on a random mutation differential evolution algorithm and a Advanced Randles Circuit model as an example, an automatic parameter identification process is briefly described. The Advanced Randles Circuit model is formed by connecting an inductor, an ohmic resistor, two resistor-constant phase elements and a Warburg element in series, and is used for describing parasitic inductance, ohmic impedance, load transmission process impedance, SEI film impedance and diffusion process impedance of the battery respectively, has good universality and can describe the internal electrode process of most lithium ion batteries. The random abrupt differential evolution algorithm is a variant of the differential evolution algorithm, the objective of which is to find the best equivalent circuit model parameters so that the residual of the calculated impedance of this equivalent circuit model and the actual measured impedance is as small as possible. The essence of the differential evolution algorithm is that in a parameter search space formed by the circuit parameters, a plurality of parameter vectors are randomly initialized, then in each iteration, the parameter vectors are randomly moved in a certain step length in the search space in a random combination mode, a random abrupt change mode and the like, then the impedance of a corresponding model is calculated by using the iterated parameters, residual calculation is carried out on the impedance obtained by measurement, and therefore model parameters which can enable the model impedance to be closest to the impedance obtained by measurement in the whole parameter search space are found within a certain iteration number limit. The fitting effect is shown in fig. 3.
The random abrupt differential evolution algorithm can generally calculate the optimal equivalent circuit model parameters within one minute, so that ohmic impedance and inflection point impedance of a diffusion process are calculated through the equivalent circuit model. The 2 norms of these two impedance data are simply referred to as impedance features in the following description. The first impedance characteristic and the second impedance characteristic obtained by the method are respectively used as the input of the multiple regression model and the training set in the training process of the multiple regression model when the capacity is actually measured. The following is the training process of the multiple regression model:
through the above steps, a set of capacity and impedance characteristics, temperature, open circuit voltage data for a certain model lithium battery is known. And repeating the steps to obtain the impedance characteristic data of the lithium ion batteries with different capacities and the same model at different temperatures and open circuit voltages. The relationship between the capacity of the lithium ion battery and the electrochemical impedance spectrum at the same temperature and open circuit voltage is shown in fig. 4, and the relationship between the capacity of the battery and the impedance characteristic after the characteristic extraction is shown in fig. 5; the relationship between electrochemical impedance characteristics of a lithium ion battery and temperature at the same capacity and open circuit voltage is shown in fig. 6; the electrochemical impedance characteristics of a lithium ion battery at the same capacity and temperature are plotted against the open circuit voltage (illustrated as a battery state of charge indicator equivalent to the open circuit voltage) as shown in fig. 7. For example, the impedance characteristics of 10 batteries with different capacities at 5 groups of different temperatures and 5 groups of different open circuit voltages are collected in one experiment, so that a data set with wider coverage for the type of batteries is formed. A multiple regression model between the capacity and impedance characteristics, temperature, and open circuit voltage of the model cell can be built and trained by multiple regression analysis.
For example, a multiple regression model may be obtained by analysis using a gaussian process regression algorithm. For the regression problem, the impedance characteristics, temperature, and open circuit voltage in the data set can be taken as independent variables, the capacity as dependent variables, and the following equation can be established
Wherein x= [ x ] 1 x 2 x 3 ]Representing an independent variable vector consisting of a battery impedance characteristic, a temperature, and an open circuit voltage; y represents battery capacity, which is a dependent variable; omega= [ omega ] 1 ω 2 ω 3 ]Representing an independent variable coefficient vector; epsilon represents noise conforming to a gaussian distribution and is a random variable. Due to x T Omega is a constant and epsilon is a mean of 0 and variance of
So y is a random variable with average value x T Omega, variance->
Y may be rewritten as a form of random variable:
from the expression of Y, a random variable expression of Y can be derived when the argument of the dataset is X and the argument is Y:
then the likelihood expression for Y is:
wherein x is i And y i Respectively representing the ith data component in the dataset.
The essence of regression is to estimate the argument coefficient vector ω. According to the Bayesian inference formula, there is posterior distribution of the independent variable coefficient vector omega:
where p (Y|X) is an edge likelihood that is independent of the argument coefficient vector ω, and can be regarded as a constant when the argument coefficient vector ω is found. p (ω) is an a priori distribution of the argument coefficient vector ω, here we consider it also obeying a gaussian distribution, i.e.:
wherein, sigma p A covariance matrix representing a gaussian distribution to which the independent coefficient vector is subjected.
It can be shown that the posterior distribution of the argument coefficient vector ω is also a gaussian distribution, which can be expressed as
In fact, the posterior distribution of the independent variable coefficient vector ω is a process of inputting measured data to perform multiple regression analysis. That is, as long as p (ω|y, X) is obtained, training corresponds to obtaining a multiple regression model.
Next, the trained model, i.e., the argument coefficient vector omega, is applied to the actual capacity measurement, and the impedance characteristics, temperature and open circuit voltage of the unknown capacity battery are measured first, denoted as x * Data can be input into this multiple regression model
Thereby obtaining the capacity y of the battery to be tested * . In fact, this y * Also a gaussian distribution, can be expressed as
The above procedure is Bayesian linear regression. In general, bayesian linear regression does not work well in low-dimensional space, because the relationship between features in low-dimensional space is likely to be nonlinear. The Gaussian process regression algorithm introduces kernel skills based on Bayesian linear regression, and can map low-dimensional data to a high-dimensional space through kernel functions. For example, the low-dimensional data described above may be mapped to a high-dimensional space using a gaussian function in a radial basis function as a kernel function. The gaussian radial basis function is as follows
φ(x)＝exp(-(ε||x-x i || 2 )
Wherein epsilon is an adjustable parameter of a Gaussian radial basis function, and the width of the basis function is controlled; x is x i Representing the radial center and also an adjustable parameter.
After kernel function mapping, the whole independent variable in the data set is mapped from X to phi, and the measured value of the battery with unknown capacity is mapped from X * Mapping to phi (x) * ) Then the final expression of the Gaussian process regression can be written
And finally, calculating the capacity of the battery through the impedance characteristic, the current temperature and the open-circuit voltage of the battery by using the Gaussian process regression model.
The foregoing describes in detail preferred embodiments of the present invention. It should be understood that numerous modifications and variations can be made in accordance with the concepts of the invention by one of ordinary skill in the art without undue burden. Therefore, all technical solutions which can be obtained by logic analysis, reasoning or limited experiments based on the prior art by the person skilled in the art according to the inventive concept shall be within the scope of protection defined by the claims.
Claims (10)
1. The rapid measurement method for the capacity of the lithium ion battery is characterized by comprising the following steps of:
1) Connecting a lithium ion battery with impedance measurement equipment, and obtaining the model of the battery;
2) Using impedance measurement equipment to measure and obtain the impedance spectrum of the battery in a set frequency range, and simultaneously measuring the current temperature and open-circuit voltage of the battery;
3) Performing data verification on the battery impedance spectrum by using a linear Kramers-Kronig verification algorithm, if the deviation between the battery impedance spectrum obtained by fitting and the battery impedance spectrum obtained by measurement falls within a first threshold range, determining that the data is legal, executing 4), otherwise, adjusting the connection between the battery and impedance measurement equipment, returning to 2), and carrying out impedance measurement again;
4) Performing equivalent circuit fitting on the battery impedance spectrum subjected to the verification of 3) through a battery equivalent circuit model parameter identification algorithm, and extracting ohmic impedance corresponding to the real axis intersection point of the Nyquist diagram and inflection point impedance corresponding to a battery diffusion process in the battery impedance spectrum, so as to obtain a first impedance characteristic of the battery impedance spectrum;
5) And inputting the first impedance characteristic, the current temperature of the battery and the open-circuit voltage into a trained multiple regression model, so as to obtain the battery capacity, wherein the multiple regression model corresponds to the model of the battery.
2. The rapid measurement method of lithium ion battery capacity according to claim 1, wherein the multiple regression model is trained based on a gaussian process regression model, and the training process is specifically as follows:
acquiring second impedance characteristics of lithium ion batteries with known capacities of specific models at different temperatures and different open-circuit voltages respectively, wherein the second impedance characteristics, the temperatures and the open-circuit voltages form a training set;
taking the second impedance characteristic, the temperature and the open-circuit voltage in the training set as first independent variables, taking the capacity as first dependent variables, and establishing a relational expression between the capacity and the impedance characteristic, the temperature and the open-circuit voltage;
and (3) rewriting the relation into a random variable form, determining posterior distribution of the independent variable coefficient vector based on the random variable form, determining a Bayesian regression model, introducing a Gaussian function in a radial basis function into the Bayesian regression model as a kernel function, and determining a multiple regression model.
3. The method for rapidly measuring capacity of a lithium ion battery according to claim 2, wherein the independent variable coefficient vector is subjected to gaussian distribution, and the expression is:
wherein, sigma p Is the covariance matrix of the independent variable coefficient vector.
4. The method for rapidly measuring capacity of a lithium ion battery according to claim 3, wherein the posterior distribution of the independent coefficient vector is gaussian, expressed as:
wherein X is the whole training set of independent variables, Y is the whole training set of independent variables, ω is the independent variable coefficient vector,
to treat the dependent variable Y as the variance of gaussian noise with respect to the independent variable X, a is the inverse of the independent variable coefficient vector posterior distribution covariance matrix,
the expression of the inverse a of the independent variable coefficient vector posterior distribution covariance matrix is:
∑ p is the covariance matrix of the independent variable coefficient vector.
5. The rapid measurement method of lithium ion battery capacity according to claim 3, wherein the expression of the battery capacity is:
wherein y is * For battery capacity, ω is the independent coefficient vector obtained by Gaussian process regression training, x * Is the first impedance characteristic and the current temperature and open circuit voltage of the battery.
6. The method for rapidly measuring the capacity of a lithium ion battery according to claim 1, wherein the step of performing data verification of the battery impedance spectrum by using a linear Kramers-Kronig verification algorithm comprises the following steps:
and (3) carrying out linear fitting on the battery impedance spectrum based on the multi-element series equivalent RC circuit, wherein the linear fitting process is to continuously increase the number M of series RC pairs from 1, the fitting result gradually goes from under fitting to over fitting along with the increase of the number M, and when the proportion of negative resistance in the multi-element series equivalent RC circuit to the total series RC pair exceeds a second threshold value, calculating whether the deviation between the battery impedance spectrum at the moment and the measured battery impedance spectrum falls within a first threshold value range or not, if yes, executing the step (4), otherwise, returning to the step (2).
7. The method for rapidly measuring capacity of a lithium ion battery according to claim 6, wherein the expression of the linear fitting is:
wherein Z (omega) is the alternating current impedance of the battery when the angular frequency is equal to omega, M is the number of RC pairs connected in series, R Ohm R is ohmic resistance component in AC impedance of battery k Resistance value of the kth RC pair, τ k Is the time constant of the kth RC pair, i.e. the product of the resistance and capacitance.
8. The method of claim 6, wherein the first threshold is 10%.
9. The method for rapidly measuring the capacity of a lithium ion battery according to claim 1, wherein a battery equivalent circuit model parameter identification algorithm is realized based on a random abrupt differential evolution algorithm and an improved landel circuit model, and the landel circuit model is formed by connecting an inductor, an ohmic resistor, two resistor-constant phase elements and a Warburg element in series;
the specific process of performing equivalent circuit fitting by adopting a battery equivalent circuit model parameter identification algorithm is as follows:
and randomly initializing a plurality of parameter vectors in a parameter search space formed by parameters of circuit elements of the Rankine circuit model by adopting a random abrupt change differential evolution algorithm, randomly moving the parameter vectors in the search space by a random combination or random abrupt change mode in each iteration, calculating the impedance of the Rankine circuit model by utilizing the iterated parameters, carrying out residual calculation with the impedance of the measured battery impedance spectrum, finding out model parameters which can lead the impedance of the Rankine circuit model to be closest to the impedance of the measured battery impedance spectrum in the whole parameter search space within a certain iteration limit, bringing the model parameters into the Rankine circuit model to obtain a battery equivalent circuit model, and calculating the ohmic impedance of the battery equivalent circuit model and inflection point impedance corresponding to a battery diffusion process.
10. The method for rapidly measuring the capacity of a lithium ion battery according to claim 1, wherein the specific process of measuring the battery impedance spectrum of the battery in a set frequency range is as follows:
the impedance measurement equipment generates a voltage or current disturbance signal through a DAC excitation generation circuit, wherein the voltage or current disturbance signal is a sine wave or a multi-sine-wave superposition wave, and the voltage or current disturbance signal is used as an excitation signal;
the method comprises the steps that through a synchronous sampling circuit, voltage and current data of a battery are collected simultaneously, and response of the lithium ion battery under a given excitation signal is obtained;
after voltage and current data are obtained, impedance corresponding to excitation signals with different frequencies is calculated in a set frequency range through Fourier transformation, and a battery impedance spectrum is obtained, wherein the set frequency range is 0.01 Hz-5 kHz.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310224928.7A CN116224109A (en) | 2023-03-07 | 2023-03-07 | Method for rapidly measuring capacity of lithium ion battery |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310224928.7A CN116224109A (en) | 2023-03-07 | 2023-03-07 | Method for rapidly measuring capacity of lithium ion battery |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116224109A true CN116224109A (en) | 2023-06-06 |
Family
ID=86582196
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310224928.7A Pending CN116224109A (en) | 2023-03-07 | 2023-03-07 | Method for rapidly measuring capacity of lithium ion battery |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116224109A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116774051A (en) * | 2023-06-28 | 2023-09-19 | 上海炙云新能源科技有限公司 | Battery capacity quick estimation method considering time-frequency domain multidimensional data characteristics |
-
2023
- 2023-03-07 CN CN202310224928.7A patent/CN116224109A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116774051A (en) * | 2023-06-28 | 2023-09-19 | 上海炙云新能源科技有限公司 | Battery capacity quick estimation method considering time-frequency domain multidimensional data characteristics |
| CN116774051B (en) * | 2023-06-28 | 2024-02-02 | 上海炙云新能源科技有限公司 | Battery capacity quick estimation method considering time-frequency domain multidimensional data characteristics |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Zhu et al. | A state of charge estimation method for lithium-ion batteries based on fractional order adaptive extended kalman filter | |
| Yang et al. | State-of-health estimation for the lithium-ion battery based on support vector regression | |
| Liu et al. | State-of-health estimation of lithium-ion batteries based on electrochemical impedance spectroscopy: a review | |
| Sheng et al. | Electric vehicle state of charge estimation: Nonlinear correlation and fuzzy support vector machine | |
| Ouyang et al. | Improved parameters identification and state of charge estimation for lithium-ion battery with real-time optimal forgetting factor | |
| US11422194B2 (en) | Battery diagnosis apparatus and battery diagnosis method based on current pulse method | |
| Su et al. | Fast capacity estimation for lithium-ion battery based on online identification of low-frequency electrochemical impedance spectroscopy and Gaussian process regression | |
| CN112069739B (en) | Battery fractional order model parameter identification method | |
| Sheng et al. | Small sample state of health estimation based on weighted Gaussian process regression | |
| Ren et al. | Parameter identification of a lithium‐ion battery based on the improved recursive least square algorithm | |
| CN110632528A (en) | Lithium battery SOH estimation method based on internal resistance detection | |
| Ouyang et al. | A novel state of charge estimation method for lithium-ion batteries based on bias compensation | |
| Wang et al. | A comparative study of fractional-order models for lithium-ion batteries using Runge Kutta optimizer and electrochemical impedance spectroscopy | |
| Xiao et al. | Rapid measurement method for lithium‐ion battery state of health estimation based on least squares support vector regression | |
| Luo et al. | An online state of health estimation technique for lithium-ion battery using artificial neural network and linear interpolation | |
| CN109613446A (en) | A kind of aging of lithium battery detection method based on time series analysis | |
| Chen et al. | State of health estimation for lithium-ion battery based on long short term memory networks | |
| CN116224109A (en) | Method for rapidly measuring capacity of lithium ion battery | |
| CN111257753A (en) | Battery system fault diagnosis method | |
| Zhang et al. | A fractional-order model of lithium-ion batteries and multi-domain parameter identification method | |
| Zhou et al. | Research on online capacity estimation of power battery based on EKF-GPR model | |
| Li et al. | A framework for states co-estimation of hybrid energy storage systems based on fractional-order theory | |
| Zhou et al. | Battery state of health estimation using the generalized regression neural network | |
| Narayanan et al. | Machine learning-based model development for battery state of charge–open circuit voltage relationship using regression techniques | |
| Chen et al. | A novel battery health indicator and PSO-LSSVR for LiFePO4 battery SOH estimation during constant current charging |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |