CN102930173A - SOC(state of charge) online estimation method for lithium ion battery - Google Patents

Abstract

The invention discloses an SOC (state of charge) online estimation method for a lithium ion battery. The SOC online estimation method comprises the following steps: recognizing characteristic parameter values of the battery online; recognizing real-time SOC of the battery online; revising the battery SOC through ampere-hour integral estimation to obtain revised SOC; and weighing the real-time SOC and the revised SOC to obtain final SOC of the battery. By the SOC online estimation method, online estimation capability of battery characteristic parameters can be improved; evaluation self-adaptive capability of the SOC is stronger, and reliability is higher.

Description

A kind of charge states of lithium ion battery estimation on line method

Technical field

The present invention relates to the technical field of lithium ion of new-energy automobile, more particularly, relate to a kind of charge states of lithium ion battery estimation on line method.

Background technology

Along with global climate progressively worsens, urban atmospheric pollution aggravates and petroleum resources consume excessively, automotive field increasing people thrown into sight on the new-energy automobile.The premium properties such as lithium ion battery has that volume is little, energy density is high, storage life is long, memory-less effect, high voltage and self-discharge rate are low becomes the winner of new-energy automobile power battery gradually.How making good use of battery has become battery and gordian technique of integration field thereof, and wherein accurate monitoring battery state-of-charge (State of Charge, SOC) has become difficult point and hot issue that battery management system and even electric automobile are studied.Therefore the practicality that improves SOC estimation precision and SOC algorithm has very high theory value and practical significance.

SOC evaluation method commonly used mainly contains at present: ampere-hour integral method, open-circuit voltage method, neural network etc.Wherein the ampere-hour integral method is the most general method.But several main problems below existing in application: (1) is affected greatly by initial SOC value, current acquisition precision, efficiency for charge-discharge; (2) the method is of serious failure behind the cell degradation capacity attenuation, so normal and additive method is united use in the practical application.The open-circuit voltage method mainly utilizes open-circuit voltage (OCV, Open Circuit Voltage) and the corresponding relation of SOC to estimate, but owing to open-circuit voltage in the electrokinetic cell course of work is difficult to measure, so that the method use is limited.Neural network need to use a large amount of data samples that algorithm is carried out off-line simulation, and method is complicated to be difficult for realizing and application, and can not be used for the on-line monitoring of SOC.

Summary of the invention

In view of this, the invention provides a kind of charge states of lithium ion battery estimation on line method, improve the estimation on line ability of battery characteristics parameter with realization, so that stronger to the estimation adaptive ability of state-of-charge, have higher reliability.

For solving the problems of the technologies described above, the technical solution used in the present invention is: the online estimation on line method of a kind of charge states of lithium ion battery comprises:

The characteristic ginseng value of on-line identification battery;

The real-time state-of-charge of on-line identification battery;

Revise the state-of-charge by the battery of ampere-hour integration estimation, draw revised state-of-charge;

The described real-time state-of-charge of weighting and described state-of-charge through obtaining after revising obtain the final state-of-charge of battery.

Preferably, described characteristic parameter comprises: open-circuit voltage, Ohmage, polarization resistance and polarization capacity.

Preferably, the characteristic ginseng value of described on-line identification battery is specially:

Obtain voltage signal V (k), the current signal I (k) of cell load and the temperature signal T of battery;

Set up the least square model according to the equivalent-circuit model of battery, the model relationship is:

V=V _oc-IR ₀-V _p (1)

$C_p{R}_p\frac{{\mathrm{dI}}_p}{\mathrm{dt}}=I-I_p---\left(2\right)$

Wherein, V _OcBe battery open circuit voltage, R ₀Be Ohmage, I is total discharge current, I _pFor by the electric current on the polarization resistance, V is the load voltage of battery, R _pBe polarization resistance, C _pBe polarization capacity;

Above-mentioned (1) formula and (2) formula are carried out the discretize arrangement to be obtained:

V(k)=θ ₁V(k-1)+θ ₂I(k)+θ ₃I(k-1)+θ ₄V _ocU(k) (3)

Wherein, θ is the function of battery characteristics parameter, and k is the current time value, and k-1 is a upper moment value;

By improving Recursive Least Squares described characteristic parameter is carried out iteration:

The compute vector matrix:

φ(k)=[V(k-1)I(k)I(k-1)1] ^T；

Gain matrix:

$K=\frac{P(k-1)\mathrm{\φ}\left(k\right)}{1+{\mathrm{\φ}}^T\left(k\right)P(k-1)\mathrm{\φ}\left(k\right)};$

Estimation error:

α=V(k)-θ ^Tφ(k)；

Estimation θ:

θ(k)=θ(k-1)+Kα；

The evaluated error covariance matrix:

P(k)=P(k-1)-Kφ ^T(k)P(k-1)+f(α)；

Wherein, f (α) is the function of estimation residual error α;

Described $f\left(\mathrm{\α}\right)=\underset{n=1}{\overset{m}{\mathrm{\Σ}}}{A}_n{\mathrm{\α}}^n$ Determine the value of m and An by off-line simulation;

Draw the battery characteristics parameter value:

$R_p=-\frac{({\mathrm{\&theta;}}_1{\mathrm{\&theta;}}_2+{\mathrm{\&theta;}}_3)\ln {\mathrm{\&theta;}}_1}{{({\mathrm{\&theta;}}_1-1)}^2},$ $C_p=\frac{-\frac{\mathrm{\&Delta;t}}{\ln {\mathrm{\&theta;}}_1}}{R_p},$ ${\mathrm{<figure-callout\; id="0"\; label="R"\; filenames="HDA00002415512800022.png"\; state="\{\{state\}\}">R</figure-callout>}}_0=-(1+\frac{1-{\mathrm{\&theta;}}_1}{\ln {\mathrm{\&theta;}}_1})R_p-{\mathrm{\&theta;}}_2,V_{\mathrm{oc}}=\frac{{\mathrm{\&theta;}}_4}{1-{\mathrm{\&theta;}}_1}.$

Preferably, the real-time state-of-charge of described on-line identification battery is specially:

According to described open-circuit voltage V _Oc, temperature T and state-of-charge curve, tabling look-up draws real-time state-of-charge SOC _Rls

Preferably, described correction draws revised state-of-charge and is specially by the state-of-charge of the battery of ampere-hour integration estimation:

The state-of-charge SOC (k-1) that previous moment is estimated by the ampere-hour integration adds current integration

Obtain revised state-of-charge SOC _Ah(k):

${\mathrm{SOC}}_{\mathrm{ah}}\left(k\right)=\mathrm{SOC}(k-1)+K_c*K_d*K_T*\frac{\mathrm{\&eta;I\&Delta;t}}{C};$

Wherein, K _c, K _dBe charging and discharging currents modifying factor, K _TBe temperature correction factor.

Preferably, the described real-time state-of-charge of described weighting and described state-of-charge through obtaining after revising obtain the final state-of-charge of battery and are specially:

With ampere-hour integral method estimation SOC _AhWith identification method estimation SOC _RlsWeighting obtains final SOC estimated value:

SOC=wSOC _ah+(1-w)SOC _rls；

Wherein, dynamic adjustable and 0≤w≤1 of w.

Can find out from above-mentioned technical scheme, a kind of charge states of lithium ion battery estimation on line method disclosed by the invention, characteristic ginseng value by the on-line identification battery, the real-time state-of-charge of estimating battery, by revising the state-of-charge by the battery of ampere-hour integration estimation, at last real-time state-of-charge and revised state-of-charge are weighted, obtain the final state-of-charge of battery, improved the estimation on line ability of battery characteristics parameter, so that stronger to the estimation adaptive ability of state-of-charge, had higher reliability.

Description of drawings

In order to be illustrated more clearly in the embodiment of the invention or technical scheme of the prior art, the below will do to introduce simply to the accompanying drawing of required use in embodiment or the description of the Prior Art, apparently, accompanying drawing in the following describes only is some embodiments of the present invention, for those of ordinary skills, under the prerequisite of not paying creative work, can also obtain according to these accompanying drawings other accompanying drawing.

Fig. 1 is the process flow diagram of the disclosed a kind of charge states of lithium ion battery estimation on line method of the embodiment of the invention;

Fig. 2 is the battery equivalent circuit diagram that the present invention uses;

Fig. 3 is the identification result figure of battery characteristics parameter value of the present invention;

Fig. 4 is real vehicle test result figure of the present invention.

Embodiment

Below in conjunction with the accompanying drawing in the embodiment of the invention, the technical scheme in the embodiment of the invention is clearly and completely described, obviously, described embodiment only is a part of embodiment of the present invention, rather than whole embodiment.Based on the embodiment among the present invention, those of ordinary skills belong to the scope of protection of the invention not making the every other embodiment that obtains under the creative work prerequisite.

The embodiment of the invention discloses a kind of charge states of lithium ion battery estimation on line method, improve the estimation on line ability of battery characteristics parameter with realization, so that stronger to the estimation adaptive ability of state-of-charge, have higher reliability.

As shown in Figure 1, a kind of charge states of lithium ion battery estimation on line method comprises:

The characteristic ginseng value of S1, on-line identification battery;

Concrete, the characteristic parameter of battery comprises open-circuit voltage, Ohmage, polarization resistance and polarization capacity;

At first, by voltage signal V (k), the current signal I (k) of battery management system Real-time Obtaining cell load and the temperature signal T of battery;

Then, the equivalent electrical circuit of basis battery is as shown in Figure 2 set up the least square model again, and wherein, the model relationship is:

V=V _oc-IR ₀-V _p (1)

$C_p{R}_p\frac{{\mathrm{dI}}_p}{\mathrm{dt}}=I-I_p---\left(2\right)$

Wherein, V _OcBe battery open circuit voltage, R ₀Be Ohmage, I is total discharge current, I _pFor by the electric current on the polarization resistance, V is the load voltage of battery, R _pBe polarization resistance, C _pBe polarization capacity;

Above-mentioned (1) formula and (2) formula are carried out the discretize arrangement to be obtained:

V(k)=θ ₁V(k-1)+θ ₂I(k)+θ ₃I(k-1)+θ ₄V _ocU(k) (3)

Wherein, θ is the function of battery characteristics parameter, and k is the current time value, and k-1 is a upper moment value;

By improving Recursive Least Squares described characteristic parameter is carried out iteration:

The compute vector matrix:

φ(k)=[V(k-1)I(k)I(k-1)1] ^T；

Gain matrix:

$K=\frac{P(k-1)\mathrm{\&phi;}\left(k\right)}{1+{\mathrm{\&phi;}}^T\left(k\right)P(k-1)\mathrm{\&phi;}\left(k\right)};$

Estimation error:

α=V(k)-θ ^Tφ(k)；

Estimation θ:

θ(k)=θ(k-1)+Kα；

The evaluated error covariance matrix:

P(k)=P(k-1)-Kφ ^T(k)P(k-1)+f(α)；

Wherein, f (α) is the function of estimation residual error α;

Described $f\left(\mathrm{\&alpha;}\right)=\underset{n=1}{\overset{m}{\mathrm{\&Sigma;}}}{A}_n{\mathrm{\&alpha;}}^n$ Determine m and A by off-line simulation _nValue;

Draw the battery characteristics parameter value:

$R_p=-\frac{({\mathrm{\&theta;}}_1{\mathrm{\&theta;}}_2+{\mathrm{\&theta;}}_3)\ln {\mathrm{\&theta;}}_1}{{({\mathrm{\&theta;}}_1-1)}^2},$ $C_p=\frac{-\frac{\mathrm{\&Delta;t}}{\ln {\mathrm{\&theta;}}_1}}{R_p},$ ${\mathrm{<figure-callout\; id="0"\; label="R"\; filenames="HDA00002415512800022.png"\; state="\{\{state\}\}">R</figure-callout>}}_0=-(1+\frac{1-{\mathrm{\&theta;}}_1}{\ln {\mathrm{\&theta;}}_1})R_p-{\mathrm{\&theta;}}_2,V_{\mathrm{oc}}=\frac{{\mathrm{\&theta;}}_4}{1-{\mathrm{\&theta;}}_1}.$

The real-time state-of-charge of S2, on-line identification battery;

Concrete, as to estimate by System Discrimination algorithm battery open circuit voltage V _Oc, temperature T and state-of-charge SOC curve table look-up and draw the real-time state-of-charge SOC of battery _Rls

The state-of-charge of the battery of ampere-hour integration estimation is passed through in S3, correction, draws revised state-of-charge;

Concrete, the state-of-charge SOC (k-1) that previous moment is estimated by the ampere-hour integration adds current integration

Obtain revised state-of-charge SOC _Ah(k):

${\mathrm{SOC}}_{\mathrm{ah}}\left(k\right)=\mathrm{SOC}(k-1)+K_c*K_d*K_T*\frac{\mathrm{\&eta;I\&Delta;t}}{C};$

Wherein, K _c, K _dBe charging and discharging currents modifying factor, K _TBe temperature correction factor;

S4, the described real-time state-of-charge of weighting and described state-of-charge through obtaining after revising obtain the final state-of-charge of battery;

Concrete, with ampere-hour integral method estimation SOC _AhWith identification method estimation SOC _RlsWeighting obtains final SOC

Estimated value: SOC=wSOC _Ah+ (1-w) SOC _Rls

Wherein, dynamic adjustable and 0≤w≤1 of w.

In the above-described embodiments, the present invention uses System Discrimination algorithm estimating battery open-circuit voltage, and tables look-up with SOC curve (OCV, T)-SOC by open-circuit voltage, temperature and to estimate SOC, and adaptive ability is stronger; And the present invention proposes a kind of improvement Recursive Least Squares, and the calculating of covariance matrix P is introduced dynamic correction term on the basis of classic method, automatically adjusts correction term according to the identification residual error, has greatly improved battery characteristics parameter estimation on line ability; Simultaneously the present invention compares with traditional ampere-hour integration, introduces in real time correction of initial value, charging and discharging currents, temperature correction, has reduced the cumulative errors of ampere-hour integration; Adopt weighting algorithm to take into full account traditional ampere-hour integral algorithm and Recursive Least-square relative merits separately, have higher reliability.

As shown in Figure 3 and Figure 4, be the battery capacity constantly real vehicle test case that occurs decaying, can find out because current acquisition precision or the impact such as aging, the final estimation error of tradition ampere-hour integral method is 33%, the SOC of the present invention estimation has avoided traditional ampere-hour integral method to be subjected to initial SOC value and current acquisition precision, wearing out etc. affects, can converge to rapidly near the actual value, adaptive ability is strong, has good estimation effect.

Each embodiment adopts the mode of going forward one by one to describe in this instructions, and what each embodiment stressed is and the difference of other embodiment that identical similar part is mutually referring to getting final product between each embodiment.

To the above-mentioned explanation of the disclosed embodiments, make this area professional and technical personnel can realize or use the present invention.Multiple modification to these embodiment will be apparent concerning those skilled in the art, and General Principle as defined herein can be in the situation that do not break away from the spirit or scope of the present invention, in other embodiments realization.Therefore, the present invention will can not be restricted to these embodiment shown in this article, but will meet the widest scope consistent with principle disclosed herein and features of novelty.

Claims (6)

1. a charge states of lithium ion battery estimation on line method is characterized in that, comprising:

The characteristic ginseng value of on-line identification battery;

The real-time state-of-charge of on-line identification battery;

Revise the state-of-charge by the battery of ampere-hour integration estimation, draw revised state-of-charge;

The described real-time state-of-charge of weighting and described state-of-charge through obtaining after revising obtain the final state-of-charge of battery.

2. method according to claim 1 is characterized in that, described characteristic parameter comprises: open-circuit voltage, Ohmage, polarization resistance and polarization capacity.

3. method according to claim 2 is characterized in that, the characteristic ginseng value of described on-line identification battery is specially:

Obtain voltage signal V (k), the current signal I (k) of cell load and the temperature signal T of battery;

Set up the least square model according to the equivalent-circuit model of battery, the model relationship is:

V=V _oc-IR ₀-V _p (1)

$C_p{R}_p\frac{{\mathrm{dI}}_p}{\mathrm{dt}}=I-I_p---\left(2\right)$

Wherein, V _OcBe battery open circuit voltage, R ₀Be Ohmage, I is total discharge current, I _pFor by the electric current on the polarization resistance, V is the load voltage of battery, R _pBe polarization resistance, C _pBe polarization capacity;

Above-mentioned (1) formula and (2) formula are carried out the discretize arrangement to be obtained:

V(k)=θ ₁V(k-1)+θ ₂I(k)+θ ₃I(k-1)+θ ₄V _ocU(k) (3)

Wherein, θ is the function of battery characteristics parameter, and k is the current time value, and k-1 is a upper moment value;

By improving Recursive Least Squares described characteristic parameter is carried out iteration:

The compute vector matrix:

φ(k)=[V(k-1)I(k)I(k-1)1] ^T；

Gain matrix:

$K=\frac{P(k-1)\mathrm{\&phi;}\left(k\right)}{1+{\mathrm{\&phi;}}^T\left(k\right)P(k-1)\mathrm{\&phi;}\left(k\right)};$

Estimation error:

α=V(k)-θ ^Tφ(k)；

Estimation θ:

θ(k)=θ(k-1)+Kα；

The evaluated error covariance matrix:

P(k)=P(k-1)-Kφ ^T(k)P(k-1)+f(α)；

Wherein, f (α) is the function of estimation residual error α;

Described $f\left(\mathrm{\&alpha;}\right)=\underset{n=1}{\overset{m}{\mathrm{\&Sigma;}}}{A}_n{\mathrm{\&alpha;}}^n$ Determine m and A by off-line simulation _nValue;

Draw the battery characteristics parameter value:

$R_p=-\frac{({\mathrm{\&theta;}}_1{\mathrm{\&theta;}}_2+{\mathrm{\&theta;}}_3)\ln {\mathrm{\&theta;}}_1}{{({\mathrm{\&theta;}}_1-1)}^2},$ $C_p=\frac{-\frac{\mathrm{\&Delta;t}}{\ln {\mathrm{\&theta;}}_1}}{R_p},$ $R_0=-(1+\frac{1-{\mathrm{\&theta;}}_1}{\ln {\mathrm{\&theta;}}_1})R_p-{\mathrm{\&theta;}}_2,V_{\mathrm{oc}}=\frac{{\mathrm{\&theta;}}_4}{1-{\mathrm{\&theta;}}_1}.$

4. method according to claim 3 is characterized in that, the real-time state-of-charge of described on-line identification battery is specially:

According to described open-circuit voltage V _Oc, temperature T and state-of-charge curve, tabling look-up draws real-time state-of-charge SOC _Rls

5. method according to claim 4 is characterized in that, described correction draws revised state-of-charge and is specially by the state-of-charge of the battery of ampere-hour integration estimation:

The state-of-charge SOC (k-1) that previous moment is estimated by the ampere-hour integration adds current integration

Obtain revised state-of-charge SOC _Ah(k):

${\mathrm{SOC}}_{\mathrm{ah}}\left(k\right)=\mathrm{SOC}(k-1)+K_c*K_d*K_T*\frac{\mathrm{\&eta;I\&Delta;t}}{C};$

Wherein, K _c, K _dBe charging and discharging currents modifying factor, K _TBe temperature correction factor.

6. method according to claim 5 is characterized in that, the described real-time state-of-charge of described weighting and described state-of-charge through obtaining after revising obtain the final state-of-charge of battery and be specially:

With ampere-hour integral method estimation SOC _AhWith identification method estimation SOC _RlsWeighting obtains final SOC estimated value:

SOC=wSOC _ah+(1-w)SOC _rls；

Wherein, dynamic adjustable and 0≤w≤1 of w.

Images