CN103235266A - Charging state estimation method and charging state estimation device of power batteries - Google Patents

Abstract

The invention discloses a charging state estimation method and a charging state estimation device of power batteries. The method includes conforming power battery parameters of the power batteries, conforming spatial vectors of initial battery system state taking the power battery parameters as vector elements, acquiring initial corrected matrixes corresponding to the power batteries, calculating estimated battery system state spatial vectors, estimated error covariance matrixes and filtering gains corresponding to the power batteries, correcting the estimated battery system state spatial vectors to acquire corresponding target battery system state spatial vectors, and confirming corresponding charging states from the target battery system state spatial vectors. The charging state estimated errors can be reduced by increasing correcting procedures when the power battery parameters are inaccurate, charging state estimated errors in follow-up preset estimated time can be further reduced, and estimated accuracy of the power battery charging states can be increased.

Description

State-of-charge evaluation method and the device of electrokinetic cell

Technical field

The present invention relates to the state-of-charge estimation field of electrokinetic cell, particularly relate to a kind of state-of-charge evaluation method and device of electrokinetic cell.

Background technology

Along with becoming increasingly conspicuous of the energy and environmental issue, pure electric automobile and hybrid-electric car are subjected to the great attention of countries in the world.(Battery Management System, BMS), the quality of its performance has restricted the overall performance of electric automobile to a great extent as the battery management system of one of gordian technique of development electric automobile.And the core of BMS is to electrokinetic cell SOC(State Of Charge, state-of-charge) accurate estimation.Wherein, SOC can embody the dump energy of electrokinetic cell.

At present, the SOC evaluation method of available dynamic battery is the ampere-hour integral method.The calculating principle of this method is:

$\mathrm{SOC}={\mathrm{SOC}}_0-\frac{1}{C_N}\underset{0}{\overset{t}{\∫}}\mathrm{\ηId\τ}---\left(1\right)$

Wherein, SOC ₀For electrokinetic cell discharges and recharges the SOC value of initial state, C _NBe the electrokinetic cell rated capacity, I is the electrokinetic cell electric current, and η is efficiency for charge-discharge, and SOC is the t SOC value of electrokinetic cell constantly.

When using the SOC of this method estimation current time battery, the SOC value that need calculate the current time electrokinetic cell with the account form of (1) formula according to SOC initial value and all electrokinetic cell current values that collect constantly before the current time of electrokinetic cell.When inaccurate the or initial SOC value of current value that collects when a certain moment is inaccurate, not only can influence the accuracy of the SOC estimated value of this moment electrokinetic cell, also can have influence on the accuracy of the SOC estimated value of electrokinetic cell constantly of afterwards each of this moment, that is, make that this moment SOC estimated value of each moment electrokinetic cell afterwards is all inaccurate.

As seen, in the process of using the ampere-hour integral method, if electrokinetic cell current measurement or initial SOC value are inaccurate, will cause the electrokinetic cell SOC value that calculates to have error, long-term accumulated, error can be increasing, finally causes the estimation of battery SOC inaccurate.

Summary of the invention

For solving the problems of the technologies described above, the embodiment of the invention provides a kind of state-of-charge evaluation method and device of electrokinetic cell, and to improve the accuracy to the state-of-charge estimation of electrokinetic cell, technical scheme is as follows:

A kind of state-of-charge evaluation method of electrokinetic cell comprises:

In this operational process of electrokinetic cell, when current time is estimated constantly for presetting, determine the electrokinetic cell parameter of described electrokinetic cell, wherein, described electrokinetic cell parameter is the parameter that obtains from the battery model corresponding with described electrokinetic cell; Described default estimation constantly for moment of this time at intervals n second Preset Time section that bring into operation of described electrokinetic cell, and the non-working time of described electrokinetic cell before this task bring into operation constantly less than this of first Preset Time, n 〉=1 and n are integer;

Determine with the initial cells system state space vector of described electrokinetic cell parameter as vector element;

Obtain the initial correction matrix corresponding with described electrokinetic cell, wherein, described initial correction matrix is the default matrix that obtains according to described battery model;

Gather the electrokinetic cell system coefficient of regime of described electrokinetic cell, and according to described electrokinetic cell system coefficient of regime and described initial cells system state space vector, calculate the preestimating battery system state space vector corresponding with described electrokinetic cell, wherein, described battery system coefficient of regime is the parameter that obtains according to described battery model;

According to the initial error covariance matrix in described electrokinetic cell system coefficient of regime and the described initial correction matrix, calculate the predictor error covariance matrix corresponding with described electrokinetic cell;

According to described predictor error covariance matrix and described electrokinetic cell system coefficient of regime, calculate and the corresponding filter gain of described electrokinetic cell;

According to the predictor error covariance matrix in described filter gain, the described initial correction matrix, described preestimating battery system state space vector is revised, to obtain the target battery system state space vector corresponding with described electrokinetic cell;

From described target battery system state space vector, determine the state-of-charge corresponding with described electrokinetic cell.

Wherein, the electrokinetic cell parameter of described definite electrokinetic cell comprises:

When described default estimation constantly be the non-working time of described electrokinetic cell before this task less than this operation of first Preset Time during zero hour, with the electrokinetic cell parameter of end of run last time described electrokinetic cell during the moment electrokinetic cell parameter as described electrokinetic cell;

When described default estimation constantly for operation zero hour of described electrokinetic cell, n(n 〉=1 and n were integer at interval) during moment of the individual second Preset Time section, with described electrokinetic cell in the electrokinetic cell parameter in the last one default estimation moment electrokinetic cell parameter as described electrokinetic cell.

Wherein, also comprise:

When not reaching default estimation during the moment, be that this that be not less than first Preset Time non-working time of described electrokinetic cell before this task is when bringing into operation constantly judging current time, gather the open-circuit voltage of described electrokinetic cell, according to default open-circuit voltage and the mapping relations of state-of-charge, obtain the state-of-charge of described electrokinetic cell.

Wherein, described battery model is the Thevenin battery model, and is corresponding, and the electrokinetic cell parameter of end of run last time described electrokinetic cell during the moment, last one default estimation electrokinetic cell parameter constantly comprise respectively:

The polarizing voltage of electrokinetic cell, ohmic internal resistance, polarization resistance, polarization capacity and state-of-charge.

Wherein, described initial correction matrix comprises:

Initial error covariance matrix, starter system noise equation matrix, initial observation noise variance matrix and initial gain correction factor matrix.

Wherein, described electrokinetic cell system coefficient of regime comprises: the terminal voltage of electrokinetic cell, current signal and temperature signal.

Wherein, according to described electrokinetic cell system coefficient of regime and described initial cells system state space vector, calculate the formula that the preestimating battery system state space vector corresponding with described electrokinetic cell adopts, comprising:

$\hat{X}(k+1|k)=\left[\begin{array}{ccccc}1-\frac{T_s}{x_3\left(k\right)x_4\left(k\right)}& 0& 0& 0& 0\\ 0& 1& 0& 0& 0\\ 0& 0& 1& 0& 0\\ 0& 0& 0& 1& 0\\ 0& 0& 0& 0& 1\end{array}\right]\left[\begin{array}{c}{x}_1\left(k\right)\\ x_2\left(k\right)\\ x_3\left(k\right)\\ x_4\left(k\right)\\ x_5\left(k\right)\end{array}\right]+\left[\begin{array}{c}\frac{T_s}{x_4\left(k\right)}\\ 0\\ 0\\ 0\\ \frac{{\mathrm{\ηT}}_s}{C_N}\end{array}\right]I\left(k\right),$

Wherein, X (k) is the last corresponding initial cells system state space vector of a moment of described current time, when described current time is the non-working time of described electrokinetic cell before this task less than this operation of first Preset Time during zero hour, described last one constantly is the end of run moment of described electrokinetic cell, when described current time for operation zero hour of described electrokinetic cell, n(n 〉=1 and n were integer at interval) during moment of the individual second Preset Time section, described last one constantly be last one defaultly to estimate the moment;

Preestimating battery system state space vector for the current time that goes out according to described last one constantly initial cells system state space vector calculation; T _SBe the current signal sampling period of electrokinetic cell, C _NBe the rated capacity of electrokinetic cell, η is enclosed pasture efficient and C _N, η is the function of temperature signal;

Wherein, according to the initial error covariance matrix in described electrokinetic cell coefficient of regime and the described initial correction matrix, calculate the formula that the predictor error covariance matrix corresponding with described electrokinetic cell adopts, comprising:

$P(k+1|k)=\hat{A}\left(k\right)P\left(k\right){\hat{A}}^T\left(k\right)+Q$

Wherein, P (k) is the last one initial error covariance matrix in the corresponding initial correction matrix constantly of described current time, P (k+1/k) was the predictor error covariance matrix according to the current time that calculates of initial error covariance matrix in described a lasted moment, when described current time is the non-working time of described electrokinetic cell before this task less than this operation of first Preset Time during zero hour, described last one constantly is the end of run moment of described electrokinetic cell, when described current time for operation zero hour of described electrokinetic cell, n(n 〉=1 and n were integer at interval) during moment of the individual second Preset Time section, described last one constantly be last one defaultly to estimate the moment; Q is the system noise variance matrix; Be the state transformation matrix, its calculating formula is:

$\hat{A}\left(k\right)=\left[\begin{array}{ccccc}1-\frac{T_s}{{\hat{x}}_3\left(k\right){\hat{x}}_4\left(k\right)}& 0& \frac{T_s{\hat{x}}_1\left(k\right)}{{\hat{x}}_3^2\left(k\right){\hat{x}}_4\left(k\right)}& \frac{T_s({\hat{x}}_1\left(k\right)-{\hat{x}}_3\left(k\right)I\left(k\right))}{{\hat{x}}_3\left(k\right){\hat{x}}_4^2\left(k\right)}& 0\\ 0& 1& 0& 0& 0\\ 0& 0& 1& 0& 0\\ 0& 0& 0& 1& 0\\ 0& 0& 0& 0& 1\end{array}\right]$

Wherein, according to described predictor error covariance matrix and described electrokinetic cell system coefficient of regime, calculate the formula that the filter gain corresponding with described electrokinetic cell adopts, comprising:

L(k)=ζ*K(k)，

Wherein, ζ is gain correction factor matrix;

$K\left(k\right)=P(k+1|k){\hat{C}}^T(k+1){[\hat{C}(k+1)P(k+1|k){\hat{C}}^T(k+1)+R]}^{-1},$ R is the observation noise variance matrix;

Be the observation model matrix, its calculating formula is: $\hat{C}\left(k\right)=\left[\begin{array}{ccccc}-1& -I\left(k\right)& 0& 0& \frac{\∂F({\hat{x}}_5\left(k\right),T\left(k\right))}{\∂{\hat{x}}_5\left(k\right)}\end{array}\right],$ F (x ₅(k), T (k)) be state-of-charge (SOC) the battery open circuit voltage OCV function corresponding with battery temperature (T (k)) of battery.

Wherein, according to the predictor error covariance matrix in described filter gain, the described initial correction matrix, described preestimating battery system state space vector is revised the formula that adopts, comprising:

$\hat{X}(k+1)=\hat{X}(k+1/k)+L\left(k\right)\{V(k+1)-g[\hat{X}(k+1/k),I(k+1)\left]\right\};$

Wherein, g[X (k), I (k)] for measuring equation, its calculating formula is:

G[X (k), I (k)]=F (x ₅(k))-x ₂(k) I (k)-x ₁(k); V (k+1) is the terminal voltage of the electrokinetic cell of current time collection; I (k+1) is the end electric current of the electrokinetic cell of the current collection time of running;

$P(k+1)=[I-L\left(k\right)C\left((\hat{k}+1)\right]P(k+1/k),$ Wherein I is unit matrix;

Be the observation model matrix, its calculating formula is: $\hat{C}\left(k\right)=\left[\begin{array}{ccccc}-1& -I\left(k\right)& 0& 0& \frac{\∂F({\hat{x}}_5\left(k\right),T\left(k\right))}{\∂{\hat{x}}_5\left(k\right)}\end{array}\right].$

A kind of state-of-charge estimating device of electrokinetic cell comprises:

Electrokinetic cell parameter determination module, be used for this operational process at electrokinetic cell, when current time is estimated constantly for presetting, determine the electrokinetic cell parameter of described electrokinetic cell, wherein, described electrokinetic cell parameter is the parameter that obtains from the battery model corresponding with described electrokinetic cell; Described default estimation constantly for moment of this time at intervals n second Preset Time section that bring into operation of described electrokinetic cell, and the non-working time of described electrokinetic cell before this task bring into operation constantly less than this of first Preset Time, n 〉=1 and n are integer;

Initial cells system state space vector determination module is used for determining with the initial cells system state space vector of described electrokinetic cell parameter as vector element;

Initial correction matrix acquisition module is used for obtaining the initial correction matrix corresponding with described electrokinetic cell, and wherein, described initial correction matrix is the default matrix that obtains according to described battery model;

Preestimating battery system state space vector calculation module, be used for gathering the electrokinetic cell system coefficient of regime of described electrokinetic cell, and according to described electrokinetic cell system coefficient of regime and described initial cells system state space vector, calculate the preestimating battery system state space vector corresponding with described electrokinetic cell, wherein, described battery system coefficient of regime is the parameter that obtains according to described battery model;

The filter gain computing module is used for according to described predictor error covariance matrix and described electrokinetic cell system coefficient of regime, calculates and the corresponding filter gain of described electrokinetic cell;

Target battery system state space vector acquisition module, be used for the predictor error covariance matrix according to described filter gain, described initial correction matrix, described preestimating battery system state space vector is revised, to obtain the target battery system state space vector corresponding with described electrokinetic cell;

The state-of-charge determination module is used for determining the state-of-charge corresponding with described electrokinetic cell from described target battery system state space vector.In the technical scheme that the embodiment of the invention provides, by initial correction matrix and filter gain preestimating battery system state space vector is revised, thereby obtain corresponding target battery system state space vector, and finally from target battery system state space vector, obtain corresponding state-of-charge.As seen, in this programme, when the electrokinetic cell parameter is inaccurate, reduced the estimation error of this state-of-charge by the makeover process that increases, and further reduced the follow-up default estimation error of estimating state-of-charge constantly, thereby improved the accuracy to the state-of-charge estimation of electrokinetic cell.

Description of drawings

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

First kind of process flow diagram of the state-of-charge evaluation method of a kind of electrokinetic cell that Fig. 1 provides for the embodiment of the invention;

Second kind of process flow diagram of the state-of-charge evaluation method of a kind of electrokinetic cell that Fig. 2 provides for the embodiment of the invention;

The structural representation of the state-of-charge estimating device of a kind of electrokinetic cell that Fig. 3 provides for the embodiment of the 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 the present invention's part embodiment, 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.

In order to have improved the estimation accuracy to the state-of-charge of electrokinetic cell, the embodiment of the invention provides a kind of state-of-charge evaluation method and device of electrokinetic cell.

Need to prove, the method that invention provides was the SOC value according to the electrokinetic cell parameter estimation current preset estimation electrokinetic cell constantly in a last moment, wherein, be somebody's turn to do the last one default estimation moment and this current time at intervals one first Preset Time, and when described current time is the non-working time of described electrokinetic cell before this task less than this operation of first Preset Time during zero hour, described last one constantly is the end of run moment of described electrokinetic cell, when described current time for operation zero hour of described electrokinetic cell, n(n 〉=1 and n were integer at interval) during moment of the individual second Preset Time section, described last one constantly be last one defaultly to estimate the moment.

The state-of-charge evaluation method of a kind of electrokinetic cell that at first embodiment of the invention is provided is introduced below.

A kind of state-of-charge evaluation method of electrokinetic cell as shown in Figure 1, can comprise:

S101, in this operational process of electrokinetic cell, when current time constantly the time, is determined the electrokinetic cell parameter of electrokinetic cell for default estimation;

Need to prove that the state of electrokinetic cell is corresponding with the state of the electric vehicle at its place, that is, when electric vehicle was in running status, corresponding electrokinetic cell also was in running status; When electric vehicle was in dead ship condition, corresponding electrokinetic cell was in static condition.Wherein, in actual applications, this electric vehicle can be electric automobile, electric bicycle, battery-operated motor cycle etc.

In addition, because electrokinetic cell mostly is the jumbo chemical cell that energy is provided according to chemical reaction, for example, hydrogen-nickel battery, lithium battery, lead-acid accumulator etc., and chemical reaction has instability, uncontrollability more, so among the present invention, obtain the electrokinetic cell parameter of electrokinetic cell according to the battery model corresponding with this electrokinetic cell.Different types of chemical cell is when being applied to different electric vehicles, and according to user's demand, the battery model of its foundation can be different, and the corresponding electrokinetic cell parameter that obtains also can be different, and this all is rational.

In actual applications, on the one hand, default estimation constantly can for moment of this time at intervals n second Preset Time section that bring into operation of electrokinetic cell, n 〉=1 and n are integer, at this moment, electrokinetic cell is preset estimation electrokinetic cell parameter constantly as the electrokinetic cell parameter of current time last one, wherein, because the restriction of Computer Processing speed at present, second Preset Time among the present invention can be 10ms, certainly, be not limited to this;

On the other hand, should default estimation can also bring into operation constantly less than this of first Preset Time for the non-working time of electrokinetic cell before this task constantly, at this moment, with the electrokinetic cell parameter of end of run last time electrokinetic cell constantly time electrokinetic cell parameter as this current moment, wherein, by to the test findings of the test of many times of lithium battery as can be known, when this first Preset Time is 2h, and the non-working time of electrokinetic cell before this task, the terminal voltage of the electrokinetic cell SOC value current with it was linear when being not less than 2h.

Also need to prove, when current time is not that default estimation is during the moment, be that this that be not less than first Preset Time non-working time of electrokinetic cell before this task is when bringing into operation constantly judging current time, also be, the time of repose of current time electrokinetic cell is greater than the first Preset Time 2h, gather the open-circuit voltage of electrokinetic cell, according to default open-circuit voltage and the mapping relations of state-of-charge, obtain the state-of-charge of described electrokinetic cell, wherein, open-circuit voltage is the parameter that obtains from the battery model corresponding with electrokinetic cell.

S102, definite with the initial cells system state space vector of this electrokinetic cell parameter as vector element;

After the electrokinetic cell parameter of determining electrokinetic cell, just can be according to this electrokinetic cell parameter by the estimation of certain algorithm realization to electrokinetic cell SOC.

Among the present invention, the model corresponding with electrokinetic cell that adopts is the Thevenin battery model, adopting on this model based, electrokinetic cell parameter, the last one default estimation electrokinetic cell parameter constantly of end of run last time described electrokinetic cell constantly the time, difference can be the polarizing voltage of electrokinetic cell, ohmic internal resistance, polarization resistance, polarization capacity and state-of-charge.

Need to prove that for the different dynamic battery model of different types of electrokinetic cell institute foundation, the kind of the corresponding electrokinetic cell parameter of acquisition can be different with quantity.But, because original intention of the present invention is the SOC of estimation electrokinetic cell, so, no matter the electrokinetic cell of which kind of type is according to which kind of electrokinetic cell model, and the electrokinetic cell parameter of the current time of acquisition must comprise the state-of-charge of end of run last time electrokinetic cell constantly time the or the state-of-charge of last one default estimation electrokinetic cell constantly.

In addition, according to the electrokinetic cell parameter determine initial cells system state space vector formula can for:

${\hat{X}}_0=\left[\begin{array}{c}{V}_{c,0}\\ R_{\mathrm{0,0}}\\ R_{\mathrm{1,0}}\\ C_{\mathrm{1,0}}\\ {\mathrm{SOC}}_0\end{array}\right],$

Wherein,

Be initial cells system state space vector, V _cBe the polarizing voltage of electrokinetic cell, R ₀Be the ohmic internal resistance of electrokinetic cell, R ₁Be the polarization resistance of electrokinetic cell, C ₁Be the polarization capacity of electrokinetic cell, SOC is the state-of-charge of current time electrokinetic cell.

Need to prove that generally, the polarizing voltage of electrokinetic cell, ohmic internal resistance, polarization resistance, polarization capacity are constant.

S103, obtain the initial correction matrix corresponding with this electrokinetic cell;

Same, in the present invention, according to the battery model of selecting, the initial correction matrix corresponding with electrokinetic cell that obtains can be for: initial error covariance matrix, starter system noise variance matrix, initially observe the noise variance matrix.

Starter system noise variance matrix, initially observe the noise variance matrix and record by corresponding electrokinetic cell is tested, it has been generally acknowledged that it all is constant, in the process of estimation electrokinetic cell SOC, above-mentioned matrix can be done correction to calculating in calculation process.

Initial error covariance correction matrix also records by corresponding electrokinetic cell is tested, and it also is a constant usually, but in follow-up computation process, the error covariance correction matrix is to change according to certain account form.

S104, gather the electrokinetic cell system coefficient of regime of this electrokinetic cell, and according to this electrokinetic cell system coefficient of regime and this initial cells system state space vector, calculate the preestimating battery system state space vector corresponding with this electrokinetic cell;

Need to prove that the electrokinetic cell system coefficient of regime can comprise: the terminal voltage of electrokinetic cell, current signal and temperature signal.This electrokinetic cell coefficient of regime also is the parameter that obtains according to corresponding electrokinetic cell model.

In addition, according to described electrokinetic cell system coefficient of regime and described initial cells system state space vector, calculate the formula that the preestimating battery system state space vector corresponding with described electrokinetic cell adopts, can comprise:

$\hat{X}(k+1|k)=\left[\begin{array}{ccccc}1-\frac{T_s}{x_3\left(k\right)x_4\left(k\right)}& 0& 0& 0& 0\\ 0& 1& 0& 0& 0\\ 0& 0& 1& 0& 0\\ 0& 0& 0& 1& 0\\ 0& 0& 0& 0& 1\end{array}\right]\left[\begin{array}{c}{x}_1\left(k\right)\\ x_2\left(k\right)\\ x_3\left(k\right)\\ x_4\left(k\right)\\ x_5\left(k\right)\end{array}\right]+\left[\begin{array}{c}\frac{T_s}{x_4\left(k\right)}\\ 0\\ 0\\ 0\\ \frac{{\mathrm{\ηT}}_s}{C_N}\end{array}\right]I\left(k\right),$

Wherein, ${\hat{X}}_0=\left[\begin{array}{c}{x}_1\left(k\right)\\ x_2\left(k\right)\\ x_3\left(k\right)\\ x_4\left(k\right)\\ x_5\left(k\right)\end{array}\right],$

Be the initial cells system state space vector corresponding with a last moment, x ₁(k) be the polarizing voltage of a last moment electrokinetic cell, x ₂(k) be the ohmic internal resistance of a last moment electrokinetic cell, x ₃(k) be the polarization resistance of a last moment electrokinetic cell, x ₄(k) be the polarization capacity of a last moment electrokinetic cell, x ₅(k) be the state-of-charge of a last moment electrokinetic cell;

Preestimating battery system state space vector for the current time that goes out according to last one constantly initial cells system state space vector calculation, I (k) is the current signal of last one electrokinetic cell that constantly collects, and when described current time is the non-working time of described electrokinetic cell before this task less than this operation of first Preset Time during zero hour, described last one constantly is the end of run moment of described electrokinetic cell, when described current time for operation zero hour of described electrokinetic cell, n(n 〉=1 and n were integer at interval) during moment of the individual second Preset Time section, described last one constantly be last one defaultly to estimate the moment; T _SBe the current signal sampling period of electrokinetic cell, C _NBe the rated capacity of electrokinetic cell, η is enclosed pasture efficient and C _N, η is the function of temperature signal.

S105, according to the initial error covariance matrix in this electrokinetic cell system coefficient of regime and this initial correction matrix, calculate the predictor error covariance matrix corresponding with this electrokinetic cell;

Carry out formula that this step adopts can for:

Wherein, P (k) is and the last one initial error covariance matrix in the corresponding initial correction matrix constantly, the predictor error covariance matrix of the current time that P (k+1/k) calculated for the initial error covariance matrix according to a last moment; Q is the system noise variance matrix;

Be the state transformation matrix,

For The state transformation transposed matrix,

Calculating formula be:

$\hat{A}\left(k\right)=\left[\begin{array}{ccccc}1-\frac{T_s}{{\hat{x}}_3\left(k\right){\hat{x}}_4\left(k\right)}& 0& \frac{T_s{\hat{x}}_1\left(k\right)}{{\hat{x}}_3^2\left(k\right){\hat{x}}_4\left(k\right)}& \frac{T_s({\hat{x}}_1\left(k\right)-{\hat{x}}_3\left(k\right)I\left(k\right))}{{\hat{x}}_3\left(k\right){\hat{x}}_4^2\left(k\right)}& 0\\ 0& 1& 0& 0& 0\\ 0& 0& 1& 0& 0\\ 0& 0& 0& 1& 0\\ 0& 0& 0& 0& 1\end{array}\right]$ Accordingly, Calculating formula be:

${\hat{A}}^T\left(k\right)=\left[\begin{array}{ccccc}1-\frac{T_s}{{\hat{x}}_3\left(k\right){\hat{x}}_4\left(k\right)}& 0& 0& 0& 0\\ 0& 1& 0& 0& 0\\ \frac{T_s{\hat{x}}_1\left(k\right)}{{\hat{x}}_3^2\left(k\right){\hat{x}}_4\left(k\right)}& 0& 1& 0& 0\\ \frac{T_s({\hat{x}}_1\left(k\right)-{\hat{x}}_3\left(k\right)I\left(k\right))}{{\hat{x}}_3\left(k\right){\hat{x}}_4^2\left(k\right)}& 0& 0& 1& 0\\ 0& 0& 0& 0& 1\end{array}\right]$

Wherein, x ₁(k) be the polarizing voltage of a last moment electrokinetic cell, x ₃(k) be the polarization resistance of a last moment electrokinetic cell, x ₄(k) be the polarization capacity of a last moment electrokinetic cell; T _SBe the current signal sampling period of electrokinetic cell.

S106, according to this predictor error covariance matrix and this electrokinetic cell system coefficient of regime, calculate and the corresponding filter gain of this electrokinetic cell;

Carry out formula that this step adopts and can be L (k)=ζ * K (k),

Wherein, L (k) is filter gain, and ζ is gain correction factor matrix;

$K\left(k\right)=P(k+1|k){\hat{C}}^T(k+1){[\hat{C}(k+1)P(k+1|k){\hat{C}}^T(k+1)+R]}^{-1},$

Wherein, R is the observation noise variance matrix;

Be the observation model matrix, its calculating formula is: $\hat{C}\left(k\right)=\left[\begin{array}{ccccc}-1& -I\left(k\right)& 0& 0& \frac{\∂F({\hat{x}}_5\left(k\right),T\left(k\right))}{\∂{\hat{x}}_5\left(k\right)}\end{array}\right];$

F (x ₅(k), T (k)) be state-of-charge (SOC) the battery open circuit voltage OCV function corresponding with battery temperature (T (k)) of battery; P (k+1/k) is the predictor error covariance matrix of the current time that calculates according to last one constantly initial error covariance matrix.

S107, according to the predictor error covariance matrix in this filter gain, this initial correction matrix, this preestimating battery system state space vector is revised, to obtain the target battery system state space vector corresponding with this electrokinetic cell;

Carry out formula that this step adopts can for:

$\hat{X}(k+1)=\hat{X}(k+1/k)+L\left(k\right)\{V(k+1)-g[\hat{X}(k+1/k),I(k+1)\left]\right\};$

Wherein, L (k) is filter gain, g[X (k), I (k)] for measuring equation, its calculating formula is:

G[X (k), I (k)]=F (x ₅(k))-x ₂(k) I (k)-x ₁(k); V (k+1) is the terminal voltage of the electrokinetic cell of current time collection; I (k+1) is the end electric current of the electrokinetic cell of current time collection;

Preestimating battery system state space vector for the current time that goes out according to last one constantly initial cells system state space vector calculation;

$P(k+1)=[I-L\left(k\right)C\left((\hat{k}+1)\right]P(k+1/k),$ Wherein I is unit matrix;

Be the observation model matrix, its calculating formula is: $\hat{C}\left(k\right)=\left[\begin{array}{ccccc}-1& -I\left(k\right)& 0& 0& \frac{\∂F({\hat{x}}_5\left(k\right),T\left(k\right))}{\∂{\hat{x}}_5\left(k\right)}\end{array}\right].$

S108, from this target battery system state space vector, determine the state-of-charge corresponding with this electrokinetic cell.

Because target battery system state space vector is corresponding with the electrokinetic cell parameter of current time, after obtaining this target battery system state space vector, can therefrom obtain the state-of-charge of the electrokinetic cell in this current moment.

In the present embodiment, by initial correction matrix and filter gain preestimating battery system state space vector is revised, thereby obtain corresponding target battery system state space vector, and finally from target battery system state space vector, obtain corresponding state-of-charge.As seen, in this programme, when the electrokinetic cell parameter is inaccurate, reduced the estimation error of this state-of-charge by increasing makeover process, and further reduced the follow-up default estimation error of estimating state-of-charge constantly, thereby improved the accuracy to the state-of-charge estimation of electrokinetic cell.

In the production application of reality, for more accurate arbitrary default estimation SOC value of electrokinetic cell constantly that obtains, in the another embodiment of the present invention, a kind of state-of-charge evaluation method of electrokinetic cell as shown in Figure 2, can comprise:

S201: in this operational process of electrokinetic cell, judge current time whether be default estimation constantly, if, execution in step S202 then, if not, execution in step S203 then;

S202: judge that whether this default estimation is and moment of n the second Preset Time section at interval operation zero hour of electrokinetic cell constantly, if, execution in step S206 then, if not, execution in step S205 then;

Wherein, n 〉=1 and n are integer.

S203: judge that whether current time is that this that be not less than first Preset Time non-working time of electrokinetic cell before this task brings into operation constantly, if then execution in step S204 if not, does not then process;

S204: gather the open-circuit voltage of electrokinetic cell, according to default open-circuit voltage and the mapping relations of state-of-charge, obtain the state-of-charge of electrokinetic cell;

S205: with the electrokinetic cell parameter of end of run last time electrokinetic cell constantly time electrokinetic cell parameter as electrokinetic cell;

S206: electrokinetic cell is preset estimation electrokinetic cell parameter constantly as the electrokinetic cell parameter of electrokinetic cell last one;

S207: determine with the initial cells system state space vector of this electrokinetic cell parameter as vector element;

S208: obtain the initial correction matrix corresponding with this electrokinetic cell;

Wherein, this initial correction matrix is the default matrix that obtains according to described battery model.

S209: gather the electrokinetic cell system coefficient of regime of this electrokinetic cell, and according to this electrokinetic cell system coefficient of regime and this initial cells system state space vector, calculate the preestimating battery system state space vector corresponding with this electrokinetic cell;

Wherein, this battery system coefficient of regime is the parameter that obtains according to battery model.

S210: the initial error covariance matrix according in this electrokinetic cell system coefficient of regime and this initial correction matrix, calculate the predictor error covariance matrix corresponding with this electrokinetic cell;

S211: according to this predictor error covariance matrix and this electrokinetic cell system coefficient of regime, calculate and the corresponding filter gain of this electrokinetic cell;

S212: according to the predictor error covariance matrix in this filter gain, this initial correction matrix, this preestimating battery system state space vector is revised, to obtain the target battery system state space vector corresponding with this electrokinetic cell;

S213: from this target battery system state space vector, determine the state-of-charge corresponding with this electrokinetic cell.

In the present embodiment, by initial correction matrix and filter gain preestimating battery system state space vector is revised, thereby obtain corresponding target battery system state space vector, and finally from target battery system state space vector, obtain corresponding state-of-charge as seen, in this programme, when the electrokinetic cell parameter is inaccurate, reduced the estimation error of this state-of-charge by increasing makeover process, and further reduced the follow-up default estimation error of estimating state-of-charge constantly, thereby improved the accuracy to the state-of-charge estimation of electrokinetic cell.

Corresponding to above-mentioned method embodiment, the embodiment of the invention also provides a kind of state-of-charge estimating device of electrokinetic cell, and as shown in Figure 3, the state-of-charge estimating device of this electrokinetic cell can comprise:

Electrokinetic cell parameter determination module 310 is used for this operational process at electrokinetic cell, when current time is estimated constantly for presetting, determines the electrokinetic cell parameter of described electrokinetic cell;

Wherein, described electrokinetic cell parameter is the parameter that obtains from the battery model corresponding with described electrokinetic cell; Described default estimation constantly for moment of this time at intervals n second Preset Time section that bring into operation of described electrokinetic cell, and the non-working time of described electrokinetic cell before this task bring into operation constantly less than this of first Preset Time, n 〉=1 and n are integer.

Initial cells system state space vector determination module 320 is used for determining with the initial cells system state space vector of described electrokinetic cell parameter as vector element;

Initial correction matrix acquisition module 330 is used for obtaining the initial correction matrix corresponding with described electrokinetic cell;

Wherein, described initial correction matrix is the default matrix that obtains according to described battery model;

Preestimating battery system state space vector calculation module 340, be used for gathering the electrokinetic cell system coefficient of regime of described electrokinetic cell, and according to described electrokinetic cell system coefficient of regime and described initial cells system state space vector, calculate the preestimating battery system state space vector corresponding with described electrokinetic cell;

Wherein, described battery system coefficient of regime is the parameter that obtains according to described battery model;

Filter gain computing module 350 is used for according to described predictor error covariance matrix and described electrokinetic cell system coefficient of regime, calculates and the corresponding filter gain of described electrokinetic cell;

Target battery system state space vector acquisition module 360, be used for the predictor error covariance matrix according to described filter gain, described initial correction matrix, described preestimating battery system state space vector is revised, to obtain the target battery system state space vector corresponding with described electrokinetic cell;

State-of-charge determination module 370 is used for determining the state-of-charge corresponding with described electrokinetic cell from described target battery system state space vector.

In the present embodiment, the filter gain that the initial correction matrix that obtains by initial correction matrix acquisition module and filter gain computing module calculate is revised the preestimating battery system state space vector in the preestimating battery system state space vector calculation module, thereby obtain corresponding target battery system state space vector, and the final corresponding state-of-charge of target battery system state space vector acquisition from the state-of-charge determination module.As seen, in this programme, when the electrokinetic cell parameter is inaccurate, reduced the estimation error of this state-of-charge by increasing makeover process, and further reduced the follow-up default estimation error of estimating state-of-charge constantly, thereby improved the accuracy to the state-of-charge estimation of electrokinetic cell.

Description by above method embodiment, the those skilled in the art can be well understood to the present invention and can realize by the mode that software adds essential general hardware platform, can certainly pass through hardware, but the former is better embodiment under a lot of situation.Based on such understanding, the part that technical scheme of the present invention contributes to prior art in essence in other words can embody with the form of software product, this computer software product is stored in the storage medium, comprise that some instructions are with so that a computer equipment (can be personal computer, server, the perhaps network equipment etc.) carry out all or part of step of the described method of each embodiment of the present invention.And aforesaid storage medium comprises: various media that can be program code stored such as ROM (read-only memory) (ROM), random-access memory (ram), magnetic disc or CD.

Claims (11)

1. the state-of-charge evaluation method of an electrokinetic cell is characterized in that, comprising:

In this operational process of electrokinetic cell, when current time is estimated constantly for presetting, determine the electrokinetic cell parameter of described electrokinetic cell, wherein, described electrokinetic cell parameter is the parameter that obtains from the battery model corresponding with described electrokinetic cell; Described default estimation constantly for moment of this time at intervals n second Preset Time section that bring into operation of described electrokinetic cell, and the non-working time of described electrokinetic cell before this task bring into operation constantly less than this of first Preset Time, n 〉=1 and n are integer;

Determine with the initial cells system state space vector of described electrokinetic cell parameter as vector element;

Obtain the initial correction matrix corresponding with described electrokinetic cell, wherein, described initial correction matrix is the default matrix that obtains according to described battery model;

Gather the electrokinetic cell system coefficient of regime of described electrokinetic cell, and according to described electrokinetic cell system coefficient of regime and described initial cells system state space vector, calculate the preestimating battery system state space vector corresponding with described electrokinetic cell, wherein, described battery system coefficient of regime is the parameter that obtains according to described battery model;

According to the initial error covariance matrix in described electrokinetic cell system coefficient of regime and the described initial correction matrix, calculate the predictor error covariance matrix corresponding with described electrokinetic cell;

According to described predictor error covariance matrix and described electrokinetic cell system coefficient of regime, calculate and the corresponding filter gain of described electrokinetic cell;

According to the predictor error covariance matrix in described filter gain, the described initial correction matrix, described preestimating battery system state space vector is revised, to obtain the target battery system state space vector corresponding with described electrokinetic cell;

From described target battery system state space vector, determine the state-of-charge corresponding with described electrokinetic cell.

2. method according to claim 1 is characterized in that, the electrokinetic cell parameter of described definite electrokinetic cell comprises:

When described default estimation constantly be the non-working time of described electrokinetic cell before this task less than this operation of first Preset Time during zero hour, with the electrokinetic cell parameter of end of run last time described electrokinetic cell during the moment electrokinetic cell parameter as described electrokinetic cell;

When described default estimation constantly for operation zero hour of described electrokinetic cell, n(n 〉=1 and n were integer at interval) during moment of the individual second Preset Time section, with described electrokinetic cell in the electrokinetic cell parameter in the last one default estimation moment electrokinetic cell parameter as described electrokinetic cell.

3. method according to claim 1 and 2 is characterized in that, also comprises:

When not reaching default estimation during the moment, be that this that be not less than first Preset Time non-working time of described electrokinetic cell before this task is when bringing into operation constantly judging current time, gather the open-circuit voltage of described electrokinetic cell, according to default open-circuit voltage and the mapping relations of state-of-charge, obtain the state-of-charge of described electrokinetic cell.

4. method according to claim 2, it is characterized in that described battery model is the Thevenin battery model, corresponding, the electrokinetic cell parameter of end of run last time described electrokinetic cell during the moment, last one default estimation electrokinetic cell parameter constantly comprise respectively:

The polarizing voltage of electrokinetic cell, ohmic internal resistance, polarization resistance, polarization capacity and state-of-charge.

5. method according to claim 4 is characterized in that, described initial correction matrix comprises:

Initial error covariance matrix, starter system noise equation matrix, initial observation noise variance matrix and initial gain correction factor matrix.

6. method according to claim 5 is characterized in that, described electrokinetic cell system coefficient of regime comprises: the terminal voltage of electrokinetic cell, current signal and temperature signal.

7. method according to claim 6, it is characterized in that, according to described electrokinetic cell system coefficient of regime and described initial cells system state space vector, calculate the formula that the preestimating battery system state space vector corresponding with described electrokinetic cell adopts, comprising:

$\hat{X}(k+1|k)=\left[\begin{array}{ccccc}1-\frac{T_s}{x_3\left(k\right)x_4\left(k\right)}& 0& 0& 0& 0\\ 0& 1& 0& 0& 0\\ 0& 0& 1& 0& 0\\ 0& 0& 0& 1& 0\\ 0& 0& 0& 0& 1\end{array}\right]\left[\begin{array}{c}{x}_1\left(k\right)\\ x_2\left(k\right)\\ x_3\left(k\right)\\ x_4\left(k\right)\\ x_5\left(k\right)\end{array}\right]+\left[\begin{array}{c}\frac{T_s}{x_4\left(k\right)}\\ 0\\ 0\\ 0\\ \frac{{\mathrm{\ηT}}_s}{C_N}\end{array}\right]I\left(k\right),$

Wherein, X (k) is the last corresponding initial cells system state space vector of a moment of described current time, when described current time is the non-working time of described electrokinetic cell before this task less than this operation of first Preset Time during zero hour, described last one constantly is the end of run moment of described electrokinetic cell, when described current time for operation zero hour of described electrokinetic cell, n(n 〉=1 and n were integer at interval) during moment of the individual second Preset Time section, described last one constantly be last one defaultly to estimate the moment;

Preestimating battery system state space vector for the current time that goes out according to described last one constantly initial cells system state space vector calculation; T _SBe the current signal sampling period of electrokinetic cell, C _NBe the rated capacity of electrokinetic cell, η is enclosed pasture efficient and C _N, η is the function of temperature signal.

8. method according to claim 6, it is characterized in that, according to the initial error covariance matrix in described electrokinetic cell coefficient of regime and the described initial correction matrix, calculate the formula that the predictor error covariance matrix corresponding with described electrokinetic cell adopts, comprising:

$P(k+1|k)=\hat{A}\left(k\right)P\left(k\right){\hat{A}}^T\left(k\right)+Q$

Wherein, P (k) is the last one initial error covariance matrix in the corresponding initial correction matrix constantly of described current time, P (k+1/k) was the predictor error covariance matrix according to the current time that calculates of initial error covariance matrix in described a lasted moment, when described current time is the non-working time of described electrokinetic cell before this task less than this operation of first Preset Time during zero hour, described last one constantly is the end of run moment of described electrokinetic cell, when described current time for operation zero hour of described electrokinetic cell, n(n 〉=1 and n were integer at interval) during moment of the individual second Preset Time section, described last one constantly be last one defaultly to estimate the moment; Q is the system noise variance matrix;

Be the state transformation matrix, its calculating formula is:

$\hat{A}\left(k\right)=\left[\begin{array}{ccccc}1-\frac{T_s}{{\hat{x}}_3\left(k\right){\hat{x}}_4\left(k\right)}& 0& \frac{T_s{\hat{x}}_1\left(k\right)}{{\hat{x}}_3^2\left(k\right){\hat{x}}_4\left(k\right)}& \frac{T_s({\hat{x}}_1\left(k\right)-{\hat{x}}_3\left(k\right)I\left(k\right))}{{\hat{x}}_3\left(k\right){\hat{x}}_4^2\left(k\right)}& 0\\ 0& 1& 0& 0& 0\\ 0& 0& 1& 0& 0\\ 0& 0& 0& 1& 0\\ 0& 0& 0& 0& 1\end{array}\right].$

9. method according to claim 8 is characterized in that, according to described predictor error covariance matrix and described electrokinetic cell system coefficient of regime, calculates the formula that the filter gain corresponding with described electrokinetic cell adopts, and comprising:

L(k)=ζ*K(k)，

Wherein, ζ is gain correction factor matrix;

$K\left(k\right)=P(k+1|k){\hat{C}}^T(k+1){[\hat{C}(k+1)P(k+1|k){\hat{C}}^T(k+1)+R]}^{-1},$ R is the observation noise variance matrix;

Be the observation model matrix, its calculating formula is: $\hat{C}\left(k\right)=\left[\begin{array}{ccccc}-1& -I\left(k\right)& 0& 0& \frac{\∂F({\hat{x}}_5\left(k\right),T\left(k\right))}{\∂{\hat{x}}_5\left(k\right)}\end{array}\right],$ F (x ₅(k), T (k)) be state-of-charge (SOC) the battery open circuit voltage OCV function corresponding with battery temperature (T (k)) of battery.

10. method according to claim 9 is characterized in that, according to the predictor error covariance matrix in described filter gain, the described initial correction matrix, described preestimating battery system state space vector is revised the formula that adopts, and comprising:

$\hat{X}(k+1)=\hat{X}(k+1/k)+L\left(k\right)\{V(k+1)-g[\hat{X}(k+1/k),I(k+1)\left]\right\};$

Wherein, g[X (k), I (k)] for measuring equation, its calculating formula is:

G[X (k), I (k)]=F (x ₅(k))-x ₂(k) I (k)-x ₁(k); V (k+1) is the terminal voltage of the electrokinetic cell of current time collection; I (k+1) is the end electric current of the electrokinetic cell of the current collection time of running;

$P(k+1)=[I-L\left(k\right)C\left((\hat{k}+1)\right]P(k+1/k),$ Wherein I is unit matrix;

Be the observation model matrix, its calculating formula is: $\hat{C}\left(k\right)=\left[\begin{array}{ccccc}-1& -I\left(k\right)& 0& 0& \frac{\∂F({\hat{x}}_5\left(k\right),T\left(k\right))}{\∂{\hat{x}}_5\left(k\right)}\end{array}\right].$

11. the state-of-charge estimating device of an electrokinetic cell is characterized in that, comprising:

Electrokinetic cell parameter determination module, be used for this operational process at electrokinetic cell, when current time is estimated constantly for presetting, determine the electrokinetic cell parameter of described electrokinetic cell, wherein, described electrokinetic cell parameter is the parameter that obtains from the battery model corresponding with described electrokinetic cell; Described default estimation constantly for moment of this time at intervals n second Preset Time section that bring into operation of described electrokinetic cell, and the non-working time of described electrokinetic cell before this task bring into operation constantly less than this of first Preset Time, n 〉=1 and n are integer;

Initial cells system state space vector determination module is used for determining with the initial cells system state space vector of described electrokinetic cell parameter as vector element;

Initial correction matrix acquisition module is used for obtaining the initial correction matrix corresponding with described electrokinetic cell, and wherein, described initial correction matrix is the default matrix that obtains according to described battery model;

Preestimating battery system state space vector calculation module, be used for gathering the electrokinetic cell system coefficient of regime of described electrokinetic cell, and according to described electrokinetic cell system coefficient of regime and described initial cells system state space vector, calculate the preestimating battery system state space vector corresponding with described electrokinetic cell, wherein, described battery system coefficient of regime is the parameter that obtains according to described battery model;

The filter gain computing module is used for according to described predictor error covariance matrix and described electrokinetic cell system coefficient of regime, calculates and the corresponding filter gain of described electrokinetic cell;

Target battery system state space vector acquisition module, be used for the predictor error covariance matrix according to described filter gain, described initial correction matrix, described preestimating battery system state space vector is revised, to obtain the target battery system state space vector corresponding with described electrokinetic cell;

The state-of-charge determination module is used for determining the state-of-charge corresponding with described electrokinetic cell from described target battery system state space vector.

Images