CN205898985U - Power lithium cell SOC estimates system for pure electric vehicles - Google Patents

Abstract

The utility model provides a power lithium cell SOC estimates system for pure electric vehicles, data acquisition module gathers terminal voltage, electric current and the temperature data who obtains power battery through voltage sensor, current sensor and temperature sensor, then passes to SOC with the data of gathering and estimate the module, the utility model discloses use forgetting factor's the real -time on -line identification battery model parameter of recursion least squares method to estimate to obtain the SOC value in inputing the SOC estimator with real -time model parameter, the output of final SOC value is confirmed according to whole turner condition to the logical decision module, control module judges the SOC estimate condition, sends corresponding information to whole car and instructs, and the while shows estimation results on the LCD display.

Description

A kind of pure electric automobile dynamic lithium battery soc estimating system

Technical field

This utility model belongs to electric automobile power battery technical field and in particular to a kind of pure electric automobile power lithium Battery soc estimating system.

Background technology

In power battery management system, the prediction of battery charge state soc (state of charge) has important meaning Justice, the accuracy of its prediction, directly affect the control strategy of battery management system, thus affecting performance and the battery of battery performance The length in life-span.

Meanwhile, accurate battery model has great importance for the assessment algorithm of state-of-charge soc, due to battery tool There is the non-linear behavior of height, battery model is very good with the concordance of electrokinetic cell, just can draw more accurately prediction knot Really.

At present, conventional electrokinetic cell model has three classes: electrochemical model, artificial nerve network model and equivalent circuit mould Type.Electrochemical model can be described in more detail the electrochemical reaction process of inside battery, but its structure is extremely complex, uncomfortable Close battery soc to estimate；Artificial nerve network model has the features such as non-linear, fault-tolerance of height, property learnt by oneself, it is possible to achieve essence True soc estimates, but it is disadvantageous in that the substantial amounts of experimental data of needs to predict the performance of battery, and to battery history number According to dependency larger；Equivalent-circuit model forms circuit network by components such as traditional resistance, electric capacity, constant pressure sources The external characteristics of description electrokinetic cell, because its structure is simple and can describe battery behavior well, is often used by researcher.

Electrokinetic cell soc method of estimation mainly has: open circuit voltage method, ampere-hour integration method, Kalman filtering method.Open-circuit voltage Method can only realize in laboratory conditions estimating offline it is impossible to real-time estimation；Ampere-hour integration method classics are easy-to-use, but its estimated accuracy Affected larger by the precision of soc initial value and current measurement value；Kalman filter method has EKF, no mark karr again The species such as graceful filtering, adaptive Kalman filter, Kalman filtering method can more accurately predict soc value, is that battery soc estimates Meter research is using most methods.

Utility model content

This utility model purpose is for the problems referred to above, provides one kind by battery model parameter and soc state value joint The system estimated, to reduce the impact that battery model lax pair soc estimates accuracy.

The pure electric automobile dynamic lithium battery soc estimating system that this utility model provides, this system includes power lithium battery Pond 1, voltage detection module 2, current detection module 3, temperature sensor 4, soc estimation module 5, control module 6；

Wherein, dynamic lithium battery 1 connects voltage detection module 2, current detection module 3, temperature sensor 4 respectively；

Voltage detection module 2, current detection module 3, temperature sensor 4 are respectively used to gather terminal voltage, the electricity of lithium battery Stream and temperature data；And voltage detection module 2, current detection module 3, temperature sensor 4 connect to soc estimation module 5, will The data message collecting sends to soc estimation module 5；

Described soc estimation module 5 adopts using open circuit voltage method, ampere-hour integration method and kalman filter method to battery Soc value is estimated and is corrected, and exports to control module 6；

Described control module 6 is connected on dynamic lithium battery 1, and the output valve according to soc estimation module 5 controls power lithium battery The charge and discharge output in pond 1.

This system also includes a lcd display 7, and described lcd display 7 is connected with control module 6, current for showing Soc value.

Described soc estimation module 5 specifically includes model parameter on-line identification module 8, soc estimator 9 and logical judgment mould Block 10；Three module unidirectional connections successively, and logic judgment module 10 is connected to model parameter on-line identification module 8；Described Logic judgment module 10 is exported as final soc value after soc estimated value being corrected by the use of kalman filter method.

In soc estimation module 5, model parameter on-line identification module 8 combines On-line Estimation with soc estimator 9.

Beneficial effect

During traditional soc estimates, battery model parameter is often off-line identification gained, because battery changes with external environment With itself capacity attenuation, model parameter will appear from very big error, and soc estimated value is also no longer accurate.This utility model uses and contains The on-line identification of the least square method of recursion implementation model parameter of forgetting factor, soc estimator is gone based on online model parameter Estimate soc value, the discreet value of output, as the input value of model parameter on-line identification, forms closed loop system.

When logic judgment module detects that electric automobile does not run for a long time, estimate that soc value is made using open circuit voltage method For final estimated value.Electric automobile run after a period of time within, using ampere-hour integration method estimate soc, when the time exceed longer When, ampere-hour integration method occurs cumulative errors, now utilizes extended Kalman filter to obtain soc value relatively accurately, interacts afterwards Ampere-hour integration method and extended Kalman filter are used as soc estimated result.

Control module, according to the output result of soc estimation module, when soc estimated value is less than 15%, controls battery discharge Output size, and result is exported lcd display screen, remind driver's electricity low；When soc value is for 100%, control mould Block sends instruction to charger, terminates charging.

Brief description

Fig. 1 is a kind of structural representation of pure electric automobile dynamic lithium battery soc estimating system；

Fig. 2 is the ekf algorithm flow chart in a kind of pure electric automobile dynamic lithium battery soc method of estimation；

Fig. 3 is a kind of logic judgment module flow chart of pure electric automobile dynamic lithium battery soc estimating system；

In figure, 1- dynamic lithium battery, 2- voltage detection module, 3- current detection module, 4- temperature sensor, 5-soc estimate Meter module, 6- control module, 7-lcd display, 8- model parameter on-line identification module, 9-soc estimator, 10- logical judgment Module.

Specific embodiment

This utility model provides a kind of pure electric automobile dynamic lithium battery soc estimating system, of the present utility model for making Purpose, technical scheme and effect are clearer, clearly, and referring to the drawings and give an actual example to this utility model further specifically Bright.It should be appreciated that described herein be embodied as, only in order to explain this utility model, being not used to limit this utility model.

As shown in figure 1, a kind of pure electric automobile dynamic lithium battery soc estimating system that this utility model provides, including Dynamic lithium battery 1, voltage detection module 2, current detection module 3, temperature sensor 4, soc estimation module 5, control module 6 He Lcd display 7, module 5 includes model parameter on-line identification module 8, soc estimator 9 and logic judgment module 10.

Voltage detection module 2, current detection module 3 and temperature sensor 4 collect the corresponding data of dynamic lithium battery, and Pass it to soc estimation module 5, control module 6 detects soc estimated value and dynamic lithium battery is sent with corresponding instruction, prevents Only over-charging of battery is put with crossing, and soc estimated result is shown on lcd display 7 simultaneously, reminds driver's battery electric quantity to use feelings Condition.

Based on said system, specific lithium battery soc method of estimation is done and supplements introduction:

Step one: by voltage detection module 2, the dynamic lithium battery phase that current detection module 3 and temperature sensor 4 collect Answer voltage, electric current and temperature data for model parameter on-line identification, the recursive least-squares parameter identification based on forgetting factor The concretely comprising the following steps of method:

Battery model is using second order rc equivalent-circuit model:

${\stackrel{\·}{u}}_1=i/c_1-u_1/\left(r_1{c}_1\right);---\left(1\right)$

${\stackrel{\·}{u}}_2=i/c_2-u_2/\left(r_2{c}_2\right);---\left(2\right)$

Ut=uoc-u₁-u₂-irs； (3)

Wherein, u_ocFor open-circuit voltage；u_tFor terminal voltage；I is electric current；r_sFor ohmic internal resistance；r₁、c₁Indicated concentration difference polarization Reflection, r₁For concentration difference polarization resistance, c₁For concentration difference polarization capacity, u₁For concentration difference polarizing voltage；r₂、c₂Represent electrochemistry pole Change reflection, r₂For activation polarization internal resistance, c₂For activation polarization electric capacity, u₂For activation polarization voltage.

Step one Chinese style 1,2,3 sliding-model control is obtained with model difference equation:

Ut (k)=m₀+m₁ut(k-1)+m₂ut(k-2)+m₃i(k)+m₄i(k-1)+m₅i(k-2) (4)

In formula, m₀、m₁、m₂、m₃、m₄、m₅For model difference equation undetermined coefficient, its value and parameter Cheng Han to be identified in model Number relation.

Formula (4) can be write asForm, wherein

θ=[m₀, m₁, m₂, m₃, m₄, m₅] (6)

Determine least square covariance p₀With the initial value of parameter matrix θ, set up least square gain matrix k_k:

$k_k=p_{k-1}{h}_k{(h_k^t{p}_{k-1}{h}_k+\mathrm{\υ})}^{-1}---\left(7\right)$

In formula, υ is least square weighter factor, takes υ=0.98.

Obtain calculating parameter estimated matrix θ after time-varied gain matrix_k:

${\mathrm{\θ}}_k={\mathrm{\θ}}_{k-1}+k_k(y_k-h_k^t{\mathrm{\θ}}_{k-1})---\left(8\right)$

Y in formula_kFor the terminal voltage measured value in k moment, θ_kFor θ_k-1In the estimates of parameters to the k moment for the k-1 moment.

The renewal of covariance matrix:

$p_k=(i-k_k{h}_k^t)p_{k-1}---\left(9\right)$

So, just complete a step recursion of the RLS based on forgetting factor, repeat this process, identification Go out m₀、m₁、m₂、m₃、m₄、m₅Value, and then draw r_s、r₁、c₁、r₂、c₂Value.

Step 2: such as Fig. 2 EKF algorithm for estimating:

Estimate that equation obtains based on second order rc equivalent-circuit model:

x_k=a_k-1x_k-1+b_k-1u_k-1+ω_k-1(10)

y_k=c_kx_k+d_ku_k+v_k(11)

Wherein, x_kIt is k moment state variable；y_kIt is k moment observational variable；u_kIt is the input control variable in k moment；ω_k、v_k It is orthogonal system noise；

Battery status equation is listed according to ampere-hour integral formula and formula (1), (2):

$\left[\begin{array}{c}soc\left(k\right)\\ u_1\left(k\right)\\ u_2\left(k\right)\end{array}\right]=\left[\begin{array}{ccc}1& 0& 0\\ 0& 1-\frac{t}{r_1{c}_1}& 0\\ 0& 0& 1-\frac{t}{r_2{c}_2}\end{array}\right]*\left[\begin{array}{c}soc(k-1)\\ u_1(k-1)\\ u_2(k-1)\end{array}\right]+\left[\begin{array}{c}-\frac{\mathrm{\η}t}{q_n}\\ \frac{t}{c_1}\\ \frac{t}{c_2}\end{array}\right]*i(k-1)---\left(12\right)$

Know that battery observational equation is by formula (3):

Ut (k)=uoc [soc (k)]-u₁(k)-u₂(k)-i(k)rs (13)

In formula (12), (13), order:

$a=\left[\begin{array}{ccc}1& 0& 0\\ 0& 1-\frac{t}{r_1{c}_1}& 0\\ 0& 0& 1-\frac{t}{r_2{c}_2}\end{array}\right],b=\left[\begin{array}{c}-\frac{\mathrm{\η}t}{q_n}\\ \frac{t}{c_1}\\ t\\ c_2\end{array}\right],c=\left[\begin{array}{ccc}uoc\left(soc\right)& -1& -1\end{array}\right],d=\[-r_s\]$

U in observational equation_oc(soc) it is nonlinear function with regard to soc, take the first order Taylor of equation to launch to carry out linearisation Process, obtain observing matrix

Determine the state error covariance matrix initial value p of EKF₀, system covariance q₀And r₀, start extension Kalman filtering algorithm.

EKF predictive equation:

The estimation of state variable: x_k=a_k-1x_k-1+b_k-1u_k-1+ω_k-1(14)

State covariance is estimated:

Kalman gain matrix: k_k=p_kh^t(hp_kh^t+r_k)^-1(16)

State-updating: x_k+1=x_k+k_k(y_k-hx_k) (17)

State covariance is estimated to update: p_k+1=(i-k_kh)p_k(18)

In EKF, terminal voltage estimates that belonging to closed loop estimates, after some step iteration update, terminal voltage u_t Gradually approaching to reality value；U simultaneously₁And u₂Value calculated by systematic parameter, estimate electricity further according to observation equation (3) Pond electromotive force u_ocAnd then estimate battery soc, it is updated in state equation, calculate new state estimation using ampere-hour integration method Value, realizes the line closed loop method of estimation of soc.

Step 3: as shown in figure 3, with the addition of Logic control module, when battery does not work for a long time, using open circuit electricity Platen press corrects soc value；Estimate soc using ampere-hour integration method, the time is longer to occur cumulative errors so that estimated value soc_ahInaccurate Really；The soc that the ekf method of estimation of model parameter on-line identification obtains_ekfRelatively accurate, but its estimated accuracy is still not so good as in a short time Accurately, in addition the method takies running space greatly to ampere-hour integration, under some conditions, can waste battery management system operation empty Between.This utility model combines advantage between the two, is obtained after soc initial value using ekf method, is used interchangeably two methods, fixed Phase utilizes the soc that the ekf method of estimation of model parameter on-line identification obtains_ekfGo to calibrate the estimation obtaining using ampere-hour integration method Value soc_ah, the precision of soc estimation can be met, running space can be saved again, meet practical application.

Step 4: control module, according to the output result of soc estimation module, when soc estimated value is less than 15%, controls electricity Tank discharge output size, and result is exported lcd display screen, remind driver's electricity low；When soc value is for 100%, Control module sends instruction to charger, terminates charging.

Claims (4)

1. a kind of pure electric automobile dynamic lithium battery soc estimating system is it is characterised in that this system includes dynamic lithium battery (1), voltage detection module (2), current detection module (3), temperature sensor (4), soc estimation module (5), control module (6)；

Wherein, dynamic lithium battery (1) connects voltage detection module (2), current detection module (3), temperature sensor (4) respectively；

Voltage detection module (2), current detection module (3), temperature sensor (4) be respectively used to gather lithium battery terminal voltage, Electric current and temperature data；And voltage detection module (2), current detection module (3), temperature sensor (4) connect to soc estimation Module (5), the data message collecting is sent to soc estimation module (5)；

Described soc estimation module (5) adopts using open circuit voltage method, ampere-hour integration method and kalman filter method to battery soc Value is estimated and is corrected, and exports to control module (6)；

Described control module (6) is connected on dynamic lithium battery (1), and the output valve according to soc estimation module (5) controls power lithium The charge and discharge output of battery (1).

2. a kind of pure electric automobile according to claim 1 with dynamic lithium battery soc estimating system it is characterised in that should System also includes a lcd display (7), and described lcd display (7) is connected with control module (6), current for showing Soc value.

3. a kind of pure electric automobile according to claim 2 with dynamic lithium battery soc estimating system it is characterised in that institute State soc estimation module (5) and specifically include model parameter on-line identification module (8), soc estimator (9) and logic judgment module (10)；Three module unidirectional connections successively, and logic judgment module (10) is connected to model parameter on-line identification module (8)； Described logic judgment module (10) is defeated as final soc value after soc estimated value being corrected by the use of kalman filter method Go out.

4. a kind of pure electric automobile according to claim 2 with dynamic lithium battery soc estimating system it is characterised in that soc In estimation module (5), model parameter on-line identification module (8) combines On-line Estimation with soc estimator (9).