CN103033752B - Battery of electric vehicle life-span prediction method and prolongation method - Google Patents

Abstract

The invention provides a kind of electric vehicle lithium battery life-span prediction method, according to the working temperature of the battery obtained, the ratio of the energy after being full of electricity when the energy content of battery and battery just dispatch from the factory needed for discharge and recharge frequency and driver drive for a day is to predict battery life.The present invention also provides a kind of electric vehicle lithium battery life-span prediction method, comprises the character parameter of driver according to obtaining, charging strategy, environment temperature, working day and weekend the driving distance on highway and urban road predict battery life scope.The present invention also provides electric vehicle lithium battery life method according to above-mentioned formula.The Accurate Prediction that said method is service life of lithium battery provides basis; Consider that the behavioural characteristic of driver predicts the service life of lithium battery of electric motor car, predict the outcome truer; Do not need hardware just can predict service life of lithium battery, cost is very low.

Description

Battery of electric vehicle life-span prediction method and prolongation method

Technical field

The present invention relates to electric automobiles, particularly relate to a kind of battery of electric vehicle life-span prediction method and prolongation method.

Background technology

Current environmental problem and the oil production rate problem that may exist facilitate the development of various electric automobile.Compared with orthodox car, electric motor car can play a significant role in the discharge of decreasing pollution thing and energy resource consumption.In electric motor car, the life-span of on-vehicle battery and cost directly can affect the performance of electric motor car, life-span and cost, predict that the life-span of battery has become a current major issue.

First, in existing research, researchist only with testing the qualitative relationships that demonstrates between battery life and correlative factor, not for battery life provides mathematical quantitative forecast.

In addition, the existing research about battery life concentrates on the physics and chemistry process of battery mostly, and such as state-of-charge (SOC), working temperature, discharge and recharge number of times, but ignore the impact of driver.Because driver is as the operator of electric automobile, its behavior directly affects the life-span of battery, so, not consider or the behavioural characteristic of insufficient consideration driver can not predict the battery life of electric motor car comprehensively, really.

In sum, lack the mathematical battery of electric vehicle life-span quantitative forecasting technique considering battery life correlative factor in prior art, in addition, in prior art, also lack the battery of electric vehicle life-span prediction method considering Characteristics of drivers' behavior.

Summary of the invention

The technical problem to be solved in the present invention is to provide a kind of battery of electric vehicle life-span prediction method, provides electric motor car lithium electric life quantitative forecast mathematically.

For solving the problem, the invention provides a kind of electric vehicle lithium battery life-span prediction method, comprising:

The working temperature of step one, acquisition battery, is full of the ratio of the energy after electricity when the energy content of battery and battery just dispatch from the factory needed for discharge and recharge frequency and driver drive for one day; With

Step 2, according to following manner prediction battery life:

$\mathrm{Capacity}(\%)=\frac{1}{100}\·(167.583-1.264T-0.097\·t\·f)\·(84.1+3.71\·f-0.0788\·t\·f)$

Wherein, T represents the working temperature of battery, and unit is DEG C; T represents elapsed time, and be elapsed time length battery calculates battery cycle number during from dispatching from the factory to, unit is sky; Electricity after Capacity (%) represents battery charging completely and battery are full of the ratio of the energy after electricity when just appearing on the scene; F represents discharge and recharge frequency;

Suppose when battery be full of the electricity after electricity cannot provide car day travel institute's energy requirement time, represent that battery life finishes, the ratio C apacity (%) of the energy after electricity is full of when the energy content of battery and battery just dispatch from the factory needed for then within one day, travelling according to charge frequency f during this critical point, battery operating temperature T, driver, obtain elapsed time t, namely this value represents the battery predictive life-span.

Optionally, the working temperature of battery calculates according to following manner:

$T=T_{\mathrm{initial}}+0.0339+0.004t^{\′}+0.001\·C+2.385(\frac{\mathrm{\Σ}\stackrel{\·}{q}\mathrm{\Δt}}{t^{\′}\·N}-\mathrm{\Δheat})-0.004\·C_p$

Wherein, T _initialrepresent the initial temperature of cell operations, unit is DEG C; T ' expression battery one-time continuous the working time, unit is second; C represents discharge rate of battery; represent the heat production rate of electric battery, unit is watt; Δ t represents the time interval of pertinent instruments record driving data, and unit is second; N represents the cell number in an electric battery; Δ heat represents relative atmospheric refrigeration module, the heat fade difference of the refrigerating module generation of other types, and unit is W/cell; C _prepresent the thermal capacitance of battery, unit is J/kg/K.

Another technical matters that the present invention will solve is to provide a kind of battery of electric vehicle life-span prediction method, considers the behavioural characteristic of driver, the service life of lithium battery of comprehensive, real prediction electric motor car.

For solving the problem, the invention provides a kind of electric vehicle lithium battery life-span prediction method, comprising:

Step one, obtain the character parameter of driver, charging strategy, environment temperature, working day and weekend the driving distance on highway and urban road;

Step 2, according to following manner prediction battery life scope:

Lifetime＝pa+pb×Personality+pc×Charging+pd×T _initial

+pe×D _hd+pf×D _ud+pg×D _he+ph×D _ue

Wherein, Personality represents the character parameter of driver, and the quantification manner of the character parameter of driver is: represent non-impulsive style successively with-1,0 and 1, normal and impulsive style; Charging represents charging strategy, i.e. the minimum electricity of the front remaining battery of battery charging, and value is a number percent; T _initialrepresent environment temperature; D _hdand D _udrepresent working day highway and urban road on driving distance, D _heand D _uerepresent weekend highway and urban road on driving distance, unit is mile; Pa is constant, and pb ~ ph represents coefficient;

Wherein, pa scope is between 393.7864 ± 147.2502, pb scope is between-35.1552 ± 21.9882, pc scope is between-2.85 ± 2.145, pd scope is between-1.3822 ± 1.1946, and pe scope is between-9.716 ± 2.7264, and pf scope is between-2.5916 ± 3.2112, pg scope is between-4.3606 ± 2.7264, and ph scope is between-11.533 ± 4.9872.

Optionally, the formula of step 2 is:

Lifetime＝393.768-35.1552Personality-2.85Charging-1.382T _initial。

-9.716D _hd-2.592D _ud-4.361D _he-11.533D _ue

Another problem that the present invention will solve is to provide a kind of method extending the electric vehicle lithium battery life-span, is driven and charging behavior extending battery life by cell arrangement optimization and optimum.

For solving the problem, the invention provides a kind of electric vehicle lithium battery life method, comprising:

Step one, the formula in method described in claim 2 to be out of shape as follows:

$\left\{\begin{array}{c}\frac{1}{100}\·(167.583-1.264T-0.097\·t\·f)\·(84.1+3.71\·f-0.0788\·t\·f)*\mathrm{CAPACITY}-E=0\\ T=T_{\mathrm{initial}}+0.0339+0.004t^{\′}+0.001\·C+2.385(\frac{\mathrm{\Σ}\stackrel{\·}{q}\mathrm{\Δt}}{t^{\′}\·N}-\mathrm{\Δheat})-0.004\·C_p\end{array}\right.$

Wherein, CAPACITY is the energy after electric battery is full of electricity when just dispatching from the factory, and unit is J; E is total power consumption in the driving procedure of a day, and unit is J; The time that t is through, represent the target life objective of battery herein, unit is sky;

Step 2, acquisition drive the energy content of battery E consumed, decision references value DMR, charge frequency every day;

Step 3, solve above-mentioned formula, obtain battery allocation optimum, namely arrive C _p, the value of Δ heat and N.

For solving the problem, the present invention also provides a kind of electric vehicle lithium battery life method, comprising:

Step one, the formula in method described in claim 2 to be out of shape as follows:

$\left\{\begin{array}{c}\frac{1}{100}\·(167.583-1.264T-0.097\·t\·f)\·(84.1+3.71\·f-0.0788\·t\·f)*\mathrm{CAPACITY}-E=0\\ T=T_{\mathrm{initial}}+0.0339+0.004t^{\′}+0.001\·C+2.385(\frac{\mathrm{\Σ}\stackrel{\·}{q}\mathrm{\Δt}}{t^{\′}\·N}-\mathrm{\Δheat})-0.004\·C_p\end{array}\right.$

Wherein, CAPACITY is the energy after electric battery is full of electricity when just dispatching from the factory, and unit is J; E is total power consumption in the driving procedure of a day, and unit is J; The time that t is through, represent the target life objective of battery herein, unit is sky;

Step 2, acquisition drive the energy content of battery E consumed, cell arrangement N, C every day _p, Δ heat, initial temperature T _initial;

Step 3, solve above-mentioned formula, obtain optimum C, and f;

Step 4, obtain optimum I by optimum C, then obtain optimum P by following formula:

$\stackrel{\·}{q}=I(U_{\mathrm{ocv}}-U_{\mathrm{op}})={\mathrm{IU}}_{\mathrm{ocv}}-P$

Wherein, P represents the power consumption of vehicle battery packs, and unit is W; represent the heat production rate of electric battery, unit is W; I represents the electric current of electric battery, and unit is A; U _ocvrepresent the open-loop voltage of electric battery, U _oprepresent electric battery operating voltage under that loading condition, unit is volt;

Step 5, obtain optimum drive speed v and acceleration a by following formula:

$p=(\mathrm{ma}+\frac{1}{2}{\mathrm{\ρv}}^2{C}_{d}A+C_{\mathrm{rr}}\mathrm{mg})0.4v;$ $p=\frac{(\mathrm{ma}+\frac{1}{2}{\mathrm{\ρv}}^2{C}_{d}A+C_{\mathrm{rr}}\mathrm{mg})v}{0.8}$

Wherein, p represents electric vehicle motion power consumption, and unit is W; M represents quality, and unit is Kg; A represents vehicle acceleration, and unit is m/s ², ρ sea level air every cubic metre roughly air weight, unit is Kg/m ³; V represents speed, and unit is m/s; C _drepresent the resistance coefficient of vehicle; A represents vehicle front face area, and unit is m ², C _rrrepresent the dimensionless factor of tire drag, g represents acceleration of gravity, and unit is m/s ²,

According to the charge frequency f tried to achieve, speed v and the behavior of acceleration a direct drivers, thus extend the electric vehicle lithium battery life-span.

Compared with prior art, the invention has the advantages that:

The first, propose people-electric motor car model, this model has considered behavioural characteristic and the factor in other influences on-vehicle battery life-span of driver, and the Accurate Prediction for service life of lithium battery provides basis;

The second, consider that the behavioural characteristic of driver predicts the service life of lithium battery of electric motor car, predict the outcome truer;

Three, do not need hardware just can predict service life of lithium battery, cost is very low;

Four, the method for extending battery life mainly concentrates on battery design itself in the past, and the present invention is carried out extending battery life by drive speed, acceleration, charge frequency and obtained the optimal design of battery.

Accompanying drawing explanation

Fig. 1 is the derivation schematic diagram of the theoretical model provided in one embodiment of the invention;

Fig. 2 is the analysis schematic diagram of human factor and the battery life correlativity provided in one embodiment of the invention.

Embodiment

Definition:

(1) the life schedule of driver: working day, weekend driving time on highway and city and distance.

(2) charging strategy: the minimum electricity of remaining battery before battery charging, value is a number percent.

(3) decision references value: in driving experiment, driver needs to select oneself to want the speeds with higher than speed limit how many miles per hours, and what present to option is corresponding " if receiving the monetary cost of penalty note " and " safety that penalty note obtains if do not receive and temporal income " simultaneously.By considering these two factors, driver determines that the difference of this new drive speed and speed limit is " decision references value ", and unit is miles per hour at appearance speed(-)limit sign stylish drive speed.

(4) people-electric motor car empirical model: the formula embodying Characteristics of drivers' behavior, relation between cell arrangement and battery life.

In order to make object of the present invention, technical scheme and advantage clearly understand, below in conjunction with accompanying drawing, the present invention is described in more detail.Should be appreciated that specific embodiment described herein only in order to explain the present invention, be not intended to limit the present invention.

theorize model

According to above-mentioned discovery, in one embodiment of the invention, provide a kind of theoretical model to calculate the life-span of electric vehicle lithium battery.As described in Figure 1, wherein the change of the factor value at junction symbol top can cause the change of junction symbol end factor value.The process of establishing of this model is as follows:

1. connect (H), the relation of drive routine data and energy ezpenditure:

When the power of acceleration needed for vehicle movement when negative value and nonnegative value can use Peterson (Peterson respectively, S.B., J.Apt, andJ.Whitacre, Lithium-ionbatterycelldegradationresultingfromrealisticv ehicleandvehicle-to-gridutilization.JournalofPowerSource s, 2010.195 (8): the formula 6 and 7 that p.2385-2392.) etc. people proposes describes, as following formula 6,7:

$p=(\mathrm{ma}+\frac{1}{2}{\mathrm{\ρv}}^2{C}_{d}A+C_{\mathrm{rr}}\mathrm{mg})0.4v---\left(6\right)$

$p=\frac{(\mathrm{ma}+\frac{1}{2}{\mathrm{\ρv}}^2{C}_{d}A+C_{\mathrm{rr}}\mathrm{mg})v}{0.8}---\left(7\right)$

Wherein, P represents that (unit is w) to electric vehicle motion power consumption, and m represents quality (unit is Kg), and a represents that (unit is m/s to vehicle acceleration ²), ρ sea level air every cubic metre roughly air weight (unit is Kg/m ³), v represents speed, C _drepresent the resistance coefficient of vehicle, A represents that (unit is m to vehicle front face area ²), C _rrrepresent the dimensionless factor of tire drag, g represents that (unit is m/s to acceleration of gravity ²).

Above-mentioned small letter p electrical representation electric motor car campaign power consumption, when pickup is for time negative, uses formula 6, when acceleration is non-negative, uses formula 7.In following content, capitalization P represents automobile batteries group output power.Wherein cell stack power output comprises motor racing and consumes p, and the power consumption of the equipment had nothing to do with vehicle movement.

According to the research of the people such as Peterson, can obtain:

m＝1590kg，ρ＝1.23kg/m ³，C _d＝0.28，A＝2.67m ²，C _rr＝0.01，g＝9.8m/s ²。

Then can the consumption of rated output and energy after the speed/acceleration data of known vehicle based on formula 6/7.

2. connect (E), the relation of energy ezpenditure and SOC:

SOC represents state-of-charge or charged state, equals electricity when surplus electricity in battery is full of electricity divided by battery.

3. connect (D), the relation of SOC and voltage:

Battery open-loop voltage comprises the voltage of cell output voltage and internal battery impedance consumption.

Battery open-loop voltage very major part depends on the SOC with battery, and open-loop voltage can be calculated by formula 18:

U _ocv′＝-1.031e ^-35×SOC+3.658+0.2156×SOC

-0.1178×SOC ²+0.321×SOC ³(18)

Cell output voltage can represent with formula 19, and monocell operating voltage is under that loading condition:

U _op′＝U _ocv′-I′Z _eq′(19)

Wherein, U _op' represent monocell operating voltage under that loading condition (unit is volt), U _ocvthe open-loop voltage (unit is volt) of ' expression monocell, the electric current (during electric discharge I ' > 0) of I ' expression monocell, Z _eq' represent that the internal resistance of cell resists.

So, electric battery output voltage can represent with formula 9:

U _op＝U _ocv-IZ _eq(9)

Wherein, U _oprepresent electric battery operating voltage under that loading condition (unit is volt), U _ocvrepresent the open-loop voltage (unit is volt) of electric battery, I represents the electric current (during electric discharge I ' > 0) of electric battery, Z _eqrepresent electric battery internal impedance.

In formula 19, the impedance of battery equivalent inner is by a resistance in series (R _series), two RC network (comprise R _{transient_s}, C _{transient_s}, R _{transient_L}and C _{transient_L}) composition, these values depend on the SOC of battery and by calculating by testing the experimental formula drawn, can refer to formula 8:

$\left\{\begin{array}{c}{R}_{\mathrm{series}}=0.1562e^{-24.37\mathrm{SOC}}+0.07446\\ R_{\mathrm{Transient}\_s}=0.3208e^{-29.14\mathrm{SOC}}+0.0469\\ C_{\mathrm{Transient}\_s}=-752.9e^{-13.51\mathrm{SOC}}+703.6\\ R_{\mathrm{Transient}\_L}=6.603e^{-155.2\mathrm{SOC}}+0.04984\\ C_{\mathrm{Transient}\_L}=-6056e^{-27.12\mathrm{SOC}}+4475\end{array}\right.---\left(8\right)$

From the known Z of formula 8 _eq' depend on the SOC of battery.In addition, known from the following description I ' is determined by the power of battery and operating voltage.To sum up, U _op' namely monocell operating voltage under that loading condition rely on the SOC of battery.

4. connect (F), the relation of energy ezpenditure, voltage and electric current:

From following formula 9, we can find that the electric current of battery is determined by the power of battery and operating voltage:

p′＝U _op′I′(9)

Wherein, the output power of parameter p ' expression monocell, U _op' represent monocell operating voltage under that loading condition, the electric current (during electric discharge I ' > 0) of I ' expression monocell.

If we suppose that the cell number of connecting in electric battery is n, so current value can be obtained by formula 10:

P＝U _op·I＝n·U _op′·I(10)

Wherein, P represents the output power of electric battery, U _oprepresent electric battery operating voltage under that loading condition, I represents the electric current (during electric discharge I > 0) of electric battery, and n represents series-connected cell number in electric battery, U ' _oprepresent monocell operating voltage under that loading condition.

5. connect (C), the relation of the heat of voltage, electric current and generation:

The heat that battery produces depends on the discharge process of battery, sees formula 5:

${{\stackrel{\·}{q}}_{\mathrm{gen}}}^{\′}=I^{\′}({U_{\mathrm{ocv}}}^{\′}-{U_{\mathrm{op}}}^{\′})-I^{\′}T\frac{d{U_{\mathrm{ocv}}}^{\′}}{\mathrm{dT}}---\left(5\right)$

Wherein, represent the heat (unit is Kg/m3) that monocell produces, electric current (during electric discharge I ' > 0) (unit is A) of I ' expression monocell, U _ocvthe open-loop voltage (unit is volt) of ' expression monocell, U _op' monocell operating voltage under that loading condition (unit is volt), T represents the working temperature (unit be DEG C) of battery.

Therefore we can obtain the heat production rate of electric battery when the open-loop voltage interconversion rate of electric battery is difficult to obtain, the differential part of formula 5 can be left in the basket.Like this, the electric battery heat production rate after simplification can obtain from formula 11:

$\stackrel{\·}{q}=I(U_{\mathrm{ocv}}-U_{\mathrm{op}})={\mathrm{IU}}_{\mathrm{ocv}}-P---\left(11\right)$

Wherein, represent the heat production rate of electric battery, I represents the electric current (during electric discharge I > 0) of electric battery, U _ocvrepresent the open-loop voltage of electric battery, U _oprepresent electric battery operating voltage under that loading condition.

6. connect (I), the relation between electric current and discharge rate:

The ratio of size of current when discharge rate C is used for representing battery discharge, i.e. multiplying power.As the battery of 1200mAh, 0.2C represents that discharge current is 240mA (0.2 multiplying power of 1200mAh), and 1C represents 1200mA (1 multiplying power of 1200mAh).

7. connect (G), the relation of the heat that battery produces and battery operating temperature:

The lithium battery of high-power output can produce a large amount of heat, and this can cause the violent rising of battery temperature, thus works the mischief to battery, and therefore battery comprises thermal management module.The effect of thermal management provides a best medial temperature and equally distributed temperature to electric battery.This module in electric motor car also can make the temperature of battery remain in an optimal average, thus can not shorten battery life.

Pesaran (Pesaran, A.A., BATTERYTHERMALMANAGEMENTINEVANDHEVS:ISSUESANDSOLUTIONS.B ATTERYMAN, 2001.43 (5): the transient temperature that p.34-49.) have studied the heat constantly produced in a module rises, he employs air cooled module, and experimental result shows the impact that the heat, discharge rate and the thermal capacity that produce rise on temperature.The heat of continuous generation, discharge rate or little thermal capacitance battery can produce larger temperature and raise.These explain the connection (G) from square frame " thermal capacitance " to " working temperature ", and from square frame " discharge rate " and " heat of generation " to the connection (G) of square frame " working temperature ".

Therefore we can obtain between these correlative factors from the data of Pesaran linear relationship, and namely inventor sums up formula 12:

$T=T_{\mathrm{initial}}+\mathrm{\ΔT}=T_{\mathrm{initial}}+t_a+t_t\·t^{\′}+t_R\·C+t_H\frac{\mathrm{\Σ}\stackrel{\·}{q}\mathrm{\Δt}}{t^{\′}\·N}+t_C\·C_p$

(12)

$=T_{\mathrm{initial}}+0.0339+0.004t^{\′}+0.001\·C+2.385\frac{\mathrm{\Σ}\stackrel{\·}{q}\mathrm{\Δt}}{t^{\′}\·N}-0.004\·C_p$

Wherein, T represents the working temperature of battery, T _initialrepresent the initial temperature of cell operations, environment temperature can be reduced to, the temperature raised when Δ T represents that electric battery works, t ' expression battery one-time continuous the working time, C represents discharge rate of battery, Δ t represents the time interval of pertinent instruments record driving data, and N represents the cell number in an electric battery, C _prepresent the thermal capacitance of battery.

By the experimental result that Pesaran provides, we have estimated regression coefficient t _a, t _t, t _r, t _hand t _cvalue, they are respectively 0.0339,0.004,0.001,2.385 and-0.004.The R square value adjusted in the linear regression analyses of SPSS is 0.842.

Due to Pesaran only application of air refrigerating module test, we suppose that different refrigerating module can produce different heat fades.We increase the heat fade difference that a Δ heat is used for representing that dissimilar refrigerating module produces relative to air cooled module in formula 12, see formula 13.The unit of Δ heat is W/cell.

$T=T_{\mathrm{initial}}+\mathrm{\ΔT}=T_{\mathrm{initial}}+t_a+t_t\·t^{\′}+t_R\·C+t_H(\frac{\mathrm{\Σ}\stackrel{\·}{q}\mathrm{\Δt}}{t^{\′}\·N}-\mathrm{\Δheat})+t_C\·C_p$

$=T_{\mathrm{initial}}+0.0339+0.004t^{\′}+0.001\·C+2.385(\frac{\mathrm{\Σ}\stackrel{\·}{q}\mathrm{\Δt}}{t^{\′}\·N}-\mathrm{\Δheat})-0.004\·C_p---\left(13\right)$

8. connect (J), the relation between the working temperature of battery and battery life:

Ramadass (Ramadass, P., etal., CapacityfadeofSony18650cellscycledatelevatedtemperatures:: PartI.Cyclingperformance.JournalofPowerSources, 2002.112 (2): p.606-613.) (Ramadass, P., etal., CapacityfadeofSony18650cellscycledatelevatedtemperatures:: PartII.Capacityfadeanalysis.JournalofPowerSources, p.614-620.) etc. 2002.112 (2): people did the complete capacity attenuation analysis of charge and discharge cycles when Sony18650 (1800mAh) lithium battery temperature raises 45-55 DEG C, after they illustrate 800 charging-discharging cycle under study for action, monocell can lose the initial capacity of 31% and 36% when 25 DEG C and 45 DEG C.Battery can lose the initial capacity more than 60% after 50 DEG C of discharge and recharges 600 times, can lose the initial capacity of 70% after 55 DEG C of discharge and recharges 500 times.

We may safely draw the conclusion, and longer elapsed time and the working temperature of Geng Gao can cause the more capacitance loss of battery.If we are full of with battery the life-span that the electric capacity after electricity represents it, so the life-span of battery can lose under higher discharge temp faster.This just can explain the connection (J) from square frame " working temperature " to " battery life ".

According to the research of the people such as Ramadass, we establish formula 14 and 15.

Capacity(％)＝C _b+C _T·T+C _N2N _cycle(14)

LT(％)＝C _b+C _TT+C _N2N _cycle(15)

Wherein, C _b, C _tand C _n2inderminate coefficient, Capacity (%) represent monocell be filled electricity after energy and battery be full of the ratio of the energy after electricity when just appearing on the scene, T represents the working temperature of battery, N _cyclerepresent charging-discharging cycle number, the ratio in life-span when LT (%) represents that battery life and firm battery just dispatch from the factory.

The data provided according to Ramadass we have estimated C _b, C _tand C _n2value, and use tf in formula 16 and 17 below _chargingreplace N _cycle, obtain formula 16 and 17.In the linear regression analyses of SPSS, R square of adjustment is 0.805, P value is 0.000.

capacity(％)＝167.583-1.264T-0.097·t·f(16)

LT(％)＝167.583-1.264T-0.097·t·f(17)

9. connect (A), the relation of energy ezpenditure and discharge and recharge frequency:

Battery electric quantity will be its charging when reducing to a certain predetermined threshold value (relevant with the personality of driver, different drivers has different threshold values).Because we can calculate the energy ezpenditure in traveling based on the speed of driver in driving procedure and acceleration information, the dump energy in battery after travelling every day also just can be obtained.Be less than predetermined threshold value once dump energy, driver can be chosen as battery charging.Like this, the discharge and recharge frequency in a special time period can just be obtained.

Can calculate charging-discharging cycle number by elapsed time length and discharge and recharge frequency, namely time span is multiplied by discharge and recharge frequency and equals charging-discharging cycle number.Wherein, elapsed time is elapsed time length battery calculates battery cycle number during from dispatching from the factory to.

10. connect (B), the relation of discharge and recharge frequency and battery life:

Takeno (Takeno, K., etal., p.298-305.) etc. Influenceofcyclecapacitydeteriorationandstoragecapacityd eteriorationonLi-ionbatteriesusedinmobilephones.Journalo fPowerSources, 2005.142 (1-2): people studied the circulating energy decline of lithium battery in mobile phone and stored energy fails.They find to reduce when the charging-discharging cycle length of battery, and the slip of the energy content of battery increases.That is pass through the same time, the decline of the energy content of battery can grow proportionately along with charging-discharging cycle number.See formula 1:

capacity(％)＝C _a+C _f·f+C _N1·N _cycle(1)

Wherein, capacity represent monocell be filled electricity after energy and battery be full of the ratio (%) of the energy after electricity when just appearing on the scene, f represents discharge and recharge frequency, N _cyclerepresent charging-discharging cycle number.

By the experimental data of Takeno, we have estimated parameter C in formula 1 _a, C _f, C _n1value.In the linear regression analyses of SPSS, R after adjustment ²value be 0.920, P value be 0.000.Wherein, R ²for coefficient of determination or the coefficient of determination, be defined as the business of regression sum of square and total sum of squares, represent that regressor is to the quality of the fitting degree of dependent variable.P value is exactly the probability that sample view result when null hypothesis obtains for true time or more extreme result occur, P value is less, shows that result is more remarkable.P < 0.05, regression model has statistical significance.Because this linear model can well describe the energy content of battery, relation between discharge and recharge time and charging-discharging cycle number, this linear model is applied on battery of electric vehicle by we.

If we replace N with tf _cycle, assuming that the electricity after battery charging completely can represent the life-span of remaining battery, formula 1 just can be rewritten as formula 2:

capacity(％)＝C _a+C _f·f+C _N1·N _cycle(2)

＝84.1+3.71·f-0.0788·t·f

Wherein, the ratio (%) in life-span when capacity (%) represents that battery life and firm battery just dispatch from the factory, f represents discharge and recharge frequency, and t represents elapsed time (being elapsed time length battery calculates battery cycle number during from dispatching from the factory to).

11. connect (K) (not shown), electric current has been charge/discharge protection, voltage be charge/discharge protection, SOC to battery life, overcharge/discharge prevention is to battery life:

Along with popularizing of electric motor car, avoid the charge and discharge SOC enough with ensureing battery that cross of battery to become the principal element ensureing that battery effectively uses, the SOC of battery estimates before extending battery life and user charge again, show its vital role in available power.Mills (Mills, A.andS.Al-Hallaj, Simulationofpassivethermalmanagementsystemforlithium-ion batterypacks.JournalofPowerSources, p.307-315.) etc. 2005.141 (2): what people overcharged technical research lithium battery by " soft " crosses charge/discharge.They find after the overcharge reaching 150% capacity, and lithium battery can produce violent, irreversible change.Research display battery operated overcharging/discharge scenario under, the heat of generation and heat radiation unequal time can there is thermal runaway and exceed charge/discharge rate and can cause battery thermal runaway produce a large amount of heat within the short reaction time.

12. battery life predicting theoretical models:

Can find from figure, be discharge and recharge frequency, working temperature and elapsed time to the direct acting factor of battery life.The initial temperature of battery, open-loop voltage, operating voltage, working current, thermal management module and thermal capacitance can affect the working temperature of battery.And open-loop voltage, operating voltage and working current again affect by SOC.

To sum up, the electricity of battery can describe with the formula 3 that aggregative formula 2 and formula 16 obtain, and wherein T can be obtained by formula 13.The dump energy of battery can describe with following formula 3.

For formula 3, when charge frequency be f, battery operating temperature be T, after elapsed time t, can calculate battery be filled electricity after energy and battery be full of the ratio of the energy after electricity when just dispatching from the factory.We think when battery be full of the electricity after electricity tested (i.e. experiment in subject) one day cannot be provided to drive institute's energy requirement time, battery life finishes; The ratio of the energy after being full of electricity when the energy content of battery and battery just dispatch from the factory needed for certain driver drives for a day will become critical point, and can carry out other parameters of reverse as known quantity.So, if known charge frequency be f, battery operating temperature is T, certain driver drives the ratio C apacity (%) of the energy after being full of electricity when the required energy content of battery and battery just dispatch from the factory for one day, can obtain battery life t.

$\left\{\begin{array}{c}\mathrm{Capacity}(\%)=\frac{1}{100}\&CenterDot;(167.583-1.264T-0.097\&CenterDot;t\&CenterDot;f)\&CenterDot;(84.1+3.71\&CenterDot;f-0.0788\&CenterDot;t\&CenterDot;f)\\ T=T_{\mathrm{initial}}+0.0339+0.004t^{\&prime;}+0.001\&CenterDot;C+2.385(\frac{\mathrm{\&Sigma;}\stackrel{\&CenterDot;}{q}\mathrm{\&Delta;t}}{t^{\&prime;}\&CenterDot;N}-\mathrm{\&Delta;heat})-0.004\&CenterDot;C_p\end{array}\right.---\left(3\right)$

In formula 3, T represents the working temperature of battery, and t represents elapsed time (being elapsed time length battery calculates battery cycle number during from dispatching from the factory to), and f represents discharge and recharge frequency, T _initialrepresent the initial temperature (being reduced to environment temperature) of cell operations, the t ' expression battery one-time continuous working time, C represents discharge rate of battery, represent the heat production rate of electric battery, Δ t represents the time interval of pertinent instruments record driving data, and N represents the cell number in an electric battery, and Δ heat represents relative atmospheric refrigeration module, the heat fade difference of the refrigerating module generation of other types, C _prepresent the thermal capacitance of battery.

So far, theoretical model (i.e. formula 3) is just set up and is completed.

Derive shown in above-mentioned 1-11 step all for lithium battery, summarize the objective law of each parameter in the serviceable life about lithium battery, therefore there is versatility, so reflect the objective law in lithium battery serviceable life according to the formula 3 of above-mentioned 1-11 step institute theory deduction, the application for lithium battery has universality.In the application, this formula is applied in lithium battery motor-car.

In an experiment, drive speed, acceleration, driving time can be obtained.By these values, each tested energy ezpenditure to battery of electric vehicle under the condition of set highway and urban road distance can be obtained according to formula 6,7.Like this, we just can obtain each tested average power (unit: W/mile) of driving on highway and urban road.The tested true driving distance answering every day oneself in ensuing questionnaire, comprises highway distance and urban road distance.By the driving average power calculated before, can calculate and drive the energy content of battery needing to consume each tested every day.After driving terminates, allly testedly also to answer " battery electric quantity how many lower than percent after, you can charge for battery ".By the answer of this problem, we consume how many energy before can calculating each tested charging each time, and this energy value can draw each tested discharge and recharge frequency f divided by the energy consumed tested every day.We hypothesis, when battery be full of the electricity after electricity cannot provide tested one day drive institute's energy requirement time, the end-of-life of this battery.Within tested one day, drive required energy value and be capaicity (%) in formula 3 divided by the battery energy value after being full of electricity that just dispatched from the factory, assuming that this ratio can represent that remaining battery can the ratio of life-span and battery battery life when just dispatching from the factory.Like this, as cells known configuration C, N, Δ heat, C _p, environment temperature T _initial, each driving time t ', driving simulator real time record driving data continuously the time interval Δ t time, we obtain battery life t by formula 3 or 4.

set up people-electric motor car empirical model

analyze empirical model:

The impact of personality on leg muscle (controlling brake and throttle) speeds control of people is obtained for good foundation in research different before, in multiple personality storehouse, the relation between muscle speed and personality is basically identical from E and the N gage in Exxon people's case system in different behavior normal form.Bachorowski and Newman (Bachorowski, J.A.andJ.P.Newman, Impulsivemotorbehavior:Effectsofpersonalityandgoalsalien ce.JournalofPersonalityandSocialPsychology, 1990.58 (3): p.512-518.) once reported at one, people's (having the feature of E+ and N+) of impulsion requires that (all mark is lower on two yardsticks) people than non-impulsion in the task of slow controlled motion has muscle movement speed faster.And when the result of behavior is demonstrated out, this phenomenon is more obvious.At Wu (C.Wu.andZhao, G, MathematicalModelingofAverageDriverSpeedControlandIndivi dualDriverDifferenceswiththeIntegrationofQueuingNetwork-ModelHumanProcessorandRule-BasedDecisionFieldTheory (Part1.ModelDevelopment) .unpublished) people such as (2010) research in, they use a character feature variable to embody the impulsion degree of people, do not have deep thinking familiar route in other words with regard to the tendency of prompt action.That is the personality (impulsive style, normal type, non-impulsive style) of driver can affect the driving behavior of driver.

In experiment, allly testedly all to complete a series of questionnaire.First questionnaire is used to obtain tested demography basic condition (such as age, sex etc.) and drives experience (such as accumulation driving range roughly, the time etc. of acquisition driving license).Then they will set up a subjective value matrix.In matrix, the value of each element represents each drive speed to tested value.What present together with selecting with each speed is the monetary cost receiving penalty note, and if do not receive the safety and temporal income that penalty note obtains.Finally tested be divided into three classes according to Exxon personality (EPQR-S) after revision by all: normal (test result is E+ and N-, or E-and N+, amount to 6 people), impulsive style (test result is E+ and N+, amounts to 3 people), (test result is E-and N-to non-impulsive style, amount to three people) (note: Exxon Personality test questionnaire uses more than 35 years in the whole world, it is the standardised questionnaire/scale measuring mankind's personality, the personality that it measures the mankind all has higher reliability (stability successively surveyed is high) and validity (effectively reflection different characters type), result measured by it can obtain the confirmation of kinds of experiments psychological study simultaneously, as far back as existing Chinese revised version in 1981, refer to: HansJ ü rgenEysenck & SybilB.G.Eysenck (1975) .ManualoftheEysenckPersonalityQuestionnaire.London:Hodde randStoughton).

In real life, the self-determining driving speed of driver is often not equal to the speed limit of this road.Therefore, in driving experiment, driver needs to select oneself to want the speeds with higher than speed limit how many miles per hours, and what present to option is corresponding " if receiving the monetary cost of penalty note " and " safety that penalty note obtains if do not receive and temporal income " simultaneously.Driver determines the drive speed (such as 100 miles per hours) of oneself automobile determined when an appearance speed(-)limit sign by considering these two factors, the difference of the self-determining drive speed of this driver and speed limit (such as 85 miles per hours) is " decision references value " (DMR), and unit is miles per hour (is convertible into kilometer/hour).In addition, driver needs to provide the life schedule of individual in an experiment, namely working day, weekend driving time on highway and city and distance.That is the decision references value of driver also can affect the driving behavior of driver.

Except driving behavior affects battery life, charging strategy, cell arrangement and life schedule will affect the life-span of battery.Suppose that driver's working day, the driving distance on highway and urban road was D _hdand D _ud, weekend, the driving distance on highway and urban road was D _heand D _ue, in sum, following empirical model can be proposed:

Lifetime＝pa×Personality+pb×DMR+pc×Charging+pd×C _p+pd×Δheat

+pf×N+ph×T _initial+pi×D _hd+pj×D _ud+pk×D _he+pl×D _ue

Wherein, personality represents personality, and DMR represents decision references value, and Charging represents charging strategy, C _prepresent battery thermal capacitance, Δ heat represents heat fade difference, and N represents monocell number in electric battery, T _initialrepresent environment temperature, D _hdand D _udrepresent the driving distance of driver's working day on highway and urban road, D _heand D _uerepresent the driving distance of driver's weekend on highway and urban road, pa ~ pl represents coefficient.

experimental situation and experimentation:

Use in experiment driving simulator (STISIMDRIVEM100K, SystemsTechnologyInc, Hawthorne, CA).It comprises the Logitech of a force feedback bearing circle, a gas pedal and a brake pedal.The rest position of gas pedal is 38.2 ° (angles on pedal surface and ground), and maximum throttle input angle is 15.2 °.The rest position of brake pedal is 60.1 °, and maximum brake input angle is 28.6 °.In addition, Driving Scene is presented in the LCD display of 27 inches of 1920 × 1200 pixel resolutions.Driving simulator automatically can gather behavioral data in experimentation, and these data have: the time (second), speed (feet per second), acceleration (feet per second 2) and distance (foot).These data can be used for calculating the output power of battery and SOC.

Experiment scene is a track (each direction) environment of a simulation, comprises urban road and highway, does not wherein have other vehicles, pedestrian or road mark (such as stopping mark).It is reported that the average one way commuting time of the U.S. is 26 minutes (mean distance is 16 miles).Therefore subject needs to drive 16 miles in an experiment.Report that gasoline consumption in the U.S. 55% is on urban road in addition, 45% on a highway.In order to simulate general case of travelling frequently, Episode sequences is followed successively by 30% urban road, 45% highway, is finally 25% urban road.Speed limit is urban road 30mph, highway 45mph, and speed limit appears in the position of driver's 200 yds ahead.Subject is required to regulate the speed when seeing speed(-)limit sign with true driving.In addition, the scene density of whole experiment scene is consistent always.

In driving procedure, be consumed how many energy in order to calculating battery, we suppose that vehicle has the physical characteristics of Toyota Ka Meirui in 2008: quality 1588kg (3500lbs), and drag coefficient is 0.28, and vehicle front face area is 2.7m ².The dimensionless coefficient of rolling resistance of tire is 0.01, and this coefficient links up resistance as a normal direction force and motion.The energy transmitted to battery by regenerative braking is assumed to 40%, and the transfer efficiency from battery to bearing circle is 80%.Battery pack power is assumed to 16kWh (i.e. the Volt of Chevrolet company production).Atmospheric density is taken from U.S.'s sea-level standard atmospheric density.We load the constant load of a 800W in electric motor car, are used for representing the Activities irrelevant with vehicle movement, such as warm braw, air-conditioning, illumination and other annexes.When the energy of acceleration needed for vehicle movement when negative value and nonnegative value can describe with above-mentioned formula 6 and 7.

After this, subject reduces to how many times by answering the electricity of electric motor car they is bound to as battery charging.Because we can calculate each tested energy ezpenditure in commuting course based on the speed of subject in driving procedure and acceleration information, the dump energy after travelling frequently every day in battery also just can be acquired.Some dump energies are less than the charge value that subject is bound to charge, and subject is bound to be chosen as battery charging.

Each is tested all will provide their life schedule, comprise workaday Commuting Distance and time, working day and weekend the driving-activity except travelling frequently, and monthly and annual life schedule, such as whether can specific date motoring.The driving data that our collection detailed as far as possible is in an experiment tested, sets up their annual driving data framework roughly with this.

According to the energy ezpenditure calculated in case study, and the charging strategy of subject and life schedule, we can obtain the battery life of simulation.

determine experimental formula:

(1) proof theory model:

In one embodiment of the invention, in conjunction with above-mentioned experimental situation, calculate battery life t according to formula 3, the battery life of simulation is shown in Table I.

Because T _initialcan continue change with Δ Heat, the working temperature of battery is difficult to calculate, and we simply suppose that the working temperature of battery can be remained on constant level by the thermal management module of battery.T _initialbe assumed that 30 DEG C, C _p=1000.4J/kg/K, Δ heat=1.33W/cell, N=200.In an experiment, suppose that turnpike driving distance is 7.2 miles, urban road driving distance is 8.8 miles, and every tested driving distance is all identical.Suppose if battery be full of electricity after electricity can not maintain the energy ezpenditure of travelling frequently for a day, this battery just must be replaced, i.e. end-of-life.We obtain the value of Capacity (%) in formula 3 divided by the battery electricity dispatched from the factory when being full of electricity with the consume battery power E altogether of travelling frequently for a day.Cell arrangement and temperature are substituted into formula 3 and can calculate battery life t.In analytic process, nature factor is quantized, and specifically, represents non-impulsive style successively with-1,0 and 1, normal and impulsive style.

Table I .T _initialfeature tested when=10 DEG C and battery life

The experimental result of battery life is calculated by formula 3.The battery life that we compare simulation and the battery life of the expectation of 96 months provided by General Electric Co. Limited, the level of signifiance value that the experimental result of battery life obtains in independent sample T detects is 0.184 (> 0.05), and the battery life of therefore experimental result and General Electric Co. Limited's expectation is without remarkable difference.

(2) regretional analysis, empirical model is determined:

Then, every tested different life schedule takes in by we, supposes that in experiment, tested driving distance on highway and urban road is respectively d _hand d _u, the energy content of battery that highway and urban road consume is respectively e _hand e _u, have again tested providing in questionnaire to be respectively D at the driving distance on highway and urban road every day _hand D _u, the energy content of battery E that so tested actual every day consumes on highway and urban road _hand E _ube respectively

$\left\{\begin{array}{c}{E}_h=\frac{D_h}{d_h}\&CenterDot;e_h\\ E_u=\frac{D_u}{d_u}\&CenterDot;e_u\end{array}\right.---\left(20\right)$

Then there is the total electricity of actual consumption every day:

$E=E_h+E_u=\frac{D_h}{d_h}\&CenterDot;e_h+\frac{D_u}{d_u}\&CenterDot;e_u---\left(21\right)$

Suppose if battery be full of electricity after electricity can not maintain the energy ezpenditure of travelling frequently for a day, this battery just must be replaced, i.e. end-of-life.By formula 21, obtain the value of Capacity (%) in formula 3 divided by the battery electricity dispatched from the factory when being full of electricity with E.By cell arrangement C _p=707J/kg/K, Δ heat=1.33W/cell, N=200, environment temperature T _initial=10 DEG C substitute into formula 3, can calculate the value of variable t in formula 3, i.e. life-span of battery of electric vehicle in tested real life.Tested feature, life schedule and battery life the results are shown in Table II.

Table II .T _initial=10 DEG C, battery life result of calculation when considering tested life schedule

The result of calculation of most of tested battery life expects battery life higher than GM, before reason is in experiment, Commuting Distance every day (highway and urban road) is all according to U.S.'s Commuting Distance 16 miles (highway 7.2 miles, urban road 8.8 miles) setting per capita.Visible in table ii, most of tested Commuting Distance is much smaller than 16 miles, and this makes charge frequency greatly reduce, and this is that battery life GM expects the reason of battery life higher than GM.

Cell arrangement, the temperature parameter of conversion Table II, carried out 16 groups, 192 times altogether to battery life and calculated, obtained Table III.Wherein, C _p=707J/kg/K or 1019J/kg/K, Δ heat=1.33W/cell or 4.45W/cell, N=200 or 228, T _initial=10 DEG C or 30 DEG C, by software SPSS, (Regression Analysis Result is shown in Table III in linear regression analysis, in Table III, * or * * represent that this variable on battery life impact significantly) result show the factor that battery life is made a significant impact, we can set up the linear equation of battery life about these factors by this analysis result, see formula 8.

Lifetime＝393.768-35.1552Personality-2.85Charging-1.382T _initial(8)

-9.716D _hd-2.592D _ud-4.361D _he-11.533D _ue

Wherein parameter is personality Personality, charging strategy Charging, environment temperature T _initial, working day highway and urban road on driving distance be D _hdand D _ud, weekend highway and urban road on driving distance be D _heand D _ue.Wherein, in Table III, nonstandardized technique coefficient is the coefficient of each variable in formula 8.

The coefficient of formula 8 is according to known tested feature (personality, decision references value, charging strategy, life schedule), and cell arrangement (thermal capacitance, monocell quantity, initial temperature, heat fade difference) and the battery life matching that got by formula 3.Because in formula 3, q derives from tested driving behavior (speed, acceleration), unpredictable in the research that these driving behaviors are former, the research thus cannot predict battery life when not driving electric motor car.Here, we, by regretional analysis, directly can obtain battery life by tested feature and environment temperature.That is, had formula 8, we do not need tested driving electric motor car just can obtain the life-span of battery of electric vehicle, can reduce the cost of prediction battery life greatly.

In the linear regression analyses of SPSS, the R after adjustment ²value is coefficient of determination or the coefficient of determination, represents that regressor is to the quality of the fitting degree of dependent variable, R herein ²value is 0.8, P value is 0.000, and that is formula 8 can well the battery life that calculates of fitting formula 3, and can predict battery life.Personality, charging strategy, environment temperature, working day and weekend, the saliency value of the driving distance on highway and urban road was all less than 0.05, and this just illustrates that their coefficient is statistically significantly non-zero, and the change of each amount changes battery life really.

Because computation process have employed the physical characteristics of Toyota Ka Meirui in 2008, but wherein some physical characteristics such as m, ρ can change along with the change of driver and external environment.When formula 8 is applied in other vehicles, these physical characteristicss also can change to some extent, and now we tackle coefficient adjustment.Therefore, here we provide each coefficient range, when the physical characteristics making formula 8 can be applied to Toyota Ka Meirui widely changes to some extent or when formula 8 is applied to other vehicles.In Table III, nonstandardized technique coefficient (B) ± factor standard difference (Std.Deviation) income value is the scope of this coefficient, and we can obtain the scope of battery life thus.So empirical model (formula 8) can be written as further:

Lifetime＝pa+pb×Personality+pc×Charging+pd×T _initial

+pe×D _hd+pf×D _ud+pg×D _he+ph×D _ue

Statistically, in normal distribution, the sample of 99.7% drops in the scope of " mean value ± 3 times of standard deviations ", therefore according to Table III, the coefficient range of formula 8 be " nonstandardized technique coefficient (B) ± 3 × factor standard difference (Std.Deviation) " wherein, pa scope is between 393.7864 ± 147.2502, pb scope is between-35.1552 ± 21.9882, pc scope is between-2.85 ± 2.145, pd scope is between-1.3822 ± 1.1946, pe scope is between-9.716 ± 2.7264, pf scope is between-2.5916 ± 3.2112, pg scope is between-4.3606 ± 2.7264, ph scope is between-11.533 ± 4.9872.

electric vehicle lithium battery life-span prediction method

In one embodiment of the invention, a kind of electric vehicle lithium battery life-span prediction method is provided, comprises:

Step one, given personal character parameter, charging strategy, initial temperature, life schedule;

Step 2, according to formula 8, calculate battery life.

Table III. battery life Regression Analysis Result

	Life-span (moon)
		Personality
Nonstandardized technique coefficient (B)	-35.1552**
		Normalisation coefft (Beta)	-.294
Factor standard difference (Std.Deviation)	7.3294
		Decision references value (DMR) (mph)
Nonstandardized technique coefficient (B)	3.3056
		Normalisation coefft (Beta)	.104
Factor standard difference (Std.Deviation)	3.1682
		Charging strategy (%)
Nonstandardized technique coefficient (B)	-2.85**
		Normalisation coefft (Beta)	-.380
Factor standard difference (Std.Deviation)	0.715
		Thermal capacitance (J/kg/K)
Nonstandardized technique coefficient (B)	-0.0142
		Normalisation coefft (Beta)	-.022
Factor standard difference (Std.Deviation)	0.0256
		Heat fade (W/cell)

Nonstandardized technique coefficient (B)	-0.268
		Normalisation coefft (Beta)	-.004
Factor standard difference (Std.Deviation)	2.553
		Number of batteries
Nonstandardized technique coefficient (B)	-0.1484
		Normalisation coefft (Beta)	-.034
Factor standard difference (Std.Deviation)	0.181
		Initial temperature (DEG C)
Nonstandardized technique coefficient (B)	-1.3822**
		Normalisation coefft (Beta)	-.140
Factor standard difference (Std.Deviation)	0.3982
		Turnpike driving distance on working day (mile)
Nonstandardized technique coefficient (B)	-9.716**
		Nonstandardized technique coefficient (Beta)	-.726
Factor standard difference (Std.Deviation)	0.7804
		Urban road driving distance on working day (mile)
Nonstandardized technique coefficient (B)	-2.5916*
		Normalisation coefft (Beta)	-.200
Factor standard difference (Std.Deviation)	1.0704
		Turnpike driving distance at weekend (mile)
Nonstandardized technique coefficient (B)	-4.3606**
		Normalisation coefft (Beta)	-.243
Factor standard difference (Std.Deviation)	0.9088
		Urban road driving distance at weekend (mile)
Nonstandardized technique coefficient (B)	-11.533**
		Normalisation coefft (Beta)	-.437
Factor standard difference (Std.Deviation)	1.6624
		Constant
Nonstandardized technique coefficient (B)	393.7864**
		Factor standard difference (Std.Deviation)	49.0834
The R square of adjustment	0.800

The level of signifiance (P value)	0.000
		Standard error is estimated	54.376
Sample number	192

*p＜.01；**p＜.001.

according to formula 3, given driver's feature and the optimum cell arrangement of battery target life prediction

According to above-mentioned analysis, in one embodiment of the present of invention, provide a kind of method, under given driver's feature and the condition in battery target life-span, obtain the allocation optimum of battery.

If first suppose battery be full of electricity after electricity can not maintain automobile and to travel frequently in one day required energy ezpenditure, this battery just must be replaced.From case study, we can calculate each driver and drive the energy content of battery E consumed every day, therefore obtain the value of Capacity (%) in formula 3.And then " fminsearch " order in Matlab software can be utilized to input formula 3, when known driver's feature and decision references value DMR and charging strategy Charging, obtain the allocation optimum of battery, namely arrive C _p, the value of Δ heat and N, makes the Capacity in formula 3 (%) value just to have dispatched from the factory close to the energy ezpenditure of middle battery of travelling frequently given tested every day and battery as much as possible the ratio of energy.In other words, the approximate zero point of battery allocation optimum and solution formula 22 when solving given driver's feature and battery target life-span.

$\left\{\begin{array}{c}\frac{1}{100}\&CenterDot;(167.583-1.264T-0.097\&CenterDot;t\&CenterDot;f)\&CenterDot;(84.1+3.71\&CenterDot;f-0.0788\&CenterDot;t\&CenterDot;f)*\mathrm{CAPACITY}-E=0\\ T=T_{\mathrm{initial}}+0.0339+0.004t^{\&prime;}+0.001\&CenterDot;C+2.385(\frac{\mathrm{\&Sigma;}\stackrel{\&CenterDot;}{q}\mathrm{\&Delta;t}}{t^{\&prime;}\&CenterDot;N}-\mathrm{\&Delta;heat})-0.004\&CenterDot;C_p\end{array}\right.---\left(22\right)$

Wherein CAPACITY is the energy after electric battery is full of electricity when just dispatching from the factory, and E is total power consumption in the driving procedure of a day, and t is the target life objective of battery.

When the value of T in computing formula 22, T _initialdifferent constants can be set to according to Various Seasonal, or be simply set as a constant constant.The value of Δ t is determined by the simulator adopted in case study.

For each tested and given 96 months battery target life-spans, we predict environment temperature T _initialthe allocation optimum of battery when being 10 DEG C, the results are shown in Table IV.

Table IV .T _initial=10 DEG C, optimum cell arrangement when given driver's feature and battery target life-span

We can find out to allow electric life close to target life objective in table iv, and the simplest method is exactly the number of batteries changed in battery module, and incident is exactly the change of old of cell module body sum.Monocell is of a size of 5 inches of x7 inches, and thickness is less than 1/4th inches, and the volume change of electric battery can obtain thus.The pyroconductivity of oil is 1.5 to 3 times of air.The water pyroconductivity be used in when indirectly cooling is 15 times of air.Visible dissimilar thermal management module will produce distinct heat fade, and therefore thermal management module is also the method for an extending battery life.The thermal management module but developing a convenience and high-efficiency needs certain fund input equally.Typical is 795J/kg/K from battery thermal capacitance, and this is determined by the chemistry of battery itself and physical material.The method of last extending battery life develops battery itself exactly.The researcher of a lot of battery material and aspect of performance wants to improve battery, and this needs a lot of fund inputs, and therefore this is also optimize the most difficult method of battery life.

Want, by changing cell arrangement extending battery life, only by the battery life value of battery target lifetime settings for originally providing higher than cell production companies, and the optimum cell arrangement making battery reach target life objective to be obtained.Use the allocation optimum of this Battery pack, original life that battery life can be provided by battery producer is to target life objective.

according to formula 3, the drivers ' behavior of given cell arrangement and battery target life prediction optimum (is namely driven time speed, acceleration, and charge frequency) and life schedule

If the electricity after same hypothesis battery is full of electricity can not maintain automobile and to travel frequently in one day required energy ezpenditure, this battery just must be replaced.From case study, we can calculate each driver and drive the energy content of battery E consumed every day, therefore obtain the value of Capacity (%) in formula 3.

And then " fminsearch " order in Matlab software can be utilized to input formula 3, at cells known configuration N, C _p, Δ heat, initial temperature T _initialtime, obtain optimum C, and f, the Capacity in formula 3 (%) value has just been dispatched from the factory close to the energy ezpenditure of middle battery of travelling frequently given tested every day and battery the ratio of energy as much as possible.Optimum C when in other words, solving given cell arrangement and battery target life-span, with the approximate zero point of f and solution formula 22.Optimum I can be obtained by optimum C, then obtain optimum P by formula 11.Optimum drive speed v can be obtained and the actual speed limit of acceleration a, v Zhi Jian highway is optimum DMR by formula 6,7.

For each tested and given 96 months battery target life-spans, we secure cell arrangement and make C _p=1000.4J/kg/K, Δ heat=43.2W/cell, N=200, and environment temperature T _initial=10 DEG C, predict the optimum driving behavior of driver and life schedule, the results are shown in Table IV.

Table V .T _initial=10 DEG C, optimum driving behavior and life schedule when given cell arrangement and target life objective

Optimizing decision reference value is 0 mean that driver should not make car speed outpace restriction, and optimum to drive acceleration be that 0m/s2 means that driver is in driving procedure, and the speed of a motor vehicle should be made after reaching desired speed to keep constant.The last driving distance of every day is urban road 5 miles, highway 4 miles.Last charge frequency is 0.2 times/day, and that is battery should fill once electricity for every 5 days.

Using optimum charge frequency, is reach battery target life-span the simplest method.Second method changes decision references value exactly, makes the drive speed of driver not higher than speed limit.Personality and life schedule are that two other affects the factor of battery life.Distance between the daily life of life schedule and driver and each destination is closely related, is difficult to change.Therefore we think given cell arrangement time, change driver charging strategy, decision references value and acceleration are three Main Means reaching target battery longevity rice function.

During current consideration driving distance, if driver's driving distance on highway or urban road is longer, they will reduce the driving distance on urban road Huo Gongsu highway.The increase of driving distance can cause the increase of charge frequency, thus reduces battery life.If driving distance reaches optimum, the increase of charge frequency also can reduce battery life.Otherwise battery life can extend.

Because life schedule is difficult to change, we are used as it is the input value the same with cell arrangement, battery target life-span.Next step, different life schedules is taken into account the optimum driving behavior and charging strategy of again predicting driver by us, the results are shown in Table VI.No matter how we can see life schedule from Table VI, the acceleration after optimizing decision reference value and target velocity reach is all 0.

Table VI .T _initial=10 DEG C, optimum driving and charging behavior when given cell arrangement, target life objective and life schedule

That is battery life will be made optimum, and driver should reduce as far as possible suddenly steps on the gas, and keep the constant of car speed, and drive speed is not higher than speed limit.General Motor corporation (GM) reports that the charging station of a 240V can be battery charging complete (domestic power supply of 120V then needs 10 hours) in about four hours.Because the quantity of charging station is little at present, people can only charge after the driving of every day terminates.Therefore the interval number of days charged in actual life is necessary for integer, and people should make average charge interval close to optimal value as far as possible.For example, if optimum charging interval is 2.5 days, the pattern of charging should be " ... 2 days-3 days-2 days-3 days ... " along with the development of quick charge station, people can be electric motor car charging near their work place, this just can make charging convenient, makes charge frequency more easily reach optimal value.

Wanting by changing drivers ' behavior and life schedule extending battery life, only by the battery life value of battery target lifetime settings for originally providing higher than cell production companies, and the optimum drivers ' behavior and the life schedule that make battery reach target life objective need be obtained.The drivers ' behavior using this group optimum and life schedule, original life that battery life can be provided by battery producer is to target life objective.

It should be noted that and understand, when not departing from the spirit and scope of the present invention required by accompanying claim, various amendment and improvement can be made to the present invention of foregoing detailed description.Therefore, the scope of claimed technical scheme is not by the restriction of given any specific exemplary teachings.

In sum, according to formula 8 (i.e. empirical model), when after the given target battery life-span, the parameters such as relevant battery allocation optimum, charging strategy, driving schedule can be obtained.Understand from another aspect, namely battery life can be made to reach desired value by optimization cell arrangement, driving behavior, charging strategy.

term parameter list

Claims (6)

1. an electric vehicle lithium battery life-span prediction method, comprising:

The working temperature of step one, acquisition battery, is full of the ratio of the energy after electricity when the energy content of battery and battery just dispatch from the factory needed for charge frequency and driver drive for one day; With

Step 2, according to following manner prediction battery life:

Wherein, T represents the working temperature of battery, and unit is DEG C; F represents charge frequency, and unit is 1/ day;

Suppose when battery be full of the electricity after electricity cannot provide car day travel institute's energy requirement time, 0 represents that battery life finishes, the ratio C apacity (%) of the energy after electricity is full of when the energy content of battery and battery just dispatch from the factory needed for then within one day, travelling according to described charge frequency f, battery operating temperature T, driver, obtain elapsed time t, namely this value represents the battery predictive life-span, and unit is sky.

2. method according to claim 1, wherein, the working temperature of battery calculates according to following manner:

Wherein, T _initialrepresent the initial temperature of cell operations, unit is DEG C; T' represents the battery one-time continuous working time, and unit is second; C represents discharge rate of battery; represent the heat production rate of electric battery, unit is watt; Δ t represents the time interval of pertinent instruments record driving data, and unit is second; N represents the cell number in an electric battery; Δ heat represents the heat fade difference that dissimilar refrigerating module relative atmospheric refrigeration module produces, and unit is the every single battery of watt; C _prepresent the thermal capacitance of battery, unit is J/kg/K.

3. an electric vehicle lithium battery life-span prediction method, comprising:

Step one, obtain the character parameter of driver, charging strategy, environment temperature, working day and weekend the driving distance on highway and urban road;

Step 2, according to following manner prediction battery life scope:

Lifetime＝pa+pb×Personality+pc×Charging+pd×T _initial

+pe×D _hd+pf×D _ud+pg×D _he+ph×D _ue

Wherein, Lifetime represents the battery predictive life-span, and unit is the moon, and Personality represents the character parameter of driver, and the quantification manner of the character parameter of driver is: represent non-impulsive style successively with-1,0 and 1, normal and impulsive style; Charging represents charging strategy, i.e. the minimum electricity of the front remaining battery of battery charging, and value is a number percent; T _initialrepresent environment temperature; D _hdand D _udrepresent working day highway and urban road on driving distance, D _heand D _uerepresent weekend highway and urban road on driving distance, unit is mile; Pa is constant, and pb ~ ph represents coefficient;

Wherein, pa scope is between 393.7864 ± 147.2502, pb scope is between-35.1552 ± 21.9882, pc scope is between-2.85 ± 2.145, pd scope is between-1.3822 ± 1.1946, and pe scope is between-9.716 ± 2.7264, and pf scope is between-2.5916 ± 3.2112, pg scope is between-4.3606 ± 2.7264, and ph scope is between-11.533 ± 4.9872.

4. method according to claim 3, wherein, the formula of step 2 is:

Lifetime＝393.768-35.1552Personality-2.85Charging-1.382T _initial

-9.716D _hd-2.592D _ud-4.361D _he-11.533D _ue。

5. an electric vehicle lithium battery life method, comprising:

Step one, the formula in method described in claim 2 to be out of shape as follows:

Wherein, CAPACITY is the energy after electric battery is full of electricity when just dispatching from the factory, and unit is J; E is total power consumption in the driving procedure of a day, and unit is J; t _mthe time be through, represent the target life objective of battery herein, unit is sky; T represents the working temperature of battery, and unit is DEG C; T _initialrepresent the initial temperature of cell operations, unit is DEG C; T' represents the battery one-time continuous working time, and unit is second; C represents discharge rate of battery; represent the heat production rate of electric battery, unit is watt; Δ t represents the time interval of pertinent instruments record driving data, and unit is second; N represents the cell number in an electric battery; Δ heat represents the heat fade difference that dissimilar refrigerating module relative atmospheric refrigeration module produces, and unit is the every single battery of watt; C _prepresent the thermal capacitance of battery, unit is J/kg/K;

Step 2, acquisition drive the energy content of battery E consumed, charge frequency f every day, and unit is 1/ day;

Step 3, by the above-mentioned formula of iterative, obtain battery allocation optimum C _p, the value of Δ heat and N.

6. an electric vehicle lithium battery life method, comprising:

Step one, the formula in method described in claim 2 to be out of shape as follows:

Wherein, CAPACITY is the energy after electric battery is full of electricity when just dispatching from the factory, and unit is J; E is total power consumption in the driving procedure of a day, and unit is J; t _mthe time be through, represent the target life objective of battery herein, unit is sky; T represents the working temperature of battery, and unit is DEG C; T _initialrepresent the initial temperature of cell operations, unit is DEG C; T' represents the battery one-time continuous working time, and unit is second; C represents discharge rate of battery; represent the heat production rate of electric battery, unit is watt; Δ t represents the time interval of pertinent instruments record driving data, and unit is second; N represents the cell number in an electric battery; Δ heat represents the heat fade difference that dissimilar refrigerating module relative atmospheric refrigeration module produces, and unit is the every single battery of watt; C _prepresent the thermal capacitance of battery, unit is J/kg/K; F represents charge frequency, and unit is 1/ day;

Step 2, acquisition drive the energy content of battery E consumed, cell arrangement N, C every day _p, Δ heat, initial temperature T _initial;

Step 3, by the above-mentioned formula of iterative, obtain optimum C, and f;

Step 4, obtain optimum I by optimum C, then obtain optimum P by following formula:

Wherein, P represents the output power of vehicle battery packs, and unit is W; I represents the electric current of electric battery, and unit is A; U _ocvrepresent the open-loop voltage of electric battery, U _oprepresent electric battery operating voltage under that loading condition, unit is volt;

Step 5, obtain optimum drive speed v and acceleration a by following formula:

Wherein, p represents electric vehicle motion power consumption, and unit is W; M represents quality, and unit is Kg; A represents vehicle acceleration, and unit is m/s ², ρ sea level air every cubic metre roughly air weight, unit is Kg/m ³; V represents speed, and unit is m/s; C _drepresent the resistance coefficient of vehicle; A represents vehicle front face area, and unit is m ², C _rrrepresent the dimensionless factor of tire drag, g represents acceleration of gravity, and unit is m/s ²,

According to the charge frequency f tried to achieve, speed v and the behavior of acceleration a direct drivers, thus extend the electric vehicle lithium battery life-span.