Title of Invention

A PROCEDURE TO DETERMINE THE CHARGE THAT IS DRAWN FROM AN ENERGY STORE DEVICE

Abstract

The present invention relates to a procedure to determine the charge that is drawn (Qe) from an energy store device, in particular a battery, up to a pre-determined end of"discharge, comprising the steps of: calculating the charge that is drawn (Qe) in the case of a predetermined discharge current pattern (IBatt, Entlade) from an energy store device with the help of a load predictor (2) on the basis of a mathematical model of an energy store device that demonstrates the electrical properties of the energy store device mathematically, and determining state variables (Z) and/or parameters (P) for the mathematical model of the energy store device from current operating values (UBatt, IBatt, T BatJ of the energy store device with the help of the state variables and parameter assessor (1), characterized in that the charge predictor (2) determines the charge that is drawn (Qe) until achieving a pre-determined end of discharge criterion.

Full Text

Description rocedure and Device to Determine the Charge that can be drawn from an Energy Store )evice "he invention is with regard to a device that determines the charge that can be drawn rom an energy store device, particularly a battery, up to a pre-determined end of iischarge in accordance with the preamble of Patent Claim 1 as well as the corresponding procedure according to the preamble of Patent Claim 2. n the case of electrical energy store devices such as e.g., batteries, the charge that can Decurrently drawn is an important quantity since it represents energy reserves that are still available right up to falling short of the minimum power capability required by an snergy store device. It is precisely in the area of automobile technology that an accurate Forecast of the charge that can be drawn is more critical than the knowledge of the current charge condition of the battery defined by the intermediary acid concentration in a lead-acid storage battery since this only provides information regarding the charge that has already been drawn compared with the full charge and not, however, regarding the charge quantity that can still be drawn. The entire charge that can still be drawn directly determines the availability of the electrical load that is attached to the energy store device. Apart from this, information regarding the charge that can be drawn can be used for control-oriented measures such as e.g., for energy management in a vehicle. This thereby enables e.g., timely initiation of load-reducing measures such as the switching off or dimming of less important loads before reaching minimum charge reserve. Determining the charge that can be drawn from an energy store device has already been established in EP 0376967 B1. The charge that can be drawn is thereby evaluated using empirically determined characteristic curve fields that are stored in a calculation unit subject to a constant discharge current, the battery temperature and onset of deterioration on the basis of the Peukert formula. It is, of course, thereby possible to determine the charge that can be drawn up to the end of discharge that is characterized by the complete discharge of the energy store device but it ist however, not possible to determine the charge that can be drawn right up to the falling short of a pre-determined minimum terminal voltage or up to falling short of the minimum power capability of the energy store device. Over and above this, the determination of the charge that can be drawn using the Peukert formula is relatively imprecise since various effects that influence the end of discharge such as e.g., an active mass loss at the electrodes due to deterioration of the battery or ice formation at the electrodes at low temperatures can not be taken into account. Thus the task of the present invention is to create a device as well as a procedure to determine the charge that can be drawn from an energy store device that, for example, enables a very precise determination of the charge that can be drawn up to a pre¬determined end of discharge criterion. This task is solved in accordance with the invention through characteristics specified in Patent Claims 1 and 9 respectively. Other designs of the invention form the subject-matter of subclaims. The essential idea of the invention consists of providing a charge predictor i.e., a device to calculate the charge that can be drawn, which calculates the charge that can be drawn taking into consideration a pre-determined discharge current pattern and temperature pattern with the help of a mathematical energy store device model. The energy store device model is thereby a mathematical model that demonstrates the electrical properties of the energy store device, based on various physical effects by using different mathematical models. The mathematical models describe functional correlations between state variables such as voltage, current, temperature etc. and incorporate various parameters. The charge calculations executed by the charge predictor take place based on the current condition of the energy store device. The mathematical models stored in the charge predictor are thereby first initialised at the current operating condition of the energy store device. A state variables and parameter assessor is provided for this purpose that determines the state variables and, if required, also the parameters of the energy store device model, based on the current operating values such as for example the voltage, the current and the temperature of the energy store device. An established Kalman filter can be used as an assessor for state variables and parameters for those state variables of the energy store device that can not be directly measured during operation. Based on this initialisation condition, the charge predictor then calculates the charge that can be drawn from the energy store device up to a pre-determined end of discharge i.e., up to one or several pre-determined end of discharge criteria which will be explained in greater detail hereinafter. The energy store device model in the case of a battery includes at least one model for the internal resistance R1of the battery, an acid diffusion resistance Rk and a transfer polarisation UD. The state variable and parameter assessor determines at least one no-load voltage UCo of the battery as state variables Z and one concentration polarisation Uk. If the battery capacity and therewith also the acid capacity C0 of the battery being used are not known, this too will have to be calculated. For this purpose, the assessor of state variables and parameters preferably determines at least the parameters Ri025, Ue, grenz. Rko25, UDo25 and C0. These parameters will be explained below in greater detail. The end of discharge criteria up to which the charge that can be drawn is calculated can, for example, be the attaining and/or falling short of a pre-determined minimum electrolyte voltage Uekrit, of a minimum terminal voltage UBattmin or the attaining of a pre¬determined minimum power capability ULastmin. In accordance with the preferred design of the invention, the charge that can be drawn up to the attaining and/or falling short of at least two, preferably all three of the end of discharge criteria will be calculated. The end of discharge criteria of the minimum electrolyte voltage Ueknt is fulfilled if the electrolyte voltage Ue falls under the pre-determined minimum electrolyte voltage Ueknt The pre-determined electrolyte voltage Ueknt thereby preferably takes the active mass loss due to battery deterioration into consideration and/or ice-formation at the electrodes in the case of low temperatures. The end-of-discharge criterion of the minimum terminal voltage UBattmin is fulfilled if the terminal voltage UBatt falls under the pre-determined minimum terminal voltage. The criterion of the minimum power capability is then achieved if the power supply voltage such as e.g. the voltage at a load supplied by the energy store device were to sink below a pre-determined threshold value if the energy store device were to be subject to a load for over the pre-determined length of time. A voltage predictor is provided in order to determine whether the load voltage at a pre-determined load current pattern were to sink below a pre-determined threshold value that determines the associated load voltage, subject to the load current pattern. With regard to a motor vehicle one can thus determine how much charge can still be drawn from a vehicle battery in the case of a pre-determined discharge current and battery temperature pattern till only that amount of charge is contained so that the power supply voltage at a load that is to be connected does not fall below the pre-determined threshold value at a pre-determined load current pattern. This is of particular importance in the case of a vehicle electrical system in order to prevent further charge being drawn from the battery such as e.g., is required at renewed start activity. Alternatively, other end of discharge criteria can also be defined. Ascertaining the charge that can be drawn from an energy store device is repeated by the charge predictor at pre-determined, chronological intervals, whereby current values for the discharge current lBatt.enti and the energy store device temperature TBatt,enti respectively are taken into account. The load predictor is preferably also in a position of being able to determine the time required to achieve the pre-determined end of discharge criterion. A state variables and parameter assessor preferably works on the basis of the same energy store device model as the charge predictor. The invention will be described in greater detail below with the help of the enclosed exemplary illustration which shows: Figure 1, a schematic illustration of a device to determine the charge that can be drawn from a battery with a charge predictor and a voltage predictor; Figure 2 is an equivalent circuit diagram for a lead-acid storage battery Figure 3a is a flow chart to illustrate the essential steps in the procedure by a charge predictor to calculate the charge that can be drawn; Figures 3b, c are flow charts illustrating the verification of various end of discharge criteria; Figure 3d is a flow chart that illustrates the essential steps by the voltage predictor in the procedure to calculate a minimum battery voltage; and Figure 4 is an illustration of the dependency of the electrolyte voltage on various physical effects. 1. Device to Determine the Charge that can be drawn Figure 1 is a block diagram of a device to calculate the charge that can be drawn from a battery, particularly from a vehicle battery. This consists of a state variables and parameter assessor 1, a charge predictor 2 and a voltage predictor 3. The device is in a position to calculate the charge that can be drawn from a battery (not illustrated), based on the current battery condition UBatt, iBatt, TBatt and a pre-determined discharge current pattern I Batumi up to arriving at a pre-determined end of discharge. The discharge current pattern lBatt,enti can thereby be any pre-determined current pattern or a constant current iBatt. The charge predictor 2 and the voltage predictor 3 comprise of a mathematical battery model which describes the electrical properties of the vehicle battery. Information such as the current operating values of the battery viz. the current battery voltage UBatt, of the prevailing battery current lBatt, and of the current battery temperature TBatt, as well as taking into consideration a pre-determined discharge current pattern lBatt.enti and a pre¬determined temperature pattern TBatt,enti . the charge that can be withdrawn from the battery Qe.uknt, Qe.uBattmin, Qe.uLastmin can be determined up to arriving at the three-way, variable end of discharge criteria (which are conjunctively combined in the present example). The discharge current pattern lBatt,enti and the temperature pattern TBattienti during the discharge can either be pre-determined by a control device (not shown) or can be ascertained from the current operating values of the battery UBatt, 'BattTBatt. The charge predictor 2 and the voltage predictor 3 comprise of a mathematical battery model that mathematically describes the electrical properties of the vehicle battery and is based on the equivalent circuit diagram for a lead-acid storage battery illustrated in Figure 2. 2. Equivalent Circuit Diagram of a Lead-Acid Storage Battery Figure 2 illustrates an equivalent circuit diagram of a lead-acid storage battery. The counting direction of the battery current lBatt, as is common practice, is positive for the charge and negative for the discharge. The individual state variables and components as follows are from left to right: Ri (Uco.Ue,TBatt) ohmic internal resistance subject to the no-load voltage UCo, the electrolyte voltage Ue and the acid temperature TBatt URi ohmic voltage drop Co acid capacity Uco no-load voltage RK (UCO. TBatt) acid diffusion resistance subject to the no-load voltage UCo (degree of discharge) and acid temperature TBatt xk = Rk *Ck response time of acid diffusion (is assumed to be constant in the magnitude order of 10 min) Uk concentration polarisation Ue = Uco + UK electrolyte voltage AUNernst (Ue, TBatt) voltage difference between the terminal voltage and the electrolyte voltage Ue subject to the electrolyte voltage Ue and acid temperature TBatt UD (IsaKi TBatt) stationary transfer polarisation subject to battery current lBatt and acid temperature TBatt UBatt terminal voltage of the battery The individual values are to be attributed to the different physical effects of the battery that are briefly described as follows: Voltage URi is the ohmic voltage drop at internal resistance Rj of the battery which in turn is subject to the no-load voltage UCo the electrolyte voltage Ue and the acid temperature TBatt- The no-load voltage UCo is proportional to the average acid concentration in the battery and is equal to the terminal voltage of the battery when the acid concentration is the same everywhere after a neutral phase in the battery. The concentration polarisation Uktakes the deviation of the acid concentrations from the average values in the battery at the location of the reaction i.e., the electrodes, into consideration. In the case of a battery discharge the lowest acid concentration is at the electrode pores since the acid there gets consumed and new acid has first to flow out from the electrolyte. The electrolyte voltage Ue takes into account the deviation of no-load voltage Uco through concentration polarisation subject to the acid concentration at reaction location. Thereby Ue =UCo + U« applies. The term AUNernst (Ue, TBatt) describes the voltage difference between the electrode potential and the electrolyte voltage which in turn is subject to local acid concentration at the reaction location and to acid temperature TBatt. The stationary transfer polarisation UD (iBatt, TBatt) takes into consideration an electrical transition resistance between the first electrode of the battery and the electrolyte and between the electrolyte and the second electrode of the battery and is, in turn, dependent on the battery current lBatt and the acid temperature TBalt. The diffusion of acids from the electrolyte at the reaction location i.e., to the electrodes during discharge is depicted by the acid diffusion resistance RK (UCo, TBatt) which in turn is subject to the no-load voltage UCo and the acid temperature TBatt. 3. The Mathematical Energy Store Device Model The mathematical energy store device model comprises of several models that describe the ohmic internal resistance of the battery Rj (UCo, Ue, TBatt), the acid diffusion resistance RK (UCo, TBatt), the voltage difference AUNernst (Ue, TBatt) between the electrode potential and the electrolyte voltage and the stationary transfer polarisation UD (IBatt, TBatt)- Alternatively, fewer or more mathematical models can be taken into consideration. Other mathematical models can also be applied for the individual values listed below. 3.1 Ohmic internal resistance: Thereby the following are: RJO25 ohmic internal resistance at full charge and TBatt = 25°C TKLfakt temperature co-efficient of the battery electric conductivity Ri( fakt characteristic curve parameter Ucomax maximum no-load voltage of the fully charged battery Ue,grenz electrolyte voltage in the case of end of discharge (subject to deterioration) 3.2 Acid diffusion resistance The following model can be applied for example to approximate the acid diffusion resistance RK: RK, fakt3 polynomal co-efficients 3.3 Voltage difference AUNernst between the electrode potential and the electrolyte voltage Ue The following model can be applied for example, for the voltage difference between the electrode potential and the electrolyte voltage: AUNernst(Ue, TBatt) = alpha*exp (- (Ue-Uekn) / beta) + TKuoo* (TBatt -25°C); with Alpha, beta, Uekn characteristic curve parameter TKuoo temperature co-efficient of the electrode potentials 3.4 Stationary transfer polarisation The following model can be applied for the stationary transfer polarisation UD : TKUD02 temperature co-efficients of the first, second and third order of transfer polarisation 3.5 Action of acid layering in the battery Acid layering is built with fluid electrolyte particularly in the case of a lead-acid storage battery if the battery is charged with a high current, starting from a low charge state i.e., lower average acid concentration. Acid in high concentration builds up in the region of the electrodes (reaction location) due to the high charge current which drops downwards as a result of its high specific gravity so that acid of lower concentration remains in the upper region. In the case of acid layering, the battery behaves like a battery with reduced capacity (and thus reduced charge that can be drawn) since only the lower battery region with higher acid concentration participates in the reaction. Apart from this, the electrode potential is increased above the value of a non-layered battery due to the increased acid concentration in the lower region. Since the no-load voltage UCoand the acid capacity C0 are determined and adapted by the state variables and parameter assessor 1, the impact of the acid layering on the charge that can be drawn in the case of the charge prediction by the charge predictor 2 is already implicitly taken into consideration. The procedure thus also takes into consideration the reduction of the charge that can be drawn in the case of batteries with acid layering. 4. Calculation of the Charge that can be drawn from the Energy Store Device Figure 3a illustrates the calculation of the charge Qe that can be drawn from a vehicle battery. For this purpose, the charge predictor 2 executes a numerical calculation and determines the state variables Uco, UK, Ue, AUNemst. URi and UBatt of the battery model from Figure 2. Details of the manner in which the calculation is executed are as follows: In Block 10, the charge qk drawn from the battery in a time unit interval tsampie in the case of an assumed discharge current pattern lBatt, emi is calculated and added in an iterative manner. The discharge current pattern lBatt, enti can, for example, be constant and correspond to the battery current lBatt or be any pre-determined current pattern. The following applies: This iterative calculation is executed till a pre-determined end of discharge criterion is fulfilled. The charge that can be drawn from the battery then is Qe = qk+i' and the time that continues to remain up to the attainment of the end of discharge criterion in the case of the pre-determined discharge current iBattemi is te = W . The stationary transfer polarisation UD (lBatt.entiT TBattemi), the no-load voltage Uco, k+i\ the concentration polarisation Uk,k+1\ the electrolyte voltage Ue,k+i\ the value A^emst, k+i\ the ohmic voltage drop URil k+1' and the battery voltage UBatt,k+i' in Blocks 11 to 15 are calculated. The equations in detail are thereby as follows: UBatt, k+i' with the index k+1 is thereby a new value after iteration. The iteration will be executed till a pre-determined end of discharge criterion, simultaneously three different end of discharge criteria as in the example at hand, is fulfilled. The comparison of state variables with the various end of discharge criteria is illustrated in Figures 3a and 3c. The first end of discharge criterion is the achieving of a critical electrolyte voltage Ue,krit that is determined by the acid concentration in the electrolyte, battery temperature TBatt. enti and a voltage limitation through active mass loss at battery electrodes AUe,grenz. A check is carried out in Step 21 of Figure 3b for every iteration step k to see whether the electrolyte voltage Ue, k+i* is less than or equal to the critical electrolyte voltage. In case it is, then the flag flagUe,krit is set at logical "1" (TRUE) in Step 22. The charge Qe that can be drawn in the case of this end of discharge criterion is thus Qe,uekht = qk+i' and the duration up to achieving the end of discharge criterion is teiUekrit = tk+1*- A check is conducted in Step 24, preferably parallel to Step 21 to see whether a second end of discharge criterion has been achieved. A verification is thereby made to see whether the battery voltage UBatt, k+i' is less than or equal to a pre-determined minimum battery voltage UBatt,min. In case it is, a specific flag with the term flagUBattmin is again set at TRUE. The charge that can be drawn Qe,ubattmin = q^i' and the time te]UBattmin up to achieving this end of discharge criterion is te,uBattmin = W- A final check is conducted in Step 26 (refer Figure 3c) to see whether the third end of discharge criterion, namely a required minimum power capability of the battery has been achieved. A check is hereby conducted to see if a dropping load voltage ULastat a load that can be pre-set would become less or equal to a minimum load voltage ULastimin during a pre-determined load current pattern lLast if the load were to be switched on at a pre-set time. The load voltage ULast is thus that voltage that adjusts itself to the load or, e.g., to the battery if the load were to be switched on with a pre-determined load current pattern lLast for a pre-determined time tLast. The background for this calculation is that it should be ensured that power supply voltage (or load voltage) is not to drop below a pre¬determined minimum value for the duration tLast and the load is to be adequately powered for the duration of its operation tLast. Voltage predictor 3 is provided to calculate the load voltage ULast that gets set after a pre-determined turn on time TLast Using the established models for state variables UCo, Uk, Ue, AU^st, URi and UD, the predictor 3 calculates the battery voltage UBatt (Step 36) in the case of a pre-determined load current pattern lLastand for a pre-determined load-turn-on time TLast. The minimum value of the battery voltage UBatt from all iteration steps (Step 37) after the termination of the load-turn-on time tLast(Step 38) is equal to the load voltage ULast(Step 39). The voltage predictor 3 uses the same calculation model in Blocks 30 to 36 as the charge predictor for the calculation of battery state variables, the only difference being that a load current pattern lLast forms the basis of the calculation. The load current pattern lLast is e.g., the current that a load such as, for example, a starter motor in a vehicle requires for operation. The load current pattern lLastturn on duration tLast can e.g., be pre-determined by a control unit. The following applies: C|k+l" = qk" + ll.ast*tsample tk+1 =tk +tsample The minimum battery voltage ULast in Block 26 that occurs in the case of load simulation is compared with a threshold value ULast,min and it ascertains whether the minimum load voltage ULast is less or equal to the voltage ULast,min. The calculation of the minimum voltage Umjn in the case of a pre-determined load current Last by voltage predictor 3 is executed in every iteration step of the charge predictor 2. When the simulation indicates that the minimum power capability has been arrived at (ULast In the case of a pre-determined discharge current iBattemi. the minimum power capability of the battery is arrived at in a period In case the end of discharge criteria in Steps 21, 24 and 26 have not been achieved then a check is run to see whether all three end of discharge criteria are simultaneously fulfilled just as was done in Step 28 after Blocks 22, 25 and 27. If yes, then the minimum value of the charges Qe,uekm, Qe, uBattmin, Qe.uLastmin that can be drawn is issued as the maximum load that can be drawn. The associated duration te can also be issued at the same time. If no, the calculation will be continued. In the case of a constant discharge current iBattenti = constant and constant temperature TBatt,enti = constant, the state variables UCo' and Uk' as well as the battery voltage UBattcan also be calculated analytically so that the iterative, time-intensive calculation by the charge predictor 2 can be dispensed with according to Figure 3a. 5. Determination of the First End of Discharge Criterion The charge that can be drawn from a battery essentially depends upon the acid contained in the electrolyte. Apart from this, the end of discharge is also secondly dependent upon the available active mass (Pb, Pb02 in the case of lead-acid storage batteries) in the electrodes of the battery during the discharge process and thirdly upon the electrolyte icing at low temperatures. The precision of the charge that can be drawn can be considerably improved by taking at least one of the above mentioned effects into consideration. 5.1 Acid restriction Discharge of the battery in the case of new batteries and batteries with a low active mass loss is essentially restricted by the acids contained in the electrolyte (acid restriction). The proportional electrolyte charge Ue is used by the charge predictor for the calculation of the charge that can be drawn for the acid concentration at the reaction location (electrodes). Typical limit values for the new batteries are e.g. Ue.krit, acid = 11.5 V in the case of end of discharge (refer branch b in Figure 4). 5.2 Active mass restriction In the case of batteries with a high active mass loss, discharge end (the battery does not supply a charge any longer) already occurs at higher voltages due to depletion of the active mass (Pb, Pb02) available for the discharge reaction. Figure 4 illustrates this displacement of the critical electrolyte voltage Ue,kirit by one value AUe,grenz to higher voltages (of 11.5 to 12V; from branch b to branch c). The following correlation can be applied taking the active mass restriction into consideration: Ue,krit»Masse = 11 -5 V + AUe,grenz 5.3 Electrolyte icing Electrolyte icing can occur if temperatures are lower than -10°C, particularly in the case of low acid concentration. Supply of acid to the reaction location at the electrodes is thereby inhibited, thus resulting in the prevalence of low acid concentration at the electrodes (refer branch a in Figure 4). The following temperature-dependent correlations can thereby be applied for a critical electrolyte voltage: Ue.knt.Ei3 (TBatt) = 11.423V-0.0558V*(TBatt/°C)-0.0011V*(TBatt/°C)2-1.0*e-5V*(TBa„/°C)3 The following correlation can be used for the first end of discharge criterion, (achieving a minimum electrolyte voltage), taking all three effects into consideration: Ue = Ue.krit = max (Ue .kritiSaure. i krit.Masse. Ue, kriti Eis) Figure 4 once again shows the resulting pattern of critical electrolyte voltage Ue,knt subject to the battery temperature TBatt and AUe,grenz. ir.rz.zuuz ROBERT BOSCH LTD; 70442 Stuttgart Legend 1 State variables and parameter assessor 2 Charge predictor 3 Voltage predictor 10-15 Calculation steps of the charge predictor 20-28 Verification of end of discharge 30-39 Calculation steps of the voltage predictor Z State variables P Parameter UBatt Battery voltage lBatt Battery current TBatt Battery temperature lBatt,enti Discharge current pattern TBatt,enti Temperature pattern Qe,Ue,krit Charge that can be drawn till reaching the critical electrolyte voltage Qe.uBattmin Charge that can be drawn till reaching the minimum battery voltage Qe.uLastmin Charge that can be drawn up to reaching the minimum power capability te Duration till arriving at end of discharge I Last Load current ULast Load voltage Ri Ohmic internal resistance UCo No-load voltage Uk Concentration polarisation URi Voltage drop at ohmic resistance Rk Acid diffusion resistance AUNemst Voltage difference between electrode potential and electrolyte voltage Ue Electrolyte voltage UD Transfer polarisation Ue,krit Critical electrolyte voltage UBatt,min Minimum battery voltage ULast,min Minimum load voltage Patent Claims 1. Device to determine the charge that can be drawn (Qe) from an energy store device, a battery in particular, up to a pre-determined end of discharge, characterised in that a charge predictorA (2) that calculates the charge (Qe) that can be drawn from an energy store device in the case of a pre-determined end of discharge pattern (lBatt,enti) on the basis of a mathematical energy store device model that demonstrates the electrical properties of the energy store device mathematically and a state variables and parameter assessor (1) that determines state variables (Z) and/or parameters (P) from current operating values (UBatt, I Baft, TBatt) of the energy store device for the mathematical energy store device model. 2. Device according to Claim 1 characterised in that the energy store device model is a battery model that incorporates at least one mathematical model for internal resistance (Rj), an acid diffusion resistance (Rk) and a transfer polarisation (UD). 3. Device according to Claim 1 or 2 characterised in that the state variables and parameter assessor (1) determines at least one no-load voltage (UCo) and one concentration polarisation (Uk) as state variables (Z). 4. Device according to Claim 3 characterised in that the state variables and parameter assessor (1) also determines a transfer polarisation (UD). 5. Device according to one of the above claims characterised in that the charge predictor (2) determines the charge that can be drawn (Qe) up to arriving at a pre¬determined minimum electrolyte voltage (Uemm) that represents a first end of discharge criterion. 6. Device according to one of the preceding claims characterised in that the charge predictor (2) determines the charge that can be drawn (Qe) up to arriving a minimum voltage (UBattmin) of the energy store device that represents a second end of discharge criterion. 7. Device according to one of the above claims characterised in that the charge predictor (2) determines the charge that can be drawn (Qe) up to arriving at a pre¬determined minimum power capability (ULastmin) that represents a third end of discharge criterion. 8. Device according to one of the preceding claims characterised in that a voltage predictor is provided, which can be preset to a load current pattern (lLast) and which, subject to the load current (lLast). determines an associated load voltage (ULast) that would be set on the basis of the pre-determined load current pattern (li_ast)- 9. Procedure to determine the charge that can be drawn (Qe) from an energy store device, a battery in particular, up to a pre-determined end of discharge characterised by the following steps: Calculation of the charge that can be drawn (Qe) in the case of a pre-determined end of discharge pattern (lBatt. Entiade) from an energy store device with the help of a load predictor (2) on the basis of a mathematical energy store device model that demonstrates the electrical properties of the energy store device mathematically, and Determination of state variables (Z) and/or parameters (P) for the mathematical energy store device from current operating values (UBatt. lean. TBatt) of the energy store device with the help of the state variables and parameter assessor (1). 10. Procedure according to Claim 9 characterised in that the load predictor (2) calculates the charge (Qe) that can be drawn up to arriving at a pre-determined minimum power capability (Uustmin), whereby a load voltatge (ULast) is taken into account that is supplied to the load predictor (2) from a voltage predictor (1) that ascertains the load voltage (ULast) subject to a pre-determined load current pattern (lLast).

Documents:

1641-chenp-2005-abstract.pdf

1641-chenp-2005-claims.pdf

1641-chenp-2005-correspondnece-others.pdf

1641-chenp-2005-correspondnece-po.pdf

1641-chenp-2005-description(complete).pdf

1641-chenp-2005-drawings.pdf

1641-chenp-2005-form 1.pdf

1641-chenp-2005-form 18.pdf

1641-chenp-2005-form 3.pdf

1641-chenp-2005-form 5.pdf

1641-chenp-2005-pct.pdf