Method and apparatus for charging or discharging an energy store

This application claims priority to PCT Application No. PCT/EP2016/055574, having a filing date of Mar. 15, 2016, based off of German application No. 10 2015 205171.4 having a filing date of Mar. 23, 2015, the entire contents both of which are hereby incorporated by reference.

The following relates to a method and an apparatus for charging or discharging an electrochemical energy store, the aging behavior of which is modeled by a description of its depth of discharge (DoD). The following furthermore relates to an electrical system and to a computer program product.

To determine the economical use of an electrochemical energy store in an energy system, the operation-dependent aging of the energy store must be taken into consideration. The so-called use-dependent lifetime loss of a lithium-ion battery can be modeled by the following formula

ΔSOH cycle,rel =f (DoD)· f ( T )· f ( I )·Δ Q   (1).

In this formula, f(DoD) is a non-linear factor that is dependent on the depth of discharge, f(T) is a non-linear factor that is dependent on the temperature that the battery cells exhibits, f(I) is a non-linear factor that is dependent on current, and ΔQ represents the charge throughput through the battery. The so-called state of health (SOH) represents the state of health of the battery and describes to what extent the measure of battery aging has already advanced. Aging of the SOH results in capacitive and/or resistive aging of the battery, as a result of which the performance of the battery is lowered. Ultimately, the battery needs to be replaced at some time.

The above-mentioned factors are determinable for different battery chemistries in similar ways. In formula (1), f(T) and ΔQ are considered to be known. The temperature-dependent non-linear factor f(T) is known from examinations of different battery types (i.e. battery chemistries) carried out in the past. The temperature-dependent factor f(T) is given by Arrhenius' law:

f ⁡ ( T ) = exp ⁢ - E a RT , ( 2 )

in which E a and R are known parameters, and T is the temperature in Kelvin.

The charge throughput ΔQ is a variable that is given by the battery operation and is metrologically determinable.

The current-dependent factor f(I) can be determined from aging experiments. A small factor for discharge currents and small charge currents and a large factor for large charge currents is typical.

The factor f(DoD) is likewise determined from aging experiments. The aging differs in dependence on stroke of the state of charge (SOC). The greater the SOC stroke, i.e. the depth of discharge, the greater the factor f(DoD). Generally, a factor f(DoD) which increases monotonously with the depth of discharge is typical.

In the experiments that are to be carried out in order to determine the factors, the energy store is cyclically charged and discharged with a constant SOC stroke, as is illustrated by way of example in FIG. 1 . FIG. 1 shows the profile of the state of charge SOC over time t. The state of charge SOC is 100% if the energy store is fully charged. The state of charge SOC is 0% if no more charge can be drawn from the energy store, i.e. the energy store is empty. The time tin the present exemplary embodiment is plotted normalized from 0 to 100. The profile of the state of charge in FIG. 1 is such that, starting from an SOC 1 , which is approximately 90%, the charging store is discharged to an SOC value of approximately 30% with a constant current. Next, a constant, yet different current is used to effect charging from SOC 2 to SOC 1 . The process subsequently repeats itself several times. The difference between SOC 2 and SOC 1 is the SOC stroke or the DoD value (DoD=SOC 1 −SOC 2 ), as is illustrated by way of the arrow DoD in FIG. 1 .

One disadvantage of using cyclic discharging and charging of the energy store, illustrated in FIG. 1 , is that it is not possible to determine from these experiments what influence the exact form of the SOC profile has on aging. In particular, it is not possible to determine whether the description of the DoD with respect to the observed battery chemistry is correct.

For this reason, the so-called “rainflow algorithm” is alternatively used to calculate the DoD value. Here, an SOC profile as shown in FIG. 2 is interpreted as different, interleaved partial cycles having a different stroke. The use of the rainflow algorithm is a good approximation for many battery chemistries. For some battery chemistries, the rainflow algorithm is an inaccurate method for determining the DoD and thus results in a false model of aging. This consequently leads to incorrectly derived charging strategies.

Since the DoD according to formula (1) has an influence on the aging of an energy store, the correct description of the DoD is necessary for correct modeling of the aging behavior.

An aspect relates to a method and an apparatus with which it is possible in a more precise manner to charge or discharge an electrochemical energy store, the aging behavior of which is modeled by a specific description of its depth of discharge. A further aspect is to specify a corresponding computer program product.

Proposed is a method for charging or discharging an electrochemical energy store, the aging behavior of which is modeled by describing its depth of discharge, at which a depth-of-discharge value is associated with each time period between two adjacent local maxima and minima in a time-dependent state-of-charge profile, which depth-of-discharge value corresponds to the charging or discharging stroke between the state-of-charge values of the two adjacent local maxima and minima.

The local maximum and minimum of the time-dependent state-of-charge profile are also considered to be points at which the charging or discharging current has the value 0 for a short time, i.e. for a short period of time. This type of modeling the depth of discharge (also referred to as DoD modeling below) means that a large discharging process is interrupted by brief stopping or brief charging, and is divided into two small partial cycles. The depth of discharge (DoD) of this discharging process is thus reduced to smaller values which are given by the SOC stroke before and after the interruption. The same is true of a charging operation which is interrupted by brief stopping or short-term discharging, and is divided into two small partial cycles.

The described modeling is not suitable for all types of battery chemistries. However, it has been found that such a DoD behavior is precise in lithium-ion batteries, in particular lithium-titanate batteries.

In the method according to embodiments of the invention, the energy store is transitioned from a first state of charge at a first time to a second state of charge at a second time by way of the charging or discharging process being interrupted at least once within the time period formed between the first and the second time to form respective charging or discharging subblocks with a respective charging or discharging stroke, wherein a respective charging or discharging stroke is smaller than the total stroke between the first and the second state of charge.

Compared to uninterrupted complete charging or discharging from the first state of charge to the second state of charge, shorter charging or discharging subblocks having a respective (smaller) charging stroke are obtained with this method. With this method, battery aging can be significantly reduced compared to constant charging without interruption from the first state of charge to the second state of charge. Since battery aging results in reduced capacity and increased internal resistance, the proposed method increases the lifetime of the battery. At the same time, the efficiency of the battery is improved owing to the comparatively low increase of internal resistance.

Interruption between two charging or discharging subblocks can comprise a charging or discharging current having the value 0. In order to thus divide a charging or discharging process into at least two charging or discharging subblocks, it is sufficient if the charging or discharging process is interrupted for a specific time. The interruption thus means that the charging or discharging current comprises the value 0.

An interruption between two charging subblocks can comprise a discharging section. Alternatively, an interruption between two discharging subblocks can comprise a charging section. With these configuration variants, energy stores are taken into consideration that have such battery chemistries in which an interruption where the charging or discharging current has the value 0 is not sufficient to be “recognized” as an interruption. However, such energy stores are also able to “recognize” the interruption in the manner according to embodiments of the invention, because a charging or discharging current with an opposite sign to the preceding charging and discharging process is used.

In accordance with a further configuration, charging or discharging the energy store can be interrupted at regular intervals. This results in the charging or discharging subblocks having identical lengths. Alternatively, the charging or discharging of the energy store can also be interrupted at irregular intervals.

Provision may furthermore be made for the time period of the interruptions to be of equal length. The time periods of the interruptions can alternatively also have different lengths.

In accordance with a further configuration, charging or discharging the energy store can be effected in the charging or discharging subblocks at least in sections with a constant current. That means the charging or discharging current has the same absolute value within a charging or discharging subblock. Alternatively, a plurality of sections with different charging or discharging currents can be provided within a charging or discharging subblock. Provision can also be made for the charging or discharging current that is used for one discharging subblock to also be used in at least one other or even all other charging or discharging subblocks. Which of the variants mentioned is best depends not only on the extent of the charging stroke between the first and the second state of charge, but also on the battery chemistry of the relevant energy store.

In accordance with a further expedient configuration, the magnitude of the charging current for each charging or discharging subblock can be adapted in dependence on the length of the previous interruption, with the result that it is ensured that the second state of charge is reached at the second time.

Furthermore, a computer program product is proposed, which can be loaded directly into the internal memory of a digital computer for controlling the charging and/or discharging of the electrochemical energy store and which comprises software code sections with which the steps of the above-described method are carried out when the product runs on the computer.

In accordance with a further aspect of the invention, an apparatus ( 702 of FIG. 7 ) for charging or discharging an electrochemical energy store is proposed, the aging behavior of which is modeled by describing its depth of discharge, at which a depth-of-discharge value is associated with each time period between two adjacent local maxima and minima in a time-dependent state-of-charge profile, which depth-of-discharge value corresponds to the charging or discharging stroke between the state-of-charge values of the two adjacent local maxima and minima. The apparatus ( 701 - 704 of FIG. 7 ) is configured to transition the energy store from a first state of charge at a first time to a second state of charge at a second time, in which the charging or discharging process is interrupted at least once within the time period formed between the first and the second time to form respective charging or discharging subblocks with a respective charging or discharging stroke, wherein a respective charging or discharging stroke is smaller than the total stroke between the first and the second state of charge.

The apparatus can comprise further means for carrying out the method.

The apparatus has the same advantages as have been described previously in connection with the method according to embodiments of the invention.

Embodiments of the invention finally propose an electrical system that comprises an apparatus of the above-mentioned type and an energy store, wherein the energy store is a lithium-titanate battery or an energy store having a comparable battery chemistry. The energy store can be for example a traction battery of an electrically operated vehicle or of a hybrid vehicle, which not only has an electrical drive but also an internal combustion engine drive. The energy store can also be used in a self-sufficient energy supply system, e.g. for intermediate storage of regeneratively generated energy. Such an energy supply system can also have a connection to a conventional composite energy supply system.

Some of the embodiments will be described in detail, with reference to the following figures, wherein like designations denote like members, wherein:

FIG. 1 shows an already described cyclic charging and state-of-discharge profile of an electrochemical energy store,

FIG. 2 shows an already described charging curve of an electrochemical energy store for use of the known “rainflow algorithm” for ascertaining a depth-of-discharge value,

FIG. 3 shows the charging curve known from FIG. 2 , by way of which a description of its depth of discharge, used in accordance with embodiments of the invention, is explained,

FIG. 4 shows a state-of-charge profile, proposed in accordance with embodiments of the invention, during charging of an electrochemical energy store, the aging behavior of which is able to be modeled by describing its depth of discharge according to embodiments of the invention, and

FIG. 5 shows a state-of-discharge profile, proposed in accordance with embodiments of the invention, during discharging of an electrochemical energy store, the aging behavior of which is able to be modeled by describing its depth of discharge in accordance with embodiments of the invention.

FIG. 6 is a flowchart of the method for charging or discharging.

FIG. 7 is a device for charging or discharging.

FIG. 3 shows an exemplary, time-dependent state-of-charge profile. The normalized time t from 0 to 100 is plotted on the x-axis. The state of charge (SOC) is plotted in percent from 0 to 100 on the y-axis. Here, an SOC value of 100% corresponds to a fully charged energy store, while an SOC value of 0% corresponds to a completely empty energy store.

In the exemplary state-of-charge profile shown in FIG. 3 , the curve starts at an SOC value SOC 1 of approximately 90%, and is initially discharged to an SOC value SOC 3 of approximately 50%. Then follows a charging process to an SOC value SOC 4 of approximately 65%. This is followed by a discharging process to an SOC value SOC 5 of approximately 25%, which increases to an SOC value SOC 6 of approximately 30%, until finally a discharge to the SOC value SOC 2 of approximately 10% takes place. Subsequently, a charging process to the SOC value SOC 1 takes place, wherein the charging process is interrupted by two temporary discharging processes (not shown in more detail in the Figure).

The DoD value associated with this exemplary state-of-charge profile is determined by initially determining the local maxima and minima of the state-of-charge curve. In the state-of-charge curve shown, local minima M 3 , M 5 and M 2 , and local maxima M 4 and M 6 are thus marked in particular. The local minima M 3 , M 5 and M 2 are associated with the SOC values SOC 3 , SOC 5 and SOC 2 . The local maxima M 4 and M 6 are associated with the SOC values SOC 4 and SOC 6 .

It should be noted that also considered to be local maxima and minima are such points on a state-of-charge curve at which the charging current takes the value 0 for a previously defined time. In the state-of-charge profile shown in FIG. 3 , however, no such “plateaus” at which the charging or discharging current assumes the value 0 for a specific time are shown.

In a next step, a DoD value, which corresponds to the charging stroke between the SOC values that are associated with the local maxima and minima, is associated for each time period between two points of reversal, i.e. a local maximum and the adjacent local minimum. The corresponding DoD values between two points of reversal are shown by way of example with arrows. Here, the corresponding charging stroke between the local minimum M 3 and the local maximum M 4 corresponds to the DoD value DoD 1 , which is obtained from the absolute value of the difference of the SOC values SOC 4 −SOC 3 . In a similar way, a charging stroke DoD 2 between the local maximum M 4 and the local minimum M 5 was ascertained, which is obtained from the absolute value of the difference of the SOC values SOC 4 −SOC 5 . Analogously, a DoD value DoD 3 between the local minimum M 5 and the local maximum M 6 is obtained from the absolute value of the difference of the SOC values SOC 6 −SOC 5 . Finally, the DoD value DoD 4 between the local maximum M 6 and the local minimum M 2 was ascertained, the stroke of which is obtained from the absolute value of the difference of the SOC values SOC 6 −SOC 2 . For the sake of simplicity, the corresponding DoD values for the further local maxima and minima contained in the charging curve were not plotted.

Since the described procedure is a continuous calculation method, no initial point has been associated with the SOC value SOC 1 for the sake of simplicity. However, this is also possible in principle.

This type of DoD modeling means that the mere discharging process from the state-of-charge value SOC 1 to the state-of-charge value SOC 2 is interrupted once or a plurality of times by stopping or by a charging or discharging process in the opposite direction, and is divided into a plurality of small partial cycles. The DoD of the discharging process from the state-of-charge value SOC 1 to the state-of-charge value SOC 2 is thus reduced to the smaller values between the adjacent local maxima and minima (points of reversal), which are given by the charging stroke before and the charging stroke after the interruption.

Such modeling applies in particular to lithium-ion batteries, and here in particular to lithium-titanate batteries.

Starting from the above-described modeling, the exemplary charging and discharging strategies which are described below are obtained, which increase the lifetime of the battery and improve the efficiency of the battery over time as compared to charging and discharging strategies that are based, for example, on modeling of the DoD in accordance with the rainflow algorithm.

FIG. 4 shows a state-of-charge curve in accordance with the method ( FIG. 6 ) on which embodiments of the invention is based. In order to transition the energy store (step 601 of FIG. 6 ) at time t 0 from the state of charge SOC 1 to the state of charge SOC 2 at time t 7 , charging with a constant current is carried out, wherein this charging process is interrupted at regular intervals for a particular short time period dt. The charging current is obtained by

Here, SOC 1 is the state of charge at time t 0 , SOC 2 is the state of charge at time t 7 , Cap is the capacity of the energy store in ampere hours, T is the time period between t 7 and t 1 (T=t 7 −t 1 ), dt is the time period of the interruption of the charging process, and n is the number of interruptions.

As a result of this charging method (step 602 of FIG. 6 ), a number of charging subblocks are obtained (step 603 of FIG. 6 ) with a respective charging stroke

In the exemplary embodiment shown in FIG. 4 , the SOC value SOC 1 =10% at time t 0 . At time t 7 , the SOC value SOC 2 =90% is intended to be reached. By way of example, the charging process comprises four charging subblocks 10 L, 20 L, 30 L, 40 L. Shown between two adjacent charging subblocks 10 L, 20 L, and 20 L, 30 L, and 30 L, 40 L are the already mentioned interruptions 12 , 22 , 32 having the length dt, during which the charging current I=0. The charging stroke DoD is the same owing to the constant charging current I in all charging subblocks 10 L, 20 L, 30 L, 40 L and the respective constant charging periods. In FIG. 4 , this is marked by the reference signs 11 L for the charging subblock 10 L, 21 L for the charging subblock 20 L, 31 L for the charging subblock 30 L, and 41 L for the charging subblock 40 L. Overall, a charging stroke of SOC 2 −SOC 1 is obtained, wherein the entire charging stroke is marked with the reference sign 50 .

Similarly, FIG. 5 shows a state-of-discharge curve in accordance with which a discharge (step 602 of FIG. 6 ) of the SOC value SOC 2 at time t 7 is intended to be reached starting from an SOC value SOC 1 =90% at time t 0 . Analogously with the preceding exemplary embodiment, the discharging process is divided (step 604 of FIG. 6 ), merely by way of example, into four discharging subblocks 10 E, 20 E, 30 E, 40 E of equal lengths, wherein respective interruptions 12 , 22 , 32 are provided between two discharging subblocks 10 E, 20 E, and 20 F, 30 E, and 30 E, 40 E, during which the discharging current I=0. Merely by way of example, the interruptions 12 , 22 , 32 are here also of equal time period dt. Owing to the constant discharging current in each discharging subblock 10 E, 20 E, 30 E, 40 E (which results from the constant negative increase of the respective sections) and the interruptions 12 , 22 , 32 , which are in each case of equal length, a discharging stroke 11 E, 21 E, 31 E, 41 E of the same magnitude is obtained for each discharging subblock 10 E, 20 E, 30 E, 40 E.

In principle, the charging breaks or interruptions 12 , 22 , 32 during a charging process should be as short as possible in order to keep the required charging current low in order to reach the second state of charge SOC 2 from the first state of charge SOC 1 within the time period T. The longer the selected interruptions 12 , 22 , 32 are, the greater the charging current required in the charging subblocks becomes, i.e. the steeper the sections of the charging subblocks 10 L, 20 L, 30 L, 40 L are. Associated damage to the energy store during charging can thus also be avoided.

In principle, the magnitude of the charging current is obtained in dependence on the boundary condition that the state of charge SOC 2 to be reached must be reached within the time period T. Another boundary condition is the battery chemistry of the energy store that permits a maximum charging current without being negatively influenced in the process.

In contrast, longer interruptions and correspondingly greater discharging currents are also acceptable during a discharging process.

Although the exemplary embodiments in accordance with FIGS. 4 and 5 illustrate charging and discharging subblocks 10 L, . . . , 40 L and 10 E, . . . , 40 E of equal lengths, it is possible for the charging and discharging subblocks to have an individually different length with an individually different charging or discharging current. The length dt of the interruptions 12 , 22 , 32 between respective charging and discharging subblocks can likewise be selected to be of different length. It is furthermore also possible for the charging and discharging currents to be selected individually for each charging and discharging subblock during a charging process or a discharging process. It is even possible to vary the charging and discharging currents within one charging and discharging subblock. The variation takes place with the boundary condition that charging or discharging from the first state of charge SOC 1 to the second state of charge SOC 2 is possible within the time period T.

In the case of regular interruptions, the interruption interval dt, i.e. the time period between t 3 and t 1 , and t 5 and t 3 , and t 7 and t 5 should be chosen such that the resulting DoD value sufficiently reduces the aging in contrast to an uninterrupted charging process. In addition, the interval dt should not be selected to be so small that the charging current as a result becomes unnecessarily large with a fixed time duration dt of the interruptions 12 , 22 , 32 , and remains within specified limits.

The battery aging in a first approximation changes on account of the proposed charging strategy by the factor

This factor is typically less than 1, which means that the energy store ages less, because the DoD-dependent aging for small DoD values is typically less than for large DoD values. If a desired reduction in aging is specified for an application or is obtained by way of an optimization calculation, the interruption interval can be calculated by way of this factor, possibly taking into consideration a specified limit of the charging or discharging current.

If properties of the battery chemistry of the energy store necessitate a short counter-charging process in order to terminate a charging or discharging subblock, then the interruption can also be replaced by such a charging process. A counter-charging process is understood to mean, for example, a brief charging process when a discharging process is carried out. In the case of a charging process, a brief discharging process is carried out in the interruption. The strength of the current during counter-charging must be adapted such that the stroke SOC 2 −SOC 1 is still reached after the time period T.

Although the invention has been illustrated and described in greater detail with reference to the preferred exemplary embodiment, the invention is not limited to the examples disclosed, and further variations can be inferred by a person skilled in the art, without departing from the scope of protection of the invention.

For the sake of clarity, it is to be understood that the use of “a” or “an” throughout this application does not exclude a plurality, and “comprising” does not exclude other steps or elements.