EP3479134A1

EP3479134A1 - Method for monitoring a battery

Info

Publication number: EP3479134A1
Application number: EP17719609.4A
Authority: EP
Inventors: Patrick MUENZING; Oliver Dieter Koller
Original assignee: Robert Bosch GmbH
Current assignee: Robert Bosch GmbH
Priority date: 2016-06-30
Filing date: 2017-04-27
Publication date: 2019-05-08
Also published as: WO2018001602A1; CN109313240B; US20190212391A1; DE102016211898A1; CN109313240A

Abstract

The invention relates to a method and an arrangement for monitoring a battery in a motor vehicle. In the method, a first module (180) determines operating parameters of the battery. Variables, which represent the operating parameters, are compared with a load capacity model in order to determine in this way reliability variables of the battery, so that a future behavior of the battery can be predicted.

Description

Description title

Method for monitoring a battery

The invention relates to a method for monitoring a battery,

in particular a battery in a motor vehicle, and an arrangement for carrying out the method.

State of the art

An electrical system or vehicle electrical system represents the entirety of the electrical components or consumers of a motor vehicle. This has the task of supplying the electrical consumers with energy. When

Energy storage in on-board networks, for example, batteries are used. Falls in today's vehicles, the power supply due to an error, eg. Due to aging, in the electrical system or in a vehicle electrical system component, so omitted important functions, such as the power steering. Since the steerability of the vehicle is not compromised, but only becomes stiff, the failure of the electrical system in today's mass-produced vehicles is generally accepted. In addition, in today's vehicles the driver as

Fallback level is available.

It should be noted, however, that due to the increasing electrification of aggregates and the introduction of new driving functions higher

Requirements for the safety and reliability of electrical

Power supply in the motor vehicle are provided. At future

highly automated driving functions, such as, for example, a highway pilot, the driver are allowed to do activities that are outside the driver to a limited extent. It follows that until the completion of the highly automated driving function, the driver Function as a sensory, regulatory technical, mechanical and energetic fallback can perceive only limited.

For the above reason has the electrical supply at

highly automated driving to ensure sensory,

control technical and actuatory fallback a previously in

Motor vehicle unknown security relevance. Errors or aging in the electrical system must therefore be reliably and as completely as possible in the sense of product safety.

In order to predict the failure of components, reliability approaches have been developed to monitor

Vehicle components developed. For this purpose, the on-board network components are monitored during operation and their damage is determined.

The document DE 10 2013 203 661 AI describes a method for operating a motor vehicle with an electrical system. This vehicle electrical system has a semiconductor switch, based on a determination

past load events an actual load is determined. In the method, the actual applied to the semiconductor switch load is detected.

Disclosure of the invention

Against this background, a procedure with the characteristics of the

Claim 1 and an arrangement according to claim 11 presented.

Embodiments result from the dependent claims and the description.

The presented method takes into account that in the future automated and autonomous driving operation in the motor vehicle, the driver is no longer available, as is known from the prior art, as a sensory, control-technical, mechanical and energy-related fallback level. Rather, the vehicle takes over the functions of the driver, such as the Environmental recognition, the trajectory planning and the trajectory implementation, which include, for example, the steering and braking.

If the power supply of the safety-relevant components fails, the vehicle can no longer be controlled by the highly or fully automated function, since all the functions described above, eg.

Ambient detection, trajectory planning and implementation are no longer available. This results in very high demands on the vehicle electrical system from the point of view of product safety. This also means that the function of automated or autonomous driving the user may only be available when the electrical system is in perfect condition and at least in the near future remains.

The battery or batteries is or are one of the most important components in the energy supply system, which ensure the energy supply in the vehicle. It has been recognized that this special position in the on-board network requires the analysis of the battery to be extended by predictive approaches.

The presented method can be structurally arranged in four consecutive modules, the total, individually or in any combination, for example. In the battery sensor, in another control unit or in a

comparable device, eg. A cloud, can be implemented or implemented. The basic first module is a prerequisite for all other modules. These can be combined in any combination with the first module. The following four modules are discussed below: first module:

The object of the first module is to determine the load of the battery by using the data of the battery sensor or a similar device that is used to determine the battery sizes and / or condition monitoring, and to compare them with a load capacity model, which determines reliability characteristics of the battery can be. Possible extensions are:

Implementing limit values of the reliability characteristics that cause the battery to change, the operating modes to be disabled, the transition to the safe state and / or the driver to take over,

- Further processing of the determined reliability characteristics for determining the system reliability characteristics, z. B. On-board network failure probability; Here, too, operating modes can be blocked via limit values, for example, and / or the transition to the safe state and / or the driver transfer can be triggered or triggered.

The second module, which is an extension to the first module, has the tasks of an online prediction of the load on the battery:

- Shares for certain scenarios, such. B. operating modes or

To give operating strategies

- Safe-stop scenarios, which are still feasible with the (aged) battery to select and

- Battery change, typically on the basis of the previous load to predict.

This data can be transmitted to a higher-level control unit for further processing.

The third module, which is an extension to the first module, has the task, by balancing the resilience model with the extrapolation of the actual SOH (state of health), for example characterized by the capacity loss, of the battery to the quality of the Adapt battery. The resilience model is subject to statistical dispersion. By comparison with the determined SOH, the quality of the battery or the shift of the load capacity model can be taken into account. The fourth module, which is an extension to the first module, has the task of matching the SOH and the previously experienced load of the battery with central databases, such as a cloud, in order to - the load-bearing models due to the large number of in-field

To improve batteries,

- to be able to adapt the load capacity models in the vehicle online and - to design future components / systems in the vehicle better.

In known methods, a system control or a system control, which performs an overall status monitoring of all relevant components or vehicle functions in the vehicle, has hitherto not been present. From the point of view of product safety, such a system appears to be necessary for safety-critical new applications with changed basic assumptions, such as, for example, automated driving.

It should be noted that wear failure of components is the root cause of much of the electrical system conditions that exist in the context of the new

Areas of application are safety relevant. Therefore, they must be preventively identified in the vehicle and countermeasures initiated. Since the battery is one of the most important components in the energy on-board network, measures for preventive battery analysis, which are indispensable for realizing the new applications, are presented in the context of the present application.

The presented method and arrangement have, at least in some of the embodiments, a number of advantages, which are listed below:

- Support for the approval and release decision for automated driving functions:

Aging effects or the exceeding of a predetermined, accepted aging in the battery leads or leads to the withdrawal of the release or to Leaving the driving functions, such. As automated driving, or to withdraw the release or to leave certain operating modes, eg. B. sailing, or to the transition to the safe state to avoid safety-critical conditions.

- Increased reliability through adapted driving strategies:

Driving situations that lead to a strong aging behavior of the battery during operation are avoided, if possible from the system point of view.

- availability increase:

Preventive battery replacement can be timely before an uncontrolled

Battery failure, for example, be carried out at regular maintenance intervals.

- Safety gain when transferring from automated driving to manual driving:

By early warning of a battery failure, the vehicle transfer can be carried out in a more easily controllable for the driver situation.

- A compelling need to bring the vehicle into safe condition even in the event of a component failure without driver intervention in fully automated driving:

Gain time in the initiation of the fallback strategy by early warning or no release of the driving functions with imminent battery failure and avoidance of unwanted electrical system failure by checking which safe-stop scenario is still permissible from the battery point.

- Increase the reliability and safety even of non-automated vehicles by early detection of upcoming failures:

As a result, even "lying down" on lanes, eg. On highways, can be avoided. As already stated, it was recognized that it is indispensable for automated or autonomous vehicles, in addition to the current state of the safety-relevant components, to predict their future behavior as well. In order to be able to evaluate and forecast the state of the energy grid, as the basis of all safety-relevant vehicle functions, forecasting units are necessary for each component. The presented method provides the necessary procedure for analyzing the battery, which is classified as an important component of the energy on-board network. One possible implementation of the procedure is in steps and with it

related effects or benefits outlined below:

The battery sensor or a comparable device, which serves to determine the battery sizes and / or to monitor their condition, transmits load-relevant parameters which were recorded at times of the respective measurements, such as, for example, SOC (state of charge) and temperature. Each characteristic is thus assigned to a point in time.

- The method determines from the load-relevant parameters the previously seen load, combined with the load capacity

Reliability characteristics of the battery, such as the

Probability of default, calculated.

- The method is based on the prediction of the reliability characteristics in a position to continue to identify possible safe-stop scenarios, taking into account electrical system errors and operating strategies.

- The method is based on the prediction of the reliability characteristics in a position to grant releases of operating modes in consideration of operating strategies, to grant or prevent time-limited.

- The method is based on the prediction of the reliability characteristics suitable to go in the event of an imminent battery failure on time in the safe state. - The method is based on the prediction of the reliability characteristics in a position to predict the failure of the battery and thus schedule a timely change.

The method is suitable for optimizing the prognosis model of the battery by the actual aging, which is determined, for example, in the battery sensor.

- The method transmits the calculated data to a central

Data storage, allowing further optimization of the forecasting model.

Further advantages and embodiments of the invention will become apparent from the description and the accompanying drawings.

It is understood that the features mentioned above and those yet to be explained below can be used not only in the particular combination indicated, but also in other combinations or in isolation, without departing from the scope of the present invention.

Brief description of the drawings

Figure 1 shows in a block diagram a battery sensor according to the prior art.

FIG. 2 shows a block diagram of a battery sensor for carrying out the method.

FIG. 3 shows in a flowchart steps which are executed in the algorithm of an embodiment of the presented method one after the other.

FIG. 4 shows in a graph a Wöhler curve.

FIG. 5 shows in a graph the Weibull distribution. Embodiments of the invention

The invention is based on embodiments in the drawings

schematically and will be described below with reference to the

Drawings described in detail.

Figure 1 shows a battery sensor according to the prior art, which is generally designated by the reference numeral 10. Input variables in a unit 12, in particular a measuring unit, are the temperature T 14 and the current 1 16, output is the voltage U 18.

In a block 20, the estimation of parameters and states takes place.

Herein, a feedback unit 22, a battery model 24 and an adaptation 26 of the parameters are provided. There will be a variable ü 28,

State variables ^ x 30 and model parameters ^ JD 32 output.

A node 29 serves to adapt the battery model 24 to the battery. The current 1 16 goes directly and the temperature T 14 is indirectly in the

Battery model 24 on. This calculates ü 28 and compensates this with the real voltage U 18. In case of deviations, the battery model 24 is corrected via the feedback unit 22.

Furthermore, a block 40 is provided for sub-algorithms. This includes a battery temperature model 42, a rest voltage 44, a

Peak current measurement 46, an adaptive starting current prediction 48 and a battery size detection 50.

In addition, charge profiles 60 are provided, which are in a block 62 with

Enter predictors. These are a charge predictor 64, a

Voltage predictor 66 and an aging predictor 68. Outputs of block 62 are an SOC 70, waveforms of current 72 and voltage 74, and an SOH 76.

The battery sensor 10 thus determines the current SOC (state of charge) 70 of the battery and the current SOH 76 (State of Health, capacity loss in the battery)

Compared to the initial state) of the battery. About the predictors 64, 66, 68 For example, the battery sensor 10 is capable of predicting the SOC 70 and the SOH 76 according to a plurality of predefined load scenarios. These can now also be adapted to automated driving or to the respective application.

FIG. 2 shows a battery sensor for carrying out the present invention

Method, indicated generally by the reference numeral 100. This battery sensor 100 is an extension to the battery sensor 10 of Figure 1. The battery sensor 100 is shown in simplified form, in principle, all components of the battery sensor 10 of Figure 1 in the

Battery sensor 100 of Figure 2 is provided.

The illustration shows a block 120 for the estimation of parameters and states. Herein, a feedback unit 122, a battery model 124 and an adaptation 126 of the parameters are provided. In a block 162 with

Predictors include a charge predictor 64, a voltage predictor 66, and a first module 180. The first module 180 is here representative of all modules. The first module is compulsory, the other modules can be placed here in any combination.

In the first module 180, the calculation of the current

Reliability characteristic (s) of the battery, such as the

Probability of failure, trigger for battery change, trigger transition to the safe state or driver acceptance.

In order to determine the load on the battery, the battery sensor 100 transfers the current SOC and temperature values to the first module 180 in the battery sensor 100 or in another control device (arrow 190). There, the values are saved as SOC and temperature curves. At the same time, the time points of the SOC and temperature measurements are recorded as a time course.

The SOC history is online in the control unit or battery sensor using

Rainflow count classified according to time. The rainflow count is a procedure in which amplitudes, their centers, their starting time and their duration are determined from the course of a measurement. This causes a conversion of the course in strokes with the characteristics amplitude, Center of lift, start time of the stroke and duration of the stroke. In addition to the Rainflow count, there are other suitable methods.

Over the time that the respective stroke has taken place, a temperature can be assigned to the hub. About the slope of the Wöhler curve, as shown in Figure 4, the respective stroke is set to the

Reference level, e.g. B. ÄSOC 30% and 25 ° C, converted, on which the

Resilience data available. The temperature can be considered, for example, via an Arrhenius approach.

FIG. 4 shows, in a graph 400, on the abscissa 402 of which the number of cycles and on whose ordinate 404 A SOC [%] is plotted, the course of the Wöhler curve N _f 406.

The Wöhler curve Nf indicates which number of cycles can be sustained at which stroke from the battery until the failure criterion is reached. The Wöhler curve can be described, for example, by equation 1:

N = a (ASOLy (1)

By transforming this equation 1, all battery strokes determined by the Rainflow count can be converted to a reference level.

The load capacity model of the battery, which in this case is represented by a Weibull distribution, plots the number of battery cycles at reference level that leads to the failure probability of the battery. By loading the battery on reference level and load capacity model on

Reference level can thus be calculated online the probability of failure of the battery at the current time. The Weibull distribution is the

most likely distribution, theoretically other distributions could better describe failure behavior. The Weibull distribution is shown in FIG.

FIG. 5 shows in a graph 500 at its abscissa 502 the number of cycles and at its ordinate 504 the probability of failure [%] Weibull distribution 506 is shown with a bottom line 508 showing the lower confidence interval, an upper line 510 representing the upper one

And a line 512 representing a probability that 50% of the components have failed.

Possible extensions or adjustments are:

Implementation of limit values of the reliability parameters, the change of the battery or the blocking of operating modes, eg. As automated driving, sailing, recuperation, the transition to the safe state and / or the driver initiation initiate

- Further processing of the determined reliability characteristics for determining the system reliability characteristics, z. B. On-board network failure probability; also here can z. B. blocked via limits operating modes and / or initiated a transition to the safe state.

In Figure 2, a second module 200 is further shown. This is used to predict exceeding of the required reliability characteristic (s) of the battery, the release of scenarios, the selection of the safe-stop scenario, the trigger for battery replacement, the trigger of a transition to the safe state or the driver transfer.

For this purpose, a release request 202, which comes from the control unit, is shown in FIG. For this, as inputs for block 202 are provided an allowable failure probability 204, a current time tst 206 and a period Atintervaii 208, which is scheduled for the battery change, the so-called change interval of the battery task of the second module 200 is the reliability characteristics of

Predict battery and make release decisions or select safe-stop scenarios. In this case, the permissible size of the reliability parameter is communicated by the higher-level control unit or it is already stored in the control unit or in the battery sensor. An example of the allowable reliability characteristic is one specific Failure probability of the battery or the maintenance of non-failure time in a three-parameter Weibull distribution.

In the second module 200, the load capacity model of the battery of

Failure probability across battery cycles at reference level

Failure probability over operating time converted. For this purpose, the quotient of previously seen load and previous operating time is formed.

In this second module 200, the battery change time

be predicted. It is assumed that the ratio of

Load and operating time is constant and with this approach a linear forecast of the remaining operating life of the battery is performed. Conceivable also approaches with non-constant ratio of load and

Operating time.

If the predicted remaining operating time falls below a certain limit value, the transition into the safe state or the driver transfer can be initiated at an early stage so that a critical vehicle state is avoided.

FIG. 3 illustrates in a flowchart a possible sequence of the method using all four modules.

To the first module:

In a memory element 300, traces of SOC 302 and temperature T 304 are stored over time. These gradients are classified by means of Rainflow count 306. A resulting rainflow matrix 308 is generated by means of a

Wöhler curve 310 converted to a reference level. This results in the number of reference cycles. By means of a resilience model 312, in this

Case of the Weibull distribution, the calculation is done

Failure probability F (n) 314.

To the second module: A number of possible errors 320 may be combined with possible scenarios 324, in particular start-stop scenarios, and conditions 326, resulting in reference cycles 330 that are added to the number 311. From the Weibull distribution 312, a prediction 334 of different scenarios additionally results. The output is done as a vector.

In addition, the time t, _s t 340 and the time interval until the next change in interval 342 are entered in a block 346 in which a conversion of battery cycles into time cycles takes place. This allows the Weibull distribution of failure probability across battery cycles to reference level in

Default probability over time to be converted. Furthermore, the probability of failure results until the next change interval after F (t '= Atintervaii + t) at the output 348.

To the third module:

In this, the Weibull distribution or the resilience model 312 can be adapted. For this purpose, damage to reference level 360 based on SOC is subjected to extrapolation 362. Furthermore, SOH 364 is taken into account by the battery sensor 366. This results in a new failure-free Teit to 370 or a correction factor for the Weibull distribution or the resilience model 312.

Lines illustrate the fourth module 380, which states at what times or after which step a cloud could be pulled in.

Release decisions are explained below:

On-line, the calculating control unit or the battery sensor checks which scenarios are admissible from a reliability point of view and which are not. It can for each scenario, the number of required

Reference cycles per operating period be stored. Alternatively, this value can also be determined online by simulating the respective scenarios and calculating according to "first module, load". Depending on the result, the release granted, issued for a specified period or not issued. The result is communicated, for example, in the form of a release vector to the higher-level control unit.

Exemplary scenarios that influence the damage to the battery and whose release is checked are:

- Operating modes (manual drive, automated drive, sailing, recuperation, ...)

- Operating strategies

When releasing, the following cases can be distinguished:

Case I: higher-level control unit inquires operating mode and duration, d. H. the operating strategy is known

As an example: The driver enters a destination in the navigation device and the system control then makes a request for the release of operating modes and their duration.

For the requested parameters, namely duration, operating mode and

Operating strategy, the "required" reference cycle number is determined and added to the previously seen load on reference level. It is now checked whether the specified reliability limit value is adhered to. If this is adhered to, the requested case is released, otherwise this is not done.

Case II: higher-level control unit continuously interrogates battery sensor or calculating control unit continuously or battery sensor or calculating control unit continuously reports remaining time for all operating modes to the higher-level control unit

In case II is used for all possible combinations of operating modes and

Operating strategies the duration until reaching the specified

Reliability limit determined and passed to the higher-level control unit. Thus, the periods are available, how long each may be driven and there is a time-limited release of Functions. If the vehicle is in an operating mode operating strategy combination in which the battery failure is imminent, it is possible to switch to a battery-saving combination or to initiate the transition to the safe state or a driver transfer.

To select the Safe Stop scenario, execute:

On-line, the calculating control unit or battery sensor checks which safe-stop scenarios are admissible from a reliability point of view and which are not. It can for each scenario, the number of required

Reference cycles are stored. Alternatively, this value can also be determined online by simulating the respective scenarios and calculating according to "Module I, load".

Possible influencing parameters on the choice of the safe-stop scenario are:

- Safe stop scenario (stopping in the lane, driving to the right)

Roadside, ...)

- fault detected in the energy on-board network (electrical system fault)

- Operating strategy

Case I: higher-level control unit queries safe-stop scenario / scenarios, with known operating strategy and identified error

The required reference cycle number is determined for the requested combination of safe-stop scenario, vehicle electrical system failure and operating strategy. This is added to the previous load at reference level and checks whether the defined reliability limit value is adhered to. If this is the case, the combination, z. B. as a result vector to the higher-level control unit, released.

Case II: higher-level control unit continuously interrogates battery sensor or calculating control unit continuously or battery sensor or calculating control unit continually reports possible safe-stop scenarios combined with Operating modes and errors in the electrical system, so will be the results of

Received fault injection simulation at the board level

The required reference cycle number is determined for all possible combinations of safe-stop scenario, vehicle electrical system failure and operating strategy. For each combination, the required reference cycle number is added to the previous load at reference level and it is checked whether the defined

Reliability limit is met. If this is the case, the

Combination released. This procedure is repeated for each combination and the result sent to the higher-level control unit, eg. In the form of a solution vector.

The third module has the task to write down the reference value of the actual aging of the battery (SOH - capacity loss) and its course until reaching the failure criterion, eg. B. capacity loss of 20%, to extrapolate over the service life or the experienced load. By the value thus obtained, the quality of the battery compared to the population of batteries can be taken into account and the previously used

Resilience model, z. B. by redefining the failure-free time or a correction factor to the battery grade, to be adjusted.

The fourth module is to use the prediction to correct the

Capacity model. For this, the fourth module supplies the damage experienced by the battery (SOH) via load and the extrapolated value (see third module) to a cloud storage. There, based on the large number of damage via load data or extrapolated values, the

Resilience model optimized and sent back to the fourth module. In this way, the underlying resilience model is continuously improved.

Optionally, it can be provided:

- The fourth module now knows the quality of the built-in battery compared to the population and can the quality of the battery z. B. over a

Take into account "correction factor", - For battery errors that occur only in the field, the release of error-causing operating modes can be denied through the cloud until the error z. B. is repaired by a replacement and thus the failure and the resulting critical vehicle condition can be avoided.

Other benefits of sharing with the cloud include:

- Realized battery life for future component / system developments / designs,

- Adaptation of the operating strategy via the cloud (objective: optimal

Components utilization)

- If battery replacement due to regular maintenance is imminent (predicted battery life is not enough for the subsequent maintenance measure), but the battery can still bear load, the operating strategy is chosen so that the battery is more heavily loaded to other components, such. B. DC / DC converter, to spare,

- automated communication with the workshop, if predicted

Battery life is nearing the end to swap components,

- Prediction of the load by knowing the route profile based on navigation data (start-destination route).

The proposed method enables the cloud-based derivation of changes to the operating strategies, if necessary, in order to reduce failures of the battery. This enables a balanced operating strategy with consideration of all relevant on-board network components.

Thus, an improvement of the component and system development and their interpretation can be achieved by field data acquisition. Also, an improvement in the resilience models due to a large number of components in the field, eg. B. by Deep-learning, possible. In addition, can an improvement of the stress models due to known, real

Component loads are achieved.

The method and the arrangement may be in any vehicle in which the probability of failure of the components and / or a

System reliability analysis is to be implemented. Basically, an application in any vehicle in which the release of certain functions or the choice of error response (safe-stop scenario) depending on the predicted behavior (based on the previous load) is to be granted, possible.

An insert may be provided in all vehicles in which the

Vehicle electrical system has high safety relevance, such. B. vehicles with sailing operation, recuperation or automated vehicles. Furthermore, vehicles with electrical brake booster (iBooster, IPB) are conceivable as the place of use. It should be noted that there are currently efforts to move away from kilometer or time interval based maintenance to condition based maintenance. The presented method can also be used for such state-based maintenance.

The evaluation algorithm described here, which is implemented by the method, can be implemented in a battery sensor, a control unit or in a computer in the vehicle or outside the vehicle, for example in a cloud. As the battery temperature has a great influence on the battery damage, reliability and life, for example

Destination information of the navigation device, the local temperature and other temperature forecasts are integrated into the analysis, so as to predict a failure of the battery can more accurately. The analysis of the battery damage can in sections, eg.

depending on the month, in order to be able to determine the damage in the respective section and to be able to better predict service intervals and failures. In this way influences, such as the temperature, are taken into consideration. Also, comparisons of the forecasts for the next few days can be taken into account in this way.

Claims

claims

A method for monitoring a battery in a motor vehicle, wherein a first module (180) determines operating quantities of the battery and compares magnitudes representing the operating quantities with a load capacity model (312) to determine reliability characteristics of the battery in this way that a future behavior of the battery can be predicted.

2. The method of claim 1, wherein the sizes are determined by converting the operating variables.

3. The method of claim 1, wherein the sizes correspond to the operating variables.

4. The method according to any one of claims 1 to 3, wherein the operating variables are at least partially provided by a battery sensor (100, 366).

5. The method according to any one of claims 1 to 4, are implemented in the limit values for the reliability characteristics.

6. The method according to any one of claims 1 to 5, wherein the determined

Reliability characteristics are further processed to

Determine system reliability characteristics.

7. The method according to any one of claims 1 to 6, wherein a second module (200) evaluated different scenarios and possibly released.

8. The method of claim 1, wherein a third module performs a balancing of the loadability model with extrapolation of an actual SOH, and optionally adjusts the loadability model.

The method of any one of claims 1 to 8, wherein a fourth module (380) balances an SOH (364) with at least one central database.

10. The method of claim 9, wherein a comparison is made with a cloud.

11. Arrangement for monitoring a battery in a motor vehicle, which is adapted to carry out a method according to one of claims 1 to 10.

12. The assembly of claim 11 implemented in a battery sensor (100, 366).

13. Arrangement according to claim 11 or 12, which is adapted to carry out a processing of data cloud-based.